Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Source controls for runoff treatment : hydrologic response and water quality attenuation in urban catchments Stime, Samuel Eugene 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_february_stime_samuel.pdf [ 11.4MB ]
Metadata
JSON: 24-1.0166099.json
JSON-LD: 24-1.0166099-ld.json
RDF/XML (Pretty): 24-1.0166099-rdf.xml
RDF/JSON: 24-1.0166099-rdf.json
Turtle: 24-1.0166099-turtle.txt
N-Triples: 24-1.0166099-rdf-ntriples.txt
Original Record: 24-1.0166099-source.json
Full Text
24-1.0166099-fulltext.txt
Citation
24-1.0166099.ris

Full Text

  SOURCE CONTROLS FOR RUNOFF TREATMENT: HYDROLOGIC RESPONSE AND WATER QUALITY ATTENUATION IN URBAN CATCHMENTS by Samuel Eugene Stime  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2014 © Samuel Eugene Stime, 2014   ii  Abstract The lower Fraser Valley receives considerable rainfall, the vast majority of which percolated to groundwater for summer recharge of streams that supported healthy ecosystems and robust salmon populations. But as the cities in the valley grew and expanded, land use rapidly changed and impervious surface area from roads, parking lots, and buildings, like elsewhere, have contributed to increased flooding events, reduced groundwater recharge, lower sum-mer base flow in urban streams, and reduced water quality that has led to significantly compromised biotic integrity and reduced biodiversity in the streams. Low impact development best management practices that promote infiltra-tion of rainfall where it falls, reduce peak runoff flows and total event runoff volumes, and offer a level of treatment of contaminated waters, are slowly gaining recognition and use. This investigation was performed in two parts. First, using a paired catchment approach in the Township of Langley, it examines whether reductions in peak runoff flow and total event runoff volumes are proportionate to the effective reduction in directly connected impervious area, by installation of a suite of on-lot best management practices in one catchment. Second, using a treatment effectiveness approach in the District of Maple Ridge, it examines whether a roadside infiltration swale is removing contaminants and improving water quality both as predicted by the literature, and as required by Canadian guidelines for the pro-tection of aquatic life. By disconnecting roof areas from direct connection to storm sewer, amending soil depths, and installing on-lot infiltration trenches in the Routley catchment in the Township of Langley, the hydraulic benefits are nearly as great as the scope of practices applied. Years after construction of a progressive suite of practices in Maple Ridge, the Foreman Drive bioswale appears to be removing contaminants as designed, and providing significant wa-ter quality benefits.  iii  Preface This dissertation is original, unpublished work by the author, Samuel Stime. The research described was undertaken under the supervision of J. Atwater, and H. Schreier, D. Zabil, and C. Johnston provided input on the design and methods of the study. The flow analysis in Chapter 2 makes use of a method developed by the engineering firm Kerr Wood Leidal Associ-ates Ltd. for calculating and removing the portion of stream flow originating from groundwater and interflow. Origi-nally used on annual hydrographs, its modified form is here used on dozens of individual storms. Water quality analysis described in Chapter 3 was performed in UBC’s Environmental Engineering laboratory. T. Ma and P. Parkinson performed nutrient, metals, and organic carbon analysis on all of the samples that I had col-lected and prepared.   iv  Table of Contents Abstract......................................................................................................................................................................... ii Preface ......................................................................................................................................................................... iii Table of Contents..........................................................................................................................................................iv List of Tables ................................................................................................................................................................. x List of Figures ............................................................................................................................................................. xii List of Abbreviations & Nomenclature ....................................................................................................................... xv Acknowledgements ................................................................................................................................................. xviii Dedication.................................................................................................................................................................... xx Chapter 1: Introduction .................................................................................................................................................. 1 1.1 Background ..................................................................................................................................................... 1 1.2 Study goals & objectives ................................................................................................................................. 2 1.3 Scope & general methodology of research ...................................................................................................... 2 Chapter 2: Selected BMPs’ impact on hydrology .......................................................................................................... 3 2.1 Introduction & project description .................................................................................................................. 3 2.2 Objectives ........................................................................................................................................................ 3 2.3 Hypothesis ....................................................................................................................................................... 3 2.4 Literature review ............................................................................................................................................. 4 2.4.1 Urban impacts on hydrology .................................................................................................................. 4 2.4.1.1 Fate of precipitation ..................................................................................................................... 4 2.4.1.2 Ways land use changes ................................................................................................................. 5 2.4.1.3 Impacts on Hydrology .................................................................................................................. 5 2.4.1.3.1 Less rainfall infiltrated ............................................................................................................ 5 2.4.1.3.2 Lower summer base flow ........................................................................................................ 6 2.4.1.3.3 Higher peak flows & cumulative runoff .................................................................................. 6 2.4.1.3.4 Reduced time to peak .............................................................................................................. 7 2.4.1.3.5 Increased return frequency of runoff of a given size ............................................................... 7 2.4.1.3.6 Sediments associated with construction .................................................................................. 8 2.4.1.3.7 Changes to stream geometry .................................................................................................... 8 2.4.1.3.8 Sediments associated with bed & bank erosion ....................................................................... 9 v  2.4.1.3.9 Total vs. effective impervious area .......................................................................................... 9 2.4.1.4 Impacts on biota ......................................................................................................................... 10 2.4.1.4.1 Deposition of sediments ........................................................................................................ 10 2.4.1.4.2 Increased flow speeds ............................................................................................................ 10 2.4.1.4.3 Compromised biological integrity ......................................................................................... 11 2.4.2 Stormwater management evolution & impact on streams ................................................................... 11 2.4.2.1 Evolving stormwater management focus ................................................................................... 12 2.4.2.2 Low impact development best management practices ............................................................... 12 2.4.2.2.1 Disconnected roof leaders ..................................................................................................... 13 2.4.2.2.2 Amended soil depths ............................................................................................................. 14 2.4.2.2.3 Infiltration trenches ............................................................................................................... 15 2.4.2.2.4 Roadside bioswales ............................................................................................................... 16 2.4.3 Regulatory framework ......................................................................................................................... 18 2.4.3.1 National ...................................................................................................................................... 18 2.4.3.2 Provincial ................................................................................................................................... 19 2.4.3.3 Metro Vancouver........................................................................................................................ 20 2.5 Methodology ................................................................................................................................................. 21 2.5.1 Study sites in the Township of Langley ............................................................................................... 21 2.5.1.1 Infrastructure & context of flow monitoring sites ...................................................................... 22 2.5.1.2 Build-out conditions & time frame ............................................................................................ 23 2.5.1.3 Drainage catchment area ............................................................................................................ 24 2.5.1.4 Total impervious area ................................................................................................................. 26 2.5.1.5 Soils & infiltration rates ............................................................................................................. 31 2.5.1.6 Slope & time of concentration .................................................................................................... 35 2.5.2 Rainfall ................................................................................................................................................ 39 2.5.2.1 Rainfall statistics ........................................................................................................................ 41 2.5.2.1.1 Annual average total .............................................................................................................. 41 2.5.2.1.2 Intensity-duration-frequency ................................................................................................. 42 2.5.3 Flow measurement ............................................................................................................................... 44 2.5.4 Manipulation of flow data .................................................................................................................... 45 2.5.5 Normalization by area .......................................................................................................................... 48 2.5.6 Removal of groundwater & interflow .................................................................................................. 48 2.5.7 Identification of discrete rainfall events ............................................................................................... 50 2.5.8 Rainfall event grouping by magnitude ................................................................................................. 51 vi  2.5.9 Discrete event runoff analysis .............................................................................................................. 53 2.5.10 Ratio of catchments’ peak runoff flows .......................................................................................... 54 2.5.11 Unitize catchments’ cumulative runoff volumes ............................................................................. 54 2.5.12 Comparison of cumulative event runoff .......................................................................................... 55 2.5.13 Statistical comparisons of magnitude-parsed storms ...................................................................... 56 2.5.14 Peak flow ........................................................................................................................................ 56 2.5.15 Cumulative total runoff ................................................................................................................... 56 2.6 Results & discussion ..................................................................................................................................... 56 2.6.1 Rainfall ................................................................................................................................................ 56 2.6.2 Runoff .................................................................................................................................................. 60 2.6.3 Peak flow comparisons ........................................................................................................................ 63 2.6.4 Event runoff comparisons .................................................................................................................... 64 2.6.5 Seasonal differences ............................................................................................................................ 66 2.6.6 Performance evaluation ....................................................................................................................... 68 2.7 Conclusions ................................................................................................................................................... 72 Chapter 3: Selected BMPs’ impact on water quality ................................................................................................... 74 3.1 Introduction ................................................................................................................................................... 74 3.2 Objectives ...................................................................................................................................................... 75 3.3 Hypothesis ..................................................................................................................................................... 75 3.4 Literature review ........................................................................................................................................... 75 3.4.1 Regulatory framework ......................................................................................................................... 75 3.4.2 Water quality impacts of urbanization ................................................................................................. 77 3.4.2.1 Solids .......................................................................................................................................... 77 3.4.2.2 Nutrients ..................................................................................................................................... 78 3.4.2.2.1 Nitrogen ................................................................................................................................. 78 3.4.2.2.2 Phosphorus ............................................................................................................................ 81 3.4.2.3 Organic carbon ........................................................................................................................... 82 3.4.2.4 Chloride ...................................................................................................................................... 82 3.4.2.5 Hardness ..................................................................................................................................... 82 3.4.2.6 Alkalinity.................................................................................................................................... 83 vii  3.4.2.7 Metals ......................................................................................................................................... 83 3.4.2.7.1 Copper ................................................................................................................................... 84 3.4.2.7.2 Cadmium ............................................................................................................................... 85 3.4.2.7.3 Iron ........................................................................................................................................ 85 3.4.2.7.4 Lead ....................................................................................................................................... 86 3.4.2.7.5 Nickel .................................................................................................................................... 86 3.4.2.7.6 Zinc ........................................................................................................................................ 87 3.4.2.7.7 Aluminum .............................................................................................................................. 87 3.4.2.8 Temperature ............................................................................................................................... 88 3.4.2.9 First flush ................................................................................................................................... 88 3.4.3 Infiltration swales improvements to water quality ............................................................................... 89 3.4.3.1 Sediments ................................................................................................................................... 91 3.4.3.2 Nutrients ..................................................................................................................................... 91 3.4.3.2.1 Nitrogen ................................................................................................................................. 91 3.4.3.2.2 Phosphorus ............................................................................................................................ 92 3.4.3.3 Organic carbon ........................................................................................................................... 92 3.4.3.4 Chloride ...................................................................................................................................... 92 3.4.3.5 Metals ......................................................................................................................................... 93 3.4.3.5.1 Copper ................................................................................................................................... 93 3.4.3.5.2 Cadmium ............................................................................................................................... 93 3.4.3.5.3 Iron ........................................................................................................................................ 93 3.4.3.5.4 Lead ....................................................................................................................................... 94 3.4.3.5.5 Nickel .................................................................................................................................... 94 3.4.3.5.6 Zinc ........................................................................................................................................ 94 3.4.3.5.7 Aluminum .............................................................................................................................. 94 3.4.3.6 Temperature ............................................................................................................................... 94 3.4.3.7 First flush ................................................................................................................................... 94 3.5 Methodology ................................................................................................................................................. 95 3.5.1 Study sites in the District of Maple Ridge ........................................................................................... 96 3.5.1.1 Build-out conditions & timeframe ............................................................................................. 96 3.5.1.2 Best management practices installed & conditions .................................................................... 97 3.5.1.3 Roadside bioswale details .......................................................................................................... 97 3.5.1.4 Rainfall statistics ........................................................................................................................ 99 3.5.2 Rainfall measurement techniques ...................................................................................................... 100 3.5.3 Treatment effectiveness sampling plan .............................................................................................. 101 viii  3.5.4 Stormwater sampling equipment & methods ..................................................................................... 102 3.5.4.1 Sampling locations ................................................................................................................... 102 3.5.4.2 ISCO autosamplers ................................................................................................................... 106 3.5.4.3 Water sensor & sampling control ............................................................................................. 107 3.5.4.4 Datalogger with telemetry ........................................................................................................ 107 3.5.4.5 Collection & transport .............................................................................................................. 108 3.5.4.6 Quality control & trip blanks.................................................................................................... 108 3.5.5 Laboratory analysis of water quality parameters ............................................................................... 109 3.5.5.1 Compositing & preservation .................................................................................................... 109 3.5.5.2 Analytical methods ................................................................................................................... 110 3.5.6 Statistical analysis on data ................................................................................................................. 110 3.6 Results & discussion ................................................................................................................................... 111 3.6.1 Accuracy & precision ........................................................................................................................ 111 3.6.1.1 Laboratory blanks ..................................................................................................................... 111 3.6.1.2 Field blanks .............................................................................................................................. 112 3.6.2 Rainfall & sampling ........................................................................................................................... 113 3.6.3 Sediments ........................................................................................................................................... 116 3.6.4 Nutrients ............................................................................................................................................ 121 3.6.4.1 Nitrogen.................................................................................................................................... 121 3.6.4.2 Phosphorus ............................................................................................................................... 124 3.6.5 Organic carbon ................................................................................................................................... 127 3.6.6 Chloride ............................................................................................................................................. 128 3.6.7 Hardness ............................................................................................................................................ 130 3.6.8 Alkalinity ........................................................................................................................................... 131 3.6.9 pH ...................................................................................................................................................... 133 3.6.10 Metals ............................................................................................................................................ 134 3.6.10.1 Copper ...................................................................................................................................... 134 3.6.10.2 Cadmium .................................................................................................................................. 134 3.6.10.3 Iron ........................................................................................................................................... 136 3.6.10.4 Lead .......................................................................................................................................... 138 3.6.10.5 Nickel ....................................................................................................................................... 138 3.6.10.6 Zinc .......................................................................................................................................... 139 3.6.10.7 Aluminum ................................................................................................................................ 140 ix  3.6.11 Temperature .................................................................................................................................. 142 3.6.12 Performance evaluation ................................................................................................................. 143 3.7 Conclusion ................................................................................................................................................... 146 Chapter 4: Conclusions & recommendations ............................................................................................................ 147 4.1 Source controls for hydrologic response ..................................................................................................... 147 4.2 Source controls for improved runoff water quality ..................................................................................... 148 Bibliography .............................................................................................................................................................. 150 Appendices ................................................................................................................................................................ 162 Appendix A Hydrologic data for 128 rainfall events in the Township of Langley ............................................... 163 Appendix B Raw water quality data ..................................................................................................................... 292 Appendix C Precipitation ...................................................................................................................................... 297 C.1 Routley & Langley Meadows & Township of Langley rain .............................................................. 297 C.2 Foreman Drive & Malcolm Knapp Research Forest .......................................................................... 302 Appendix D Silver Valley rainfall and sampling times ........................................................................................ 307 Appendix E Water sensor circuit & code .............................................................................................................. 318 Appendix F Township of Langley letter of introduction ...................................................................................... 327 Appendix G District of Maple Ridge letters & permissions ................................................................................. 328 Appendix H Letters from the researcher and UBC ............................................................................................... 334 Appendix I Reflections on the work ..................................................................................................................... 337    x  List of Tables Table 2.1 Source controls’ contribution to flow mitigation objectives ........................................................................ 13 Table 2.2 Percent rainfall capture by 300mm absorbent soils, I/P ratio of 2 ............................................................... 14 Table 2.3 Summary of impervious areas in paired TOL catchments ........................................................................... 31 Table 2.4 Soil drainage in catchments under investigation ......................................................................................... 34 Table 2.5 Hydraulic conductivity combinations for soils comparison......................................................................... 34 Table 2.6 Mean saturated hydraulic conductivity calculated for three Ks combinations ............................................ 35 Table 2.7 Topography of study catchments ................................................................................................................. 38 Table 2.8 Routley and Langley Meadows drainage network comparison ................................................................... 39 Table 2.9 Rainfall intensity (mm) for given storm durations and return frequencies .................................................. 43 Table 2.10 Cumulative monthly rainfall depths compared .......................................................................................... 58 Table 2.11 Monthly rainfall peak intensities compared ............................................................................................... 58 Table 2.12 Assumed and range of values for sensitivity analysis................................................................................ 70 Table 3.1 BC Environment, CCME, and USEPA aquatic life water quality guidelines .............................................. 76 Table 3.2 Residential urban runoff TSS concentrations, from the literature ............................................................... 77 Table 3.3 Residential urban runoff TDS concentrations, from the literature ............................................................... 78 Table 3.4 Average annual atmospheric deposition loadings of nutrients (kg/ha/year) ................................................ 79 Table 3.5 Residential urban runoff nitrogen concentrations, from the literature ......................................................... 80 Table 3.6 Residential urban runoff total phosphorus concentrations, from the literature ............................................ 81 Table 3.7 Residential urban runoff chloride concentrations, from the literature ......................................................... 82 Table 3.8 Residential urban runoff hardness, from the literature ................................................................................ 83 Table 3.9 Sources of heavy metals found in stormwater runoff (Whalen & Cullum 1998) ........................................ 84 Table 3.10 Average annual atmospheric deposition loadings of metals (kg/ha/year) .................................................. 84 Table 3.11 Residential urban runoff copper concentrations, from the literature ......................................................... 85 Table 3.12 Residential urban runoff cadmium concentrations, from the literature ..................................................... 85 Table 3.13 Residential urban runoff iron concentrations, from the literature .............................................................. 86 Table 3.14 Residential urban runoff lead concentrations, from the literature .............................................................. 86 Table 3.15 Residential urban runoff nickel concentrations, from the literature........................................................... 87 Table 3.16 Residential urban runoff zinc concentrations, from the literature .............................................................. 87 Table 3.17 Residential urban runoff aluminum concentrations, from the literature .................................................... 88 Table 3.18 Guideline, typical concentration, and expected reduction through bioswales for water quality parameters ..................................................................................................................................................................................... 96 Table 3.19 Preservation requirements for analytical parameters ............................................................................... 109 Table 3.20 Analytical methods of water quality parameters...................................................................................... 110 Table 3.21 Laboratory blank samples water quality results ....................................................................................... 112 Table 3.22 Field blank samples water quality results ................................................................................................ 113 Table 3.23 Discrete samples collected for each of the 22 rainfall events in Silver Valley ........................................ 114 xi  Table 3.24 TKN measured in road runoff and BMP effluent in Silver Valley .......................................................... 122 Table 3.25 NOX concentrations for 20 rainfall events in Silver Valley ..................................................................... 123 Table 3.26 TP measured in road runoff and BMP effluent in Silver Valley .............................................................. 125 Table 3.27 Chloride concentrations measured in each rainfall event ........................................................................ 129 Table 3.28 Hardness calculated from calcium and magnesium for each sample ....................................................... 130 Table 3.29 Measured pH in road runoff and BMP effluent in Silver Valley ............................................................. 133 Table 3.30 Cadmium concentrations in Silver Valley runoff & treated effluent samples ......................................... 135 Table 3.31 Iron concentrations in Silver Valley runoff & treated effluent samples .................................................. 136 Table 3.32 Nickel concentrations in Silver Valley runoff & treated effluent samples .............................................. 139 Table 3.33 Aluminum concentrations in Silver Valley runoff & treated effluent samples ........................................ 140 Table 3.34 Runoff and treated effluent temperature in Silver Valley ........................................................................ 142 Table 3.35 Summary water quality benefits of the Foreman Drive infiltration swale ............................................... 144  xii  List of Figures Figure 2.1 Fate of precipitation as a catchment is urbanized ......................................................................................... 5 Figure 2.2 Typical cross section of Foreman Drive bioswale ...................................................................................... 17 Figure 2.3 Paired watersheds in the Township of Langley .......................................................................................... 22 Figure 2.4 Typical self-draining lot with front & rear lot drainage outside of driveway ............................................. 24 Figure 2.5 Langley Meadows (bottom) & Routley (top) catchments .......................................................................... 25 Figure 2.6 Langley Meadows and Routley catchments ............................................................................................... 27 Figure 2.7 Langley Meadows and Routley drainage networks .................................................................................... 27 Figure 2.8 Langley Meadows and Routley roof contributions to TIA ......................................................................... 28 Figure 2.9 Langley Meadows and Routley road contributions to TIA ........................................................................ 28 Figure 2.10 Langley Meadows and Routley sidewalk contributions to TIA ............................................................... 29 Figure 2.11 Langley Meadows and Routley driveway and other on-lot contributions to TIA .................................... 29 Figure 2.12 Langley Meadows and Routley pervious and impervious areas ............................................................... 30 Figure 2.13 Soil map of the catchments under investigation (Luttmerding 1980) ....................................................... 33 Figure 2.14 Collection network branches defining longest time to outlet in the study catchments ............................. 37 Figure 2.15 Time of channel flow as a function of pipe-full depth ............................................................................. 38 Figure 2.16 Existing rain gauge relative to study catchment locations. ....................................................................... 40 Figure 2.17 Rain gauges installed in Langley Meadows (left) and Routley (right) neighborhoods ............................ 41 Figure 2.18 Cumulative annual rainfall at TOL Municipal Hall for last eight years ................................................... 42 Figure 2.19 Monthly rainfall at TOL Municipal Hall .................................................................................................. 42 Figure 2.20 Intensity-duration-frequency curves for the study area (FlowWorks 2014) ............................................. 43 Figure 2.21 Illustration of SFE’s custom compound weirs at various flows ............................................................... 44 Figure 2.22 Flow chart of methodology to compare flow from the catchments .......................................................... 47 Figure 2.23 Estimation of the GWI contribution to event runoff ................................................................................ 49 Figure 2.24 Removal of GWI from flow to determine direct runoff contribution to stream flow ............................... 50 Figure 2.25 Distribution of rainfall events by magnitude (with 6-hour ADWP) ......................................................... 51 Figure 2.26 Distribution of three-year rainfall volume by event magnitude (with 6-hour ADWP)............................. 52 Figure 2.27 Distribution of three-year rainfall volume in events of 2-60mm .............................................................. 52 Figure 2.28 Partitioning of rainfall events by approximate contribution quartile ........................................................ 53 Figure 2.29 Comparison of catchments’ runoff-induced peak flows (03-04 December 2012) .................................... 54 Figure 2.30 Unitized axes to compare cumulative rainfall and runoff (04 December 2012) ....................................... 55 Figure 2.31 Rain gauges compared for mid-December 2013 ...................................................................................... 57 Figure 2.32 15 September 2013 storm as measured by the three rain gauges ............................................................. 59 Figure 2.33 27 December 2013 storm as measured by the three rain gauges .............................................................. 59 Figure 2.34 09 January 2014 storm as measured by three rain gauges ........................................................................ 60 Figure 2.35 Peak runoff compared for 12-13 January 2011 event (55% reduction) .................................................... 61 Figure 2.36 Peak runoff compared for 25 December 2012 event (18% reduction) ..................................................... 61 xiii  Figure 2.37 Peak runoff compared for 28 September 2013 event (45% reduction) .................................................... 62 Figure 2.38 Cumulative runoff compared for 12-13 January 2011 event .................................................................... 62 Figure 2.39 Cumulative runoff compared for 25 December 2012 event ..................................................................... 63 Figure 2.40 Cumulative runoff compared for 28 September 2013 event ..................................................................... 63 Figure 2.41 Mean Routley peak flow reduction and 95% CI for rainfall size groups ................................................. 64 Figure 2.42 Mean & 95% CI of ratio of runoff to rainfall, by catchment and event size ............................................ 65 Figure 2.43 Reductions of event cumulative runoff volume in the Routley catchment ............................................... 66 Figure 2.44 Mean and 95% CI of peak runoff reduction between wet and dry months .............................................. 67 Figure 2.45 Mean & 95% CI of ratio of runoff to rainfall, by season ......................................................................... 67 Figure 2.46 Routley cumulative runoff reductions compared by season ..................................................................... 68 Figure 2.47 On-lot BMPs to capture and infiltrate rainfall .......................................................................................... 69 Figure 2.48 Sensitivity analysis on variables affecting rainfall capture ...................................................................... 71 Figure 2.49 Peak flow and event runoff volume reductions compared to connected impervious area reduction ........ 72 Figure 3.1 CCME total ammonia maximum concentrations as a function of pH and temperature ............................. 81 Figure 3.2 Typical longitudinal profile of road-side infiltration swale ........................................................................ 91 Figure 3.3 Upper Foreman Drive, with roadside swale, parkette, and detention ponds .............................................. 98 Figure 3.4 Typical longitudinal cross-section of roadside infiltration swale along Foreman Drive ............................ 98 Figure 3.5 Average monthly rainfall and temperature at MKRF, nearby Silver Valley .............................................. 99 Figure 3.6 Tipping-bucket rain gauge installed near sampling site along Foreman Drive ........................................ 100 Figure 3.7 MKRF and Foreman Drive hourly rainfall comparison ........................................................................... 101 Figure 3.8 Road Runoff and Treated Effluent sampling sites along Foreman Drive ................................................. 102 Figure 3.9 Road runoff stilling bucket & collection site along Foreman Drive ......................................................... 103 Figure 3.10 PVC pipe protecting road runoff sampling hose along Foreman Drive ................................................. 104 Figure 3.11 Treated effluent stilling bucket & collection site in the parkette ............................................................ 105 Figure 3.12 PVC pipe housing sampling tubing & control wire in the parkette ........................................................ 105 Figure 3.13 Programmable autosampler in protective box along Foreman Drive ..................................................... 106 Figure 3.14 Discrete sampling schedule for road runoff and treated effluent ............................................................ 106 Figure 3.15 Sampler control unit and water level sensor circuit ................................................................................ 107 Figure 3.16 Autosampler, sensor/controller, datalogger, and battery deployed in the field ...................................... 108 Figure 3.17 Rainfall, and timing of road runoff & treated effluent samples for 14 February event .......................... 115 Figure 3.18 Rainfall & sampling times over spring 2014 .......................................................................................... 116 Figure 3.19 TSS concentrations (linear scale) for 22 rainfall events in Silver Valley ............................................... 117 Figure 3.20 TSS concentrations (log scale) for 22 rainfall events in Silver Valley ................................................... 117 Figure 3.21 TSS concentration reduction for 17 paired events in Silver Valley ....................................................... 118 Figure 3.22 Mean TSS & 90% CI for road runoff and treated effluent ..................................................................... 118 Figure 3.23 TDS concentrations (linear scale) for 22 rainfall events in Silver Valley .............................................. 119 Figure 3.24 TDS concentrations (log scale) for 22 rainfall events in Silver Valley .................................................. 119 xiv  Figure 3.25 TDS effluent lag and steady-state concentration .................................................................................... 120 Figure 3.26 Mean TDS concentrations & 90% CI for road runoff and treated effluent ............................................ 121 Figure 3.27 NOX concentrations for 20 rainfall events in Silver Valley .................................................................... 123 Figure 3.28 Mean NOX concentrations & 90% CI for road runoff and treated effluent ............................................ 124 Figure 3.29 Road runoff and BMP effluent total phosphorus concentrations for 14 rainfall events ......................... 126 Figure 3.30 Mean total phosphorus concentrations & 90% CI for runoff and treated effluent samples .................... 127 Figure 3.31 Total organic carbon concentrations for 22 rainfall events in Silver Valley .......................................... 127 Figure 3.32 Mean total organic carbon & 90% CI for road runoff and treated effluent ............................................ 128 Figure 3.33 Chloride concentrations measured in road runoff and BMP effluent ..................................................... 129 Figure 3.34 Hardness calculated for road runoff and BMP effluent during 22 rainfall events .................................. 131 Figure 3.35 Mean hardness with 90% CI in road runoff and BMP effluent .............................................................. 131 Figure 3.36 Alkalinity measurements for 22 rainfall events in Silver Valley ............................................................ 132 Figure 3.37 Mean alkalinity & 90% CI for road runoff and treated effluent ............................................................. 132 Figure 3.38 Mean pH, with 90% CI, of runoff and treated effluent in Silver Valley ................................................ 134 Figure 3.39 Total iron concentrations in runoff and treated effluent, compared to CCME guidelines ...................... 137 Figure 3.40 Calculated mean total iron concentration and 90% CI for runoff and BMP effluent ............................. 138 Figure 3.41 Total aluminum concentrations in runoff and treated effluent (log scale), and CCME guideline .......... 141 Figure 3.42 Calculated mean total aluminum concentrations & 90% CI for runoff and BMP effluent ..................... 141 Figure 3.43 Runoff and treated effluent temperatures over spring 2014 in Silver Valley ......................................... 143    xv  List of Abbreviations & Nomenclature ‘  - foot µg/L  - microgram per liter AC  - asbestos concrete ADWP  - antecedent dry weather period Al  - aluminum BC  - British Columbia BMP  - best management practice BOD  - biochemical oxygen demand CCME  - Canadian Council of Ministers of the Environment Cd  - cadmium CI  - confidence interval Cl-  - chloride cm  - centimeter COD  - chemical oxygen demand CSO  - combined sewer overflow Cu  - copper DFO  - Department of Fisheries and Oceans DL  - detection limit DMR  - District of Maple Ridge DR  - depth of rock reservoir DS  - depth of soil EIA  - effective impervious area ET  - evapotranspiration Fe  - iron GIS  - geographic information systems GWI  - groundwater and interflow ha  - hectare xvi  I/P  - ratio of Impervious to Pervious area IDF  - intensity-duration-frequency km  - kilometer KS  - saturated hydraulic conductivity KWL  - Kerr Wood Leidal (Associates Ltd.) L  - liter LID  - low impact development LM  - Langley Meadows m  - meter m2  - square meters m3  - cubic meters mg/L  - milligram per liter MKRF  - Malcolm Knapp Research Forest mL  - milliliter mm  - millimeter mV  - millivolt n  - Manning’s n, or rock porosity N  - nitrogen NH3  - ammonia NH4+  - ammonium Ni  - nickel NO2-  - nitrite NO3-  - nitrate NOX  - nitrogen oxides (NO2- + NO3-) P  - phosphorus PE  - preparation error PO4  - organophosphate PVC  - polyvinyl chloride xvii  R  - design rainfall capture depth RH  - hydraulic radius RN  - Routley Neighborhood ROW  - right of way SO  - channel Slope t  - time TDP  - total dissolved phosphorus TDS  - total dissolved solids TIA  - total impervious area TKN  - total Kjeldahl nitrogen TN  - total nitrogen TOC  - total organic carbon TOL  - Township of Langley TP  - total phosphorus TSS  - total suspended solids UBC  - University of British Columbia USEPA  - United States Environmental Protection Agency V  - volt WQ  - water quality Zn  - zinc  xviii  Acknowledgements For the design and successful completion of both the hydrological analysis and the water quality sampling and anal-ysis, I have many people to recognize for the significant contributions that they made. This project could not have been completed without the guidance and support of my supervisor Jim Atwater, who urged me to challenge my as-sumptions at every turn. I would like to thank Timothy Ma and Paula Parkinson for the many technical and non-technical contributions they made to this research in the Environmental Laboratory, Hans Schreier for his generous contribution of two autosamplers and rain gauges, and Sietan Chieng for also generously lending me two more au-tosamplers for this work. I am grateful to Ionut Aron, with the Malcolm Knapp Research Forest, who went out of his way to provide me with climate data as I needed it. Art Kastelein with the Township of Langley, and Joe Dingwall, Stephen Judd, Ian Rennie, George Irwin, and David Pollock with the District of Maple Ridge all warmly welcomed my research effort and provided me with letters of introduction, GIS information, and permissions that allowed the research to take place at unique points in the Fraser Valley. Their letters of introduction and as-built maps of the neighborhood augmented my understanding of the drainage system. Marlo in the Routley Neighborhood and Mark in the Langley Meadows neighborhood kindly hosted rain gauges in their yard to support the research, and in Silver Valley, Ron & Linda and Ian & Donna welcomed me to their com-munity with open arms, making every effort to ensure that the research equipment was protected and that the re-search was fruitful. In fact, the water quality portion of this research simply would not have worked without the warmth of the community members that I met in Silver Valley, who are proud of their unique neighborhood’s rela-tionship to the idyllic creek that flows to the north of the neighborhood, who have taken ownership of the bioswale’s aesthetics and maintenance, and who took a personal interest in the process, success, and results of this project. That I was approached numerous times by parents wanting me to explain to their children what my research and the bioswales were about, was truly inspiring. One February evening when an install was taking longer than I had antici-pated, one neighbor brought me a mug of hot tea. Others welcomed me to investigate storm connections in their yards, and again and again showcased the hospitality of their neighborhood. This runoff water quality monitoring experience in Silver Valley was a privileged experience. This research was supported by an NSERC Industrial Postgraduate Scholarship and the generous partnership of Kerr Wood Leidal Associates Ltd., where I am indebted to David Zabil, Chrystal Campbell, Chris Johnston, Caroline Gort, Alan Tse, and Jason Vine, who over the last two years made liberal contributions of their time, vast experi-ence, and material & financial resources to ensure that my research was supported. All of the hydrologic data used in the first part of this thesis was made accessible by the folks at KWL, and their client, the Township of Langley, orig-inally gathered and quality controlled for a multi-year runoff monitoring program. I am grateful for access to the dataset. Thanks to my fellow students at UBC, including Greg Reynen who helped design a working model for the water sensor, Patricia Oka for her Excel expertise and encouragement during challenging decisions, and Zaki Abdullah for xix  amplifying an appreciation for the water quality benefits of this research to rivers and fish. I owe a deep debt of grat-itude also to Bjorn Stime and Guy Polden, who did not hesitate to accompany me for challenging fieldwork associ-ated with this research, even at 7 a.m. on New Year’s Day. And most of all, I am grateful for the inspiration, encouragement, and patience of my wife, Andrea Vasquez, who put up with my middle of the night trips to pick up samples, and late nights writing my thesis. I promise I’ll be around a little more now.   xx  Dedication                    Para mi esposa, Andrea Milagros Vásquez Fernández.1  Chapter 1: Introduction 1.1 Background Urban settlements both expand over geographic area, and densify once initially developed. Land use changes from natural forest or pasture, to annually-disturbed farmland, to urban development and a 40-60 year cycle of intensify-ing redevelopment. With these changes comes removal of vegetation that once provided macropores through which rainfall could penetrate to aquifers, evapotranspired moisture from the ground, and shade streams. The hard and im-pervious surfaces of concrete and asphalt roads, and of the roofs of homes and apartment buildings and industrial complexes and commercial districts, and all the connecting sidewalks and decks and driveways, take the place of porous native soils with thick and absorbent organic layers containing the riot of biota constantly cycling nutrients and minerals. On sloped land, shallow groundwater that once slowly moved laterally to feed streams during dry summer months, is blocked by incised building foundations, is shortcutted to the streams when intercepted by foot-ing drains. In urbanized watersheds providing home to vast communities of people, the life-sustaining hydrological cycle has been interrupted and rushed. Precipitation that is needed for all life processes is too-quickly marshalled toward the underground storm sewers and urged away from humanity and our material accumulation. Decades ago, people observed and measured the detrimental effects of these practices on the watershed. All the hard surfaces connected directly to the streams caused quick rises in stream flows from even the smallest rainfall – a dis-tinct break from stream flow patterns in forested catchments. Downstream flooding became a problem, since water was no longer retained and used where it fell, but instead rushed across smooth, hard surfaces to smoother channels and pipes, to converge in the lowlands and floodplain – the fertile, but precarious floodplains. What once was infre-quent flooding returned with more regularity. The higher flows rush at higher speeds, scouring stream beds and banks, causing massive movement of sediments and changes to the course of the streams. Fish and other biota suf-fered, and still do, because of the flow regimes. Add to that non-point sources of contamination that enter the aquatic ecosystems from all points across the hydraulically-connected watershed. Excess inputs of nitrogen and phosphorus – both necessary nutrients for a healthy stream, but also the limiting nutrients to unchecked algal growth – lead to eutrophic conditions: summer blooms of algae that soon die, and decomposing, deplete dissolved oxygen to levels below which oxygen-sensitive salmonids and other aquatic species need for life and to stay competitive. In Canada as elsewhere, indices of the biotic health of streams plummet with urban growth in their catchment. Grass-roots citizens’ organizations, scientists, engineers, and municipal planners have responded to these observa-tions, and are always improving the state-of-the-art set of best management practices to steer development away from harmful designs and toward less impactful human use. The philosophy around how rainfall is treated has evolved over time, with progressives now treating it as a resource to be held on site, protected, treated to assure high quality, directed to recharge the precious stores of ground water, and, if entering a receiving water body, allowed to do so through a treatment train of practices that will slow its egress, reduce volumes of surface flow, and remove pollutant concentrations and loadings to protect aquatic health. 2  In the literature reviews that begin each chapter, I place this investigative effort within the framework of the healthy and rigorous body of knowledge around (a) urban impacts on hydrology and water quality, and (b) how well the cur-rent paradigm’s precipitation “source controls” are used to effectively address the hydrology and water quality chal-lenges facing aquatic life in urban streams, human populations made vulnerable by the changes to the hydrologic regimes. 1.2 Study goals & objectives Two very separate studies were performed that together take a full-spectrum view of the effectiveness of exemplary models of low impact development source controls in the Lower Fraser Valley. I say ‘exemplary’ because despite general understanding that these practices are necessary, and despite federal, provincial, and municipal guidelines to install them, there are yet very few examples of partial- or full-suite low impact development practices in the lower Fraser Valley. A cost associated with this research taking on a broad scope, of course, is the sacrifice in study depth. The objectives of the study are twofold. In Chapter 2, I pursue the objective to determine, using a paired catchment approach, whether the on-lot source con-trols installed in the extensive Routley neighborhood reduce peak flows and event total runoff volumes, over that of the conventionally-developed Langley Meadows neighborhood. Then in Chapter 3, I attempt to determine, using a treatment effectiveness approach, whether a vegetated bioswale is providing water quality benefits to the receiving body. 1.3 Scope & general methodology of research The first part of this research effort was done primarily by a desktop analysis of data collected over past years as part of a monitoring strategy between paired catchments with dissimilar stormwater management practices. On-site visits and installation/maintenance of rain gauges accompanied this desktop research. Aside from actually analyzing rain-fall and runoff data to determine whether peak flow and event cumulative total runoff were significantly mitigated in the catchment with progressive stormwater management, the bulk of the work was to determine that the two catch-ments were appropriately paired for the analysis. The results of this analysis are compared to how such a suite of source controls should perform, according to the modern designs and the literature. The second part of the research was considerably more hands-on: after initial setbacks, I installed a set of two auto-matic samplers – capable of taking 24 discrete samples each, on individually-programmed timing schedules – along with a custom-built water sensing circuit designed to start both autosamplers during a runoff event. Generously-lent telemetry-equipped datalogging equipment and a subscription to online data hosting enabled a fully automated sys-tem that simply sent me an email when a sampling sequence had begun. After collecting the samples and changing bottles, I and UBC Environmental Laboratory staff performed analysis on a suite of water quality parameters of in-terest, including sediments, nutrients, alkalinity, hardness, and chloride, as well as total and dissolved metals and total organic carbon. Concentrations of these parameters in road runoff is compared to the literature and aquatic health water quality guidelines. And finally, the reductions of the pollutants analyzed are compared to that predicted by the literature, and processes are discussed where the four months of spring sampling pointed to trends. 3  Chapter 2: Selected BMPs’ impact on hydrology 2.1 Introduction & project description The first part of this urban runoff research is a close examination of the well-known impacts of urbanization to the hydrology of a catchment through extensive literature review. The literature review also addresses the compromised biological integrity caused by both increased flow speeds and increased sediment loading, and the maturing relation-ship between cities and urban rainfall. The review touches on both low impact development practices in use on the site- and neighborhood-scales, as well as the regulatory framework within which municipalities, developers, and homeowners must work. I describe a paired catchment approach followed in the methodology, in which I access a detailed database (of 5-minute resolution rainfall and storm sewer flow data collected by Kerr Wood Leidal Associates, Ltd. for a period of nine years for the Township of Langley), identify a set of 128 unique storms over the 2011-2012-2013 three-year period, then normalize storm sewer flows by catchment area, remove groundwater and interflow volumes to deter-mine specifically flows associated with precipitation runoff from urban surfaces, and perform standard statistical analyses on peak flows and total runoff from each storm examined. Finally, these results are compared to the physi-cal differences between the two catchments – specifically the roofs (about 30% of the area) disconnected from the drainage network and amended soils and rock trenches designed to absorb and infiltrate 90% of the year’s rainfall. 2.2 Objectives The objectives of this hydrological study were specifically to address two questions: 1. Does the suite of lot-only source controls, as installed in the Routley neighborhood of the Township of Langley, effectively reduce peak storm sewer flows during rainfall events up to the design precipitation event, as com-pared to a control catchment with conventional stormwater management practices? 2. Do these practices in the Routley neighborhood effectively reduce cumulative volumes of runoff from individ-ual events, as compared to the Langley Meadows control catchment with conventional stormwater management practices? 2.3 Hypothesis The Langley Meadows neighborhood was built out in the mid-eighties with no intentional source controls. The nearby Routley neighborhood, however, was developed in the mid-2000s under an integrated stormwater manage-ment plan with progressive bylaws mandating a suite of on-lot source controls designed to meet urban runoff guide-lines. It is hypothesized that during rainfall events up to the ‘design storm,’ the Routley neighborhood is behaving as if it is ‘less impervious’ than its counterpart, exhibited in significantly lower peak flows and significantly reduced event runoff. 4  2.4 Literature review I discuss first the urban impacts on both hydrology and biota, next how cities’ relationship with rainfall has matured over time, then explore four low impact development ‘best management practices’ that are in use in the Township of Langley and the District of Maple Ridge, and conclude the literature review with a summary of the multi-jurisdic-tion regulatory framework in Canada. 2.4.1 Urban impacts on hydrology For health, peace of mind, and ease of access, we prefer impermeable surfaces above our heads and solid surfaces beneath our feet. As human populations grow and cities densify, an unintended consequence of the build-out of roofs and roads is a great change to natural hydrological processes and once-healthy streams. 2.4.1.1 Fate of precipitation Depending on the local soil infiltration rates of an undisturbed forested catchment, the vast majority of a given year’s precipitation is intercepted, evapotranspired, infiltrated to shallow groundwater, or infiltrated to deep ground-water, allowing a minimal fraction of the year’s rainfall – that which arrives in exceptional precipitation events – to leave site as surface runoff (Booth 1991, Stephens 2002). Although the proportions vary greatly, and may be consid-erably different in the Vancouver area, according to Livingston & McCarron (1992), approximately 40% of the pre-cipitation in a forested watershed may be ‘lost’ to the streams through either leaf interception or evapotranspired from the soils or trees; 50% of the precipitation is intercepted to both deep and shallow groundwater, also effectively ‘lost’ to the streams, at least in rainfall-event scale; and 10% is lost as surface runoff. Of course this is not evenly distributed among rainfall events: surface runoff may only occur from the largest storms – those on a return period of greater than six months. In the lower Fraser Valley, 90% of annual rainfall is from the 6-month 24-hour or smaller event (KWL et al. 2012, pp.1-4). Precipitation from events on shorter return periods (with a higher return frequency) falling on undisturbed forested catchments should typically be completely evapotranspired or infiltrated. The fate of precipitation falling on a forested catchment is illustrated in Figure 2.1, based on values from Livingston & McCarron (1992). 5   Figure 2.1 Fate of precipitation as a catchment is urbanized As land use in a catchment is changed, and as pervious surfaces are paved, the fate of precipitation is altered, as il-lustrated in Figure 2.1. with the right side illustrating the worst-case scenario: a fully urbanized, highly impervious catchment in which only about 15% of the annual rainfall is allowed to infiltrate to recharge aquifers and summer base flows, while roughly 55% is discharged as runoff (Livingston & McCarron 1992). With climatic and soil differ-ences, these values will vary greatly across Canada and the United States, but the general trend is increased site run-off at the expense of evapotranspiration and infiltration. 2.4.1.2 Ways land use changes As human populations grow, residential development expands and densifies, and as agricultural and industrial activ-ities expand spatially from the urban centers, there are many ways that changes in land use in a catchment affects the hydrology. The first is the removal of natural vegetation, which diminishes at a rate at least as fast as impervious surfaces replace it, followed by replacement of pervious land with impervious surfaces, encroachment on streams (reduction of forested buffer), channelization or culverting of streams, and incision of basements into subsurface flow (Klein 1979, Booth 1991). 2.4.1.3 Impacts on Hydrology 2.4.1.3.1 Less rainfall infiltrated The proportion of precipitation allowed to infiltrate to groundwater is inversely proportional to the impervious sur-face area in the catchment (Klein 1979) by two mechanisms: first, that rainfall cannot pass through the surface (the 6  total area remaining to infiltrate is thus reduced), and second, that these impervious surfaces are usually sloped to-ward conveyance infrastructure that removes precipitation from where it fell. 2.4.1.3.2 Lower summer base flow A direct result of the increased wet-weather runoff, and reduced infiltration to groundwater and subsurface flow is the reduction in the dry-weather base flow (Arnold & Gibbons 1996). In one study comparing fifteen streams in catchments of varying levels of urban development, it was found that 65% of the reduction in stream base flow could be explained by catchment percent imperviousness (Klein 1979). In a study in the Lower Fraser Valley in British Columbia, Finkenbine et al. (2000) found that dry-season base flow was “extremely low” in catchments with greater than 20-40% TIA (a strong negative correlation between base flow velocity and catchment imperviousness), and noted that stream velocity dropped far more than did depth. The results of this survey of several creeks identi-fied poor riparian integrity and the notable lack of large woody debris in urban streams as impediments to a healthy fish habitat. In a comparison of ten streams in New York, Spinello & Simmons (1992) found that in forested catch-ments and those with the lightest urban development, up to 96% of the annual runoff was from base flow, but dropped to as low as 14% of the annual total in catchments that were highly urbanized with direct (combined) sewer connections. The authors attribute the reduction in base flow to a lowered water table caused by urbanization’s inter-ception of rainfall and disallowing it to recharge to groundwater. 2.4.1.3.3 Higher peak flows & cumulative runoff Closely linked to the increased return frequency of floods (notable runoff events) of a given magnitude, (or the in-crease in magnitude of floods on a given return frequency), is the demonstrated increase in magnitude of peak flows (Leopold 1968). With a loss of absorption capacity by natural soils, and connection of hard surfaces to efficient drainage pathways, precipitation in a catchment can reach the outlet not only faster, but also can converge from all parts of an urban catchment within the same timeframe. Peak runoff from urban catchments is typically higher than that from natural catchments: a direct (though not necessarily linear) relationship between %TIA and ratio of peak increase. Seaburn (1969) found runoff peak flows in a highly developed catchment to be 2.5 times that of pre-urban conditions, though ranging from 1.1 times in large storms to 4.6 times in small events. This very phenomenon – that the ratio of post-development to pre-development runoff peaks are greater for small events on short return periods than for large events on long return periods – is said to be because of the rainfall volumes of large events overwhelm forest soils’ capacity to retain or infiltrate the volumes, so runoff occurs as if the surfaces were effectively impervi-ous (Hollis 1975). Simulations by Holman-Dodds et al. (2003) confirm this: the 2-year 24-hour rainfall produces roughly seven-fold increase in peak flow from a 50% TIA “high impact development” urbanized catchment over pre-development, but the 100-year 24-hour storm causes a 2.5-fold increase in peak flow. Comparing ten streams in catchments of varying imperviousness, Spinello & Simmons (1992) found a direct correlation between magnitude of high flows and imperviousness. In a unique study performed in Miami, Lee & Heaney (2003) made a more sophisti-cated link: identifying that it is the directly connected impervious area that contributes to the surface runoff in small events. Using modelling software, they identified a linear relationship between total event volume and directly con-7  nected impervious area (change in imperviousness from 13% to 36% (a 177% increase) yielded also a 177% in-crease in total volume), and a direct relationship between peak discharge and directly connected impervious area (161% increase in peak discharge for the same change in directly connected impervious area). In a paired catchment study performed in Connecticut, peak flows from a conventionally-built catchment (32% TIA) increased by 16 times over pre-development conditions (Bedan & Clausen 2009). In addition to high peak flows, the total runoff volume from a catchment as a response to rainfall events also in-creases with imperviousness: a function of reduced infiltration and thus a strong indicator of whether urbanization is reducing rainfall recharge to groundwater (Klein 1979). Using compiled data from six previous studies, Leopold (1968) demonstrated that both imperviousness and sewerage (drainage) connection both increase total runoff. At 40% impervious and 40% sewered, post-urbanization discharge may be 2.5 times that of pre-urbanization; at 70% impervious and 70% sewered, the ratio can be as high as 4 times pre-urbanization runoff volume. Thus comparing event runoff volumes from a conventionally-built urban catchment to that from an un-touched catchment we would expect to see not only significantly higher peaks, but also notably higher total runoff from each event – the ratio likely decreasing with magnitude of storm. 2.4.1.3.4 Reduced time to peak Another documented effect of urbanization on stream response to a rainfall event is a reduction in the delay from rainfall center of mass to peak runoff center of mass. The increased conveyance efficiency of the connected surfaces and drainage network in an urbanized catchment allow runoff to reach the outlet more rapidly (5-20 minutes) than in a forested catchment (several hours) – contributing to flashy stream responses as a result of a shorter time to higher peak flows (Hirsch et al. 1990, p.344, Leopold 1968).  2.4.1.3.5 Increased return frequency of runoff of a given size Bank-full flow, the most important flow magnitude for the geomorphology of streams, is typically on a 1- to 2-year return period in a forested catchment. However, as urbanization increases the percent imperviousness of a catch-ment, these important floods may occur more frequently. One study (compiling data from several others) identified that the frequency of the bank-full flood may double when the catchment is 20% impervious, quadruple at 50% im-pervious, and may occur nearly 6 times per year at 60% impervious and 100% sewered (Leopold 1968). Disruption of natural infiltration processes through paving of a catchment increases the return period of floods of a given mag-nitude. Hollis (1975) published (non-linear) increases of flood events after 20% urbanization of a catchment: dis-charge during 10-year floods was doubled. One-year floods increased in volume by ten times, and smaller flow surges increased by up to twenty times. In another study examining changes in return period for a flood event of a given size, the 100-year post-development flow peak was found to be twice that of pre-development, and the 2-year post-development peak was as high as 57 times that of pre-development (WEF 1998 via Roesner et al 2001).  8  2.4.1.3.6 Sediments associated with construction During urban build-out of a catchment, natural vegetation is removed and soils are disturbed, compacted, and left exposed to rainfall and erosive runoff (Arnold & Gibbons 1996). During the active build-out phase of urbanization, the construction-associated sediment contributions to the streams are far above pre-construction levels, though they usually drop again after construction ceases (Klein 1979, Wolman & Schick 1967). One study examining the sedi-ment loading and concentrations before, during, and after construction of a road crossing found that background TSS concentrations of less than 5 mg/L jumped to above 1300mg/L during construction, but dropped to less than 5mg/L afterward (Barton 1977). In a study examining the impact of construction and urbanization on stream flow and sediments, a range of 7-100 tons/acre were calculated from construction sites (compared to 0.03-0.2 tons/acre from forests/grasslands, and 3.7 tons/acre from urban areas) (Yorke & Herb 1978). In a Maryland study that ulti-mately attempted to make estimates of the (public) cost of urban construction-associated sediment contributions to rivers, Wolman & Schick (1967) identified a two to five order of magnitude increase in sediment yield (on a tons/square mile/year basis) from construction sites over that of natural or agricultural areas. The study also found that the majority of sediment transported from construction sites was associated with storms, generally with spring runoff and intense summer events (Wolman & Schick 1967). The detrimental effects of high sediment loading from construction sites to waterways and fish habitat are known, and several best management practices to limit sediment losses from these sites have been mandated and used (to varying success) for decades, including temporary vegeta-tion of disturbed areas, mulching or matting, plastic covering of disturbed areas and stockpiles, maintaining buffer zones around waterways, preserving natural vegetation, stabilization of construction roads, control of dust, roughen-ing of construction site surface, terrace gradients, provision of bioengineered solutions (planting of vegetation) for steep slopes, interception dykes & swales, check dams in drainage ditches, filter fence, brush, or straw bale barriers, storm drain inlet filters, sediment traps, and temporary sediment (settling) ponds/basins (Gibb et al. 1999). As early as 1978, a 60-80% reduction in construction-associated sediment loading was reported, as a result of ordinances mandating such sediment control measures (Yorke & Herb 1978). 2.4.1.3.7 Changes to stream geometry With increased runoff, higher peak flows, more total volume of runoff at high flow, and increased frequency of flooding, the geomorphology of streams is affected. Initial bed aggradation of sediments and overbank deposition (associated with construction phase) gives way to both bed incision and bank erosion caused by the higher flow ve-locities (Leopold 1973, Paul & Meyer 2001). The incised channels often have sandier beds, shallower base flow, and offer less habitat diversity including fewer pools (Shields et al. 1994). A study documenting the “radical” geomor-phologic changes of a Connecticut brook through urbanization of its catchment identified an increase in channel width, decrease in depth, increase in slope as a result of a decrease in sinuosity, causing a wave of sediment to move slowly downstream (Arnold et al. 1982). A comparison of rural and urban streams in Maryland and Pennsylvania indicate significant increase in bankfull channel width in the urban streams, and a maximum rate of widening of 30 cm/year (Wolman 1967): the source of significant sediment contributions to the stream and changes to the benthic habitat. Hammer (1972) found that channels in catchments with high impervious area experienced up to seven times the enlargement effects than streams in forest or agricultural catchments. The widening and straightening of urban 9  stream channels through these natural responses to anthropogenic inputs in turn causes an increase in the efficiency of precipitation transport (and loss) out of the catchment. To mitigate the erosion of the stream banks themselves, common practice has been to channelize the streams and armor the banks. This, however, increases stream flow ve-locities and contributes to further flooding in the lowlands. 2.4.1.3.8 Sediments associated with bed & bank erosion As construction activities give way to a steady-state urban catchment, the increased flows and frequency of high flows associated with the impervious areas erodes both the stream bed and the bank, contributing to both (a) changes in the cross section (depth, width, area) of the streams, and (b) a continued sediment load downstream (Arnold & Gibbons 1996). Yorke & Herb (1978) reported an increase of 18 to 123 times the sediment loading from urbanized catchments over forested catchments. In the case of construction of a single bridge, Barton (1977) found that high construction suspended sediment concentrations dropped to pre-construction levels afterward, a case in which the construction led to no appreciable increase in impervious surface in the catchment. Olley et al. (1996, not accessed) had identified that most of the sediment in lower reaches of an Australian river were subsoil- rather than topsoil-sourced (indicating erosion of stream bed and banks may be contributing significantly to the sediment load). Green et al. (1999) linked sediment transport rates in large catchments to bed erosion and stream flow. Thus, after con-struction activities have ceased and development has reached its full build-out, stream flow changes continues to cause sedimentation of downstream reaches as upstream beds and banks experience continued erosion. Trimble (1997) attributed one-half to two-thirds of annual sediment load of a stream in an urbanized catchment in southern California to channel erosion. Six stages of geomorphic change resulting from urbanization are summarized as pre-modified (stable condition before urbanization, with some erosion on outside stream bends and deposition on inside bends), constructed (with loss of vegetation and trapezoidal channel cross section), degradation (with over-bank deposition, erosion of sand bars, removal of in-stream vegetation, and low flow depths), threshold (channel widen-ing, vertical banks, block failures on cut bank), aggradation (reduction of bank angle, new flood plain, fallen riparian vegetation, failed banks), and eventual restabilization (stable condition with meandering thalweg, alternating sand bars, deep flows again) as vegetation returns to banks and halts widening (Simon 1989). 2.4.1.3.9 Total vs. effective impervious area Total impervious area (TIA) has been identified as the single greatest watershed factor affecting stream health, and is used because it is reliable, integrative, and measurable (Klein 1979, Arnold & Gibbons 1996). As detailed further in this literature review, the biotic integrity of streams seems to degrade rapidly after a given threshold catchment TIA (Klein 1979, Limburg & Schmidt 1990, Booth & Jackson 1997).  It is apparent, however, that some impervious surfaces do not directly contribute to runoff leaving the catchment given physical separation from other hard sur-faces or conveyance infrastructure. Thus, it is the ‘directly connected impervious area’ or ‘effective impervious area’ EIA that is the direct contributor of runoff – especially from the smaller storms (Booth & Jackson 1997, Roy & Schuster 2009, Lee & Heaney 2003, Walsh 2004). Dinicola (1989) presumed a set of %TIA and (%EIA) values for different land uses: low density residential 10% (4%); medium density residential 20% (10%); suburban density 35% (24%); high density residential 60% (48%), and commercial-industrial 90% (86%). The relationship, though it 10  may not be broadly applicable, demonstrates the significant difference in EIA vs. TIA estimates, especially at low catchment imperviousness and introduces uncertainty to the results of studies if the definition of imperviousness is not described (Booth & Jackson 1997). 2.4.1.4 Impacts on biota When we speak of ‘stream health’ in Pacific watersheds, images of salmon swimming upstream against all odds to spawning grounds in gravelly streams is the often-conjured image. Hydrological consequences of urbanization, and non-point sources of pollution challenge the returning adults, the more sensitive life stages of young, and the other aquatic species that make up the complex ecosystem they rely on. 2.4.1.4.1 Deposition of sediments As summarized by Kemp et al. (2011), Greig et al. (2005), and Jones et al. (2012), anthropogenic increases in fine sediment negatively impact freshwater ecosystems by many pathways. First, increases in suspended sediment in-crease turbidity and limit penetration of light, which reduces photosynthesis and lower trophic productivity, thus diminishing the productivity of all higher trophic levels (Quinn et al. 1992, Izagirre et al. 2009). These suspended sediments also increase the abrasive and scouring effects of higher flows: damaging plant, invertebrate, and fish tis-sues, inhibiting attachment to benthic surfaces (causing wash-out), and ultimately diminishing the density and diver-sity of benthic invertebrates (Luce et al. 2010, Wood & Armitage 1999). Second, the increased deposition of sedi-ments leads to smothering or burial of habitat and biota and in-filling of bed gravel, which leads to both habitat loss (Lancaster & Hildrew 1993) and a reduction in intragravel dissolved oxygen by inhibiting the flow of oxygenated water into the spaces between the gravel (Sear et al 2008). This dissolved oxygen reduction in salmonid redd gravels has been found to cause hypoxia, weight reduction, and mortality of incubating salmonid embryos (Heywood & Walling 2007): a causal link between urbanization and depletion of salmon populations. Third, increases in bio-chemical oxygen demand (BOD) associated with increased sediment loading lead to depressed oxygen levels in the freshwater ecosystem (Ryan 1991, Whitman & Clark 1982), with consequences to species’ competitiveness and sur-vivability (Greig et al. 2007, Servizi & Martens 1991). Ryan (1991) advocates legislation that limits anthropogenic increases to stream sediment loadings or turbidity levels relative to (perhaps a percent increase above) background levels. 2.4.1.4.2 Increased flow speeds Flood-associated elevated stream velocities occur infrequently in natural catchments, but higher-frequency flood flows with high velocities may overwhelm fish species’ and life stages’ ability to keep up. Compiled data from Katopodis & Gervais (1991) in the Government of Canada’s DFO’s Land Development Guidelines for the Protec-tion of Aquatic Habitat lists the sustained, prolonged, and burst swimming speeds for various life stages of several fish (Chilibeck et al. 1993). For example, adult Sockeye salmon – some of the stronger freshwater swimmers – may sustain a swimming speed of 1.0 m/s indefinitely, an elevated 3.1 m/s for maximum 200 minutes, and up to 6.3 m/s for periods no longer than 165 seconds. Fifty-centimeter juveniles, however, may sustain a swimming speed of only 0.2 m/s indefinitely, 0.4 m/s for 200 minutes, and 0.6 m/s burst speed for 165 seconds (Chilibeck et al. 1993). 11  Flashy stream flows have an especially strong impact on juveniles. Carline & McCullough (2003) document 84% reduction in year-plus Brook Trout density (from previous year) in West Virginia headwater streams after a signifi-cant flood, and 98% reduction in younger juveniles. Similarly, Elwood & Waters (1969) link drastically reduced invertebrate and Brook Trout populations to a series of extreme flows in an urban stream that decimated food sup-plies, and Hoopes (1975) documented 96% reduction in the number of juvenile Brook Trout in a Pennsylvania stream after hurricane-associated flood flows many times that of summer base flow. Sudden and elevated stream flows during sensitive life stages may completely wash out a (keystone) member of the aquatic habitat, thereby re-ducing biodiversity (Poff et al. 1997). 2.4.1.4.3 Compromised biological integrity In a landmark publication, Klein (1979) first pointed out that once a threshold value of approximately 12% total im-pervious area of a catchment is crossed, irreversible stream quality impairment can be observed, with severe impair-ment (absence of fish) occurring at around 30% TIA. Steedman (1988) confirmed that urbanization was negatively correlated to the (benthic) index of biotic integrity (IBI) – less so where the riparian forest was maintained. Limburg & Schmidt (1990) found a significant inflection point as well at 10% TIA: mean densities of fish species signifi-cantly dropped between 10% and 20% TIA. In a more recent study correlating the effects of urban development to the magnitude of land use change in western Washington, Booth & Jackson (1997) found degradation of a water-shed to be observable also at the threshold of 10% total impervious area. Wang et al. (1997) also indicate a 10% TIA threshold, above which the IBI is significantly reduced, but more generally confirm that as urban land cover in-creases, and as forested cover decreases, the IBI drops as well. A further important relationship is the distribution of urban and forested areas, with an emphasis on maintaining a healthy ecosystem outside of the streams themselves, including a forested buffer (Wang et al. 1997, Walton et al. 2007). In a comprehensive study in the Puget Sound lowland of Washington, May et al. (1997) drew several relationships between metrics of urbanization and those of stream health, including %TIA to road density; %TIA to 2-year storm to baseflow discharge ratio; %TIA to riparian buffer; %TIA to mature riparian forest; %TIA to large woody debris frequency; upstream riparian buffer to large woody debris frequency; %TIA to the intragravel DO to DO ratio; and %TIA to IBI score, all of which indicated  rapidly decreasing stream health after only about 5-10%TIA. 2.4.2 Stormwater management evolution & impact on streams Urbanization has many negative effects on stream health and implies considerable liability to safety and built invest-ments in the forms of erosion and flooding. Land-use planning and stormwater management philosophies have evolved to meet these challenges. While initial priorities were to protect cities from flooding (a pragmatic effort for public safety and to minimize acute liabilities to public & private investments) and to keep streets dry (perhaps more an aesthetic choice), rainwater was first removed from site by ditches, and then combined storm & sanitary pipes (Bradford & Gharabaghi 2004, Debo & Reese 2002). 12  2.4.2.1 Evolving stormwater management focus This combined sewer system solved some problems, but led to the emergence of others. Because of health concerns downstream with municipal sewage directly entering waterways, the combined sewers were separated, with sanitary sewerage directed to treatment plants, and storm sewers directed urban runoff to the nearest receiving water body. Evidences that this efficient removal of rainfall from urban catchments was causing downstream flooding and sedi-ment loading led to a shift to value detaining the rainfall in distributed ponds, which helped reduce instantaneous peak flows from highly impervious areas, but failed to address the total volume of runoff leaving catchments, and failed to reduce flooding downstream (Debo & Reese 2002). Rather than just focusing on maintaining the 2-year (bankfull) storm at pre-development frequency and volumes, the community came to understand that it is changes to the full spectrum of flows that contribute to stream degradation, so further efforts were made to distribute runoff control throughout the catchment with an aim to protect the stream channels (Bradford & Gharabaghi 2004). Calling it “rainwater management” instead of “stormwater management” that is being practiced, is a recent change in vocab-ulary that reflects the underlying evolution of the paradigm (KWL et al. 2012). Watershed-wide master-planning efforts emerged to coordinate flow release to manage flooding, and eventually concerns about non-point sources of pollution (especially elevated nutrient loads) and stormwater quality emerged, focusing the philosophy on both treatment and runoff reduction targets, which eventually switched to stream health or biotic integrity targets that depend on a certain quality of water flowing at predictable flows, depths, and flood frequencies. In Ontario, like many other jurisdictions in Canada and the US, phosphorus and sediment removal per-formance criteria were developed for installed BMPs in new developments (Bradford & Gharabaghi 2004). Given the integrated relationship between the stream, riparian forest, other natural areas, and urbanized parts of a catch-ment, there has emerged an appreciation of the value of watershed-wide decision-making that emphasizes the physi-cal and chemical impacts to stream health from other activities, treats stormwater as a resource (rather than just a liability), considers long-term maintenance, and mimics natural hydrology (Debo & Reese 2003). The yet-evolving philosophy is more recently considering the health of wetlands, valuing groundwater recharge but cautious of the introduction of contaminants to the aquifer by infiltration practices, addressing stormwater management simultane-ously on the site, neighborhood, and watershed scales, identifying how to monitor the long-term efficacy and im-pacts of these practices, and perhaps most importantly, learning how watershed-level decision-making can broadly implement the hard (technical) or soft (policy) solutions (Bradford & Gharabaghi 2004). 2.4.2.2 Low impact development best management practices Breaking from the conventional stormwater management approach reflected by past modes of thinking, current in-novative practices – today’s low impact development best management practices (LID-BMPs) – for management of rainfall arriving in small storms (up to the 6-month event) on the site scale, include green roofs, flat roofs, rain bar-rels, reduction of on-lot impervious surfaces, use of pervious asphalt/concrete/pavers, wheel-strip driveways, grassed modular plastic or concrete grid pavement, tilling of subgrade to reduce compaction, infiltration swales with and without underdrains, precast concrete or plastic infiltration chambers, flow-through planters, infiltration plant-ers, amended soil depths (and maintaining porosity), rainwater reuse, the planting of trees,  (Marsalek & Schreier 13  2010, KWL et al. 2012, Stephens et al. 2002, King County 2009). For managing rainfall in 6-month to 1-year events on the neighborhood scale, other innovative approaches include road width reduction, curb & gutter removal, road-side bioswales, pervious pavement, detention ponds, protect wetlands, construct wetlands, street sweeping, and con-taminant control (Marsalek & Schreier 2010, Stephens et al. 2002). On the watershed scale, Marsalek & Schreier (2010) list the creation of riparian buffer zones and wetlands, maintenance of natural stream channels and flood plains, and mitigating the cumulative effects of non-point sources of urban pollution by infiltrating & detaining all runoff, as innovative practices. Practices at all three levels are integrated to achieve the stream health and aquifer quality targets set for the watershed. Provided in a summary in Table 2.1 are the site-scale source controls impact on the objectives of peak flow reduction, total runoff reduction, and peak flow delay. Table 2.1 Source controls’ contribution to flow mitigation objectives  The low impact development best management practices under investigation in the first part of this study are those designed with the intention of infiltrating rainfall to groundwater, specifically:  disconnected roof downspouts from the storm drain (to effectively disconnect the roof’s impervious surface from direct connection to the receiving water body – reducing both peak flow and total runoff),  amended soil depths (to the optimal depth of 30cm, to abstract the initial volume of larger storms falling on the pervious area, and to absorb more of the volume directed there from the now-detached roof area, acting tempo-rarily as a sponge to hold the water so mineral subsoils can infiltrate at a perhaps lower rate),  infiltration trenches (to which is directed excess rainfall from residential lawns, driveways, and other hard sur-faces, also reducing the storm-associated direct runoff), and 2.4.2.2.1 Disconnected roof leaders Roofs make up a significant portion of the impervious area of a watershed. Hammer (1972) identified that only when residences’ downspouts are directly connected to the storm drainage system do they contribute to stream flows that cause erosion and channel enlargement. Street and sidewalks’ effects on flows and erosive forces are only sig-nificant when directly connected to the conveyance infrastructure. In the watersheds examined in this investigation, roofs contribute nearly half of the impervious area, or about 30% of the total area of the catchments. Disconnecting Source Control Peak Flow Reduction Total Runoff Reduction Flow (Peak) Delaygr n roof X X Xr in b rr l X X Xr ducti  of impervious surface X X Xrvious pavement X Xinfiltrati n swale X Xrock trench X X Xamended soil depth X X Xtrees X Xroadside bioswale X X X14  this impervious area from drainage infrastructure significantly reduces the directly connected impervious area, or EIA, of the catchment. By disconnecting rooftops from external drainage, planners can expect a 25-50% reduction in runoff (Battiata et al. (2010). This of course is dependent on where the roof runoff is directed, and what absorbance and infiltration capacity the surrounding pervious area has for the additional volumes. Typically in the fairly wet Metro Vancouver area, disconnection of residential roof leaders is not a stand-alone source control, but is coupled with either amended soil depths on remaining pervious portions of the lot, deep, localized infiltration trenches or rain gardens, or a combination of these practices to meet rainfall capture and runoff reduction targets. 2.4.2.2.2 Amended soil depths Absorbent landscaping – typically increased soil depths underlying grass – are designed to mimic the absorption and filtration functions of pre-development topsoil (KWL et al. 2012). For lawns, 150-300 mm absorbent soils (with roughly 8% organic matter) is recommended, depending on annual rainfall (300 mm soils for storage of 60 mm rain-fall), and 450 mm (with 15% organic matter) is recommended for areas with trees and shrubs planted (KWL et al. 2012).  When the contributing impervious area is less than twice the receiving area, absorbent soils installed according to design should produce no surface runoff below design rainfall (which, in Metro Vancouver is the 6-month 24-hour storm of 34-66 mm, depending on location), and only begin to allow surface runoff during the few extreme events in excess of this (KWL et al. 2012). Comparing to the Virginia Piedmont, Battiata et al. (2010) published findings of 50-57% annual reduction in site runoff if appropriate depths of soil are used in pervious areas of residential lots. KWL et al. (2012) offers a far more sophisticated and useful approach, linking percent capture to four key variables: annual rainfall depth, I/P ratio, infiltration rate of native mineral soils, and absorbent soil depth. For example, as shown in Table 2.2, holding absorbent soil depth to 300 mm and for a 2:1 ratio of contributing impervious to absorb-ing pervious area (the I/P ratio), the percent annual capture for three rainfall zones over a range of soil infiltration rates of 75-97% where rainfall is 1100 mm/year, 60-97% where rainfall is 1600 mm/year, and 50-97% where rain-fall is 2100 mm/year can be expected. Table 2.2 Percent rainfall capture by 300mm absorbent soils, I/P ratio of 2  ubs il infilt a i nrate ( m/hr)% annualcapturesubsoil infiltrationrate (mm/hr)% annualcapturesubsoil infiltrationrate (mm/hr)% annualcapture0.5 75 0.5 60 0.5 501 87 1 77 1 662 93 2 88 2 823 95 3 92 3 885 97 5 95 5 9310 97 10 97 10 9620 97 20 97 20 9730 97 30 97 30 971100m  annual rainfall 1600mm annual rainfall 2100mm annual rainfall15  According to stormwater management specialists with Kerr Wood Leidal Associates Ltd. (2012), determining the appropriate I/P ratio and depth of absorbent landscaping as a stand-alone source control for a site is a straightfor-ward process, but may require sophisticated modelling software or complex spreadsheets for continuous simulations for a treatment train of BMPs. To capture a specified depth in 24 hours, or to capture a specified percentage of an-nual rainfall – both realistic design scenarios – separate sizing strategies are used (whether for lawn areas or rain gardens): For a design targeting a 24-hour depth capture, designers in the Vancouver area calculate the soil depth necessary by Ds = [R*(I/P + 1) – 24*Ks]/0.2, where DS is the depth (mm) of the absorbent soil, R the design rainfall capture (mm), I/P the ratio of contributing impervious area to available pervious area, and KS the hydraulic conductivity (mm/hour) of saturated subsoil (KWL et al. 2012). For a design targeting percentage capture of annual rainfall, they determine the average annual rainfall that the site receives, the site’s I/P ratio, the hydraulic conductivity of saturated subsoil, and consult charts plotting curves (from data such as in Table 2.2, above), to determine the required soil depth (KWL et al. 2012). Where absorbent soils’ absorbance and infiltration capacity will be overwhelmed by con-tributing impervious area and limited subsoil infiltration rate for the design rainfall event or percent capture of an-nual rainfall, alternate solutions such as infiltration trenches, described below, are used to meet the infiltration tar-gets (KWL et al. 2012). The implication of the limits of absorbent soils is the importance of optimizing competing interests of (a) watershed planners to keep the impervious-to-pervious ratio low, and (b) municipalities and developers to increase population and building densities for tax revenues and residence sales, respectively. High TIA from buildings, roads, and drive-ways leaves minimal pervious area in which to absorb and infiltrate the concentrated runoff with the single practice of absorbent landscaping alone; high-volume infiltration trenches can aid in meeting the site’s runoff reduction tar-gets. 2.4.2.2.3 Infiltration trenches Infiltration trenches – also called rock trenches or rock pits – are significantly more complex to design and construct than is the absorbent landscaping. The core of the system is a porous-media-filled trench with a source runoff pipe, a perforated distribution pipe, and an overflow, along with a separate catch basin sump to limit sedimentation and binding off (KWL et al. 2012). Since infiltration trenches offer no treatment of surface runoff (through biofiltration as through absorbent soils), according to Stephens et al. (2002), infiltration trenches should be used only for infiltra-tion of runoff from relatively clean sources such as roofs, lawns, and parks, to prevent compromising groundwater quality by recharging contaminated runoff. The media that fills the trench (and provides structural support to the overlaying lawn) is usually clean round or coarse rock – hence also being called rock pits or rock trenches – with a porosity (void space between the rocks that can fill with water) of 35-40%. Current best practices for a well-designed infiltration trench include locating it appropriately far from buildings (pe-rimeter drains) and other infiltration schemes, installing catch basin sumps for sediment pre-treatment, providing minimum separation of 600mm between water table and base of the trench, scarifying the native soil before installa-tion of geomembrane that fully wraps the rock, providing it sufficient plan area to allow it to drain from full in 4 16  days, and providing an overflow to storm drain for extreme events (KWL et al. 2012). Typical depths are 300-2000 mm and widths 600-2400 mm. To size the infiltration trench for capture of a specified rainfall depth can be an iterative process as dimensions are negotiated within limited I/P to meet targets. Site soil hydraulic conductivity is a driving factor for sizing of the in-filtration trench: in more rapidly draining soils the infiltration trench can have a greater depth, and for a given vol-ume can have a reduced footprint area. Design depth is found by DR = 24*KS*T/n, where DR is the depth (mm) of the trench, KS the hydraulic conductivity (mm/hr) of saturated subsoil, T the drain time (days), and n the unitless porosity of the rock pit. From this and a known rainfall depth, hydraulic conductivity of the site soils, and rock po-rosity, the contributing impervious to infiltration trench area is calculated by I/P = [(24*KS + DR*n)/R]-1, where R is the design rainfall capture depth (mm). The infiltration trench footprint area is then found by dividing the contrib-uting impervious area by I/P (KWL et al. 2012). To size the infiltration trench for percent capture of the annual rain-fall, designers in Metro Vancouver use knowledge of the site’s annual rainfall and subsoil hydraulic conductivity to easily determine from available charts the I/P ratio required for the infiltration trench to capture the design percent-age of the annual rainfall. The needed footprint area is then found by dividing the contributing the contributing im-pervious area by the determined I/P (KWL et al. 2012). Where site constraints preclude building the infiltration trench to design (footprint area exceeds available geometry, or design depth cannot be achieved), the infiltration trench is instead designed with a perforated drain pipe at the level of its invert, with minimum 10mm orifice to con-strain site discharge to drain (KWL et al. 2012). Such a modified design would of course have substantially different impacts on stream flow: much of the water still drains to the conveyance network, but is just detained and released at a slower rate. This would help with reducing peak flow, but not overall runoff volumes. 2.4.2.2.4 Roadside bioswales Also in use is the group of practices known as bio(in)filtration: an aesthetically appealing engineered runoff reten-tion structure that provides both infiltration (runoff control) and filtration (water quality treatment) to a relatively large impervious catchment area (high I/P ratio), often allowing temporary surface ponding. They are effectively a “two-layer swale and infiltration trench system” with an absorbent/filtering top layer of organic soil overlaying a rock-filled bottom layer for infiltration and lateral conveyance (Stephens et al. 2002). Runoff reduction from empiri-cal studies in the Virginia Piedmont indicate that designers can expect 40-80% runoff reduction through bioretention mechanisms, or 50-90% reduction using infiltration practices (Battiata et al. 2010). Both rain gardens (on residential lots or parking lots), or vegetated roadside swales are designed in many forms for different rainfall capture or treat-ment objectives, but usually are designed with a drain rock reservoir wrapped with geotextile overlain with roughly 450mm of well-draining and absorbent topsoil, a concave surface, some form of overflow (perforated pipe to drain, with or without a flow-limiting orifice), and planted with grasses, shrubs, or trees appropriate to the local moisture conditions (KWL et al. 2012).  Of specific interest to this study (though its application for water quality treatment will be discussed more in the fol-lowing chapter) is the bioretention-biofiltration vegetated roadside swale, effectively a chain of hydraulically-linked 17  infiltration rain gardens along a roadway with surface and subsurface flow limiting devices (KWL et al. 2012). Ac-cording to Stephens et al. (2002), such devices used along roads (where the ratio of contributing impervious surface to bioswale area is low) have greater rainfall capture effectiveness than on residential lots (where the I/P ratio is high). Indeed, on single family lots, the I/P for a rain garden may be as high as 50, compared to approximately 30 for residential roadways (KWL et al. 2012).  Full design guidelines and sizing procedure commonly used by design-ers in Metro Vancouver are described in the updated Stormwater Source Control Design Guidelines 2012, with dif-ferent sizing procedures for either rainfall depth or percent annual capture targets (KWL et al. 2012). An example is shown in Figure 2.2, an illustration developed from typical details, and used with permission from KWL et al. (2012). Typically runoff is received as sheet-flow from a slightly sloping roadside (2% slope toward the swale) over a flat (flush) panel curb, over a small vertical drop into gravel or stone erosion & sediment control (replacing the turf strip of previous designs), down 2:1 slopes into the concave swale bowl and organic soils with turf or other plantings (KWL et al. 2012).  These structures have gentle (1-2%) longitudinal slopes, broken by terraces and rock splash pads if site slopes are greater, underground plugs made of undisturbed native soils to check flows, and are provided a perforated drain at the top of the rock trench for overflow to drain (KWL et al. 2012). Surface flow down the ditch, and bio-filtered runoff from the overflow drain should only occur during large events that have overwhelmed the rock trench retention volume.  Figure 2.2 Typical cross section of Foreman Drive bioswale In the illustration of the infiltration swale cross-section in Figure 2.2 above, the numbers identify (1) roadway sloped at 2% toward the swale, (2) flat panel curb, (3) 25mm drop to turf strip, (4) 25 mm-minus round drain rock, to 800 mm minimum depth filling the rock trench, (5) timber weir, (6) 150 mm diameter perforated PVC pipe, (7) filter cloth completely surrounding the rock trench and perforated pipe, (8) processed soil to 600 mm depth, (9) 2:1 slope, and (10) sidewalk also sloped 2% toward the swale. The required roadside bioswale footprint depends, of course, on local rainfall conditions, local soils’ hydraulic con-ductivity, the site slope, and local municipal construction bylaws specifying topsoil depth and drawdown time. For 18  capture of a design rainfall depth, the depth of the infiltration reservoir is sized exactly as for that in infiltration trenches (outlined in the section above), and the base area calculated by  𝐵𝑎𝑠𝑒 𝐴𝑟𝑒𝑎 =  𝐼[(24𝐾𝑆) + (𝐷𝑅n) + (0.2𝐷𝑆)𝑅 ] − 1 where I is the contributing impervious area (m2), KS the hydraulic conductivity (mm/hr) of the native subsoil, DR the depth (mm) of the rock reservoir, n the (unitless) rock porosity, DS the depth (mm) of soil, and R the design rainfall capture depth (mm) (KWL et al. 2012). Base width is calculated by dividing the base area by the linear road dimen-sion that it serves; total bioswale width is base width plus the width of the side-slopes (a function of both depth and slope), to determine the total increase in road right-of-way width that the bioswale will contribute. For capture of a percentage of annual rainfall, designers consult charts to determine the I/P from known values of the site’s annual rainfall, soils’ hydraulic conductivity, and capture target (KWL et al. 2012). Small rainfall events less than the de-sign rainfall should produce no discharge from site, but of course will during intense storms that overwhelm the ab-sorption/infiltration rate of the topsoil, overwhelm the volume of the rock trench, or over time after siltation, leaves, or other debris have bound off the surface and reduced the structure’s treatment and infiltration capabilities from design. Thus, uncertainty is introduced to estimations of BMP effectiveness over time, given changes in design, con-struction practices, the quality and frequency of maintenance, siltation, and compaction. For a properly-sized, cor-rectly-built, and well-maintained partial infiltration bioswale, or rain garden as modeled by Holman-Dodds (2003), significant site runoff reductions (over conventional stormwater management) can be achieved for small storms, but peak reduction or total volume may not be significantly different from the conventional SWM practices during storms that greatly exceed the design. For this reason, site-scale source controls and neighborhood-scale BMPs are used in conjunction with watershed-scale practices to protect streams and reduce the harmful effects of flooding (King County 2009, Stephens et al. 2002). 2.4.3 Regulatory framework In Canada, there are many overlapping jurisdictions that regulate (or explicitly do not regulate) the discharge of wa-ter, including wastewater, stormwater, industrial effluent to receiving bodies, and navigating the legislation, regula-tions, and guidelines is a bit of a challenge. Following are the relevant codes at the Canadian (national), British Co-lumbia (provincial), and Metro Vancouver (municipal) scales. 2.4.3.1 National Until November 2013, Section 35(1) of the Fisheries Act had stated that “no person shall carry on any work, under-taking or activity that results in the harmful alteration or disruption, or the destruction, of fish habitat,” though pro-vided several provisions by which such activity could be permitted, including by simply having a letter granting per-mission by the Minister (Fisheries Act 2012). However, in 2012 the Conservative-majority party tabled back-to-back omnibus bills (Bill C-38, Bill C-45), which included sweeping changes to the Fisheries Act, and by the end of the year the provisions had been given Royal Assent to Canadian law. The Bills included provisions for changing Section 35(1) to read “No person shall carry on any work, undertaking or activity that results in serious harm to fish 19  that are part of a commercial, recreational or Aboriginal fishery, or to fish that support such a fishery,” which it now does (Fisheries Act 2013). This signals a broad deregulation of the broad protection of fish habitat or water bodies for other non-economic uses, where doing so would could have hindered economic expansion. A document by the Canadian Council of Ministers of the Environment (CCME), the Canada-wide Strategy for the Management of Municipal Wastewater Effluent,  provides water quality guidelines for effluent discharged from mu-nicipalities (including TSS), and stipulates “no increase in combined sewer overflow frequency due to development or redevelopment […], no combined sewer overflow discharge during dry weather […], sanitary sewer overflow frequencies will not increase due to development or redevelopment, and sanitary sewer overflows will not occur dur-ing dry weather…” (CCME 2009). For sewerage & drainage districts that still have combined sewers, this provides significant pressure to (a) begin or accelerate a storm-sanitary separation project, (b) pass bylaws that stipulate de-velopment or redevelopment BMPs that serve to reduce rainfall runoff (especially in catchments with shared piping) by evapotranspiration, reuse, or infiltration on-site, or (c) both. The Department of Fisheries & Oceans’ (DFO’s) Urban Stormwater Guidelines and Best Management Practices for Protection of Fish and Fish Habitat (Chilibeck & Sterling 2001) gives guidelines for outcome- or target-focused BMPs. They are:  Volume Reduction: to reduce/mitigate runoff volumes from impervious urban areas, and to maximize infiltra-tion/recharge to groundwater, with a guideline of infiltrating to ground all volumes from impervious surfaces during post-development 6-month 24-hour events.  Water Quality: to mitigate the WQ impacts to fish by treating the first flush from events, with a guideline to treat everything up to the 24-hour event that is 90% of the annual rainfall.  Detention or Rate Control: to reduce post-development peak runoff rates to pre-development runoff rates, with a guideline to match the pre-development volume, shape, and instantaneous rates for 6-month, 2-year, and 5-year 24-hour rainfall events. The DFO’s guidelines to sewerage & drainage districts are central to Metro Vancouver’s stormwater management planning for both flow and quality targets.  2.4.3.2 Provincial British Columbia’s Stormwater planning: A guidebook for British Columbia (Stephens et al. 2002) provides guide-lines for management of the full spectrum of rain: small events that are less than half of the rainfall event on a one-year return period (roughly the 6-month storm); large events between the 6-month storm and the 2.33-year storm; and extreme events larger than the one-year storm. To manage this spectrum, guidelines specify targets for event runoff volume reduction, water quality, rate reduction, and peak flow conveyance:  Volume Reduction: to capture, infiltrate, evaporate, or re-use 90% of the annual rainfall – that arriving in small storms up to the 6-month storm – on-site where it falls.  Water Quality: to treat the (above-mentioned) 90% of the annual rainfall (small storms) on-site using best man-agement practices. 20   Rate Control/Reduction: to store runoff from the large storms (that arriving in events on a return period between 6-month and 2.33-year), and release at a controlled rate approximating pre-development flows.  Peak Flow Conveyance: to rapidly remove rain that falls SOUin events up to the 100-year storm (via overland and piped flow paths) with minimal damage to property. Further, the Municipal Wastewater Regulation (B.C. Reg. 87/2012), specifically addressing the discharge of treated (sanitary) sewage, and in force under British Columbia’s Environmental Management Act (SBC 2003), section 7(2) prohibits directing non-domestic waste (including stormwater) to sanitary sewers or municipal wastewater facilities, section 38(1) prohibits expansion, design, or construction of combined sewer systems, and sections 42 through 44 stipulate wastewater effluent discharged to streams does not occur during snowmelt or rainfall events, and limits inflow & infiltration-sourced increases to twice the average dry weather flow occurring on less than a 5-year return period. Also, Implications to municipalities are similar to the DFO’s directive: both phase out combined sewers, maintain both storm and sanitary sewer conveyance infrastructure (to limit the inflow & infiltration into the sanitary sewers), and infiltrate that runoff to groundwater. 2.4.3.3 Metro Vancouver The Integrated Liquid Waste and Resource Management Plan for the Greater Vancouver Sewerage and Drainage District (GVS&DD) (subsumed under Metro Vancouver) was approved in 2012 by British Columbia’s Minister of the Environment, under which each of the 18 member municipalities must develop an Integrated Stormwater Man-agement Plan (ISMP), to be submitted individually by 2014 (possibly extended to 2016) to the Minister of the Envi-ronment for individual approval (Metro Vancouver 2010). Metro Vancouver committed to eliminating all remaining CSOs by 2050, is already advancing a phased project to separate the remaining CSOs at an annual rate of 1% (Crowe 2014), and is working with member municipalities on both carrot (incentive) and stick (bylaw) approaches to encourage implementation of LID-BMPs in new development and redevelopment, through the provision of ISMP planning documents and templates (Johnston 2005). The ISMP template provides guidance to municipalities to infil-trate and treat rain where it falls in an integrated manner that considers stream health and the intent of the provincial and DFO legislation. The City of Burnaby, for example, already has in place a Guide for developing near streams, with bylaws and setback criteria in place (City of Burnaby 2014), and is preparing bylaws and municipal policies necessary for meeting stormwater flow and quality targets, including requirements for stormwater source controls on developments and redevelopments (Johnston et al. 2014). This review of the scientific literature has summarized the current knowledge of the impacts on hydrology by urban-ization-induced land-use change, its impacts on biota, and the evolving stormwater management ideology that has developed as a result. Low impact development best management practices on the site scale used as source controls are designed to reduce peak flows and total site runoff, and the requirements in place by several jurisdictions reflect the need and capability to infiltrate the majority of urban precipitation. The objective of this study reflects whether the mature source controls installed in one neighborhood significantly reduce peak flows and total runoff volumes compared to a nearby neighborhood without these intentional measures. 21  2.5 Methodology To address the question of whether the partial suite of lot-only source controls (or low impact development prac-tices) in an urbanized catchment – even nine years after installation – make a significant reduction in peak runoff flow and total runoff volume, a paired catchment approach was used. Flow from two subcatchments in the Township of Langley (TOL) was examined in this study. The following discussion on methodology addresses the study sites; examines measures/statistics used to determine their comparability; discusses the method of acquiring rainfall and flow measurements; and outlines the desktop method of normalizing flow volumes by catchment area, removing groundwater and interflow, and comparing peak flow and cumulative storm runoff for a series of frequently-occur-ring storms (on short return periods). 2.5.1 Study sites in the Township of Langley This study investigated the runoff response in storm sewers draining two neighborhoods in close proximity to each other in the Township of Langley, developed with different stormwater management strategies. The location of the two catchments are indicated in Figure 2.3. Although site verification was performed and rain gauges were installed by the researcher in both neighborhoods, this is primarily a desktop-based investigation, making use of quality-con-trolled data collected by professional hydrology and civil engineering consultants. 22   Figure 2.3 Paired watersheds in the Township of Langley 2.5.1.1 Infrastructure & context of flow monitoring sites Under a nine-year flow monitoring contract performed by Kerr Wood Leidal Associates Ltd. (KWL) for the Town-ship of Langley’s Engineering Department, flow measurement equipment had been installed and maintained since 2005; rainfall and flow measurements had been analyzed to determine the hydrological impacts of the source con-trols installed on lots as part of the neighborhood plan at build-out of the Routley neighborhood. The requirement for 23  biannual reports on monitoring for the first nine years was a requirement set out in Item 27 of the DFO’s Authoriza-tion for Work of Undertakings Affecting Fish Habitat issued in 2005 (KWL 2013). KWL’s Routley Neighborhood Area Stormwater & Environmental Monitoring annual reports for the Township of Langley were an important source of information for this investigation, as were the raw data made available via FlowWorks web-based database and analysis tool with permission from KWL and the Township of Langley. 2.5.1.2 Build-out conditions & time frame The Township of Langley is located in the lowland area of the Fraser Valley, in south-western British Columbia. With a growing Greater Vancouver population but limited north, south, and westward expansion opportunities due to mountains, the border with the USA, and the Pacific ocean, respectively, urban development and densification has been pushing eastward up the Fraser Valley for the last few decades. In 1961, the population of the Township of Langley was registered at 14,585, growing fairly consistently at a compound rate of 4.1% per year to 614% greater than its 1961 population in 2011, when the Township of Langley had a total population of 104,177 residents living in 39,114 private dwellings for an average of 2.66 people per residential unit (Township of Langley 2012).  The paired catchments examined in the Township of Langley, using study infrastructure and data from the Routley Neighborhood Area Stormwater & Environmental Monitoring project, were, specifically, a subcatchment within the Langley Meadows neighborhood, and a subcatchment within the Routley neighborhood with similar area and hy-draulically-relevant statistics. The Langley Meadows Neighborhood, roughly bounded by 197 Street, Wakefield Drive, and Willoughby Way (to-ward the bottom of Figure 2.5), was built out in the mid-1980s with conventional stormwater management practices, under a philosophy of direct connection of all pervious areas, roof downspouts included. The nearby Routley neighborhood, roughly between 196 Street and 197 Street, and between 73B Avenue and 71 Avenue (the upper portion of Figure 2.5), was constructed in the mid-2000s under guidance of the Routley Neigh-borhood Area Stormwater Management Plan and resulting bylaws that specified a comprehensive set of source con-trols for each residential lot. A 15 m3 infiltration/rock trench, minimum lawn soil depth of 300 mm, and disconnect-ing roof leaders (draining to side yards) were specified for each lot (Township of Langley 2003/2004). Other imper-vious areas, such as driveways and streets, are directly connected to stormwater conveyance infrastructure by con-ventional curb & gutter practices. As has been documented via aerial imagery by KWL, since original build-out the subcatchment has experienced continuous development, with the last lots being subdivided and re-developed in 2014. 24   Figure 2.4 Typical self-draining lot with front & rear lot drainage outside of driveway 2.5.1.3 Drainage catchment area Contributing catchments to the storm sewer (drainage) flow are delimited by both their drainage network and sur-face terrain boundaries. Since the drainage point for measuring flow from each catchment is at a fixed location in the storm sewers, the drainage network defines the primary structure of channel flow, with surface terrain contributing an overland catchment to each point of the sub-terrestrial engineered drainage network. The Langley Meadows subcatchment (at the bottom of Figure 2.5, below) has a total contributing area of 13.59 hec-tares, as initially determined by the KWL project, and subsequently verified over the course of this research project through a combination of work with global information systems (GIS) software, and ground-truthing by several field visits to determine storm sewer connections and the direction of flow of areas near to the boundaries of the catch-ment. In some cases the surface catchment based on topography added considerable complexity to the primary (storm sewer) drainage network-defined catchment boundaries. Specifically, although the school parking lot (at the 25  south-east corner of the catchment) slopes away from the road and apparently does not contribute area to the catch-ment, it is hydraulically connected by the piping network.   Figure 2.5 Langley Meadows (bottom) & Routley (top) catchments 26  The Routley subcatchment (at the top of Figure 2.5, above) has a total contributing area of 13.57 hectares, also as originally determined by the KWL project and verified by both GIS and site visits. Incongruities between the storm drain-defined catchment and the surface topography-defined catchment contributed complexity to identifying ex-actly the contributing catchment. Specifically, a section of paved 72nd Avenue overlaying the drainage network actu-ally does not contribute to the Routley catchment, but rather flows east to an adjacent catchment. The two catchments thus have similar contributing areas: not just on the same order of magnitude, but within 0.2% of each other. That the two catchments are of similar area is the first confirmation that they can fairly be compared in this paired catchment study. 2.5.1.4 Total impervious area Although much of the early literature identified TIA as the key variable driving a decline in watershed health (Klein 1979, Arnold & Gibbons 1996, Limburg & Schmidt 1990), later findings indicated that it is rather the effective im-pervious area (EIA) – the percent of the catchment that is both impervious and directly connected to the drainage network (Booth & Jackson 1997, Walsh 2004). Since in this study the source controls’ contribution to a reduction in EIA is essentially what is under investigation, (whether disconnecting the hard-surface contribution from roofs actu-ally reduces peaks and total event runoff), it would be a confounding factor in the experiment if there was a signifi-cant difference in contribution from different impervious surface types, like roads or roofs. So once the catchment perimeters were satisfactorily determined, the impervious surfaces in the catchments were manually identified on ortho-rectified high-resolution aerial imagery taken in the summer of 2013, with pixel size of 0.1 m x 0.1 m, using the software ArcMap 10.2. Total area in the two catchments was partitioned among (a) roofs, (b) driveways/walkways/decks and other paved surfaces on lots, (c) pervious lot area, (d) roads, (e) sidewalks within the road right of way (ROW), and (f) pervious area within the road ROW. These values, along with the percent TIA they represent, are summarized in Table 2.3. The series of figures below illustrate the buildup of impervious area in the catchments. First, the drainage network defines the primary bounds of the catchment, with adjustments given lot boundaries and the slope of the surface near the catchment perimeter, as in Figure 2.6. Langley Meadows (without source controls) is pictured on the left; the Routley neighborhood (with lot-only source controls) is pictured on the right. Small circles mark the point to which each catchment drains, and where instruments installed in manholes measure flow. The drainage network is traced in Figure 2.7. Impervious area from roofs in each catchment are identified in Figure 2.8; the addition of roads in Figure 2.9, sidewalks in Figure 2.10, and additional contributions from driveways and other on-lot impervious areas in Fig-ure 2.11. The complex mix of different contributions of impervious area in the two catchments is illustrated in Figure 2.12. The area left over after the on-lot and road ROW impervious areas have been subtracted is considered to be pervious and allowing the infiltration of rainfall. In the Langley Meadows catchment, it is assumed that roofs and driveways are directly connected to the stormwater conveyance network (including both roads and storm drains), but that roofs 27  and driveways in the Routley neighborhood are not directly connected to the conveyance network (but via overflow from the rock trench).  Figure 2.6 Langley Meadows and Routley catchments  Figure 2.7 Langley Meadows and Routley drainage networks 28   Figure 2.8 Langley Meadows and Routley roof contributions to TIA  Figure 2.9 Langley Meadows and Routley road contributions to TIA 29   Figure 2.10 Langley Meadows and Routley sidewalk contributions to TIA  Figure 2.11 Langley Meadows and Routley driveway and other on-lot contributions to TIA 30      Figure 2.12 Langley Meadows and Routley pervious and impervious areas In 2013, the 13.6 hectare Langley Meadows subcatchment had a total impervious area of 8.62 hectares, or 63%. The neighborhood’s impervious area is distributed between roads (2.21 Ha, or 16% of the catchment area), sidewalks (0.91 Ha / 7%), roofs (3.87 Ha / 28%), and driveways and other on-lot impervious area (1.63 Ha / 12%). This left 4.97 Ha, or 37% of the catchment as pervious, including wooded, lawn or other undeveloped areas. Table 2.3 sum-marizes these data. A total of 235 single-family residences (averaging 160 m2 each) are located within the catch-ment, for a density of 17.1 residential units per hectare. In the same year, the 13.6 hectare Routley subcatchment had a total impervious area of 9.47 hectares, or 70%. This neighborhood’s impervious area is distributed between roads (2.59 Ha, or 19% of the catchment area), sidewalks (1.04 Ha / 8%), roofs (4.25 Ha / 31%), and driveways/other on-lot impervious area (1.60 Ha / 12%). This left 4.10 Ha, or 30% of the catchment as pervious, although the remaining lot was at the time being redeveloped into seven smaller lots. The catchment contained 204 single family residences (averaging 202 m2 each), for a lower residential density of 15.5 residential units per hectare, despite a higher roof contribution to impervious area, as roofs are on average 26% larger than in the Langley Meadows catchment. 31  Table 2.3 Summary of impervious areas in paired TOL catchments  As the second check (after total catchment area), the differences in total and partitioned impervious areas does not seem to be a factor that would confound a paired catchment study. At 63% vs 70% TIA, the two catchments are within 10% of each other. As will be explored further along, it is here assumed that the effective impervious area (EIA) of Langley Meadows is close to the TIA, though in the Routley neighborhood it is significantly different. On the grounds of having similar impervious area, and similar distribution of impervious area among contributing types (roads, sidewalks, roofs, and driveways), a comparison of the two catchments in this study is deemed fair. 2.5.1.5 Soils & infiltration rates The third check is whether the mineral soils of the two catchments are similarly draining, for if one has, say, well-draining soils and the other, say, poorly draining soils or a perched water table, the soil type itself would confound a comparison between runoff responses from the two sites. (If this were the case, precipitation would preferentially infiltrate in one catchment over the other, and the impact of the source controls on runoff would be impossible to separate from the effects of differing infiltration rates. In such a case, the different infiltration rates would cause con-siderable noise in the runoff profile that would occult the source controls’ signal that we hope to examine.) Of course well-draining soils contribute positively to infiltrating precipitation volumes, and are capable of more rap-idly storing in shallow or deep groundwater that from a more intense or sustained rainfall event than poorly draining soils. From Soils of the Langley-Vancouver Map Area (Luttmerding 1980), and open-source surficial geology maps made available online on the Township of Langley’s GeoSource interactive mapping tool (http://geosource.tol.ca/), I was able to roughly triangulate soil types and infiltration rates for the two neighborhoods. The Routley neighborhood has two major soil types as identified in the Luttmerding (1980) publication of area soil surveys: Area (Ha) of Total Area (Ha) of TotalTotal 13.59 100% 13.57 100%ROW 3.80 28% 4.59 34%Roads 2.21 16% 2.59 19%Sidewalks 0.91 7% 1.04 8%Pervious 0.68 5% 0.97 7%ROW TIA 3.12 23% 3.62 27%Lot 9.79 72% 8.98 66%Roofs 3.87 28% 4.25 31%Driveways Etc. 1.63 12% 1.60 12%Pervious 4.29 32% 3.13 23%Lot TIA 5.50 40% 5.85 43%Catchment TIA 8.62 63% 9.47 70%ParamaterConventional (Langley Meadows) Source Control (Routley Neighborhood)32  Bose-Whatcom-Scat, identified by the symbology (BO-W-SC)/(c; S1-2), is made up, specifically, of Bose (BO), a Duric Humo-Ferric Podzol described as “30-60cm of gravelly lag or glacial outwash deposits over moderately coarse textured glacial till and some moderately fine textured glaciomarine deposits” that is well to moderately well draining and with telluric seepage; Whatcom (W), a Luvisolic Humo-Ferric Podzol described as “moderately fine textured glaciomarine deposits” that is moderately well draining and with telluric seepage;” and Scat (SC), an Or-thic Humic Gleysol is described as “moderately fine textured glaciomarine deposits” that is poorly draining with a perched water table.” The topography is described as gently undulating (with 0.5-2% slopes), and slightly to moder-ately stony. Approximately 65-70% of the soils in the Routley subcatchment under investigation is of the Bose-Whatcom-Scat soil group. Another 20-25% of the subcatchment soils are Scat-Boosey, identified by the symbology (SC-BY)/(bc; S1), which are made up of Scat (SC), described above, and Boosey (BY), a Rego Humic Gleysol described as “30-160cm of gravelly lag or glacial outwash deposits over moderately coarse textured glacial till and some moderately fine tex-tured glaciomarine deposits” that is poorly draining with a perched water table. The area has gently undulating to undulating topography with 0.5-5% slopes, and is slightly stony.  The remaining 5% of the soils are Bose-Boosey, identified by (BO-BY)/(CD; S2), made up of both Bose (BO), de-scribed above, and Boosey (BY), also described above, is moderately stony and has gently to moderately sloping topography. Comparatively, the Langley Meadows neighborhood has two other soil types: Bose-Nicholson, identified by the symbology (BO-N)/(DE; S2-1) is made up of Bose, described above; and Nichol-son (N), a Podzolic Gray Luvisol described as “moderately fine textured glaciomarine deposits” that is moderately well draining. The area’s topography is characterized as moderately to strongly sloping (5-15%), and is slightly to moderately stony. Approximately 35-45% of the soils in the Langley Meadows subcatchment under investigation is Bose-Nicholson. The remaining 55-65% of the soils in the catchment are identified by the symbology CD/B, made up of Cloverdale (CD) soil , a Humic Luvic Gleysol described as “moderately fine to fine-textured marine deposits” that is poorly draining with a perched water table with very gently sloping (0.5-2%) topography. 33   Figure 2.13 Soil map of the catchments under investigation (Luttmerding 1980) Remembering that this inquiry into soils is to determine whether the drainage rates of the two catchments are com-parable, or so significantly different as to disqualify the paired catchment study, Table 2.4 below summarizes the drainage narratives and contributing areas of each of the major soils groups. It must be recognized that the 34  Luttmerding (1980) soil surveys are mapped at a 1:20,000 scale, and are not accurate on the site scale, which does introduce a degree of uncertainty into the application of these soil group qualities to individual lots.  Table 2.4 Soil drainage in catchments under investigation  Both “poorly draining with perched water table” and “well to moderately well draining” soil types are distributed across both catchments. To estimate the weighted geometric mean of the saturated hydraulic conductivity (KS) for each catchment, I assign values, listed in Table 2.5, to soil groups described as “well to moderately well draining,” “moderately well draining,” and “poorly draining with perched water table,” for three different sensitivity analysis conditions. Condition A is very conservative, B is less so, and C even less so, with still a maximum hydraulic con-ductivity of 4 mm/hr. This upper limit is in line with BMP selection criteria in KWL et al. (2012), for which absor-bent soils placed to 300 mm depth should be able to capture 90% of annual rainfall, even when the impervious-to-pervious (I/P) ratio is 2:1.   Table 2.5 Hydraulic conductivity combinations for soils comparison  Then, in Table 2.6 is summarized the results of the exercise of assigning these Ks values to relative extents of each catchment, and calculating the geometric mean saturated hydraulic conductivity for each.  Soil Group Soil Name Drainage ExtentBO-W-SC Bose well to moderately well drainingWhatcom moderately well drainingScat poorly draining with perched water tableSC-BY Scat poorly draining with perched water tableBoosey poorly draining with perched water tableBO-BY Bose well to moderately well drainingBoosey poorly draining with perched water tableBO-N Bose well to moderately well drainingNicholson moderately well drainingCD Cloverdale poorly draining with perched water table 60%RoutleyLangley M.70%25%5%40%A B Cwell to moderately well draining 2.0 3.0 4.0moderately well draining 1.0 2.0 3.0poorly draining with perched water table 0.5 0.5 0.5Estimated Ks (mm/hr)35  Table 2.6 Mean saturated hydraulic conductivity calculated for three Ks combinations  For combination A, the two values of mean Ks are only 10% different; for combination B, they are 14% different, and for combination C, they are 14% different, Routley neighborhood always with the higher value. Even perform-ing this exercise with a fourth option (not shown in the table), with KS values of 0.5, 5, and 30 mm/hr, we find a maximum difference of only 20% between the two catchments. The resolution of information on the soils drainage and distribution is sufficient to argue that the hydraulic conductivity of the two catchments is small, and that the two nearby catchments do not have sufficiently different soil drainage characteristics to disallow the paired study. 2.5.1.6 Slope & time of concentration As precipitation falls on a catchment, individual droplets that fall in different parts of the watershed will take differ-ent periods of time to reach the watershed outlet. The impervious area, infiltration rates of various surfaces, slope, overland flow length and roughness, channel length and roughness, and intensity of storm will all contribute to char-acterize the runoff response from the watershed. For a given (design) rainfall intensity, the longest time that an indi-vidual droplet that falls on the catchment will take to flow to the outlet is termed the ‘time of concentration,’ tc, or the time at which all parts of the catchment are contributing to runoff at the outlet. It is to this rainfall intensity that storm sewers are designed. Rainfall intensity listed in Table 2.9 in the next section, define the intensity-duration-frequency (IDF) curves for the study area, plotted in Figure 2.20 (FlowWorks 2014). Although the design storm for storm sewer piping might be for the 15-minute duration storm on a 10 year return period (9mm over the entire watershed within 15 minutes), design storms for the source controls may be considerably smaller (less intense), such as the 24-hour duration storm on a 6-month return period (perhaps less than 50mm over the entire watershed within 24 hours). In a typical catchment, to calculate the time it takes for a unit of runoff from the furthest part of the catchment to reach the outlet, the time of channel flow is added to the time of overland flow (tconcentration = toverland flow + tchannel flow). The Kirpich (1940) method in catchments with primarily channel flow, or the Kerby-Hatheway (1959) method in Soil NameKs (mm/hr)Estimated ExtentMean Ks (mm/hr)Ks (mm/hr)Estimated ExtentMean Ks (mm/hr)Ks (mm/hr)Estimated ExtentMean Ks (mm/hr)Bose 2.0 23.3% 3.0 23.3% 4.0 23.3%Whatcom 1.0 23.3% 2.0 23.3% 3.0 23.3%Scat 0.5 23.3% 0.5 23.3% 0.5 23.3%Scat 0.5 12.5% 0.5 12.5% 0.5 12.5%Boosey 0.5 12.5% 0.5 12.5% 0.5 12.5%Bose 2.0 2.5% 3.0 2.5% 4.0 2.5%Boosey 0.5 2.5% 0.5 2.5% 0.5 2.5%Bose 2.0 20.0% 3.0 20.0% 4.0 20.0%Nicholson 1.0 20.0% 2.0 20.0% 3.0 20.0%Cloverdale 0.5 60.0% 0.5 60.0% 0.5 60.0%Langley M.CBARoutley1.000.901.501.301.991.7036  catchments with mostly short reaches of overland flow, and the Morgali & Linsley (1965) method for planar drain-age in urban catchments up to 20 hectares, can be used where appropriate. The fully-urbanized catchments under study are distinctly different from forested catchments in many ways, notably here in that drainage of precipitation from the entire catchment is accelerated by direct connection of hard surfaces, so very little overland flow extends the time of concentration above that accrued by the channel flow, especially at rainfall intensities above or below the storm sewer network’s design time of concentration. A high-altitude (and thus not very high resolution) examination of channel flow is necessary for this research, and since the roadways in both catchments are directly connected to the piping network, the overland flow component is assumed not to contribute significantly to the time of concentration, so the channel flow (in the collection piping network) is examined in absence of calculation of overland flow in either catchment. One hydrological symptom of directly connecting catchment impervious surfaces to drainage infrastructure (and thus to natural streams) is that the peak runoff flow is caused to be timed more closely coupled to that of the event’s most intense rainfall. Remembering that in a natural catchment, layered vegetation, absorbent organic soils, and un-disturbed infiltration pathways both reduce runoff volumes and delay its time to peak flow, the paving over of great portions of this area and efficiently connecting these areas to a conveyance network should accelerate the transfer of runoff to the streams, thereby reducing the time of concentration of the catchment. It could be expected then, that as source controls that effectively disconnect certain hard surfaces from the drainage network are installed, the runoff volumes should not only be reduced, but should be drawn out temporally as well. This comparison is made to determine if differences in shape, size, distribution, or composition of the catchments will lead to a significant delay in peak in one catchment as compared to the other. From the Township of Langley’s publicly accessible online Open Data catalogue (TOL 2013), storm drainage was downloaded both in shape-file format (for geospatial plotting and analysis with ArcMap ™ 10.2 software). Pipe sec-tions within the two subcatchments were isolated (from those outside the catchments), and associated data including identifier, length, slope, material, and installation year for each pipe section was exported as a spreadsheet for subse-quent processing in Microsoft Excel ™ software. For each section of pipe (with its own unique slope and diameter), using the general Manning equation for uniform flow, v = (Rh2/3So1/2)/n, where v = velocity (m/s), Rh = hydraulic radius, So = slope, and n = Manning’s n (a coeffi-cient of channel roughness), the velocity of flow in each pipe (running half-full) was calculated. Then, by dividing each pipe length by the velocity calculated using the Manning equation, a time of flow in each pipe section was cal-culated.  Then for each branch of a catchment, the calculated ‘time of flow’ values for each discrete pipe section were summed, giving a total time of channel flow. The branch with the longest total channel flow – identified in Figure 2.14, thus contributed the defining time of concentration, at that pipe-flow depth (a function of storm intensity), for that catchment. Given different slopes along the different drainage routes, in the Langley Meadows catchment the longest time to outlet (6.0 minutes) is in the second-longest pipe route to outlet (550 m); the longest pipe route to 37  outlet (684 m) from the north-west corner of the catchment takes a slightly shorter 5.6 minutes to reach outlet. In the Routley catchment, the maximum time to outlet of 9.7 minutes running half-full does occur in the longest pipe route to outlet of 801 m. These data are summarized in Table 2.8 below.  Figure 2.14 Collection network branches defining longest time to outlet in the study catchments And calculating the hydraulic radius for a range of pipe-full depths from 5% to 95% for the two defining drainage routes (from the most hydraulically distant points in each catchment – a function of slope and pipe material as much as total pipe length), it is shown in Figure 2.15 that during rainfall events below the design storm (toward the left on the plot), the time of channel flow increases in both catchments, and the magnitude of the difference in time of chan-nel flow increases (though that in the Routley neighborhood is calculated to remain at consistently 38% greater de-lay than that in the Langley Meadows neighborhood). 38   Figure 2.15 Time of channel flow as a function of pipe-full depth Table 2.7, below, shows the range of slopes within each catchment: clearly the Langley Meadows neighborhood has steeper slopes, which should, of itself, accelerate runoff toward the outlet and shorten the time of concentration. Table 2.7 Topography of study catchments  Differences between the two catchments’ average slope, distribution of drainage pipe, age and material of pipe, maximum length of drainage pipe from outlet, and, not mentioned above, distribution of directly-connected pervious areas each contribute to either shorten or lengthen the time of concentration, or the delay to peak flow. For example, that a large school parking lot is directly connected to the drainage infrastructure fairly nearby the Langley Meadows catchment outlet would cause a rapid rise in runoff flow directly linked to rainfall. Aside from directly-connected roads, the Routley catchment does not have this.  Soil Group Slope ExtentBO-W-SC 0.5-2% 65-70% of RNSC-BY 0.5-5% 20-25% of RNBO-BY 2-9% 5% of RNBO-N 5-15% 30-40% of LMCD 0.5-2% 60-70% of LM39  Table 2.8 Routley and Langley Meadows drainage network comparison  Rainfall and runoff data has 5-minute resolution: excellent for comparisons of flow volumes and linking to timing and intensity of rainfall. With the difference in calculated minimum difference in time to outlet between the two catchments (6 minutes vs. 10 minutes), the difference is approximately equal to the data resolution. Differences in time to outlet between the catchments, then, do not preclude comparison of flow data, or time-to-peak comparisons. 2.5.2 Rainfall For the TOL-sponsored nine-year study, rainfall has been measured by a tipping-bucket rain gauge located on the roof of the Township’s municipal offices located at 20338 65 Avenue: on the corner of 203 Street and 65 Avenue, 1.4km to the east of the centroid of the Langley Meadows catchment, and 2.3 km to the south-east of the centroid of the Routley catchment, as identified in Figure 2.16, below. Since the pattern of rainfall may be temporally and spa-tially inconsistent between where it is measured and the two study sites, it was therefore important to verify similar-ity (or significance of differences) between the existing rain gauge and that measured actually within the study catchments – so that inter-catchment comparisons of peak flow, total runoff, and time to peak (functions of rainfall timing and/or intensity), can be confirmed to be accurate. To do so, additional rain gauges were installed, located where illustrated in Figure 2.16, and as pictured in Figure 2.17. Parameter Routley Langley MeadowsTotal length of pipes in drainage network (m) 2,594 1,971Material PVC, Concrete AC, ConcreteYears of install 2004-2008 1981-1985Pipe sizes (mm) 250-750 250-750Slope range (%) 0.21-2.02 0.14-4.50Average slope (%) 0.74 1.96Velocity range (50% Full) 0.69-2.41 0.81-2.89Velocity range (25%, 75% Full) 0.49-1.69 0.57-2.02Velocity range (10%, 90% Full) 0.28-0.97 0.33-1.16Max pipe length to outlet (m) 684 801Pipe length to outlet for max time to outlet (m) 550 801Max time to outlet (minutes)    6.0 *    9.7 ** at 50% full40   Figure 2.16 Existing rain gauge relative to study catchment locations. I installed the tipping-bucket rain gauges at optimal locations within the two neighborhoods, both with similar expo-sure and distance from trees and buildings. After briefly introducing myself and the study to the residents, and shar-ing with them letters of introduction from UBC and the Township of Langley, as well as a short introduction to the research (all of which are included in Appendices F and H), home owners at the two locations were very willing to host rain gauges, generously allowing me to install a rain gauge on the side of a recently-constructed shed in Lang-ley Meadows, and beside the garden in Routley. The homeowners also accommodated my infrequent visits to down-load rainfall data and maintain the gauges. 41   Figure 2.17 Rain gauges installed in Langley Meadows (left) and Routley (right) neighborhoods 2.5.2.1 Rainfall statistics Rainfall data has been collected by the TOL at the site noted above since mid-2005 (FlowWorks 2014). The follow-ing paragraphs discuss the annual and monthly historical trends, as well as the intensity-duration-frequency curves developed for the area, which are used to determine design storms for sizing of infiltration practices. 2.5.2.1.1 Annual average total From eight years of historical rainfall data at the existing TOL rain gauge, the average annual rainfall is calculated as 1300 mm, (FlowWorks 2014). Annual cumulative rainfall for the past eight years are compared in Figure 2.18, below. Historical monthly average rainfall is presented in Figure 2.19, for appreciation of the distribution (Flow-Works 2014) and monthly range of distribution (max and min) around the eight-year mean. The dashed line indi-cates 2013 monthly totals, with record low July rainfall. 42   Figure 2.18 Cumulative annual rainfall at TOL Municipal Hall for last eight years  Figure 2.19 Monthly rainfall at TOL Municipal Hall 2.5.2.1.2 Intensity-duration-frequency The series of intensity-duration curves for rainfall events in the Township of Langley with return periods between two and one hundred years (see Table 2.9) based on the 37 years of rainfall monitoring (1962-1998) at Surrey – Kwantlen Park, are plotted in Figure 2.20 (FlowWorks 2014). Any large storms can be compared to these curves to determine the return frequency of a given intensity-duration combination. For the purposes of this study, however, it 43  is the smaller events – up to the 24-hour event on a 6-month return period that are of interest: those that should not be overwhelming the capacity of the designed source controls to infiltrate precipitation.  Table 2.9 Rainfall intensity (mm) for given storm durations and return frequencies   Figure 2.20 Intensity-duration-frequency curves for the study area (FlowWorks 2014) In Metro Vancouver, the 6-month, 24-hour rainfall event is typically in the range of 67-76% that of the 2-year, 24-hour event (mean 72%), and can be acceptably assumed when an IDF curve has not been developed for the higher frequency events (KWL et al. 2012). With spatially-variable annual rainfall volumes within Metro Vancouver, the 6-month event also varies by location: from White Rock to Burke Mountain, for example, the 6-month (24-hour) event varies between 34 mm and 66 mm, respectively. Here, the design storm is 72% of 2.6 mm/hour for 24 hours, or 44.9 mm. Fr qu ncy 5 min 10 min 15 min 30 min 1 hr 2 hr 6 hr 12 hr 24 hr2 Year Event 37.2 27.0 21.6 15.2 10.7 8.0 5.2 3.8 2.65 Year Event 50.4 37.2 30.4 20.8 13.7 9.9 6.3 4.7 3.410 Year Event 60.0 44.4 36.0 24.6 15.7 11.2 7.1 5.3 3.925 year Event 70.8 53.4 43.2 29.4 18.1 12.8 8.0 6.1 4.550 Year Event 79.2 59.4 48.4 33.0 20.0 13.9 8.7 6.6 5.0100 Year Event 87.6 66.0 53.6 36.4 21.8 15.1 9.3 7.0 5.5Storm Duration44  2.5.3 Flow measurement In addition to precipitation data collected through KWL’s multi-year project for the Township of Langley, for this analysis I also made use of 5-minute resolution storm sewer flow data collected under the same project. The data was also accessed via FlowWorks, the online portal to the stored data. For the flow monitoring, KWL had subcon-tracted installation of the weirs in the manholes to the local office of SFE Global (SFE), and shared regular mainte-nance and data download of the sites with the subcontractor. Processing of pressure readings from pressure transduc-ers installed upstream of the weirs provide water level, and from proprietary stage-discharge relationships (custom for each weir), storm flow over the weir is calculated. SFE’s proprietary compound-profile weirs are designed to provide accurate flow measurements over the entire range of expected flows: low flows trickle over a rectangular notch; flows in the medium range flow over a regularly-broadening opening (like a Vee-notch weir), and high flows again flow through a rectangular opening, roughly as illustrated in Figure 2.21. Illustrated are (a) the weir installed in a storm sewer, (b) low flow measured in the lower rectangular opening, (c) mid-level flow measured in the Vee-section, and (d) high flow measured in the upper rec-tangular portion. Since the weirs’ geometries and hydraulics are distinctly different from conventional weirs, the conventional stage-discharge relationships do not apply.   Figure 2.21 Illustration of SFE’s custom compound weirs at various flows The ‘custom compound weir’ was originally developed in 1983 through modelling and testing at the Canadian Cen-tre for Inland Waterways, and have since further developed the stage-discharge relationship with weir geometry (SFE Global 2008). With customized applications in use at over 2000 sites across the USA and Canada, the custom compound weir can be accurate down to approximately 1 l/s. Since ± 5% of the actual flow can be achieved, and the 45  design is supposedly self-cleaning, it is useful in both storm and sanitary sewer applications, and, coupled with an accurate level meter (such as a pressure transducer), can provide more reliable readings than does a conventional V-notch weir (SFE Global 2008). In the storm sewer manhole located in the street at the corner of Willoughby Way and Wakefield Drive, at the hy-draulic outlet of the Langley Meadows subcatchment, SFE had installed a custom compound weir with a 350 mm width upper rectangular section. The lower lip of the weir is elevated from the storm sewer invert, providing 250mm head for minimal flow over the lip. The lip of the weir is 2,165 mm below the manhole ring at street level. The flow control weir is complemented with a Teledyne ISCO 2150 area-velocity meter, with level measurements capable of ± 3 mm accuracy from a submerged pressure transducer mounted in the invert of the pipe upstream of the weir. An ultrasonic Doppler meter measures velocity, with accuracy of ± 0.03 m/s at low flow, and ± 2% of the reading above 1.5 m/s. According to Teledyne ISCO product literature, the flow module emits an ultrasonic wave that re-flects off of bubbles and particles entrained in the flow, the frequency shift of the ultrasonic waves’ echoes is then measured by the probe and digitized (Teledyne ISCO 2008). At this site, measured level, flow, and velocity values are recorded in five minute increments, and are downloaded monthly by SFE. In the storm sewer manhole at the corner of 196B Street and 71 Avenue, the hydraulic outlet of the Routley sub-catchment under study, SFE had installed a custom compound weir with a 750 mm wide upper rectangular section. A deep manhole, the lip of the weir is 5,392 mm below the manhole ring, and provides 290 mm head above the in-vert of the manhole. It is similarly fitted with Teledyne ISCO 2150 area-velocity meter, also logging 5-minute incre-ment level, flow, and velocity values, and maintained by SFE staff. The monthly level, flow, and velocity data downloaded by SFE from the Teledyne ISCO meters installed in the manholes was subsequently regularly uploaded to the FlowWorks server, available for quality control, manipulation, graphing, and download. FlowWorks is an online platform for storing, managing, analyzing, graphing, and reporting hydrometric data for clients. Now operating out of Seattle, Washington, FlowWorks originated as an offshoot of KWL, using the company’s decades of experience with hydrometric analysis and reporting to develop this sophisti-cated suite of integrated hydrometric tools, with its latest calculation engine and user interface launched in 2011 (FlowWorks 2014).  2.5.4 Manipulation of flow data With clearance to view, download, and use the data given by Mr. Art Kastelein, an engineer with the Township of Langley, and access to the password-protected FlowWorks interface provided by KWL, I chose to use rainfall and flow data from storms during only the last three years (2011, 2012, 2013), since software underwent the latest qual-ity control improvements in 2011. In addition to having the capability to graphically visualize the data online, I was also able to download comma sep-arated value files containing the data of interest that can be opened and manipulated using common spreadsheet pro-grams such as Microsoft Excel™. 46  The methodology used to identify unique storms (and calculate pertinent statistics including rainfall volume, dura-tion, intensity, antecedent dry weather period, and rainfall in previous seven- and thirty-day periods), normalize flow volumes by area, remove groundwater and interflow (GWI) contribution to flow to leave only runoff, unitize cumu-lative event runoff volumes, plot, and make comparisons among events 2-8 mm, 8-16 mm, 16-30 mm, and 30-60 mm is outlined in Figure 2.22, below. To accelerate the replicable process (and automate it for use with data from other catchments), I built a spreadsheet into which over 315,000 rows of five-minute resolution rainfall and flow data (12 readings per hour, 24 hours a day, 365 days a year, for three years) could be pasted so that I could assemble, plot, analyze, and manipulate it. These steps are given some detail in the following sections. 47   Figure 2.22 Flow chart of methodology to compare flow from the catchments 48  2.5.5 Normalization by area If and when the paired catchments are of unequal size, it is necessary to normalize the storm flows by the area of each catchment, to remove the effect of differences in catchment area. (If comparing two differently-sized catch-ments, for example, for a given depth of rain falling on both of them, the larger catchment would receive a larger total volume of water than the smaller catchment, so storm sewer flows would be appropriately larger in the larger catchment.) To match units of rainfall (for subsequent storm-by storm precipitation mass-balance for the catch-ments, the backbone of the analysis described here), raw five-minute volumes (units of m3/5minutes) are divided by catchment area (units of m2), and converted to millimeters by multiplying by 1000 mm/m, as in: Normalized Vol-ume [mm/5min] = (1000 x Raw Volume [m3/5min])/Area [m2]. This was initially identified as necessary for comparison of runoff from the Langley Meadows and Routley sub-catchments under study, because I had been relying on area measurements as performed by others. My own meas-urements using ArcMap™ 10.2, as outlined previously, identified that the two catchments are of almost equal size (13.59 Ha and 13.57 Ha, respectively). With catchments this close in size, normalization would not have been neces-sary, but is included here as part of the methodology so the process can be replicated with catchments less similarly sized. 2.5.6 Removal of groundwater & interflow During a rainfall event, stream (or in this case storm sewer) flow sharply rises and then drops off back toward the base flow. Contributions to this storm-response hydrograph from three contributing sources: surface runoff, inter-flow, and shallow groundwater. The groundwater and interflow (GWI) contributions are variable, depending on an-tecedent dry weather conditions, and depth to groundwater, and soil permeability, and rises during infiltration and ground saturation during a rainfall event, until, as the hydrograph illustrated in Figure 2.23 illustrates, the GWI again contributes entirely to the flow of the stream and drops again to pre-storm conditions (Chow 1964, Davie 2008). De-pending on antecedent conditions, and intensity and volume of the storm, the GWI may take on the order of hours or days to return to pre-storm levels (Davie 2008). In this study, in which differences in runoff contributions to stream flow during and after an event is specifically what is under investigation, fluctuations in groundwater and interflow contributions become noise that can mask the runoff signal in the flow. From the evidence of continued flow measured at the outlet of both subcatchments’ drainage networks, it can be concluded that the storm sewers have a degree of permeability, with measurable inflow of groundwater and inter-flow even during dry times. Around a storm event, it is thus important to calculate (or at least estimate) how much of the flow is from groundwater and interflow, and how much is actually runoff. Since water running in sheet flow across the surface, and in trenches and pipes responds in a similar time-scale to the rainfall, but the movement of groundwater is on a very different time-scale, their responses to rainfall and contributions to storm flow can be sepa-rated. 49  Removing the groundwater and interflow from stream flow is an iterative exercise: to determine the appropriate rate at which the groundwater and interflow increases their (combined) contribution to stream flow, one has to make an engineering judgment as to where the break in the curve on the hydrograph’s falling limb indicates a difference in the source of stream contributions. For example, referencing Figure 2.23, recreated from Chow (1964) via Spinello & Simmons (1992), the precipitous drop after peak (between points C and D), occurring within the first couple of hours after a storm peaks and passes, is attributed to runoff, after which interflow and groundwater contribute more completely to the more gentle part of the curve (between points D and E). Groundwater rises and interflow contrib-utes to stream flow at a rate roughly estimated by drawing a line on the hydrograph between point B (base flow be-fore the increases because of rainfall) and point D. The slope of this line is the rate at which GWI contributions to stream flow increase over time. To remove the noise from a storm-by-storm or an annual hydrograph, this slope must be estimated, and the flow from groundwater and interflow – shaded gray in the image - must be removed.  Figure 2.23 Estimation of the GWI contribution to event runoff During an increase in stream flow due to rainfall, the spreadsheet thus calculates the rising contribution from GWI, with the slope of the rising GWI variable by the researcher and adjusted within a narrow range so that the rising GWI intersects the falling limb where a slope break is perceived by the researcher. Johnston et al. (2002) caution that “the analysis of a single storm event may significantly over- or under-estimate performance, depending on the antecedent conditions,” and suggests that in this region, where there is significant climatic differences between a dry 50  summer and wet winter, use of this method over an annual hydrograph (using a single GWI slope value) may offer a rapid assessment, and may still be appropriately accurate to determine the effective impervious area of a catchment.  When stream flow is steady, rising at a rate lower than the ΔV/Δt due to GWI rise, or decreasing, the GWI contribu-tion to be removed is calculated as GWI [mm/5min] = stream flow [mm/5min] and when stream flow increases at a greater rate than does the GWI, it is calculated GWI [mm/5min]t = stream flow [mm/5min]t-5min  + (ΔV/Δt) [mm/5min], where GWI is the flow contribution from groundwater and interflow, t is time, ΔV is the change in vol-ume, and Δt the change in time between measurements: in this case five minutes. Thus in the absence of runoff-sourced contributions to flow, or when flow increases at a rate equal to or less than the rise in GWI, stream flow is made up completely of GWI. During a rainfall event when stream flow increases faster than can the GWI, the rising GWI is calculated to rise at a steady rate. From area-normalized flow, area-normalized GWI was calculated. Then, as exemplified in Figure 2.24 below, the GWI was then subtracted from the flow, leaving the direct runoff contributions to stream flow as the remainder. On the left, both GWI and direct runoff make up total stream flow (here multiple events illustrated). The GWI is the dark shaded area. When this GWI noise is removed, it leaves the direct runoff signal that is of interest: the highly flashy runoff response to rainfall on directly-connected, impervious surfaces.  Figure 2.24 Removal of GWI from flow to determine direct runoff contribution to stream flow 2.5.7 Identification of discrete rainfall events The raw hourly rainfall data downloaded from FlowWorks, was manipulated to determine  start time & end time,  total length of event (hours)  total depth of event rainfall (mm),  antecedent dry weather period (hours or days),  total rainfall within the previous seven days (mm), and  total rainfall within the previous thirty days (mm) 51  for several hundred discrete rainfall events in the 2011-2013 period used for the investigation. A filter was applied that would limit the list of returned storms to only those within a range of event depth (for example between 2mm and 9mm), minimum storm length (for example 2 hours), and antecedent dry weather period (ADWP) (for example 6 hours since the end of the previous event). 2.5.8 Rainfall event grouping by magnitude As would be expected, during the period January 2011 through December 2013, the number of storms was heavily skewed toward the smaller storms on a higher return frequency, with 62% of the 446 discrete events (with minimum 6-hour ADWP) had a volume between 0 and 2mm, and 10% fell in events of magnitude 2 to 4mm, as illustrated in Figure 2.25. The distribution of rainfall volume, however, was more evenly distributed among events of the full range, as illustrated in Figure 2.26.  Figure 2.25 Distribution of rainfall events by magnitude (with 6-hour ADWP) 52   Figure 2.26 Distribution of three-year rainfall volume by event magnitude (with 6-hour ADWP) A subset of the full range of rainfall was examined for this analysis, excluding events with less than 2 mm total vol-ume, since rainfall in volumes below this on dry pavement should not produce runoff (Lee & Heaney 2003). Alt-hough this removed 277 events from the analysis, it accounted for only 8% of the volume of precipitation that fell in the period.  Distribution of the remaining 169 rainfall events (with greater than 6-hour ADWP and 2mm total rainfall volume), accounting for 98% of the precipitation, is plotted in Figure 2.27.  Figure 2.27 Distribution of three-year rainfall volume in events of 2-60mm These rainfall events with greater than 6-hour antecedent dry weather period and at least 2 mm in depth were further truncated by minimum length: only storms overlapping two hours were included in this analysis. Since the resolu-tion of this data is hourly, a “minimum 2-hour storm” can be interpreted to mean that it overlaps two discrete hours, 53  so is at least ten minutes long (for example, began at 1:55 and ended at 2:05). This final filter only removed a further 13 events, leaving 156 rainfall events for analysis. These were then partitioned into four groups, each constituting approximately one quarter of the rainfall during the three-year period (that passed the ADWP, minimum rainfall, and event length filters). The four groups were 23% ‘extra small’ (XS) storms of 2-8mm, 27% ‘small’ (S) storms of 8-16mm, 25% ‘medium’ (M) storms of 16-30mm, and 24% ‘large’ (L) storms of 30-60mm. Figure 2.28 below illus-trates the quartile partitioning of the filtered storms.  Figure 2.28 Partitioning of rainfall events by approximate contribution quartile In order to select storms of interest for the storm-by-storm analyses performed in this research, the minimum ante-cedent dry weather period of six hours will be used, and only storms of minimum 2 hours will be examined. (This is hourly data, so does not indicate that the storm is at least 2 hours long, just that it spans two discrete hours: so is guaranteed to be at least ten minutes.) This leaves 49 storms of depth between 2 mm and 7 mm, 44 storms of depth between 7 mm and 15 mm, and 19 storms of depth between 15 mm and 24 mm available for comparison. 2.5.9 Discrete event runoff analysis For each of the events identified above, it was necessary to manually manipulate the spreadsheet in the following manners:  Event begin time – was adjusted to include a little more (1-2 hours) of the time before the rainfall began, to be sure that the runoff from both catchments was steady before the event (completely groundwater and interflow).  Event end time – was adjusted to include several hours more (4-8 hours) of the time after the rainfall ended, to ensure inclusion of an appropriate portion of the hydrographs’ falling limb, as these extend temporally beyond the actual period of rainfall.  GWI slope – for both catchments was adjusted – using ‘engineering judgment,’ so that the rising GWI intersects the falling limb at a point above which the sharp drop in stream flow is estimated to be the result of direct runoff from hard surfaces only. The GWI slope was adjusted independently for the catchments, but, depending on the 54  season (antecedent dry or wet conditions), the slope for both catchments’ GWI slope (the variable ΔV/Δt) would be increased or decreased by the same amount. 2.5.10 Ratio of catchments’ peak runoff flows The exercise of normalizing flow by catchment area, adjusting the GWI and removing it from total flow to leave only that from direct runoff was performed on each of the 128 storms determined by the analysis of hourly rainfall data. Then, among each grouping of event rainfall depth, the ratio of event runoff-induced peak flow from the Rout-ley subcatchment (with the source controls) to event runoff-induced peak flow from the Langley meadows subcatch-ment (without source controls) was calculated and recorded. For example, for the 04 December 2012 event graphed in Figure 2.29, the flow from the Langley Meadows sub-catchment peaks at 0.170 mm/5minutes (or 277 m3/hour) and flow from Routley subcatchment peaks at 0.13 mm/5minutes (or 213 m3/hour). Here the Routley neighborhood’s runoff-induced flow peaks at 77% that of the con-ventionally-developed Langley Meadows neighborhood.  Figure 2.29 Comparison of catchments’ runoff-induced peak flows (03-04 December 2012) It can be noted in this December 2012 example that a similar ratio exists among the other peaks: that the catchment with conventional stormwater management practices has considerably higher individual peaks. These are not in-cluded in the present analysis, but is a valuable qualitative observation. Too, not covered in the present study is a quantitative analysis of the duration of flows that may be erosive to the receiving streams, but a qualitative compari-son of each of the event hydrographs may find that storm flow from the neighborhood with source controls has a lower duration of erosive flows than its conventionally-developed counterpart. 2.5.11 Unitize catchments’ cumulative runoff volumes The second quantitative analysis performed on the runoff from the paired catchments is in the vein of a subcatch-ment water balance for each discrete event. For the volume of rainfall received by each catchment during an event, it ‘leaves’ the site by various routes: evapotranspiration, as direct runoff, through interflow or (shallow) groundwater 55  contributions to the stream, or by infiltration to deep groundwater, the mass balance described by Vrain = VET + Vrun-off + VGWI + Vinfiltration. This study focuses on the measurable Vrunoff for each event (the cumulative total volume of runoff), as a fraction or percent of the Vrain that each catchment receives. Performed on an annual basis, this is how KWL estimates the EIA of catchments, but uses a single value of GWI for the year (Jones et al. 2002). Using a sin-gle value saves considerable time, but may be confounded by anomalies in the catchments’ drainage networks and I&I that cause irreconcilable noise in an annual hydrograph. Cumulative event rainfall and subcatchment runoff-sourced flow was calculated as a function of time for each event. To permit comparisons between storms, each hydrographs was unitized: at each time the cumulative volume was calculated as a percentage of the cumulative total rainfall as V(unitized rain),t = 100 x V(cumul, rain),t /V(cumul, rain),total. Further, the time scale was unitized, with each event’s length scaled to 99.5% of the greatest time to total direct-run-off sourced stream flow. From the beginning of each storm to this end time, then, was unitized as a percent of this maximum time. With this, all the storms could be compared on a scale that had been unitized on both axes, for ex-ample as in the 04 December 2012 event illustrated in Figure 2.30.   Figure 2.30 Unitized axes to compare cumulative rainfall and runoff (04 December 2012) 2.5.12 Comparison of cumulative event runoff As in the analysis of peak flows discussed above, the normalizing of cumulative rainfall and flows was performed on all 128 storms, and the cumulative total runoff-sourced runoff from each catchment as a percent of cumulative total rainfall was recorded. For the 04 December 2012 event illustrated in Figure 2.30 above, cumulative total runoff from Langley Meadows outlet accounted for 40% of the total rainfall volume the subcatchment received, and cumulative total runoff from 56  the Routley outlet accounted for only 25% of the total rainfall volume received by its subcatchment: a notable reduc-tion. 2.5.13 Statistical comparisons of magnitude-parsed storms Among each magnitude-parsed group of rainfall events (XS, S, M, L), a comparison was made to determine whether the direct runoff-sourced flow from the Routley subcatchment is significantly less than that from the Langley Mead-ows subcatchment. 2.5.14 Peak flow For each of the 128 storms, the peak flow from the two catchments was calculated and recorded. The ratio of Rout-ley runoff peak flow to Langley Meadows peak flow was calculated, and the percent reduction in peak flow between the two subcatchments was calculated by: % Reduction = 100*[(VLangley Meadows, Peak – VRoutley, Peak)/VLangley Meadows, Peak]. A mean value of percent reduction, and 95% confidence limits, were calculated for the entire group of 128 storms, and for each of the four magnitude-parsed groups of storms. 2.5.15 Cumulative total runoff For each rainfall event, values representing the ratio of cumulative total runoff (in units of mm to match rainfall) to rainfall total depth (units of mm) were calculated for both catchments, and recorded along with other key event metadata. Mean values, and 95% confidence limits, were calculated for the group of 168 storms, and for each of the groups of storms by size. The results and their significance are discussed in section 2.6, below. 2.6 Results & discussion The effort to determine comparability of runoff results from the two catchments, including investigations into catch-ment area, distribution of impervious surfaces, rainfall, soils, slope, and runoff time to outlet, etc., and then the sub-sequent hydrologic analysis of 128 individual rainfall events turned up interesting results, largely that because of the BMPs installed, the Routley catchment (with 70% TIA) is behaving considerably less pervious than its convention-ally-built neighbor, the Langley Meadows catchment (with 63% TIA). Let’s explore the evidences encountered. 2.6.1 Rainfall Rainfall measurement in the study sites over the period of data collection was collected automatically by a tipping bucket rain gauge as described in the methodology section. As a cursory check of whether rainfall measured at this site actually represented the rainfall received at the two study catchments, two more tipping bucket rain gauges were installed in mid-2013 within the catchments of interest. For the months of overlap, the rainfall measured by the three gauges is plotted in Appendix C.1 and Figure 2.31 below. In each figure, the TOL rain gauge (to which runoff over 2011-2013 in the two catchments is compared) is plotted in the first graph, below which is that recorded in the Lang-ley Meadows catchment, and then that recorded in the Routley catchment.  57   Figure 2.31 Rain gauges compared for mid-December 2013 At the resolution observed in the five figures in Appendix C.1, it can be concluded that rainfall arrives to the three sites at roughly the same time, at similar intensities, with a few notable exceptions, including that observable in Fig-ure 2.31 for the period 10-20 December 2013. The cumulative monthly rainfall volumes are listed and compared in Table 2.10. Percent difference between Langley Meadows rain gauge monthly totals and that measured by the TOL (reference) rain gauge are small: ranging between 4% and 15% with only 5% difference overall. Similarly, between the Routley and the TOL rain gauge, differences are even smaller, with only 4% difference between 5-month cumu-lative volumes. 58  Table 2.10 Cumulative monthly rainfall depths compared  Hourly rainfall intensity – an instantaneous phenomenon that is not diminished by a monthly average – has slightly more variability: in the month of October both catchments registered 23% lower peak than the maximum hourly in-tensity in the reference catchment. See Table 2.11 below. This difference between the original, but distant TOL rain gauge, and those hosted within the catchments under study, highlights the importance of hosting rain gauging equip-ment very close to the monitored catchment for runoff studies. Table 2.11 Monthly rainfall peak intensities compared  A comparison of monthly rainfall or peak intensities, however, does not necessarily determine whether rainfall ar-rives in the two catchments at precisely the same time as that that compared to from the TOL rain gauge. As clouds move with wind currents, we expect temporal variability in the arrival of rainfall over a scattered area, so a higher resolution comparison is needed, plotted on the following three graphs with a 30-minute rainfall from 5-minute reso-lution data (5-minute time steps in the plots). Intensities here are millimeters of rainfall per five minutes: the maxi-mum resolution of the rain gauges. Figure 2.32, below, zooms into a storm on 15 September 2013, a 2mm (in 30 minutes) spike in rainfall. It depicts a pair of cells moving through the valley from the southeast: the first set of spikes represents intense rainfall at the TOL rain gauge at 18:55, then at Langley Meadows at 19:30, and at Routley by 20:00. Difference in timing of the spikes is a half-hour; the same evening the second storm spiked 45 minutes apart in the same order among the same rain gauges. TOL Langley Meadows (%diff) Routley (%diff)September 102 121 15% 119 14%October 58 63 9% 60 4%November 130 135 4% 131 1%December 116 108 8% 112 4%January 167 175 5% 172 3%5-months 115 120 5% 119 4%Cumulative Rainfall (mm)MonthTOL-60min LM-60min (%diff) R-60min (%diff)September 7.82 8.89 12% 9.65 19%October 4.37 3.56 23% 3.56 23%November 6.44 6.10 6% 5.84 10%December 3.45 3.56 3% 3.30 5%January 10.12 10.16 0.4% 9.40 8%M nthPeak Intensity (mm/hr)59   Figure 2.32 15 September 2013 storm as measured by the three rain gauges This equally-intense 27 December 2013 event, plotted in Figure 2.33 below, shows a different pattern: a stationary cell with roughly equal intensities spread out over the Township of Langley. Unlike the 15 September event, here the TOL rain gauge here very well represents rainfall received in the Langley Meadows and Routley catchments.  Figure 2.33 27 December 2013 storm as measured by the three rain gauges And finally, this 09 January 2014 event plotted in Figure 2.34, also reaching similar 30-minute intensity to the two events above, indicates a storm with equally intense rainfall throughout the area; it does not appear to be moving. 60  Here, too, it is fair to assume that TOL rain gauge represents the rainfall received in the two catchments under inves-tigation.  Figure 2.34 09 January 2014 storm as measured by three rain gauges For the purposes of peak flow analysis and event cumulative total runoff analysis between the two catchments, it is determined to be acceptable to use precipitation timing & volumes from the TOL rain gauge. However, given the distance of 1.2 km and 2.1 km from the TOL rain gauge to the centroid of the Langley Meadows and Routley catch-ments, respectively, the time lag in measurements introduces an unquantified uncertainty to an analysis comparing the delay in runoff peak after peak rainfall intensities (or time delay between rainfall centroid and runoff centroid). 2.6.2 Runoff Earlier in this chapter, I described the methodology of accessing three full calendar years of rainfall and runoff data via FlowWorks, identifying individual storms (with minimum 6-hour antecedent dry weather period and 2mm of total rainfall), normalizing 5-minute runoff volumes by catchment areas, calculating and removing a reasonable esti-mate of the groundwater and interflow component of the measured flow, plotting the resulting runoff-derived hydro-graphs for each unique event, and calculating & plotting the time- and total rain-normalized cumulative event run-off. Finally, I calculated (a) a ratio of peak flow rate from the Routley catchment (with on-lot source controls) to peak flow rate from the Langley Meadows catchment (with conventional stormwater management practices), four exam-ples of which are plotted in Figure 2.35 through Figure 2.37, and (b) the event runoff from each catchment as a per-centage of event rainfall, four examples of which are plotted in Figure 2.38 through Figure 2.40. The full set of these graphs – plotting normalized flow and GWI, runoff hydrographs, and cumulative rainfall and runoff – is included for reference in Appendix A. 61  In the following four plots, note that both X- and Y-axes are scaled for each event, and vary between events. The point here is to compare peak flow between the two catchments, not between events. Five-minute rainfall intensity is plotted along the top, with 5-minute resolution runoff-sourced flow along the bottom. Visual observation of the magnitude of the peaks among the four events indicates that runoff from the Langley Meadows catchment (solid line) consistently peaks at a higher flow than runoff from the Routley catchment (dashed line) with the effectively disconnected residential lots.  Figure 2.35 Peak runoff compared for 12-13 January 2011 event (55% reduction)  Figure 2.36 Peak runoff compared for 25 December 2012 event (18% reduction) 62   Figure 2.37 Peak runoff compared for 28 September 2013 event (45% reduction) Qualitative observations of rainfall-normalized and time-normalized cumulative runoff among the three events re-veals a similar relationship between the two catchments: significant reductions (between 18 and 64% reductions) in runoff volumes from the Routley catchment (Cumul. R Runoff) compared to that from the Langley Meadows catch-ment (Cumul. LM Runoff). Though there is a significant trend of reduced cumulative runoff from the Routley catch-ment, there is noticeable variability in runoff reduction between storms, from both catchments. Reasons may include antecedent dry weather conditions, saturation of the soils and BMPs, and rainfall intensity and pattern, none of which are indicated on the graphs.  Figure 2.38 Cumulative runoff compared for 12-13 January 2011 event  63   Figure 2.39 Cumulative runoff compared for 25 December 2012 event  Figure 2.40 Cumulative runoff compared for 28 September 2013 event 2.6.3 Peak flow comparisons The runoff hydrograph for only four storms are illustrated above, but this same process was performed for 128 unique events over the three-year period. Of interest is whether the LID-BMP source controls installed in the Rout-ley neighborhood significantly reduce runoff during rainfall events up to the design rainfall event (from the 6-month, 24-hour event, in the Township of Langley equaling approximately 45 mm). Partitioning the 128 rainfall events by range of magnitudes that contribute roughly one quarter of the rainfall, and performing statistical analysis on the 64  sixty-nine 2-8mm events, thirty 8-16mm events, twenty 20-60mm events, and eleven 30-60mm events, I confirmed the significant reduction observed in peak flows from the Routley catchment. Below, Figure 2.41 illustrates the mean peak flow reduction in Routley, calculated as [1- (Routley peak flow)/(Langley Meadows peak flow)], for each range of events and for all events pooled together. Error bars indicate the 95% confidence interval around the estimated means.   Figure 2.41 Mean Routley peak flow reduction and 95% CI for rainfall size groups There is a caveat to this analysis, however: that if during small events peak runoff from the catchments’ storm drain-age networks contribute insignificantly to peak creek flows, and do not increase peak creek flows and velocities above a threshold that (a) erodes stream bed and bank, (b) exceeds fish swimming speeds, or (c) contributes to stream geomorphology, then a decrease in this peak flow during small events has little importance. Such an analysis is beyond the scope of this investigation, and would require measurements of the flow in receiving streams, as well as further research into such thresholds. Of note, though, is a statistically significant reduction in peak flows (by an average of 41%) in the catchment with source controls, even averaging 26% during the largest rainfall events. 2.6.4 Event runoff comparisons The second major comparison made to determine the effectiveness of the source controls, is between values of total runoff between the two catchments. Since rainfall events with greater intensity or total magnitude affects the urban-ized catchments’ runoff volumes, the cumulative runoff volume from each catchment was normalized to (divided by) the total rainfall. Thus for an event during which a catchment receives 20mm of rainfall, and a runoff volume equivalent of 5mm was measured, we could calculate a 25% cumulative runoff / cumulative rainfall ratio. To relate this ratio to BMP performance, a lower value indicates that very little of the total rainfall is being lost from the catchment as runoff, so for stream health, low values indicate that the urban development is contributing less harm to the natural hydrology of the catchment. This analysis was performed for both catchments, for each of the 128 65  unique events during 2011-2013. The data was then compared separately for each group of events by size. The range of values calculated by this ratio was large; the maximum runoff as a percent of rainfall was 77%, the minimum 10%. From the Routley catchment, however, maximum runoff as a percent of rainfall was 58%, with a minimum of 5%. The calculated mean values with 95% confidence intervals are plotted in Figure 2.42 for each group of storms, as well as a pooled mean for all 128 storms. To be addressed in greater detail in the Discussion section of this chapter, two important points must be made about these results. The first is the statistically significant reduction in runoff between the two catchments for storms up to the design storm, observed in the difference in magnitude between the shaded and hollow bars, and the non-overlapping 95% confidence intervals. The second point is the statistically sig-nificant positive correlation between percent runoff/rainfall and event magnitude, for both catchments.  Figure 2.42 Mean & 95% CI of ratio of runoff to rainfall, by catchment and event size Calculating percent reduction of event rainfall for each group of rainfall events plotted above, as [1 – (Runoff/Rain-fall)Routley / (Runoff/Rainfall)Langley Meadows], we find an average 30% reduction in event runoff from the Routley catchment compared to that from the Langley Meadows catchment, dropping to only 20% for the large storms. See Figure 2.43 below. 66   Figure 2.43 Reductions of event cumulative runoff volume in the Routley catchment 2.6.5 Seasonal differences Having identified that there exist significant differences in peak flow and total runoff between the two catchments, and having found a decreasing difference with increasing storm magnitude, we now examine if differences exist be-tween peak flow and total runoff between saturated wet-season conditions, and parched dry-weather conditions.  First, a comparison of mean values of the peak flow reduction [1 – (Routley Peak Flow)/(Langley Meadows Peak Flow)] between wet and dry season was made, to determine if the saturated soils and elevated water table would make a difference in peak flow from the two catchments. A comparison of the mean values for wet season (Novem-ber through May) and dry season (June through October), along with 95% confidence intervals, is illustrated in Fig-ure 2.44, below. Mean peak runoff reduction is 41% for both seasons, indicating no dry weather vs. wet weather im-plications on peak runoff reduction by the source controls in the Routley neighborhood. 67   Figure 2.44 Mean and 95% CI of peak runoff reduction between wet and dry months A comparison of seasonal differences of the runoff/rainfall ratio (plotted in Figure 2.45) does, however, reveal a sig-nificant decrease in percent runoff from both catchments during dry months (when a greater portion of the rain from an event can be either absorbed or infiltrated, and evapotranspiration rates are high), compared to wet months (when saturated soils, elevated water table, reduce the soils’ capacity to absorb or infiltrate rainfall).  Figure 2.45 Mean & 95% CI of ratio of runoff to rainfall, by season Percent reductions, calculated by dividing Routley’s runoff/rainfall ratio by Langley Meadow’s runoff/rainfall ratio for each season, are plotted in Figure 3.28 below along with 95% confidence intervals. The detached roofs, amended soil depths, and infiltration trenches in the Routley neighborhood appear to exhibit no significant seasonal difference in the already-notable improvements over the Langley Meadows catchment. 68   Figure 2.46 Routley cumulative runoff reductions compared by season 2.6.6 Performance evaluation In the Routley neighborhood, an integrated suite of on-lot source controls are designed to infiltrate 90% of the an-nual rainfall, at this site roughly 1300 mm. According to KWL et al. (2012), 90% of the year’s rain falls in events on a return frequency of 6 months or less, so the design event is the 6-month, 24-hour rainfall, approximately 72% of the 2-year, 24-hour event, or 45 mm (72% x 2.6 mm/hour x 24 hours). So 100% of the 45 mm rainfall on an area of the average 202 m2 Routley roof and 15m2 rock trench gives a volume of 9.8 m3, plus the 6.1 m3 falling on the 150 m2 of lawn area, should be absorbed and infiltrated by the amended soils and the infiltration trench. Figure 2.47 is an unscaled sketch of a typical lot, showing amended soils of the lawn and a rock trench each receiving and infiltrating the rainfall that falls on their own area plus one-half that falling on the building’s footprint. Thus, the I/P ratio for the lawn area is 101/135 = 0.75, and the I/P ratio for the rock trench is 101/15 = 6.73. Soil depth should have been installed to a depth of 300 mm (Township of Langley 2004). First, rearranging the equation DS = [R*(I/P + 1) – 24*Ks]/0.2, repeated from section 2.4.2.2, and (originally from KWL et al. 2012) for depth of rainfall capture as R = [0.2Ds + 24Ks]/(I/P + 1) and using the above-calculated I/P ratio of 0.75, and a fairly conservative saturated hydraulic conductivity of KS = 1.0 mm/hour, we calculate rainfall captured in a 300 mm depth of absorbent topsoil to be 41.9 mm, or 93% of the design storm.  69   Figure 2.47 On-lot BMPs to capture and infiltrate rainfall That falling on the other half of the roof and the lawn must be absorbed by the lawn. According to as-built drawings, the rock trenches have a 15 m3 volume, with mean depth of 1.0 m (Township of Langley 2003/2004). Rearranging the equation I/P = [(24*KS + DR*n)/R]-1 (KWL et al. 2012) for rainfall capture, as R = [(24*KS + DR*n)]/(I/P + 1), and assuming a hydraulic conductivity of KS = 1.0 mm/hour, a void ratio of 0.3, nominal rock trench depth of 1.0 m, and the above-calculated I/P ratio of 6.73, we calculate maximum  rainfall capture of 48.1 mm, or 7% greater than the design storm. If these assumed values for the central variables are correct, together the amended soils and rock trench should capture 98% of the 45 mm design rainfall mandated by both the DFO and the BC provincial govern-ment for capture of 90% of annual rainfall. As such, during storms up to 45 mm, runoff should only be generated by the hard surfaces directly connected to the drainage infrastructure, and everything falling on the roofs and green ar-eas should be captured. A sensitivity analysis on the impacts of changing the values of the assumed variables on the calculations just per-formed is necessary before making conclusions, to give context to the results. To which variable is rain capture by the two BMPs most sensitive? Would a small change in the assumed values drastically change conclusions about the design? Table 2.12 summarizes the assumed values and changes of both 25% and 50% above and below for the var-iables trench depth, void ratio (n), saturated hydraulic conductivity (Ks), and depth of amended soil. 70  Table 2.12 Assumed and range of values for sensitivity analysis  Figure 2.48 illustrates the changes to the percent of the design rainfall of 45 mm that would be captured by changing each variable individually by the percent indicated, while holding all other variables at their originally-assumed val-ues. Percent capture of the design rainfall is most sensitive to the void ratio of the rock trenches, and least sensitive to both the saturated hydraulic conductivity of the underlying soil and the trench depth. Thus, even if the conductiv-ity of the soils is one-half this assumed value, holding all other variables constant we can still assume 90% capture of the design storm. Alternatively, as the void space between the rocks fills in over time with sediments, it will have significant impacts on the rainfall capture. At a void space of 0.15, the lot’s capture of rainfall falls to 72%. Not tab-ulated or illustrated, a combination of “worst-case” conditions of shallow trench depth, reduced void ratio, low satu-rated hydraulic conductivity of underlying soil, and shallow lawn soils would of course reduce the capacity of the on-lot source controls to capture rainfall. If all of these variables fall to 50% below the assumed value, the average lot would only be able to capture 45% of the 45 mm design depth. The corollary to this, of course, is that if the ac-tual values are larger than assumed, the on-lot source controls should be capable of capturing even more rain. If, say, the native soil’s hydraulic conductivity is 3 mm/hr, holding all other variables constant, the site should be able to collect all rainfall in all events up to 59 mm; if 6 mm/hr, it should be able to collect rainfall in events up to 80 mm. Have we seen an indication of this in the peak flow and total runoff analyses performed? From Table 2.3 we recall that the total impervious area in the 13.59 ha Langley Meadows catchment is 8.62 ha, or 63%, and that the total impervious area in the 13.57 ha Routley neighborhood is 9.47 ha, or 70%. By disconnecting the 4.25 ha roof area from the Routley catchment (31% of the catchment surface area, or 45% of the total impervi-ous area), and assuming that the rest of the area is yet connected to the drainage network, we calculate only 5.22 ha (or 38%) of the Routley catchment is connected to the drainage network, or at least not being addressed by the lot source controls. According to design, rain volumes up to 45 mm falling on these roofs should be absorbed by lot landscaping and rock trenches, and should not be contributing to peak flows and cumulative runoff volumes meas-ured during rainfall events in the storm sewers at the catchment’s outlet.  So by removing the roofs from the total impervious area of the Routley catchment, and comparing the remaining percent 38.5% imperviousness to that of the control catchment’s 63.4% imperviousness – Routley now having (63.4%/38.5%)/63% = 39.4% less directly connected impervious area than Langley Meadows – we should expect to see similar reductions in stream responses to rainfall events.  -50% -25% assumed value +25% +50%Trench Depth (m) 0.5 0.75 1.0 1.25 1.5Void Ratio, n 0.15 0.23 0.3 0.38 0.45Ks (mm/hr) 0.5 0.75 1.0 1.25 1.5Soil Depth (mm) 150 225 300 375 45071   Figure 2.48 Sensitivity analysis on variables affecting rainfall capture Is the percent reduction of runoff-derived peak flow in the storm sewers significantly different from the percent re-duction in Routley impervious area (compared to Langley Meadows impervious area) by disconnecting the roofs and draining to the on-lot source controls? Recalling from Figure 2.41 that the average reduction in peak flow was 41% (± 3% at the 95% confidence level), we find that this is not statistically different from the 39.4% reduction as a result of the disconnected roofs.  Similarly, is the percent reduction of runoff-derived cumulative total event volumes measured in the storm sewers, calculated as [1 – (Runoff/Rainfall)Routley / (Runoff/Rainfall)Langley Meadows], significantly different from the percent reduction in Routley impervious area by disconnecting the roofs? Recalling Figure 2.43, in which it is shown that the percent reduction of cumulative event runoff volumes is 30% (± 2% at the 95% confidence level), we can con-clude that although significantly large reductions were made, in fact achieving 75% of design over the 128 storms, several years after build-out, these reductions are yet significantly different from the 39% EIA reduction as a result of disconnecting the roof areas and providing amended soils and rock trenches to absorb and infiltrate on-site rain-fall in events up to 45 mm. Reductions in EIA (Routley compared to Langley Meadows), peak flow, and event run-off volume are compared in Figure 2.49. Although peak flow reduction does not significantly differ from EIA reduc-tion, mean runoff volume reduction does differ. 72   Figure 2.49 Peak flow and event runoff volume reductions compared to connected impervious area reduction 2.7 Conclusions The source controls installed in the Routley catchment were designed to capture and infiltrate precipitation from events smaller than the 45 mm design storm. According to the analysis on 128 rainfall events between 2 mm and 60 mm during the years 2011, 2012, and 2013, the hydrologic effect strongly reflects the effort made to disconnect the roofs. The Routley neighborhood is achieving 100% benefit in reducing peak flows, and 75% benefit in reducing event cumulative runoff, as a result of disconnecting roof areas from the storm drainage network, and directing roof-sourced volumes to rock trenches and amended soils in lawn areas. The BMPs are performing as designed to delay rainfall from the roof areas, and are performing nearly to capacity to infiltrate the precipitation within the lot boundaries. This slight reduction several years after installation be due to several factors, some of which were highlighted in the key variables sensitivity analysis above, including a combina-tion of the following:  Sediments entering the rock trenches may be reducing the void space and their capacity to hold the designed volume of rain.  Given the variability of the soils in the area, some rock trenches may be built over native soils with extremely low hydraulic conductivity, or perched water tables.  Soil depths may be shallower than design, or could have been compacted over time. Additionally, it is here assumed that all hard surfaces identified in both catchments are directly connected to the storm drains, but realistically, some driveways, roofs, sheds, and other on-lot impervious areas drain to adjacent lawns, thereby reducing the effective impervious area of both catchments, perhaps either reducing or increasing the EIA ratio between them. Finally, despite identifying that the Routley catchment is not performing fully to its hydraulic design of capture 90% of the annual rainfall (100% of the 6-month, 24-hour rainfall, equivalent to 45 mm falling in 24 hours), I conclude 73  that, even several years after construction, the catchment is significantly benefiting from the effective removal of its homes’ roof area, by use of these on-lot source controls.  74  Chapter 3: Selected BMPs’ impact on water quality 3.1 Introduction The second part of this urban runoff research turns to an examination not only of the physical and chemical quality of runoff from a residential neighborhood, but also how the compromised urban runoff quality is treated through individual low impact development best management practices, and those strung together in a treatment train for both quantity and quality benefits for the watershed and aquatic life. Sophisticated weirs and level gauging equip-ment, and techniques to determine discharge from stage have been developed and are in wide use to measure stream flow and urban runoff rates. Planning, sampling, and analysis of a suite of water quality parameters to accompany flow monitoring add significant complexity and budget to a monitoring program. Water quality monitoring is often overlooked. Also, there is considerable knowledge of the hydrologic performance of BMPs, and a growing body of knowledge around the water quality benefits they offer. But as is discussed in the literature review, many of the wa-ter quality studies are on recently-built facilities, pilot-scale test plots, or bench-scale columns from which water quality is sampled. That no academic or engineering study has examined how effective are the individual BMPs in-stalled in the Silver Valley neighborhood – or the full suite as a whole – gave further impetus to pursue this water quality aspect of the research. It offers a four-month springtime glimpse into the water quality benefits of a mature roadside bioswale. In the following sections a literature review summarizes how urbanization of a natural watershed brings with it wa-ter quality impacts – often from distributed, non-point sources through atmospheric deposition and washed-off pol-lutants during rainfall. Sources of the pollutants is discussed, and more importantly is reviewed the mean and range of values of contaminants including suspended and dissolved sediments, nutrients, hardness, alkalinity, chloride, pH, temperature, and a group of heavy metals of interest in urban runoff, as reported in the literature. These typical con-centrations in urban runoff are compared to the Water Quality Guidelines for the Protection of Aquatic Life pub-lished by the Canadian Council of Environmental Ministers (CCME 2014), with some reference to British Colum-bia’s Water Quality Guidelines for Aquatic Life (BC Environment 2014, Nagpal et al. 2006) and the United States Environmental Protection Agency’s Aquatic Life Criteria (USEPA 1987/2014). The literature review then summa-rizes the published performance of infiltration swales at reducing the concentrations of urban pollutants. Of course, bioswales must be considered in terms of a mass balance, with the understanding that there may be movement be-tween particle-bound and soluble fractions due to soil processes within the infiltration swale itself. I then describe the treatment effectiveness approach followed in the methodology section, in which I install, main-tain, and collect samples using automatic samplers and periphery equipment including a custom-built water level sensor and sampler controller, and a telemetry-equipped datalogger, at two sites along Foreman Drive in the Silver Valley neighborhood of Maple Ridge. With the automated equipment, I sampled road runoff and treated effluent from the roadside infiltration swale during twenty-two unique storms, and filtered, preserved, and analyzed the sam-ples in the Environmental Laboratory at the University of British Columbia. This chapter on water quality research 75  effort concludes with a discussion of the water quality results, and addresses the performance of this eight-year-old BMP. 3.2 Objectives The objective of this water quality treatment effectiveness study is to answer the following question: Is the infiltra-tion swale along Foreman Drive, designed and installed to the best practices of the day for both infiltration and treat-ment of road runoff, removing contaminants of concern according to the levels predicted in the literature?  3.3 Hypothesis The Silver Valley neighborhood was developed in the mid-2000s under an official community plan, and the devel-oper did not receive permission from the DFO to build until a comprehensive stormwater management plan featur-ing designs for the most progressive suite of precipitation source controls in the Fraser Valley was proposed, includ-ing on-lot source controls, a roadside bioswale, and a three-tier treatment train designed to remove contaminants from urban runoff. It is hypothesized that runoff water quality is significantly improved by treatment through the infiltration swale, evidenced in lower concentrations of the urban pollutants measured. 3.4 Literature review To appreciate the hazards that runoff poses to aquatic life after precipitation has passed over the hard surfaces asso-ciated with urban development and picked up particles, dissolved and particle-bound nutrients, metals, and organic contaminants, and heat, the following section reviews the published academic and professional literature on urban water quality and its treatment through bioretention swale-type practices. 3.4.1 Regulatory framework The Canadian Council of Ministers of the Environment (the CCME) publishes a set of Water Quality Guidelines for the Protection of Aquatic Life that, though not comprehensive in their scope, include acute and long-term maximum concentration guidelines that have been developed for the protection of both marine and freshwater biota (sepa-rately). Urban runoff quality should be assessed against these guidelines as in many urban streams, much of summer flow during a rainfall event will be this contaminated runoff. In this research effort, both direct road runoff and the treated effluent are compared to the CCME guidelines. The Environmental Protection Division of British Colum-bia’s Ministry of Environment also publishes both approved and working water quality guidelines (BC Environment 2014, Nagpal et al. 2006). Acute (short term) and chronic (long term) maximum guideline concentrations from both of these jurisdictions, as well as from the United States’ Environmental Protection Agency (USEPA) are tabulated in Table 3.1. Toxicity studies on the various life stages of aquatic biota identify the water quality criteria that lead to a given effect on the organism. From the range of concentrations of a parameter and chemical conditions in the water that inhibit or enhance its effect on the many life stages of the range of trophic levels for the aquatic environment, guidelines set maximum concentrations well below that found in the studies. Guideline concentration of a given pa-rameter is often a function of water chemistry, most notably hardness and pH, which affect speciation, bioavailabil-ity, and toxicity.76  Table 3.1 BC Environment, CCME, and USEPA aquatic life water quality guidelines  Acute Chronic Acute Chronic Maximum ContinuousTSS mg/LClear Flow:Increase < 25 above backgroundHigh Flow: Increase < 25 above background | background = 25-250 mg/LIncrease <10% | Background > 250 mg/LIncrease < 5 above backgroundTN mg/L0.31 (rivers & streams)0.10 (lakes & reservoirs)NH4-N mg/L0.752 - 28.7 Range,depending on pH & Temp0.102-2.08 Range,depending on pH & temppH, Temp,life-stage dependentNOXmg/LNitrate: 32.8Nitrite: 0.06Nitrate: 3Nitrite: 0.02Nitrate: 550Nitrite: no guidelineNitrate: 13Nitrite 0.06TP µg/L47.0 (rivers & streams)8.75 (lakes& reservoirs)Cl-mg/L600 150 640 120 860 230Al mg/L0.1 | pH > 6.5. Otherwise, | pH <6.5,e (^1.209-2.426(pH) + 0.286(pH)2)Soni 0.75 0.087Cd µg/L10*exp (^0.86*log(hardness)-3.2) - 0.11 | hardness < 5.3 mg/L10 [^1.016*(log(hardness))-1.71)]        | hardness = 5.3-360 mg/L7.7   | hardness > 360 mg/L0.04 | hardness < 17 mg/L10 [^0.83*(log(hardness))-2.46)]        | hardness = 17 - 280 mg/L0.37 | hardness > 280 mg/L2 0.25Cu µg/L2      | hardness <50 mg/L0.04*(hardness)  | hardness >50 mg/L0.094(hardness) + 2 2 | hardness < 82 mg/L0.2*e [^0.8545(ln(hardness))-1.465]   | hardness 82-180 mg/L4 | hardness > 180 mg/LCriteria calculated usingBiotic Ligand ModelFe µg/L 1000 (total); 350 (dissolved) 300 1000Mn mg/L0.8  | hardness = 25 mg/L;1.1  | where hardness = 50 mg/L;1.6  | hardness = 100 mg/L;2.2  | hardness = 150 mg/L0.8 | hardness = 25 mg/L;0.8 | hardness = 50 mg/L1.0 | hardness = 100 mg/L;1.9 | hardness = 150 mg/LNi µg/L25   | hardness 0-60;65   | hardness 60-120;110 | hardness 120-180;120 | hardness >18025   | hardness < 60 mg/Le [^0.76*(ln(hardness))+1.06]       | hardness 60-180 mg/L150 | hardness ? 180 mg/L470 52Pb µg/LE 3 (total) | hardness < 8 mg/Le (^1.237*ln(hardness)-1.460)    | hardness > 8 mg/L1   | hardness < 60 mg/Le [^1.273(ln(hardness))-4.705]     | hardness 60-180 mg/L7   | hardness > 180 mg/L65 2.5Zn µg/L 33 + 0.75*(hardness - 90) 7.5 + 0.75*(hardness - 90) 30 e (^0.83*(ln(hardness))+1.95) 47pH mg/L 6.5 < pH > 9.0 6.5-9Alkalinity mg/L >20Trigger Ranges for ultra-oligotrophic (<4); oligotrophic (4-10); mesotrophic (10-20); meso-eutrophic (20-35); eutrophic (35-100); hyper-eutrophic (>100)pH & Temperature dependent. Refers to a table.Source: BC Environment 2014, Nagpal et al. 2006 Source: CCME 2014 Source: USEPA 1987, USEPA 2014UnitsParameterIncrease <20 | background <100 mg/LIncrease <20% | background >100 mg/LUSEPA - Aquatic Life CriteriaCCME - Guidelines for the Protection of Aquatic LifeBritish Columbia - Water Quality Guidelines for Aquatic LifeMust not reduce depth of compensation point for photosynthetic activity >10% above background77  3.4.2 Water quality impacts of urbanization Along with the land use changes associated with urbanization, including the disruption of soils, removal of vegeta-tion, replacement of pervious areas with impervious surfaces (roofs, roads, parking lots), and the increased human activity (yearly cycles of bare agricultural fields, irrigation and fertilization of planted areas, industrial withdrawals and discharges, damming of rivers, and increased vehicle traffic), streams and other receiving bodies take considera-ble abuse – in terms of flow extremes and impaired water quality, both of which degrade stream health (Klein 1979). In the following sections, I discuss several stormwater quality parameters that impact stream health, without pro-found detail as to chemical processes. I address the point- and non-point sources of increased concentrations of the pollutants in urban stormwater, compare the mean concentration (and range) from several studies, address the effects of elevated concentrations on stream health, and examine the typical stormwater concentrations to regulations/guide-lines related to the discharge concentrations or loadings of these substances to waterways, drawing on several stud-ies of urban runoff quality, including the US Environmental Protection Agency’s 1978-1983 National Urban Runoff Program (NURP), which collected runoff water quality at 28 urban sites across the USA (USEPA 1983). 3.4.2.1 Solids As was addressed in the review of literature on hydrologic impacts of urbanization in Chapter 2, major sources of sediments in urban streams are construction associated with the development phase, and later from erosion of the stream bed and banks (Arnold & Gibbons 1996, Klein 1979, Simon 1989). Continued urban inputs of particulates and sediments, including from wear of moving vehicle parts (engine, brakes, tires), vehicle emissions, road wear (both concrete and asphalt), and winter road sanding (Burton & Pitt 2002). Atmospheric deposition of total sus-pended solids is on the scale of 300 kg/ha/year (Marsalek & Viklander 2011), or 140 to 220 kg/ha/year for suburban or urban neighborhoods, respectively (Debo & Reese 2002), much of which can enter streams when transported by runoff. As mentioned in the previous chapter, increased sediment loading in streams can have significant impact on aquatic life. Several studies, including the USEPA’s NURP, have published data on both total suspended solids (TSS) and total dissolved solids (TDS). Table 3.2 lists the minimum, mean, and maximum values of TSS concentra-tions in urban residential stormwater, as well as the number of samples, and the season the study or studies took place. A * indicates number of studies rather than number of samples.  Table 3.2 Residential urban runoff TSS concentrations, from the literature  Par meter Units min mean max N Season Source or StudyTSS mg/L 141 796 All Brombach et al. 200543 161 476 10 Winter Taebi & Droste 2003174 2000 All (NURP) USEPA 1983101 208* All (NURP) Strassler et al. 199978.4 3047 All Smullen et al. 1999809 18 Fall, Winter, Spring Tiefenthaler et al. 199925 228 2166 1102 All (NURP) Whalen & Cullum 198835 44 59 41 All Macdonald 2003263 5 All Frederick 200378  With some significant outliers, typical urban runoff TSS concentrations are in the range of 50-200 mg/L, according to the published literature. These serve as a point of reference only, and without knowledge of the receiving bodies’ background TSS, it is impossible to compare these studies’ figures to the guidelines. Where urban runoff volumes contribute significant portions of the flow of urban streams, TSS increases above background concentrations may significantly exceed guidelines and compromise stream health. Total dissolved solids (TDS) – all molecular and ionized substances, both organic and inorganic – are differentiated from the suspended solids by passage through a 2 µm filter. TDS itself is not considered a pollutant, or given a max-imum concentration by stream protection guidelines. It is comprised of organics, nutrients, and metals that can pose a health threat to biota. KWL (2010) reported higher TDS in wet season road runoff than dry season road runoff, but the opposite seasonal differences for runoff from roofs. So few sampling points may limit the significance of these ranges, however. A mean value of 46 mg/L was reported by Frederick (2003). See the ranges of values in Table 3.3. Table 3.3 Residential urban runoff TDS concentrations, from the literature  3.4.2.2 Nutrients Nutrients are needed for growth and production at all trophic levels, but excess concentrations of these limiting nu-trients can cause significant changes to a stream. Both point- and non-point source contributions of nutrients – espe-cially nitrogen (N) and phosphorus (P) – from urban, industrial, and agricultural areas has been one of the most sig-nificant anthropogenic sources of stream ill-health, eutrophic conditions causing algal blooms, depleted oxygen lev-els, and fish kills (Galloway et al. 2003). 3.4.2.2.1 Nitrogen In the urban environment, sources of nitrogen include lawn and landscaping fertilizers, vehicle emissions, and at-mospheric deposition (Marsalek & Viklander 2011, Burton & Pitt 2002). Estimated atmospheric loading rates of total nitrogen and total phosphorus in urban catchments are 8 and 7.6 kg/ha/year, respectively (Marsalek & Vi-klander 2011). Additionally, Debo & Reese (2002) published a summary of atmospheric deposition rates in Wash-ington, DC, reproduced in Table 3.4 in metric units below. Par m er Units min mean max N Season Source or StudyTDS mg/L 46 5 All Frederick 2003110 120 130 2 Winter KWL 2010, wet, storm46 48.57 51.71 3 Summer K L 2010, dry, stor14 19 24 2 Winter KWL 2010, wet, roofs16 71.81 115.43 3 Summer K L 2010, dry, roofs79  Table 3.4 Average annual atmospheric deposition loadings of nutrients (kg/ha/year)  Nitrogen enters streams in both organic, and the more biologically-available inorganic forms as nitrate (NO3-N), nitrite (NO2-N), ammonia (NH3-N), (Galloway et al. 2003, Debo & Reese 2002). Measures of total nitrogen (TN) include all forms of both organic and inorganic nitrogen; total Kjeldahl nitrogen (TKN) is a measure of just the am-monia and organic forms; NOX identifies the sum of both nitrate and nitrite. Of the inorganic nitrogen forms, nitrate is the most soluble in receiving waters, and elevated concentrations are associated with anthropogenic sources (Tay-lor et al. 2005). Table 3.5 below summarizes the TN, TKN, NH3, and NOX concentrations from several studies. Ur-ban runoff TN concentrations are in the range of 1.5-3 mg/L. A significant portion (1.47 – 2.5 mg/L) is TKN (TN – NOX), of which a fraction (0.2-0.8 mg/L) is ammonia. A Melbourne, Australia study identified that the international literature typically identifies the TKN, TN, the NOX and ammonia, but does not partition the organic nitrogen into either dissolved or particulate fractions. In urban stormwater, they identified nitrogen in stormwater as half inorganic and half organic, and 75-85% in dissolved form, and report NOX concentrations as roughly 40% of the total (Taylor et al. 2005). Rural Suburban UrbanTN 17.8 11.4 15.2TKN 9.4 6.4 9.1NH3-N 4.9 1.0 0.9NO3-N 8.4 5.0 6.1TP 0.63 0.45 0.75O-PO4 0.25 0.23 0.31source: Debo & Reese 200280  Table 3.5 Residential urban runoff nitrogen concentrations, from the literature  In a small study performed in the Township of Langley, BC, KWL et al. (2010) report NOX values in runoff samples collected during rainfall events both during wet weather and dry weather, and in road runoff (“storm”) and roof/lawn/perimeter drains (“roofs”), with insignificant difference in NOX concentrations between the conditions or seasons, indicating perhaps similar (atmospheric) source of NOX to both roofs and roads. The CCME Guidelines for the maximum concentration of total ammonia (the sum of both NH3 and NH4), as illus-trated in Figure 3.1, are dependent on both temperature and pH. Low pH drives the chemical equation NH3 + H2O ↔ NH4 + OH to the right (more harmless ionized NH4), and a high pH drives the equation to the left (more poten-tially harmful unionized NH3) (Sawyer 2008). Also, since the activity of ammonia is directly proportional to temper-ature, sensitive species are more tolerant to ammonia at low temperatures; ammonia concentrations reflect this toler-ance, with a higher guideline concentration at lower temperatures. Guideline un-ionized ammonia maximum con-centration is limited to 19 mg/L (CCME 2014). Parameter Units min mean max N Season Source or StudyTN mg/L 2.4 796 All Brombach et al. 20051.22 6.65 22.38 10 Winter Taebi & Droste 20030.7 0.8 1 41 All Macdonald 20032.2 2.9 4.6 5 All Frederick 20030.34 1.5 20 ? All BC Environment 1992TKN mg/L 1.9 16.1 All USEPA 19830.48 2.3 11 898 All Whalen & Cullum 19881.73 2693 All Smullen et al. 19991.6 2.5 4.1 5 All Frederick 2003NH4-N mg/L 0.8 796 All Brombach et al. 20050.14 0.18 0.25 41 All Macdonald 2003NOXmg/L 0.8 796 All Brombach et al. 20051.6 13.6 81.9 28 All Tuccillo 20060.837 1234 All (NURP) USEPA 19830.658 2016 All Smullen et al. 19990.736 208* All (NURP) Strassler et al. 19990.31 1.8 9.5 583 All (NURP) Whalen & Cullum 19880.51 0.64 0.85 41 All Macdonald 20030.692 0.793 0.894 2 Winter KWL 2010, wet, storm0.271 0.428 0.624 3 Summer KWL 2010, dry, storm0.206 0.237 0.268 2 Winter KWL 2010, wet, roofs0.096 0.519 0.574 3 Summer KWL 2010, dry, roofs81   Figure 3.1 CCME total ammonia maximum concentrations as a function of pH and temperature 3.4.2.2.2 Phosphorus Similarly, phosphorus enters streams in anthropogenically modified landscapes in both organic and inorganic forms; total phosphorus (TP) is a measure of both. Biologically available orthophosphate (O-PO4) as the fraction of particu-lar interest for its impacts on stream health (Strassler et al. 1999), in one case reported at a concentration of 100 µg/L in New York urban runoff (CWP 2010). As was seen in Table 3.4, urban atmospheric loading of TP is one-twentieth that of TN. The theoretical ratio of N to P uptake by algae is roughly 7.5:1 (Debo & Reese 2002). TP con-centrations in urban residential runoff, summarized in Table 3.6 from a number of studies, are significantly lower than that of TN: roughly 0.2-0.6 mg/L. A partial result of runoff, however, is that streams in urban catchments re-ceive 5-20 times the pre-development level of this limiting nutrient, causing accelerated growth of lower trophic lev-els (Walker 1987). Table 3.6 Residential urban runoff total phosphorus concentrations, from the literature  The published values of urban runoff TP concentrations in Table 3.6 are significantly higher than background con-centrations of 10-20 µg/L for mesotrophic conditions (listed in Table 3.1). Where such stormwater runoff volumes make up a significant fraction of urban stream flow, severe nutrient gluts may cause eutrophic or hyper-eutrophic conditions, with harmful consequences. Par met r Units min mean max N Season Source or StudyTP µg/L 420 796 All Brombach et al. 200564 274 790 10 Winter Taebi & Droste 2003337 1902 All (NURP) USEPA 1983383 208* All (NURP) Strassler et al. 1999315 3094 All Smullen et al. 1999220 620 4410 1029 All (NURP) Whalen & Cullum 1988120 140 180 41 All Macdonald 2003260 440 540 5 All Frederick 200382  3.4.2.3 Organic carbon Anthropogenic increases of organic carbon in the aquatic environment are associated with synthetic sources includ-ing hydrocarbons, pesticides, detergents, and pharmaceuticals, in addition to seasonal inputs of leaves and grasses. From a study to determine urban non-point sources of organic carbon loading to the Sacramento River, Sickman et al. (2007) found median TOC concentrations of 8.9 mg/L in urban runoff, with a yield of roughly 150 kg/ha/year, approximately 17% of the loading to the river. 3.4.2.4 Chloride Introduction of chloride to watersheds in the form of sodium chloride (NaCl) as a road de-icing compound has con-tributed to elevated concentrations of Cl in streams, and in highly urban catchments, seasonal variations in concen-trations can reach two orders of magnitude above pre-development levels, with implications to stream health (Kaushal et al. 2005). Chloride concentrations, especially associated with spring freshet and snowmelt, can be high, and Kaushal et al. (2005) report peak concentrations in several Baltimore creeks as high as 4,600 mg/L in the win-ter, and peaks up to 330, 470, and 600 mg/L in spring, summer, and fall, respectively, typical mean concentrations of urban runoff are not so extreme. Table 3.7 lists published values of minimum, mean, and maximum chloride con-centrations found in residential runoff from several studies. Table 3.7 Residential urban runoff chloride concentrations, from the literature  Average values are all below the CCME (2014) Guidelines for the Protection of Aquatic Life value of 640 mg/L, with only the maximum concentration found by Tuccillo (2006) exceeding the guideline value. Kaushal et al. (2005) remind us that concentrations above 250 mg/L are harmful to freshwater life, but according to the guideline, this maximum concentration is “not intended to protect all components of aquatic ecosystem structure and function but rather to protect most species against lethality during severe but transient events (e.g. inappropriate application or disposal of the substance of concern)” (CCME 2014). A recent study on chloride concentrations in urban streams in New York identified groundwater and soils as likely reservoirs for chloride concentrations, and suggested that sig-nificant rainfall volumes are needed to wash out the accumulated loads (Kincaid & Findlay 2009). 3.4.2.5 Hardness Cadmium, copper, nickel, and lead toxicities are all dependent on the hardness of the water: even low concentrations in soft water can be toxic to living organisms; tolerance is higher in hard water. The range of values for water hard-ness in Metro Vancouver and the Township of Langley, compiled in Table 3.8, are in the range of ‘soft’ and ‘moder-ately hard’ waters (Macdonald 2003, KWL 2010).  The full range of soft to very hard water (15-327 mg/L as CaCO3) is reported in stormwater by Hall & Anderson (1988), 19-142 mg/L from just residential sources. Values Pa ameter Units min mean max N Season Source or StudyCl- mg/L 1.7 75.3 884 26 All Tuccillo 20066.9 8.4 10.7 41 All Macdonald 20035.3 5.45 5.6 2 Winter KWL 2010, wet, storm1.8 2.62 3.57 3 Summer K L 2010, dry, storm0.5 0.55 0.6 2 Winter KWL 2010, wet, roofs1.4 2.67 4.71 3 Summer K L 2010, dry, roofs83  from the recent study in the Township of Langley indicate that during wet weather, the road-derived runoff is signif-icantly harder than that from roofs and perimeter drains; little apparent difference is observable in runoff from either during the dry season (KWL 2010). Again, few samples limit confidence in the representativeness of these numbers. Table 3.8 Residential urban runoff hardness, from the literature  3.4.2.6 Alkalinity Alkalinity, water’s capacity to neutralize acid, will vary considerably in urban runoff: a range of 20-34 mg/L as CaCO3 is reported for urban runoff from residential neighborhoods, and 8-153 mg/L among all urban land uses two decades ago in the Lower Fraser Valley (Hall & Anderson 1988). A study of urban runoff quality in Nigeria found a mean alkalinity of 33.5 mg/L during summer storms, and lower 18.7 mg/L in runoff from wet-season storms (Izonfuo & Bariweni 2001), intuitive since the consistent flow of rain with low (zero) alkalinity would cause a re-duction in stormwater alkalinity. Aquatic environments with alkalinity below 10 mg/L are “highly sensitive to acid inputs,” with water having alkalinity above 20 mg/L having a “low sensitivity to acid inputs,” (Nagpal et al. 2006). The USEPA specifies a minimum alkalinity of 20 mg/L. 3.4.2.7 Metals Increased human activity in and disturbance of watersheds is increasing metals concentrations above background levels. Although not alarming on its own, the extensive NURP study detected cadmium, copper, lead, nickel, and zinc in 48, 91, 94, 43, and 93%, respectively, of the urban runoff samples collected, with copper, cadmium, lead, and zinc of primary concern (USEPA 1983, Debo & Reese 2002, Erickson et al. 2013). Urban sources of metals are mostly vehicle-related, with wear from breaks, bushings, bearings, and tires, as well as wear associated with moving parts in the engine and drivetrain, as prime contributors of cadmium, copper, iron, manganese, lead, and zinc (Mar-salek & Viklander 2011, Burton & Pitt 2002, Whalen & Cullum 1998), though concentrations of lead have de-creased since lead was removed as a gasoline additive in the 1970s (Erickson et al. 2013). Modified from Whalen & Cullum (1998), Table 3.9 summarizes the sources of metals in stormwater. Asphalt contributes further to concentra-tions of nickel; insecticides and fungicides to copper and cadmium concentrations; oil and other vehicle fluids to zinc, lead, and nickel; and exposed and buried rusting infrastructure like pipes, cables, rebar, bridges, and guardrails contribute to elevated iron concentrations (Marsalek & Viklander 2011, Burton & Pitt 2002).  Par meter Units min mean max N Season Source or StudyHardness mg/L 21 24 26 41 All Macdonald 2003(as CaCO3) 69.5 80.3 91 2 Winter KWL 2010, wet, storm24.5 28.9 31.1 3 Summer K L 2010, dry, stor8 8.2 8.3 2 Winter KWL 2010, wet, roofs6 28.8 44.9 3 Summer K L 2010, dry, roofs84  Table 3.9 Sources of heavy metals found in stormwater runoff (Whalen & Cullum 1998)  In a mid-1970s study on the contributing sources of lead in a New Jersey river, Wilber & Hunter (1977) identified the significant contribution that atmospheric deposition and river scouring can have to metals concentrations in ur-ban watersheds. This is supported by experiments conducted in a small catchment in Los Angeles, from which Sabin et al. (2005) identified that the majority (57-100%) of metals accounted for in stormwater were attributed to atmos-pheric deposition. Metric atmospheric deposition rates for five metals compiled by Debo & Reese (2002) are listed in Table 3.10, below. But although atmospheric deposition of metals is correlated to vehicle traffic intensities (Bur-ton & Pitt 2002), both cadmium and zinc have been found to have higher deposition rates in the suburban catch-ments than in the urban catchments studied in Washington. Reviews of pollutant sources remind us that metals enter the waterway in stormwater are roughly equally divided between dissolved and particle-bound forms, (Erickson et al. 2013), with particulate-bound concentrations directly proportional to the amount of organic matter comprising the suspended sediments, and inversely proportional to particle size (Paul & Meyer 2001).  Table 3.10 Average annual atmospheric deposition loadings of metals (kg/ha/year)  3.4.2.7.1 Copper Background concentrations of copper in natural waters range from 0.6 to 400 µg/L, median value of 10 µg/L, with dissolved copper concentrations rarely exceeding 5 µg/L in surface waters (Alberta Environmental Protection 1996). Minimum, mean, and maximum copper concentrations found in the literature, as well as the number of samples ana-lyzed and season of monitoring, are compiled in Table 3.11, below. Comparing these values to the CCME guidelines Source Cd Cu Fe Ni Pb ZnGasoline X X X XExhaust Emissions X XMotor Oil & Grease X X X X XAntifreeze X XUndercoating X XBrake Linings X X X XRubber X X X XAsphalt X X XConcrete X X XDiesel Oil XEngine Wear X XRural Suburban UrbanCadmium -- 0.080 0.003Copper -- 0.19 0.54Lead 0.05 0.39 0.47Iron -- 1.40 5.00Zinc 0.60 1.20 0.58source: Debo & Reese 200285  in Table 3.1, the mean values of all these studies exceed the highest CCME guideline value of 4 µg/L. Whether or not it is significant, the older publications (BC Environment 1992, Whalen & Cullum 1998, Strassler et al. 1999, and Tiefenthaller et al. 1999) reported higher mean values than the later studies.  Table 3.11 Residential urban runoff copper concentrations, from the literature  3.4.2.7.2 Cadmium Cadmium concentrations in residential runoff from several studies are summarized in Table 3.12. Some values are high, likely exceeding the CCME short-term guideline limits of 0.11 – 7.7 µg/L, especially in soft water. Mean val-ues from recent studies in the Fraser Valley, though, do not approach short term guideline maxima, and are generally close to the long-term guidelines in the region’s soft water. Table 3.12 Residential urban runoff cadmium concentrations, from the literature  3.4.2.7.3 Iron Iron concentrations in urban runoff published in the literature, summarized in Table 3.13, indicate frequent exceed-ances of the CCME Guideline 0.3 mg/L limit in stormwater. Only the residential stormwater measured in one Cali-fornia study actually met the CCME guideline (Tiefenthaler et al. 1999). As listed in Table 3.1, BC Environment and the USEPA, set the maximum concentration a little higher, at 1 mg/L. Parameter Units min mean max N Season Source or StudyCu µg/L 48 796 All Brombach et al. 20053 16.6 65 23 All Tuccillo 200633 208* All (NURP) Strassler et al. 199965.9 18 Fall, Winter, Spring Tiefenthaler et al. 199911 50 104 152 All (NURP) Whalen & Cullum 19885.5 32 232 41 All Macdonald 200310 5 All Frederick 20033.8 4.45 5.1 2 Winter KWL 2010, wet, storm7.4 13.38 24.33 3 Summer KWL 2010, dry, storm2.5 5.85 9.2 2 Winter KWL 2010, wet, roofs1.7 6.48 11.34 3 Summer KWL 2010, dry, roofsP ra et r Units min mean max N Season Source or StudyCd µg/L 2.3 796 All Brombach et al. 20051.6 18 Fall, Winter, Spring Tiefenthaler et al. 19990.2 0.24 0.31 41 All Macdonald 20030.04 0.05 0.06 2 Winter KWL 2010, wet, storm0.05 0.085 0.144 3 Summer KWL 2010, dry, storm0.02 0.035 0.05 2 Winter KWL 2010, wet, roofs0.03 0.075 0.154 3 Summer KWL 2010, dry, roofs86  Table 3.13 Residential urban runoff iron concentrations, from the literature  3.4.2.7.4 Lead Concentrations of lead in urban stormwater is frequently at levels that are harmful to aquatic environments. Table 3.14 summarizes published minimum, mean, and maximum lead concentrations in residential runoff. Exceedances of CCME guideline maximum of 1 to 7 µg/L (depending on hardness) and, the USEPA’s higher acute concentration of 65 µg/L, are regular, though the more recent studies in the Lower Fraser Valley do show reduced lead concentra-tions in urban runoff. Table 3.14 Residential urban runoff lead concentrations, from the literature  3.4.2.7.5 Nickel Mean published reported values of total nickel concentrations in urban runoff are between 12 and 44 µg/L, summa-rized in Table 3.15 below. BC Environment and CCME guideline maximum nickel concentrations are 25-120 µg/L and 25-150 µg/L respectively, depending on the water hardness. Without water hardness associated with each sam-ple published in the literature, it is difficult to conclude whether the nickel concentrations in the runoff samples were toxic (or above these guidelines), but compared to the USEPA acute concentration of 470 µg/L, typical nickel con-centrations in storm water appear to be benign. Parameter Units min mean max N Season Source or StudyFe mg/L 0.2 3.6 20.8 36 All Tuccillo 20060.061 18 Fall, Winter, Spring Tiefenthaler et al. 19990.86 1.11 1.56 41 All Macdonald 200351 88 125 2 Winter KWL 2010, wet, storm0.356 0.746 1.463 3 Summer KWL 2010, dry, storm23 24.5 26 2 Winter KWL 2010, wet, roofs0.032 0.235 0.587 3 Summer KWL 2010, dry, roofsParameter Units min mean max N Season Source or StudyPb µg/L 118 796 All Brombach et al. 20055 14.6 38.1 22 All Tuccillo 200618 278 558 0 Winter aebi & Dros e 2003175 1579 All (NURP) USEPA 9831 4 208* All (NURP) Strassler et al. 199967 2713 All Smullen et al. 199934.9 18 Fall, Winter, Spring Tiefenthaler et al. 199946 203 409 209 All (NURP) Whalen & Cullum 19889 11.7 16.4 41 All Macdonald 200317 20 29 5 All Frederick 20030.2 0.3 0.4 2 Winter KWL 2010, wet, storm1.6 3.55 7.14 3 Summer KWL 2010, dry, storm<0.2 2 Winter KWL 2010, wet, roofs0.3 1.04 2.43 3 Summer KWL 2010, dry, roofs87  Table 3.15 Residential urban runoff nickel concentrations, from the literature  3.4.2.7.6 Zinc Maximum zinc concentration for the protection of aquatic life is set at (CCME 2014). Nearly all studies reviewed reported mean concentrations of zinc in urban runoff exceeding the CCME guideline of 30 µg/L by two to nine times, as summarized in Table 3.16. Only the two samples each of wet-season flow from both roof and street runoff in a study in the Township of Langley, BC, identified zinc concentrations below this guideline, though dry-weather flows of street-generated runoff exceeded the guideline (KWL 2010). Table 3.16 Residential urban runoff zinc concentrations, from the literature  3.4.2.7.7 Aluminum Aluminum concentrations in urban runoff, too, tend to far exceed the CCME (2014) guidelines for the protection of aquatic life. Published minimum, mean, and maximum concentrations of aluminum in urban runoff are summarized in Table 3.17. Some of the lowest concentrations were found in street and roof runoff during the wet season in the Township of Langley. Toxicity studies have determined that aluminum is increasingly detrimental to aquatic life as pH drops (as solubility increases with decreases in pH, has minimum solubility at pH of 5.5, and precipitates in pres-ence of a hydroxide) (USEPA 1986) though is not necessarily the same among species or life stages (Baker 1982, Gostomski 1990). Because of this relationship to pH, aluminum guidelines are set at 5 µg/L if the pH is below 6.5 (Al3+ ion dominating), and 100 µg/L if the pH is higher than 6.5 (Al(OH)4 dominating). Parameter Units min mean max N Season Source or StudyNi µg/L 22.6 796 All Brombach et al. 200543.8 18 Fall, Winter, Spring Tiefenthaler et al. 1999<2 12 126 ? All BC Environment 1992Parameter Units min mean max N Season Source or StudyZn µg/L 275 796 All Brombach et al. 20057.8 66.8 201 24 All Tuccillo 200615 278 2386 10 Winter Taebi & Droste 200317 12 1 All (NURP) USEPA 1983162 2234 All Smullen et al. 1999229 18 Fall, Winter, Spring Tiefenthaler et al. 199937 254 1416 221 All (NURP) Whalen & Cullum 198864 73 86 41 All Macdonald 200380 5 All Frederick 20036 12 18 2 Winter KWL 2010, wet, storm31 55.6 103.7 3 Summer KWL 2010, dry, storm5 6.5 8 2 Winter KWL 2010, wet, roofs8 24.4 52.3 3 Summer KWL 2010, dry, roofs88  Table 3.17 Residential urban runoff aluminum concentrations, from the literature  Although we do not know the pH of the waters sampled in the studies (to determine which guideline to compare against), we do see that while even the minimum values all greatly exceeded the 5 µg/L guideline, most also ex-ceeded the 100 µg/L guideline. Significant aluminum enters freshwater ecosystems through stormwater, especially during low-flow periods when elevated fractions of stream flow may be associated with direct urban runoff.  3.4.2.8 Temperature Streams experience both seasonal and diurnal temperature flux, directly influencing natural dissolved oxygen varia-bility, but under urbanization, runoff passing from or across heated impervious surfaces contribute to elevated stream temperatures; modelling has indicated that this increase can be by more than 7 oC (LeBlanc et al. 1997, Nel-son & Palmer 2007). Contrary to a storm-associated cooling of summer streams in natural catchments, Galli (1990) reported a reduction in storm size needed to increase stream temperatures as catchment imperviousness increased. The surges of elevated stream temperatures have a reduced dissolved oxygen (DO) saturation level, in turn causing stress to fish and reduced capacity of sensitive species to compete (Nelson & Palmer 2007), though Galli (1990) re-ported no DO sag related to urban storms. Twenty years of monitoring in a highly urbanized city in Japan revealed an increase in both summer and winter temperatures (Kinouchi et al. 2007), but a study on five streams in New York revealed urban streams experienced a decrease of 1.3-3 oC in winter and increase of 5-8 oC in summer than their forested counterparts (Pluhowski 1970). 3.4.2.9 First flush The intuitive phenomenon that accumulated pollutants are ‘washed off’ at the beginning of a rainfall event, espe-cially in small catchments like roofs or parking lots, has long been thought to occur in urban catchments, carrying a disproportional mass load of pollutants in the initial volumes of runoff. A review of the last seventeen years of liter-ature on the concept identifies an evolving method of quantification of the phenomenon, as well as maturing under-standing of the processes, contributing factors, and strength of first flush among pollutants and dissolved versus par-ticle-bound fraction. Definitions of the first flush have evolved from Geiger’s (1984) mass first flush (early runoff volumes carry more than the mean mass of pollutant) led to the concrete though arbitrary definition by Saget et al. (1996) that 80% of pollutant load must be generated in the first 30% of volume. Sansalone & Christina (2004) plot-ted percent cumulative pollutant mass against percent cumulative flow, fitted the curves to power functions (Y=Xa) by linear regression, and confirmed a significant first flush if the coefficient a ≥ 0.185. Using similar techniques, Sansalone & Buchberger (1997) identified ‘first flush’ occurring in any event exhibiting, over the course of the run-Parameter Units min mean max N Season Source or StudyAl µg/L 100 1500 7100 36 All Tuccillo 2006451 600 885 41 All Macdonald 200349 57 65 2 Winter KWL 2010, wet, storm387 811 1638 3 Summer K L 2010, dry, storm26 26.5 27 2 Winter KWL 2010, wet, roofs44 151 341 3 Summer K L 2010, dry, roofs89  off, the percent mass greater than the percent flow (plotting above a 1:1 line of cumulative mass/total mass vs. cu-mulative volume/total volume. More recent studies of the phenomenon compare the various definitions, some even proposing their own arbitrary definitions, including the percentage of the total cumulative load carried by the first 20% of the runoff (Deletic 1998), a seasonal first flush: ratio of pollutant concentration in first storm of season vs. storm later in the season (Lee et al. 2004), and pollutant loading per quartile of storm (Flint & Davis 2007). Individually, and among the studies, one point of agreement was identified: high variability in the strength of first flush (even in small catchments and among similarly-sized rainfall events and antecedent dry weather conditions), and considerable disagreement in the strength of first flush among pollutants. Saget et al. (1996) found that only 20% of the storms monitored exhibited a strong first flush, though 65% exhibited a “moderate deviation” and con-clude scarcity in observation of the phenomenon. Sansalone & Buchberger (1997) made no attempt to quantify the distribution, but found that generally the percent mass loading preceded the event percent cumulative flow, but de-tailed metals analyses identified that copper, zinc, and cadmium (found mostly in their biologically-available dis-solved form) were more easily mobilized by small rainfall events than lead, iron, and aluminum (found mostly parti-cle-bound) were less responsive to the small events. Deletic (1998) attempted to correlate climate, rainfall, runoff quantity and quality to first flush effect, but only identified a slight effect for TSS (though highly variable), and nothing for pH or temperature. In residential neighborhoods, Lee et al. (2002) found a relative first flush strength, among the parameters measured, of TKN > O-PO4 > TP > TSS > Fe > Pb, compared to the Zn, Cd > Cu > Pb, Fe, Al identified by Sansalone & Buchberger (1997), the TSS > COD > TN > Pb > Zn identified by Taebi & Droste (2004) and the TP > NO3 > TKN > Cu > TSS > Zn > Pb identified by Flint & Davis (2007). Typically Pb and Zn have the lowest first flush strength, with TSS exhibiting a weak first flush strength as well. Of the studies consulted, none found a strong causal relationship between watershed or climate factors (size, shape, TIA, rainfall intensity or depth, antecedent dry weather period, etc.) and strength of first flush, and the weak relationships found by any one study were not confirmed by other studies. A conclusion that may be drawn from literature examining the phenomenon may be to size BMPs designed for treatment of urban runoff as if the first flush did not exist (Hathaway et al. 2012, Bertrand-Krajewski et al. 1998). 3.4.3 Infiltration swales improvements to water quality Of central concern for this part of the study is the effectiveness of source controls installed in residential neighbor-hoods at reducing contaminant concentrations (and loading) in urban runoff. The effectiveness of a specific suite of on-lot source controls installed in the mid-2000s in the Township of Langley at reducing peak flow volumes and total event runoff was examined. In this chapter, of specific interest are vegetated, engineered infiltration swales – effectively partial exfiltration biofilters – a physically attractive and effective way of meeting neighborhood runoff reduction targets. The focus is on how well these structures reduce contaminant concentrations from urban runoff, which eventually enters the stream as wet weather-associated urban flow. The following is a review of the literature on contaminant reductions through such practices. It is important to note that among the studies, there is variability among design, construction, monitoring technique, age of structure tested, season of monitoring, frequency of 90  maintenance, influent pollutant concentrations, and rainfall pattern (Debo & Reese 2002), each of which may intro-duce further uncertainty to the application of and comparisons between the published ranges of percent reductions (Strassler et al. 1999). Bioretention-type best management practices remove pollutants through mechanisms of settling/sedimentation (den-sity-exclusion of suspended but not dissolved particles, especially in low-velocity settling or detention ponds), filtra-tion (exclusion/straining of suspended particles and anything adsorbed to them through plants and soil), adsorption (binding of molecules and ions sites on mineral and organic growing/filter media; ion exchange/adsorption is di-rectly proportional to percent organics in the soil, and inversely proportional to mineral soil particle size), infiltration (dissolved/soluble pollutants infiltrating to groundwater rather than leaving site as runoff), phytoremediation (degra-dation, extraction, and assimilation, especially of nutrient, organic, and metals), physical plant resistance (slow run-off velocities, filter, degrade pollutants by their presence in the flow path), volatilization (vaporizing contaminants from liquid to atmosphere), and thermal tempering (helping to decrease high runoff temperatures from urban sur-faces before heated runoff enters streams), and solubilization (speciation of metals due to soil chemistry including pH, ORP, and microorganisms) (Davis et al. 2003, Hinman 2012, Chuan et al. 1996, Chen & Lin 2001). The roadside bioswale under investigation in this study operates both as an infiltration device (rock-filled galleries that detain runoff and allow infiltration, but do nothing for treatment), and as a biofilter (with nutrient-demanding plants in organic soils with a high ion-exchange capacity). Both federal and provincial guidelines mandate treatment of urban runoff. The DFO’s guideline is to mitigate the water quality impacts to fish by treating the first flush from events, with a guideline to treat everything up to the 24-hour event that is 90% of the rainfall (Chilibeck & Sterling 2001): effectively the 6-month 24-hour event (KWL et al. 2012). British Columbia guidelines are no different: to treat 90% of the annual rainfall on-site using best manage-ment practices (Stephens et al. 2002). The percent effectiveness of best management practices, however, is empiri-cal, and designers must base designs of individual BMPs and a treatment train of multiple practices on published data. Fortunately, 530 studies on the performance of stormwater BMPs have been compiled for the International Stormwater BMP Database, available at www.bmpdatabase.org, from which the entire database can be downloaded, the literature review accessed, and summary reports for various BMPs’ performance for removal of various pollu-tants consulted. For gross summaries of treatment effectiveness in the following paragraphs, the statistical summar-ies published by Geosyntec & Wright (2012a) of 30 BMP studies will be consulted, with nuance provided by other sources in the literature. The fine line between ‘bioretention devices’ and ‘bioswales’ is described to exist only in the geometry difference between the two: bioswales are simply linear bioretention devices, with greater than 2:1 length to width ratio, and often installed along roadways (Clark & Acomb 2008), but others describe bioswales as longitu-dinal, low-velocity filtering devices with underdrains, and bioretention as non-linear, and with a rock gallery for temporary detention overlain with a thick layer of organic growing media that absorbs and retains runoff (Geosyntec & Wright 2012b). For the purpose of clarity, the practice under investigation is a bioswale linearly connecting a series of bioreten-tion/bioinfiltration cells, called “infiltration swales” in Metro Vancouver and as illustrated in Figure 3.2, adapted 91  from a typical design drawing with permission by KWL et al. (2012). The numbers in the illustration indicate (1) rock trench with 25 mm-minus round rock, (2) driveway, (3) 350 mm diameter culvert, (4) 150 mm diameter perfo-rated PVC pipe at 2% slope, (5) processed soil, and (6) trench dam with timber weir. More detailed typical cross-sections of the practice are illustrated in Figure 2.2, and Figure 3.4 further below.  Figure 3.2 Typical longitudinal profile of road-side infiltration swale 3.4.3.1 Sediments Infiltration swales remove suspended sediments from urban runoff principally through size-exclusion infiltration. Published TSS reductions by bioretention practices and bioswales – representing diverse designs and treatments – is 78% and 37%, respectively (Geosyntec & Wright 2012a). Expected TSS removal by bioretention and vegetated bi-oswales is estimated at 90%, and 81%, respectively by the USEPA (1999a/b); and vegetative filtration ditches/swales are said to achieve a median 81% TSS reduction (Debo & Reese 2002).  3.4.3.2 Nutrients Through vegetated infiltration swales, stormwater passes through organic soils with active biota, nutrient cycling processes present complexity to the mass balance and to removal efficiencies, with a wide range of efficiencies re-ported for both soluble and particle-bound fractions of both N and P (Hinman 2012).  3.4.3.2.1 Nitrogen Over the time periods examined by researchers, published nitrogen removal rates are highly variable, in some cases with consistent increase in TN over that supplied in stormwater effluent, perhaps from leaching from the organic soils (Hatt et al. 2009). Since the majority of nitrogen in urban runoff is in biologically available dissolved fractions, bioretention facilities must reduce these fractions through nitrification in an aerobic zone and denitrification in an anaerobic, saturated zone, achieved by low flow rates and high detention, so deeper growing media with high water absorption capacity is recommended (Taylor et al. 2005). Possibly a result of older designs not considering the need for nitrification and denitrification, the reported total nitrogen (TN) removal rates vary from 235% increase to 89% reduction through these facilities (Hunt 2003). Studies examining the specific fractions, however, have found gener-ally high NH3 reduction (54-86% indicating nitrification or uptake occurring), poor NOx treatment (194% increase to 23% reduction, indicating nitrification but insufficient denitrification), and a high TKN reduction (36-80% indi-cating removal of much of the particulate and dissolved organic fractions, perhaps in part through ammonification and nitrification but poor denitrification) (Taylor et al. 2005, Hinman 2012, Geosyntec & Wright 2012, USEPA 92  1999a, Hatt et al. 2009). Confirming this, experimental results published by Taylor et al. (2005) point to the need for deep soils in bioretention facilities: achieving 68-75% reduction in TKN concentrations and 60-79% reduction in NH3 concentrations from deeper sampling depths in an experimental cell, but only 38-57% TKN reduction and neg-ligible NH3 reduction from shallower sampling depths. Results of temperature effects on nutrient removal in experi-mental cells point to no negative temperature effects for removal of phosphorus, but elevated temperatures (20oC) are linked to increases leaching of total nitrogen, with relative fractions of very high NOX and high NH3 concentra-tions indicating the lack of, and need for, denitrification in a saturated zone (Blecken et al. 2007). 3.4.3.2.2 Phosphorus High variability is found among reported values of total phosphorus (TP) concentration reductions through bios-wales, ranging from greater than a 10x increase, to 83% reduction, depending on the specific design, treatment, and depth of soil. For vegetated swales, the USEPA reports only 9% reduction (1999b), but if designed with deeper planting/filter media, 70-83% reduction (1999a). Summaries by Hinman (2012) report zero removal at a bioretention depth of only 250mm, but increases to 73% by 500mm and 81% by 1000mm soil depth. By drawing off treated wa-ter from various depths of bioretention, Davis et al. (2001) reported a 78% reduction from deeper parts of the struc-ture, but only 7% from shallow depths, and propose that the mechanism of TP removal is sorption to aluminum, iron, and clay in the soils.  Geosyntec & Wright (2012) report a statistically weak 18% reduction in TP among biore-tention facilities, and a statistically strong 73% increase in TP through the bioswales. Concentrations of soluble phosphorus – the fraction of particular concern for stream health – are reported to increase through bioswales and bioretention devices, by as much as 300%, (Geosyntec & Wright 2012). Experiments in which 15cm head was en-sured by a low-hydraulic conductivity soil underlying a higher conductivity soil exhibited 85% reductions, over the 63% reductions with media densities reversed (Hseih et al. 2007). Fractioning of the TP remaining in the soils was found to be in a form available for uptake by the plants. Depending on the specific soils used, organic and mineral content, particle distribution, depth, and design, it may be difficult to predict the exact phosphorus removal through bioretention/infiltration swales. When designing a stormwater treatment structure to best practices, deep soil depths and the presence of fine mineral (clay) particles in the soil can contribute to phosphorus removal, at least until ad-sorption sites reach saturation. 3.4.3.3 Organic carbon Organic carbon loading – including BOD, COD, and TOC – is reduced through infiltration swales by several pro-cesses, including filtration, phytoremediation, infiltration, and volatilization, and can achieve fairly high removal of roughly 70% (Winer 2000), however organic carbon is an important component of the infiltration swales’ grow-ing/filtration media’s capacity to reduce metals concentrations. 3.4.3.4 Chloride Over 15 months in Maryland, Li & Davis (2009) found an increase in chloride concentrations through bioretention devices, and found that, contrary to expected rapid wash-through of chloride from runoff, elevated concentrations persisted in the BMP effluent, and noted that it “bleeds through all year long.” 93  3.4.3.5 Metals Stormwater carries metals in both dissolved and particulate fractions; metal concentrations have been found to have significant correlations to TSS (Beck & Birch 2012), so as TSS is removed through filtration and sedimentation, par-ticle-bound metals are also temporarily removed from flow. Over time and through physical and biochemical pro-cesses in the soil and water, the accumulated metals initially adsorbed to particles may solubilize and, in the soluble state, re-enter the runoff, interflow, or percolate to deep groundwater. Chuan et al. (1996) demonstrated that solubili-zation of Cd, Pb, and Zn is directly proportional to redox potential, and inversely proportional to pH, propose the dissolution of Fe-Mn oxyhydroxides under acidic, reducing conditions to be the mechanism for the solubilization of heavy metals. Chen & Lin (2001) demonstrate a non-linear relationship between metal solubilization and pH: metal solubilization rapidly increases at low pH, with sensitivity Mn > Ni > Zn in the pH range of 6-8.   3.4.3.5.1 Copper There is great variability in published values of infiltration swales’ (or similar practices’) effectiveness at reducing copper concentrations in urban runoff. Debo & Reese (2002) report a range of 11-70% reductions (51% median) among vegetative filtration systems, the highest reduction by dry swales. From experiments in North Carolina with conventionally-drained bioretention cells with 0.6 and 1.2 m thickness of loamy sand with organic mix, Hunt (2003) found copper removal rates consistently greater than 97% at both treatment depths. In a study performed in Australia examining pollutant removal found that biofiltration systems reduced runoff volumes by 33%, partially contributing (along with other physical and biochemical processes) to the decrease effluent copper loading by an average of 78% (Hatt et al. 2009).  A recent study performed in Auckland, New Zealand, found a 50% increase in median dissolved copper concentration after treatment (though also statistically weak); its authors point out the tendency of copper to equilibrate between dissolved and solid phases, so may reflect a movement into solution of accumulated copper in the soils (Trowsdale & Simcock 2011). From data compiled from 30 studies, Geosyntec & Wright (2012) report a 27% reduction in copper concentration through bioswales (though statistically not strong).  3.4.3.5.2 Cadmium Mean values among the 30 studies reviewed by Geosyntec & Wright (2012) reveal that in the initial years after con-struction, anyway – bioswales can achieve a statistically significant 43% reduction in dissolved cadmium concentra-tions in urban runoff, and 39% reduction in total cadmium concentration. Published effectiveness of bioretention devices was lower, achieving only 5% reduction. 3.4.3.5.3 Iron Few studies examine iron reductions through infiltration swales or similar stormwater treatment devices, but Geo-syntec & Wright (2012) report poor removal rates: 43% removal of total iron through bioswales, and, over a short-term study, a 100% increase total iron concentrations through a bioretention device. This claim must be considered in context of the particular experiment, however, since, unless there is a source of iron (or any other metal) in the filter media itself, the mass balance of increased effluent concentrations is highly improbable. 94  3.4.3.5.4 Lead Similarly for lead, high removal rates are reported. Designed more as bioretention facilities, they can achieve 93-98% reduction of total lead (USEPA 1999a, Hinman 2012, Hatt et al. 2009), and if more as vegetated swales (with less of a growing/filtration media) can achieve 33-67% removal (USEPA 1999b, Hinman 2012, Geosyntec & Wright 2012). Lower removal efficiencies are reported for dissolved lead, however, at only 21% (Geosyntec & Wright 2012).  3.4.3.5.5 Nickel Though considered an important pollutant in urban runoff (both for its elevated concentrations and potential impact on aquatic life), most studies and compiled reviews do not address how well bioretention facilities or infiltration swales reduce nickel concentrations. Comparing mean bioswale inlet and outlet concentrations compiled by Geosyn-tec & Wright (2012), we find a 66% reduction in total nickel concentrations, and 59% reduction in dissolved frac-tions.  3.4.3.5.6 Zinc Infiltration swales (and similarly constructed practices) can achieve high reductions in total zinc concentrations, an-other key contaminant in stormwater: 71% (with standard deviation of 36%) among vegetative filtration systems (Debo & Reese 2002) and 30-99.5% in a conventionally drained bioretention cell (Hunt 2003). Trowsdale & Simcock (2011) published reductions in concentration of total zinc of 96% and of dissolved zinc of 93%, though only 54% reductions of dissolved zinc are reported by Geosyntec & Wright (2012). Statistically significant total zinc concentration reductions of 75% in bioretention practices and 37% in bioswales are reported by Geosyntec & Wright (2012). 3.4.3.5.7 Aluminum Little data exists on the removal efficiency of aluminum by infiltration swales. Glass and Bissouma (2005) reported only 17% decrease in total aluminum concentration over a three-year study of a bioretention facility over fifteen storms.  However, it is assumed that similar mechanisms of removal of particle-bound fractions – specifically through filtration/sedimentation of TSS will achieve notable reductions in this metal. 3.4.3.6 Temperature Especially where bioretention facilities receive runoff from parking lots, roofs, and roads, during summer months, runoff temperature can be significantly reduced (Jones & Hunt 2009), though this difference is not observed where runoff is from sources with less capacity to heat up the runoff (Dietz & Clausen 2005). 3.4.3.7 First flush Pollutant concentration peaks associated with urban runoff and, within the course of runoff, uneven distribution of pollution loading manifesting as a first flush, is mitigated through bioretention and bioswale treatment structures. Since a mass balance indicates a build-up of pollutant load in the BMPs, processes in the soil may cause particle-bound particles to solubilize, and lead to a delay in the release of accumulated metals/ions, nutrients, and carbon. 95  3.5 Methodology The literature review provide valuable information on three themes that inform the research and guide the methodol-ogy and analysis: (a) water quality guidelines for protection of aquatic environments, (b) typical concentrations of pollutants of concern in urban stormwater runoff, and (c) reductions in these concentrations that can be expected through a roadside bioswale. These values are tabulated beside each other in Table 3.18. Exceedances of guideline concentrations of the parameters TP, Cu, Cd, Fe, Pb, Zn, and Al, and potentially harmful concentrations of TSS (de-pending on background concentration) and NOX are reported in the literature on stormwater runoff from residential neighborhoods. From this literature review, we can also expect significant reductions in concentrations of the param-eters TSS, TKN, NH3, TOC, Pb, Ni, and Zn, though there is high variability within and among the studies. Concen-trations of the parameters NOX, TP, Cl-, Cu, and Fe were not reported to be significantly reduced through bioswales. To answer the research question of whether the nine year old roadside bioswale installed in Silver Valley is still per-forming as designed, paired, composited runoff and treated effluent samples were collected over the course of twenty-two rainfall events at strategic sites along Foreman Drive. The following sections outline the build-out of the neighborhood under investigation and the innovative suite of low impact development BMPs installed. As well, the details of the infiltration swale installed along the street for both flow reduction and quality treatment, the equipment installed and sampling schedule, and the laboratory analyses performed are presented.  96  Table 3.18 Guideline, typical concentration, and expected reduction through bioswales for water quality pa-rameters  3.5.1 Study sites in the District of Maple Ridge 3.5.1.1 Build-out conditions & timeframe Construction of Phase 1 of the ‘Silver Maples’ residential development broke ground in 2004, after DFO approval of the developer’s proposed plans that featured a progressive set of low impact development practices for infiltration of rainfall and treatment of runoff from both residential lots and the municipal road right-of-way (ROW). These were Parameter Units GuidelinesTSS mg/L<25 above background (acute)<10% above background (acute)<5% above background (chronic)44 to 809 37% to 90%TDS mg/L - 19 to 120TKN mg/L - 1.73 to 2.5 36% to 80%NOXmg/LNitrate: 550 (acute); 13 (chronic)Nitrite: 0.06 (chronic)0.237 to 13.6 -194% to 23%NH3mg/LNH3: 19 mg/LNH3 + NH4: ph & temp. dependent0.18 to 0.8 54% to 86%TP µg/L trigger ranges; mesotrophic at 10-20 140 to 620 -1000% to 83%PO4µg/L - 100 to 210TOC mg/L - 11.9 to 33.7Cl-mg/L 640 (acute); 120 (chronic) 0.55 to 75.3Hardness mg/L - 8.2 to 80.3 ?Alkalinity mg/L>20, low sensitivity to acid inputs; <10, highly sensitive to acid input28 to 38.2 ?pH 6.5 to 9.0 7 to 7.9 ?Cu µg/L max 2 - 4; depends on hardness 5.85 to 65.9 -50% to 97%Cd µg/L max 0.11 - 7.7; depends on hardness 0.035 to 2.3 5% to 43%Fe µg/L 300 0.061 to 88 -100% to 43%Pb µg/L max 1 - 7; depends on hardness 0.3 to 278 21% to 98%Ni µg/L max 25 - 150; depends on hardness 12 to 22.6 ~66%Zn µg/L 30 6.5 to 275 30% to 99.5%Al µg/L 5 (pH < 6.5); 100 (pH > 6.5) 26.5 to 1500 ~17%Guidelines: CCME (2014); Nagpal et al. (2006) for AlkalinityRange of Means from LiteratureExpected Reductions through Bioswale?bleed-through?~70%97  necessary as development had to be sensitive to impacts on flooding, fish spawning and rearing habitat, management of urban runoff, retention of native trees & vegetation, erosion & geotechnic failure of unstable hill slopes, and frac-tioning of wildlife corridors (District of Maple Ridge 2014). Phase 1 featured construction of fifty-four homes. Many lots back onto parkland adjacent to Anderson Creek; others have a ‘no disturbance covenant’ totaling up to a quarter of some of the lots. Under-street storm drainage carries away only runoff that overflows on-lot infiltration trenches, and that from footing drains. Foreman Drive is a quiet residential street on the extreme northern edge of Maple Ridge, and has a light traffic load, serving only the residences of the neighborhood itself. 3.5.1.2 Best management practices installed & conditions A full suite of LID-BMPs had been installed. On the residential lots, (a) roof downspouts were disconnected from the drainage network, draining instead either to surface or to rock trenches, (b) amended soils to a depth of 300 mm were installed, and (c) an infiltration trench was installed on every lot, to meet site runoff requirements by the DFO. Depending on the location in the neighborhood, runoff from the residential roads either (a) flowed to rain gar-dens/linear infiltration swales, (b) was collected in catch basins and directed to a ‘parkette’ – a large infiltration gal-lery, or (c) was collected by conventional curb & gutter practices and directed to the detention pond. From the upper portions of Foreman Drive, road runoff from intense events reaching the creek would have to exceed the capacity of and pass through the treatment train of (a) a roadside infiltration swale, then (b) a parkette, and finally (c) the deten-tion pond. This research focuses only on the upper portion of Foreman Drive, between 230 Street and 232 Street, where all run-off from the sloping road flows diagonally (down and across the road) into the infiltration swale. Twenty-four homes line this section of Foreman Drive; roof runoff drains to lawns and infiltration trenches. 3.5.1.3 Roadside bioswale details The roadside bioswale under investigation is approximately 200 m from end to end, has a footprint of roughly 600 m2, and serves an impervious area of approximately 3,900 m2. Along the 20 m wide ROW of Foreman Drive, in ad-dition to the longitudinal slope of the road (downhill from 232 Street), a 2% slope carries water across the road to the bioswale located only along one side of the road. Total bioswale width – from roadside flat-curb panel to side-walk – is 4.65 m (39% of the ROW width); an 800 mm depth of 25 mm-minus round drain rock is wrapped with a geotextile filter fabric, underlaying 600 mm of soil as both absorption and filter medium. Pressure treated timbers incised laterally control surface flow; undisturbed soil weirs between rock trenches hold back subsurface longitudi-nal flow to promote infiltration on-site. An overflow 150 mm diameter perforated PVC pipe, sloped at 2% and lo-cated at the top of the infiltration trench (but below the soil layer) allows runoff volumes in excess of trench capacity to leave site. A 300 mm grass strip along the panel curb filters road sediments, and the bioswale is planted with na-tive species including hackberry currant, salal, daylily, an imported honesuckle, and dwarf mountain pine. Under driveways a 300mm culvert allows surface flow to leave site down the bioswale toward the outlet in the parkette. See Figure 3.3 and Figure 3.4. The contributing impervious area, hatched diagonally right in Figure 3.3, (including road, driveways sloping toward the road, and the driveway on the north side of the road) has an area of 3,850 m2 (0.38 ha), 5.5 times the 700 m2 combined area of the thirteen sections of the bioswale, identified in black. Perforated 98  piping for bioswale overflow (dashed lines) and storm drain network (solid lines) are indicated. The detention ponds (cross-hatched polygon to left side of the figure), connected from the parkette (cross-hatched irregular polygon near center of the figure) by storm drain piping, allows for the settlement of sediments, and buffers peak flows before allowing runoff to flow north to the fish-bearing Anderson Creek.  Figure 3.3 Upper Foreman Drive, with roadside swale, parkette, and detention ponds  Figure 3.4 Typical longitudinal cross-section of roadside infiltration swale along Foreman Drive In Figure 3.4 above, the numbers indicate (1) rock trench (of three pictured), (2) driveway, (3) 350 mm diameter culvert, (4) filter cloth entirely wrapping the rock trench and perforated pipe, (5) 2% surface slope, (6) processed soil, 600 mm deep, (7) 25 mm-minus round drain rock, to 800 mm deep minimum, (9) trench dam, and (10) weir timbers. The sketch, not to scale, is a reproduction of a typical bioswale detail from Kerr Wood Leidal Associates, Ltd. A lateral cross-section of the infiltration swale is illustrated in Figure 2.2 in Chapter 2. 99  Briefly mentioned above, stormwater management in the Silver Valley follows a comprehensive ‘treatment train’ approach. Road runoff from upper Foreman Drive that is not infiltrated and leaves the site as overflow through the perforated PVC pipe, discharges to the subsequent group of infiltration galleries in the ‘parkette,’ overflow from which is carried by the storm sewer to the detention basin, which in turn drains to Anderson Creek. A relatively large area of the road ROW was dedicated to both infiltrating and treating road runoff. The 600 mm deep soil layer is in line with best practices (designed much more as a rain garden or bioretention device than as a roadside swale: appropriate for both absorption and treatment of runoff). Other than during extreme events, surface flow does not occur, and the only effluent discharged to the parkette is from the perforated trench overflow. Independently from the rest of the treatment train that follows the bioswale, I examine its treatment efficiency as if it were a stand-alone BMP, identifying water discharged from the perforated pipe to the parkette as the treated effluent. 3.5.1.4 Rainfall statistics The Silver Valley neighborhood receives significant annual rainfall; its suite BMPs is aggressively sized to infiltrate and treat 90% of it. Rainfall statistics, including annual totals and IDF curves drive this sizing. Upper Foreman Drive is 1.3 km from the meteorological measurement equipment hosted at UBC’s Malcolm Knapp Research Forest (MKRF). Researchers there maintain records of hourly rainfall as far back as 2011 and daily rainfall data as far back as 1946. Mr. Ionut Aron, MKRF research coordinator, supplied me with historical climate data for the site. Average annual rainfall between 1946 and 2001 is 2,190 mm ± 93 mm at 95% confidence; Foreman Drive likely receives a little less since it is at a slightly lower elevation. Using data from 1983 to 2002, cumulative monthly rainfall and mean temperature are plotted in Figure 3.5 below.  Figure 3.5 Average monthly rainfall and temperature at MKRF, nearby Silver Valley 100  According to KWL et al. (2012), capture of 90% of the annual rainfall at a site receiving 2100 mm/year means de-signing infiltration practices to capture the first 66 mm of rainfall: the 6-month 24-hour event, roughly equivalent to 72% of the 2-year 24-hour event.  3.5.2 Rainfall measurement techniques To get an indication of the intensity of road runoff, I installed a tipping bucket-type rain gauge (manufactured by Rainwise™ Inc.) on a post 1.6 m above the ground within the rain garden 5 m distant from the point of collecting road runoff samples. See Figure 3.6 below; the rain gauge with funnel cover removed is in the foreground; in the mid-ground is the plywood box containing the autosampler, with Foreman Drive in the background. The rain gauge’s datalogger was programmed to count 0.254 mm of rainfall depth per tip. Data summaries with 5-minute res-olution were downloaded on three occasions; the datalogger and battery had more than sufficient capacity for the entire four months of this investigation. RL-Loader™ 2.2.2 software – available for free from the manufacturer’s website – was used to set the datalogger’s clock and tip calibration, and to download the data to Microsoft Excel™.  Figure 3.6 Tipping-bucket rain gauge installed near sampling site along Foreman Drive Just as a check, this rainfall data was compared to MKRF hourly rainfall data supplied by Mr. Ionut Aron. In Figure 3.7 below is a plot of this comparison for the month of March as an example of the similarities and differences be-tween the two gauges. Note the differences in hourly intensity on 08 March (about 2.5 mm/hour in MKRF, up to 4.5 mm/hour at Foreman Drive), and the second spike on 14 March (up to 4 mm/hour in MKRF, and about 2.2 mm/hour at Foreman Drive). For the month of March 2014, the two sites’ cumulative volumes are reported as 312 mm in MKRF and 360 mm at Foreman Drive: significantly wetter by both measurements than average March rainfall (of 101  203 mm). On the lower chart, hollow diamonds indicate the times when the autosamplers began their cycles, started by road runoff filling the stilling bucket.  Figure 3.7 MKRF and Foreman Drive hourly rainfall comparison 3.5.3 Treatment effectiveness sampling plan To address the objective of this facet of the research, a treatment effectiveness approach was used: taking paired samples of (a) direct road runoff, and (b) infiltration swale effluent representing treated runoff. Discrete samples of road runoff collected during each rainfall event were composited, as were composited the discrete samples of treated effluent, for paired analysis of each of these composited samples. Paired values for each parameter were then plotted for the events sampled, and statistical analyses compare mean concentrations and the significance of observed trends. 102  3.5.4 Stormwater sampling equipment & methods 3.5.4.1 Sampling locations The two sampling locations are identified by the triangle (road runoff sampler) and star (treated effluent sampler) in Figure 3.8, along with contributing impervious road, driveway, and sidewalk areas (diagonally-hatched), perforated pipe that drains full rock trenches (dashed lines), and the infiltration parkette (cross-hatched) that receives overflow via the perforated pipe. Context for this portion of the site can be seen in Figure 3.3 above.  Figure 3.8 Road Runoff and Treated Effluent sampling sites along Foreman Drive Road runoff enters the bioswale along its full length; samples were collected adjacent to the last driveway on the north side of Foreman Drive, just uphill from 230 Street. A narrow, shallow channel was cut into the landscaping, lined with poly, and filled with clean rip-rap, leaving space for runoff to flow without significant contamination into a stilling bucket, from which samples were automatically drawn, as in Figure 3.9 below. The road is just beyond the dwarf mountain pine to the top of the image. Also, in Figure 3.10 is pictured the PVC pipe providing stiffness and protection to the sampling line and water sensor wires leading back to the autosampler and sensor/controller. 103   Figure 3.9 Road runoff stilling bucket & collection site along Foreman Drive 104   Figure 3.10 PVC pipe protecting road runoff sampling hose along Foreman Drive Similarly, a rock-filled, poly-lined channel directed treated effluent from the discharge point of the perforated rock trench overflow pipe. The lower end of this pipe discharged just underground at the upper end of the parkette after passing under 230 Street. Although there is a 300mm culvert under the road, it did not carry this water. In fact, this culvert remained dry in all but the most intense events. In Figure 3.11 below, the sampling line can be seen reaching the stilling bucket during a period of effluent flow, and in Figure 3.12, the PVC housing both protecting the sam-pling line, and carrying the control wire under the road from the sensor/controller located with the road runoff sam-pler. 105   Figure 3.11 Treated effluent stilling bucket & collection site in the parkette  Figure 3.12 PVC pipe housing sampling tubing & control wire in the parkette The two sampling sites were 35 m apart, the runoff sampling site selected so that it was nearby the effluent sampling site, as the two samplers were connected by a cable, so the signal to begin sampling could be sent to both samplers simultaneously. 106  3.5.4.2 ISCO autosamplers Programmable ISCO 3700 automatic samplers capable of taking 24 discrete or composited samples and pumping to a maximum head of 7.9 m were used. Powered by 12V batteries, and housed in 2’ x 2’ x 4’ boxes built with ¾” ply-wood and secured with padlocks as in Figure 3.13, the samplers were safely deployed for extended periods (up to two weeks between battery changes).  Figure 3.13 Programmable autosampler in protective box along Foreman Drive Each sampler was programmed with a unique sampling schedule (Figure 3.14). Because of the unpredictability of rainfall timing and intensity, the sampler collecting runoff from the road was programmed to collect all 24 samples at an interval of 15 minutes, sampling regularly over 5 hours 45 minutes. Since effluent flow was more predictable during/after a storm, the effluent sampler was programmed to take six samples spaced 15 minutes apart after the ini-tial sample, then at half-hour intervals for the 17 remaining samples, sampling for a total of 10 hours.  Figure 3.14 Discrete sampling schedule for road runoff and treated effluent 107  3.5.4.3 Water sensor & sampling control Since the ISCO 3700 samplers are equipped with an 8-pin connector for communication with external flow sensors, with a function to inhibit sampling as long as the internal 12V passes between two of the pins, with help from tech-nical staff at Kerr Wood Leidal Associates Ltd., I was able to use this feature to start the samplers when a water level sensor (with sensor wires located in the road runoff stilling bucket) opened the circuit. An Atmega™ 328P mi-crochip was programmed using the open-source Arduino™ programming language, user interface, and electronics prototyping platform to read the analog signal strength from a simple transistor-based water sensing circuit, take a multiple-point 2-minute running average of signal strength, and above a programmable threshold, send a digital sig-nal to a double-pole double-throw latching relay, which in turn opened the ISCO 3700 inhibit circuit and sent a “high” digital signal to the external datalogger. Photographs of the control unit and the sensor circuit can be seen in Figure 3.15, below. This control unit was supplied with a reset switch, and a test/control function, to determine from a 10-point, 12-second running average the analog signal strength.   Figure 3.15 Sampler control unit and water level sensor circuit Four cables pass through the unit’s waterproof housing: 12 V power, sensor wires from the stilling bucket, signal out to the datalogger, and inhibit line to the two autosamplers. A circuit schematic for the sensor unit, and the code pro-grammed in Arduino™, are included for reference in Appendix E. 3.5.4.4 Datalogger with telemetry A key component to the experimental set-up was a used telemetry-equipped datalogger on loan from Kerr Wood Leidal Associates, Ltd., programmed to monitor the battery voltage (and report it every four hours to the FlowWorks server), and, upon receiving the sample-start signal from the above-mentioned sensor/controller, report to Flow-Works. In turn, FlowWorks was programmed to send an email message to me, notifying me of the exact time that sampling began. The datalogger was also powered by the 12 V deep-cell battery. In Figure 3.16, below, is pictured the square box housing the telemetry-equipped datalogger, along with the level sensor/sampler controller, the 12V power source, and the autosampler within the secured, protective housing. 108   Figure 3.16 Autosampler, sensor/controller, datalogger, and battery deployed in the field 3.5.4.5 Collection & transport I found myself in a position of privilege as I pursued this research. The first luxury I had was that of the datalogger that sent me an email at time of the initial sample, Since the 24 discrete samples are taken over a period of maximum ten hours, I could program my time with ten hours advance, to ensure arrival at site for collection of up to (24 x 2 = 48) discrete samples immediately after collection, for rapid transport back to the laboratory for compositing, preser-vation, and cooling. The second luxury was having sole access to four ISCO 3700 samplers. Needing only two sam-plers in the field, I was able to easily interchange the base (holding 24 bottles), capping the recently filled sample bottles and uncapping the clean bottles before re-deployment. Sample start and collection time was recorded, as well as any samples that were missed. pH and temperature meas-urements were taken if sufficient water was flowing into the stilling buckets from road runoff or BMP effluent, and all measurements were recorded. Samples were transported directly back to the Environmental Engineering laboratory at UBC, for subsequent com-positing and preservation. The samples were not put on ice in the field, and all work with the samples was per-formed in the clean conditions of the laboratory. 3.5.4.6 Quality control & trip blanks Sample contamination was checked by rigorous cleaning of all bottles between sample runs, and by physical and chemical analysis of trip blanks, representing approximately 10% of the samples collected. After emptying the dis-crete samples into the compositing container, the bottles were washed by the following method:  Rinse with distilled water.  Wash with distilled water and 1mL of 2% solution of nitric acid. Cap, shake, uncap, and empty.  Rinse with distilled water: cap, shake, uncap, and empty. 109   Rinse with distilled water: cap, shake, uncap, and empty.  Rinse with deionized water: cap, shake, uncap, and empty.  Air dry, and cap. One-liter plastic bottles, cleaned in the same manner as above, and then filled with deionized water, were trans-ported to site with the fresh bottles, uncapped at time of bottle change, and left on-site until the next storm’s samples were collected, at which time they were capped and transported to the laboratory with the runoff and effluent sam-ples. 3.5.5 Laboratory analysis of water quality parameters All samples were transported to the Environmental Laboratory at the University of British Columbia, in Point Gray, Vancouver. For reasons mentioned elsewhere, analysis was not performed on the discrete samples, partly because of the vastly greater time and cost associated with the increased number of samples, and partly because of the high var-iability expected in any ‘first-flush’ sampling sequence. Because of varying rainfall intensity, usually less than the full set of 24 discrete road runoff samples was collected, but usually enough samples were collected to allow the full suite of laboratory analyses, described in the following sections. For each event, both of these composited samples were then separated into seven bottles with different filtering and preservation requirements for cooling until analy-sis. 3.5.5.1 Compositing & preservation Discrete road runoff samples were composited into a single aggregate event sample by shaking, uncapping, and emptying each bottle’s contents into a 10 L plastic composite container. The same was done with the effluent sam-ples. In Table 3.19 below is listed the filter and preserving agent required for each parameter, as well as the maxi-mum hold time to analysis. Temperature and pH are not included in the table because if measurements were made, they were done in the field to acquire readings representative of field conditions.  Table 3.19 Preservation requirements for analytical parameters  Bottle Para eter Hold Time Size (mL) Filter Preservative1 TSS 7 days 1000 - -2 TDS 7 daysAlk linity 14 daysChoride 28 days3 PO4 28 daysNOx 3 daysNH3 28 days4 TP 28 daysTKN 28 days5 Metals-Total 180 days 60 - HNO3 < pH 26 Metals-Dissolved 180 days 60 0.45µm HNO3 < pH 27 TOC 7 days 125 - HCl < pH 2- -H2SO4 < pH 20.45µm- H2SO4 < pH 212560250110  3.5.5.2 Analytical methods Analysis of most of the water quality parameters followed procedures in Standard Methods (APHA, AWWA, WEF 2005) and are listed in Table 3.20.  Table 3.20 Analytical methods of water quality parameters  Initially, chloride concentrations were measured following Standard Method 5400 C: Mercuric Nitrate Method, however because of the mercury-containing waste that the method produced, and that an alternative method using the Beckman 320 pH/Temp/mV probe & meter produced results within 2% of that of the mercuric nitrate method, the latter method was used for the rest of the analyses. Using the probe and meter, a calibration curve was deter-mined for each batch of samples, against which sample concentrations were calculated. Sample pH was measured in the field using a Waterproof Double Junction pHTestr® 30 digital pH meter manufac-tured by Oakton Instruments, with simultaneous display of pH (+/ 0.01 unit accuracy) and temperature (+/- 0.1oC accuracy). 3.5.6 Statistical analysis on data To facilitate comparisons to published urban runoff quality parameters, and concentration reductions, the mean value was calculated from the n samples analyzed. In cases where measured concentrations fell below detection lim-its, either the value was removed entirely from statistical calculations (thereby reducing the number of samples ex-amined), or a compromise value of one-half the detection limit was assigned. Where applicable in the discussion of the results, what was done in the particular case is addressed. To determine 90% confidence intervals, the standard Parameter MethodTSS Standard Methods 2540D: Total Suspended Solids Dried at 103-105CTDSRadiometer Analytical pIONeer30TM TDS/conductivity probe & meterTKN Standard Methods 4500-N(org) D: Block Digestion and Flow Injection AnalysisTP Standard Methods 4500-P H: Manual Digestion & Flow Injection Analysis for Total PhosphorusAmmonia Standard Methods 4500-NH3 H: Flow Injection AnalysisNOx Standard Methods 4500-NO3- I: Cadmium Reduction Flow Injection MethodPO4 Standard Methods 4500-P G: Flow Injection Analysis for OrthophosphateChloride Beckman 320 pH/Temp/mV probe & meter Metals Standard Methods 3030 E: Nitric Acid Digestion (preparation)Standard Methods 3120 B: Inductively Coupled Plasma (ICP) Method (analysis)Hardness Standard Methods 2340 B: Hardness by CalculationAlkalinity Standard Methods 2320 B: Titration MethodTOC Standard Methods 5310 B: High Temperature Combustion MethodpH Oakton Instruments Waterproof Double Junction pHTestr® 30 digital pH meter Temperature Oakton Instruments Waterproof Double Junction pHTestr® 30 digital pH meter 111  error of the mean (standard deviation divided by the square root of the number of samples considered) multiplied by the two-tailed Student’s t-statistic was added to/subtracted from the calculated mean value. 3.6 Results & discussion Twenty-two unique rainfall events were sampled between mid-February and mid-June in Silver Valley of Maple Ridge. Following is a discussion of the technical aspects of the work, from quality control, through a discussion of rainfall measurements, and finally a discussion of the concentrations of pollutants in the runoff and reductions (or measured increases) in the effluent discharged to the parkette. To support the thesis argued here, raw water quality for road runoff and treated effluent is tabulated in Appendix B, a rainfall comparison between Silver Valley and the nearby Malcolm Knapp Research Forest is graphed in Appendix C.2, twenty-two graphs of rainfall & sampling times are plotted in Appendix D, the custom-built water sensor circuit and code are included in Appendix E, and District of Maple Ridge Letters & Permissions are compiled in Appendix G. 3.6.1 Accuracy & precision 3.6.1.1 Laboratory blanks Four deionized water laboratory blanks were run through the analytical process for each water quality parameter. In the seven cases in which laboratory blanks exceeded the de facto detection limits, they were raised to accommodate the uncertainty evidenced in the elevated concentrations in the blanks. In Table 3.21 are listed the original detection limit for each parameter, and elevated detection limits where appropriate given exceedances among the blanks. These elevated detection limits were used in the subsequent parameter analyses. 112  Table 3.21 Laboratory blank samples water quality results  3.6.1.2 Field blanks Field blanks – thoroughly cleaned 1000 mL plastic bottles filled in the laboratory with deionized water – were un-capped in the field and left in the autosampler along with the sample bottles, then capped and returned with the sam-ples after autosampler collection during a rainfall event. The field blanks were then analyzed for each of the water quality parameters, exactly following the processes of the rest of the samples. Perhaps redundantly listed in Table 3.22, none of the field blanks exceeded the (elevated) detection limits. Parameter Units Detection Limit Increased to 1 2 3 4TSS mg/L 1 < D.L. < D.L. < D.L. < D.L.TDS mg/L 1 < D.L. < D.L. < D.L. < D.L.TKN mg/L 0.5 < D.L. < D.L. < D.L. < D.L.NOXmg/L 0.1 < D.L. < D.L. < D.L. < D.L.NH3mg/L 0.1 < D.L. < D.L. < D.L. < D.L.TP mg/L 0.02 → 0.05 < D.L. 0.03 < D.L. 0.04PO4mg/L 0.25 < D.L. < D.L. < D.L. < D.L.TOC mg/L 0.5 → 1 < D.L. < D.L. 0.76 0.88Cl-mg/L 1 < D.L. < D.L. < D.L. < D.L.Hardness mg/L 0.9 < D.L. < D.L. < D.L. < D.L.Alkalinity mg/L 0.5 < D.L. < D.L. < D.L. < D.L.Cu µg/L 0.003 P.E. P.E. P.E. P.E.Cd µg/L 0.002 0.002 < D.L. < D.L. < D.L.Fe µg/L 0.02 → 0.05 0.05 < D.L. 0.04 0.03Pb µg/L 0.01 → 0.03 < D.L. < D.L. 0.024 0.025Ni µg/L 0.007 → 0.010 < D.L. < D.L. < D.L. 0.009Zn µg/L 0.07 → 0.07 < D.L. < D.L. < D.L. < D.L.Al µg/L 0.018 → 0.100 0.074 < D.L. 0.048 0.102P.E. - sample preparation error P.E. - sample preparation error< D.L. - below detection limit < D.L. - below detection limit113  Table 3.22 Field blank samples water quality results  3.6.2 Rainfall & sampling Along Foreman Drive in the residential development of Silver Valley in the District of Maple Ridge, road runoff and infiltration swale effluent were sampled as described in the methodology section above. I collected samples through 22 unique storms from mid-February to mid-June 2014. The sampler collecting road runoff was programmed to take samples every 15 minutes after the first one; the sampler collecting treated BMP effluent was programmed to take six samples every 15 minutes after the first one, and then every 30 minutes for the remaining seventeen samples. Shaded circles in Table 3.23 below indicate a unique sample actually collected; dashes indicate samples not col-lected. During most storms, almost all discrete samples of treated effluent were taken; an average of 6.5 road runoff samples per rainfall event were collected. Then for each event, all discrete samples of road runoff were combined Parameter Units Detection Limit 28-May 19-Jun 28-May 19-JunTSS mg/L 1 < D.L. < D.L. < D.L. < D.L.TDS mg/L 1 < D.L. < D.L. < D.L. < D.L.TKN mg/L 0.5 P.E. P.E. P.E. P.E.NOXmg/L 0.1 < D.L. < D.L. < D.L. < D.L.NH3mg/L 0.1 < D.L. < D.L. < D.L. < D.L.TP mg/L 0.05 < D.L. < D.L. < D.L. < D.L.PO4mg/L 0.25 < D.L. < D.L. < D.L. < D.L.TOC mg/L 1 < D.L. < D.L. < D.L. < D.L.Cl-mg/L 1 < D.L. < D.L. < D.L. < D.L.Hardness mg/L 0.9 < D.L. < D.L. < D.L. < D.L.Alkalinity mg/L 0.5 < D.L. < D.L. < D.L. < D.L.Cu µg/L 0.003 P.E. P.E. P.E. P.E.Cd µg/L 0.002 < D.L. < D.L. < D.L. < D.L.Fe µg/L 0.05 < D.L. < D.L. < D.L. < D.L.Pb µg/L 0.03 < D.L. < D.L. < D.L. < D.L.Ni µg/L 0.010 < D.L. < D.L. < D.L. < D.L.Zn µg/L 0.07 < D.L. < D.L. < D.L. < D.L.Al µg/L 0.100 < D.L. < D.L. < D.L. < D.L.P.E. - sample preparation error< D.L. - below detection limitRunoff Sampler Effluent Sampler114  into a composite sample in the laboratory, and all discrete samples of treated effluent were combined into another composite sample. Table 3.23 Discrete samples collected for each of the 22 rainfall events in Silver Valley  Start Date& TimeSampler 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff - - - - - - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● - - - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● ● ● ● ● ● ● - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● - - - - - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● - - - - -Runoff ● ● ● - - - ● ● ● ● - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● - - - ● ● ● - - - - - - - ● ● ● - - - - - -Effluent - ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● - - - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff - ● ● ● ● ● ● ● - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● - - - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● - - ● ● ● - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● - ● - - - - - - - ● ● - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● - - - - - - - - - - - - - - - - - - - -Effluent - - ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● - - - - - - - - - - - - - - - - - - - - -Effluent - ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● ● ● - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● - - - - - - - -Effluent - - ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● ● ● ● - - - - - - - - - - - - - - - - -Effluent - ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● - - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● ● - - - - - - - - - - - - - - - - - - -Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● - - - - - - - - - - - - - - - - - - - - - - -Effluent - ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● - ● ● ● ● ● ● ● ● - ● ● ● ● ● ● ● ● ● ● ● ●Effluent ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Runoff ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●Effluent - - - - - - ● ● ● ● ● ● ● ● ● - - - - - - - - -Runoff - - - - - - - - - - - - - - - - - - - - - - - -15 Feb 15:2814 Feb 04:0411 Feb 18:5610 Feb 08:5828 Mar 17:3214 Mar 13:4104 Mar 13:3302 Mar 16:1525 Feb 15:4817 Feb 23:3317 Apr 00:5716 Apr 12:2403 Apr 15:2930 Mar 10:4929 Mar 20:5629 Mar 08:2019 Jun 22:0528 May 23:5426 May 03:3308 May 22:4724 Apr 01:2923 Apr 06:59115  Rainfall around each sampled event is charted in Appendix D, one example of which is repeated in Figure 3.17 be-low. The shaded circles following the plotted running hourly rainfall indicate the timing of road runoff samples (since after initial abstraction, road runoff intensity closely follows rainfall intensity), and the open circles indicate the timing of the treated effluent sampling.  Figure 3.17 Rainfall, and timing of road runoff & treated effluent samples for 14 February event Rainfall from nearby MKRF over the period February through June 2014 is plotted in Figure 3.18, below, with the beginning of runoff and treated BMP effluent sampling times plotted as hollow diamonds along the bottom of the chart. Many storms were missed, due in part to the following:  If collection of filled sample bottles occurred during a subsequent rainfall, and runoff was still occurring, it was impossible to reset the sampler without it immediately taking a back-to-back set of samples (which was not wanted). When this occurred, I had to leave the sampler turned off, and return the next day (or days later) to reset it.  Over the course of the spring, as conductivity of the road runoff decreased, analog signal strength, too, de-creased, halting sampling altogether for a time until I reprogrammed the sensor to a lower sensitivity threshold. It may not have entirely solved the problem, but it did pick up more storms.  To allow drainage after storms, the stilling buckets had ¼” holes drilled in their floor. Though hourly rainfalls intensities may have been high, perhaps sometimes the rainfall was not intense enough to cause sufficient rapid enough runoff to fill the stilling bucket as it simultaneously drained. 116   Figure 3.18 Rainfall & sampling times over spring 2014 Laboratory analyses of each of the parameters are discussed in the sections that follow, with emphasis on observed trends, contaminant reductions through the BMP, and these findings’ relationship to both the literature and the CCME Guidelines for the Protection of Aquatic Life. 3.6.3 Sediments Mean TSS concentrations in road runoff were found to be 84mg/L, median value of 35.5, with range between 6.9 mg/L and 656 mg/L. The highest concentration was measured in runoff during a rain-on-snow event on 25 February. This mean value is a little lower than the mean values published in most of the literature, with the exception of 78 mg/L reported by Smullen et al. (1999), 55 mg/L reported by CWP (2010), and the 44 mg/L reported in Macdon-ald’s (2003) Vancouver area study. Event TSS concentrations are plotted for each event in Figure 3.19 (linear scale) to appreciate the peaks, and Figure 3.20 (log scale) to visualize the range values. 117   Figure 3.19 TSS concentrations (linear scale) for 22 rainfall events in Silver Valley  Figure 3.20 TSS concentrations (log scale) for 22 rainfall events in Silver Valley TSS reductions through infiltration swales is typically high, and upward of 80% can be expected (USEPA 1999a/b). TSS values for the 22 sampled events are plotted in the figures above; significant reductions are noted for the 17 events for which both direct road runoff and treated effluent from the infiltration swale could be taken, plotted in Figure 3.21 below, usually achieving over 90% reduction. 118   Figure 3.21 TSS concentration reduction for 17 paired events in Silver Valley Mean TSS concentration in road runoff and treated effluent are plotted in Figure 3.22, with 90% confidence limits represented by the error bars. Mean effluent TSS was found to be 5.3 mg/L. Mean TSS reduction, plotted was calcu-lated to be 92%, (+/- 3.8% at 90% confidence level). This TSS reduction significantly helps to minimize TSS contri-butions to the receiving waterbody (Anderson Creek), and to minimize short- and long-term TSS increases above background concentrations.  Figure 3.22 Mean TSS & 90% CI for road runoff and treated effluent A different picture of sediment treatment emerged, however, after examination of TDS concentrations in road runoff and treated effluent. Rather than the observing decreased TSS concentrations through the treatment BMP, a strong increase in TDS is observed in the effluent from the BMP compared to the direct road runoff. Figure 3.23 and Fig-ure 3.24 plot the runoff and effluent concentrations for 22 storms, first on a linear concentration scale to illustrate peaks, and then on a log scale to highlight the lower concentrations. For every rainfall event except the 25 February rain-on-snow event, BMP effluent concentrations greatly exceeded road runoff concentrations. Of special interest 119  though is the fairly consistent concentration of TDS in the effluent in the absence of significant spikes. Excess dis-solved solids introduced during the 25 February spike appeared to wash out of the bioswale by mid-March, with only a gentle decline over time. With only a short period of observation (only four months in the latter part of the rainy season), and without flow information, it is difficult to make conclusions about mass balance and outflow loading, and processes. Consistent with findings by Chuan et al. (1996) and Chen & Lin (2001) that metals solubil-ize under acidic soil conditions, may be that since construction of the neighborhood and the bioswale, components of the accumulated suspended sediments are be solubilizing in the low-pH soils and slowly leaching from the BMP with subsequent flow.  Figure 3.23 TDS concentrations (linear scale) for 22 rainfall events in Silver Valley  Figure 3.24 TDS concentrations (log scale) for 22 rainfall events in Silver Valley 120  Plotting this same information on a line graph, as in Figure 3.25 below, we can observe peak TDS in effluent lag-ging behind peak runoff TDS by about a week, dropping back to low conditions within three weeks, and slowly de-creasing over time, though still many times higher than the runoff concentrations. Perhaps the peak (delay) in efflu-ent indicates wash-through of a spike in dissolved solids from the recent runoff, and the yet-elevated concentrations after mid-March reflect the wash-out of solids with a longer residence time in the BMP.  Figure 3.25 TDS effluent lag and steady-state concentration Finally, statistical analysis of the data indicates an increase in mean TDS concentration through the infiltration swale from mean of 36 mg/L (median 5.7 mg/L, range 1.3-462 mg/L) in road runoff to mean of 103 mg/L (median 43 mg/L, range 24-526 mg/L) in BMP effluent. Error bars in Figure 3.26 below indicate 90% confidence interval of the calculated means. 121   Figure 3.26 Mean TDS concentrations & 90% CI for road runoff and treated effluent Road runoff TDS concentrations measured here are similar to published values of TDS in residential urban runoff. Mean value of 46 mg/L was reported by Frederick (2004), and KWL (2010) reported a wet season TDS mean of 120 mg/L (range 110-130 mg/L) in residential road runoff, and a dry season mean of 49 mg/L (range 46-52 mg/L). Re-moval of TDS – or dissolved portions of most pollutants – via infiltration swale-type BMPs is not specifically ad-dressed in the literature. 3.6.4 Nutrients 3.6.4.1 Nitrogen Three analyses of nitrogen concentrations were made for each sample: total Kjeldahl nitrogen (TKN), NOX, and NH3. Total Kjeldahl nitrogen (TKN), the sum of (dissolved and particulate) organic nitrogen and ammonia, was measured in the runoff and BMP effluent samples. My own sample preparation error in the laboratory compromised the quality of the last batch of TKN samples, the results of which were rendered useless. With detection limits of 0.5 mg/L, All BMP effluent TKN concentrations fell below the detection limit of 0.5 mg/L, and only five road runoff samples had concentrations above the limit, three of which were close to the limit. Mean TKN concentration of these five road runoff samples was found to be 1.2 mg/L (median 0.8 mg/L, range 0.6-2.7mg/L), which is similar to values reported in the literature: 1.9 mg/L (USEPA 1983), 1.47 mg/L (CWP 2010), and between the values of 0.23 mg/L and 2.5 mg/L reported by Whalen & Cullum (1998) and Frederick (2003), respectively. Because of the limita-tions noted above, statistical analysis of TKN reductions was not performed, and I was unable to determine exact TKN reductions, though since all effluent concentrations are below the detection limit, it does appear that TKN is reduced through the bioswale. Whether values below the detection limit are assigned a value equal to the detection limit (0.5 mg/L) or one-half the detection limit (0.25 mg/L), treatment efficiency can be calculated as either 38% or 63%, respectively, though these numbers should not be considered reliable or statistically supported. 122  Table 3.24 TKN measured in road runoff and BMP effluent in Silver Valley  NOX, the combined measurement of nitrate (NO3-) and nitrite (NO2-), was measured in all composited stormwater samples. In stormwater, nitrate concentrations are known to typically far exceed nitrite concentrations (Taylor et al. 2005). Measurements of NOx in eight of the direct road runoff samples and two of the BMP effluent samples were below the detection limit of 0.1 mg/L, and two road runoff samples were not measured. Measured NOx concentra-tions for twenty storms are summarized in Table 3.25. Event Road Runoff BMP Effluent11-Feb < D.L.14-Feb < D.L. < D.L.15-Feb < D.L.17-Feb < D.L. < D.L.25-Feb 0.57 < D.L.02-Mar 2.67 < D.L.04-Mar < D.L. < D.L.14-Mar < D.L. < D.L.28-Mar 0.83 < D.L.29-Mar 1.15 < D.L.29-Mar < D.L.30-Mar < D.L. < D.L.03-Apr < D.L. < D.L.16-Apr 0.73 < D.L.Units are in mg/L. Detection Limit = 0.5mg/L Total Kjeldahl Nitrogen123  Table 3.25 NOX concentrations for 20 rainfall events in Silver Valley  These values are plotted in Figure 3.27; for statistical purposes NOx concentrations below detection limit are here assigned a value of one-half the detection limit: they plot to 0.05 mg/L.  Figure 3.27 NOX concentrations for 20 rainfall events in Silver Valley Event Road Runoff BMP Effluent % Increase14-Feb < D.L. 0.46 > 360%15-Feb < D.L. 0.35 > 250%17-Feb < D.L. 0.46 > 360%25-Feb 0.12 0.46 298%02-Mar 0.36 0.51 44%04-Mar 0.35 0.64 82%14-Mar 0.14 0.23 68%28-Mar < D.L. 0.25 > 140%29-Mar 0.00 0.22 > 110%29-Mar < D.L. 0.20 > 90%30-Mar < D.L. 0.18 > 70%03-Apr 0.47 0.54 15%16-Apr 0.40 0.55 37%17-Apr 0.40 0.56 41%23-Apr 0.36 < D.L.24-Apr 0.41 < D.L.08-May 0.14 0.40 182%26-May < D.L. 0.41 > 310%28-May < D.L. 0.29 > 190%19-Jun 0.33All units in mg/L. Detection Limit = 0.1 mg/LNOX124  Mean values of NOx concentrations – with 90% confidence intervals – for direct road runoff and BMP effluent are plotted in Figure 3.28 below. These mean values and descriptive statistics were calculated by assigning a value of one-half the detection limit to any values that fell below the detection limit, rather than calling them either zero or the detection limit.  Figure 3.28 Mean NOX concentrations & 90% CI for road runoff and treated effluent Mean road NOX concentration in road runoff was found to be 0.196 mg/L (median 0.127, range 0.05-0.468 mg/L), which is lower than the typical values published in the literature (see Table 3.5 in the literature review section for values from several studies); even the local Vancouver studies reported mean values 2-4 times that of what I found (Macdonald 2003, KWL 2010). These values are significantly lower than the CCME guidelines of maximum 550 mg/L for acute concentrations, and 13 mg/L for longer periods. Concentrations of ammonia in both the road runoff and the treated effluent were low, and most of the measured val-ues – especially in the BMP effluent – were below the detection limit of 0.1mg/L. Since so many of the measured values are below detection limit, any attempt to assign zero, detection limit value, one-half the detection limit value, or excluding altogether the values was determined to skew the data, so no statistical analyses on this data are pre-sented, other than mean (0.153 mg/L) and median (0.146mg/L) were calculated for road runoff. I will not attempt to make grand interpretations of any perceived ammonia reduction. Macdonald (2003) reported similarly low mean value (0.18 mg/L). Ammonia concentrations in the effluent, at all values of pH, were below CCME guidelines. The low ammonia concentration in the effluent, but increasing NOX through the BMP, may indicate insufficient resi-dence time in saturated rock zone for denitrification to occur (Taylor et al. 2005, Hatt et al. 2009).  3.6.4.2 Phosphorus Both total phosphorus (TP) and bioavailable orthophosphate (PO4) was measured for all samples. All measurements of PO4 were below the 250 µg/L detection limit, so I will make no attempt at analysis. Published concentrations of 125  PO4 in urban runoff is low: a mean value of 230 µg/L in suburban neighborhoods is reported in Debo & Reese (2002), and 100 µg/L is reported in CWP (2010). Unidentified sample preparation error compromised the quality of the final batch of TP samples, the results of which had to be excluded from analysis. Further uncertainty identified among laboratory blanks forced an increase of the reportable detection limit from 20 to 50 µg/L. TP concentrations for the remaining storm events are listed in Table 3.26 and plotted by event in Figure 3.29. In this case, the most conservative approach for dealing with values below the detection limit is to exclude them from calculation. No per-storm trends are observed, other than generally lower TP concentrations in the effluent than in the direct road runoff. Table 3.26 TP measured in road runoff and BMP effluent in Silver Valley  Event Road Runoff BMP Effluent % Reduction10-Feb 5211-Feb 77 < D.L.14-Feb < D.L.15-Feb 70 70 0%17-Feb 88 61 31%25-Feb 905 73 92%02-Mar 94 < D.L.04-Mar 57 164 -188%14-Mar 129 119 8%28-Mar 357 < D.L.29-Mar < D.L.29-Mar 58 53 9%30-Mar 86 < D.L.03-Apr 84 < D.L.Units are in µg/L. Detection Limit = 50 µg/L Total Phosphorus126   Figure 3.29 Road runoff and BMP effluent total phosphorus concentrations for 14 rainfall events Excluding values below the detection limit, a mean TP concentration of 182 µg/L (median 86 µg/L, range 57-905 µg/L) was found in road runoff, and 84 µg/L (median 70, range 52-164 µg/L) in BMP effluent. Road runoff concen-trations compare well to what can be expected in urban stormwater: Macdonald (2003) had reported a range of 120-180 µg/L in urban runoff; others reported a little higher. According to the USEPA, a TP reduction of 70-83% can be expected through bioretention structures with deep planting media (1999a); similar reductions are reported by Hin-man (2012) for deep soils in bioretention practices. The mean runoff and effluent concentrations, with error bars identifying 90% confidence intervals, are plotted in Figure 3.30 below. We can observe an average 54% reduction of TP through the infiltration swale (not statistically strong), confirming removal of TP as such a runoff treatment structure can be expected. 127   Figure 3.30 Mean total phosphorus concentrations & 90% CI for runoff and treated effluent samples 3.6.5 Organic carbon The total organic carbon (TOC) was calculated for the composited samples, with high variability among road runoff samples, and relatively steady, though slowly rising and falling, levels in the infiltration swale effluent. Despite the TOC concentration spike to 47 mg/L associated with the 25 February rain-on-snow event, runoff concentrations re-mained for a while below 2 mg/L in the effluent, as can be observed in Figure 3.31 below.  Figure 3.31 Total organic carbon concentrations for 22 rainfall events in Silver Valley Independence of effluent TOC from runoff TOC is evident. A slow increase in TOC is observed over time, but never as a direct result of an inflow spike. For comparison with the literature, mean runoff TOC, including data from the 128  25 February event, was found to be 6.7 mg/L (median 3.7 mg/L, range 1.3-47.2 mg/L); mean effluent TOC was 3.2 mg/L (median 3.0 mg/L, range 1.7-5.7 mg/L); see Figure 3.32 for these mean values compared, with 90% confi-dence intervals. This is similar to the value of 8.9 mg/L reported by Sickman et al. (2007) for urban runoff TOC concentrations in Sacramento. Winer (2000) reported roughly 70% removal of TOC through infiltration swales, a little better than the (statistically weak) 53% reductions found here. This average reduction is composited from all events; Figure 3.31, above, reveals that this mean reduction cannot be expected for any individual event.  Figure 3.32 Mean total organic carbon & 90% CI for road runoff and treated effluent 3.6.6 Chloride Consistently higher chloride concentrations in BMP effluent than that in road runoff was a surprising result. Many road runoff samples had concentrations at or near the detection limit of 1 mg/L (listed in Table 3.27 below). Assign-ing a value of one-half the detection limit to those falling below, mean chloride concentration in road runoff was calculated to be 33.2 mg/L (median 1.2 mg/L, range 0.5-472.5 mg/L), and increased through the BMP to mean con-centration of 84.4 mg/L (median 19.9 mg/L, range 7.7-523.5 mg/L). Likely de-icing efforts around the 25 February snow and then rain-on-snow event caused the chloride peak observed in Figure 3.33. Afterward, despite low chlo-ride in the road runoff, that in BMP effluent remained high, indicating a time lag as chloride percolated through the BMP, though remained an order of magnitude higher than the runoff levels, perhaps from a high backlog of chloride through the snow season, with chloride only running off at a meted rate. This correlates well to the reported findings of Li & Davis (2009), that elevated chloride levels persist in bioretention device effluent throughout the year, and that chloride peak concentrations from winter de-icing do not rapidly wash through the BMP. Runoff and effluent chloride concentration (of 640mg/L) were never exceeded, but the effluent did exceed the 24-hour to 30-day long-term limit of 120 mg/L for sixteen days. 129  Table 3.27 Chloride concentrations measured in each rainfall event   Figure 3.33 Chloride concentrations measured in road runoff and BMP effluent Event Road Runoff BMP Effluent10-Feb 9111-Feb 87 19414-Feb 16215-Feb 2 12617-Feb 3 8125-Feb 472 21602-Mar 11 52304-Mar 3 22614-Mar 8 3728-Mar < D.L. 1829-Mar 1929-Mar 1 1930-Mar 1 1803-Apr 3 2116-Apr < D.L. 2117-Apr < D.L. 1723-Apr 1 1324-Apr < D.L. 1508-May < D.L. 1326-May < D.L. 1028-May < D.L. 819-Jun 10All values in mg/L. Detection Limit = 1 mg/L.Chloride130  3.6.7 Hardness Measured total concentrations of calcium and magnesium and the calculated hardness for both runoff and effluent samples are listed in Table 3.28. These values of hardness were then used in CCME guideline calculations to deter-mine maximum limits of cadmium, copper, lead, and nickel (CCME 2014).  Table 3.28 Hardness calculated from calcium and magnesium for each sample  The values from Table 3.28 plotted over time in Figure 3.34 reveal similar results to that of TDS in effluent: a time lag to peak, and consistently higher hardness in effluent in the absence of great inputs of calcium and magnesium. Mean hardness in street runoff is 9.6 mg/L (median 4.9, range 0.7-69.6 mg/L), and in the BMP effluent is 25.9 mg/L (median 12.9, range 7.3-115.1 mg/L): a statistically significant increase (to 90% confidence), plotted in Figure 3.35.  Runoff values calculated here are lower than the 24 mg/L reported in Vancouver stormwater (Macdonald 2003) and the 80.3 mg/L (wet season) and 28.9 mg/L (dry season) reported in Langley road runoff (KWL 2010). Changes in hardness through infiltration swales or bioretention devices is not reported in the literature, so I cannot compare the significant increase identified here. Increased hardness in the effluent, however, reduces the toxicity (and raises the guideline maximum concentration) of metals in the runoff to aquatic biota.   Runoff EffluentCa Mg Hardness Ca Mg Hardness10-Feb 9.4 1.01 27.611-Feb 1.0 0.45 4.3 16.2 1.89 48.314-Feb 1.2 0.49 5.0 15.5 1.81 46.215-Feb 1.1 0.36 4.2 11.3 1.38 33.817-Feb 1.0 0.30 3.8 9.6 1.20 28.925-Feb 15.5 7.50 69.6 17.0 2.23 51.602-Mar 0.9 0.72 5.3 39.0 4.32 115.104-Mar < D.L. 0.37 1.5 19.1 2.18 56.614-Mar 2.7 0.68 9.5 4.9 0.63 14.728-Mar 7.2 3.40 32.0 4.1 0.54 12.329-Mar 3.7 0.50 11.429-Mar 2.8 0.32 8.3 4.3 0.56 13.030-Mar 2.0 0.14 5.6 4.2 0.54 12.803-Apr 2.0 0.19 5.7 4.1 0.55 12.616-Apr 1.4 0.32 4.9 3.8 0.52 11.717-Apr 1.0 0.23 3.3 3.7 0.52 11.523-Apr 1.3 0.36 4.8 3.2 0.43 9.824-Apr 1.1 0.18 3.5 2.4 0.34 7.508-May < D.L. 0.17 0.7 3.7 0.55 11.626-May 2.1 0.28 6.4 4.3 1.28 15.928-May 1.5 0.21 4.6 2.4 0.32 7.319-Jun 3.2 0.42 9.8All units in mg/L. Calcium detection limit = 0.9 mg/LMagnesium detection limit = 0.01 mg/LHardnessEvent131   Figure 3.34 Hardness calculated for road runoff and BMP effluent during 22 rainfall events  Figure 3.35 Mean hardness with 90% CI in road runoff and BMP effluent 3.6.8 Alkalinity Analysis of the composited samples revealed that the BMP effluent had significantly higher alkalinity than did road runoff (see Figure 3.36). All measured values exceeded the detection limit of 0.5 mg/L. Mean runoff alkalinity of 5.0 mg/L as CaCO3 (median 5.0, range 1.3-17.0 mg/L) was a little lower than the range of 8-53 mg/L reported by Hall & Anderson (1988), and is highly sensitive to the acidic conditions found in the soils of the bioswale (Nagpal et al. 2006).  132   Figure 3.36 Alkalinity measurements for 22 rainfall events in Silver Valley Mean alkalinity in BMP effluent was found to be 17.9 mg/L (median 18.0 mg/L, range 12.0-23.0 mg/L), a statisti-cally significant increase, at 90% confidence level, through the infiltration swale. Mean values and 90% confidence intervals are plotted in Figure 3.37. Such an increase of alkalinity through infiltration swales or bioretention struc-tures is not reported in the literature, but treated effluent with higher alkalinity offers greater resistance to changes in pH. If CO2 is off-gassing through, it would both decrease the pH and increase the alkalinity of the water (Reuss & Johnson 1985). Both of these effects are observed in the results of this research.  Figure 3.37 Mean alkalinity & 90% CI for road runoff and treated effluent  133  3.6.9 pH Since the pH of water can change over short periods of time, the pH of samples – picked up from the field hours af-ter collection – was not measured. Rather, at time of picking up the samples and changing bottles, pH was measured in the BMP effluent flow and in road runoff (if there was still runoff at time of pick-up). With this limitation, I made only six measurements of runoff pH, and 16 measurements of BMP effluent pH; values are listed by event in Table 3.29.  Table 3.29 Measured pH in road runoff and BMP effluent in Silver Valley  Mean pH of road runoff and BMP effluent was calculated as 8.00 and 6.85, respectively, a statistically significant reduction at the 90% confidence level, and is plotted in Figure 3.38. This pH of 8.0 in runoff only slightly exceeds the range of reported mean values of pH in urban runoff. The literature does not address acidification of waters pass-ing through this kind of BMP so the significant decrease in pH found here was not expected. An explanation may be that the low alkalinity of the runoff water provides little pH resistance to the acidic soil conditions (Nagpal et al. 2006). Date Runoff Effluent10-Feb11-Feb14-Feb15-Feb17-Feb 7.20 6.3725-Feb02-Mar 8.40 6.1904-Mar 8.80 6.6714-Mar 7.5828-Mar 7.5229-Mar 6.9129-Mar 7.0130-Mar 6.7903-Apr 8.36 7.0816-Apr 7.28 6.7117-Apr 8.01 6.6523-Apr 7.0124-Apr 6.5108-May 7.0526-May 7.0128-May 6.4919-JunpH134   Figure 3.38 Mean pH, with 90% CI, of runoff and treated effluent in Silver Valley 3.6.10 Metals As was identified through the literature review, runoff from urban surfaces frequently carries metals to streams at concentrations that are harmful to aquatic life. Bioretention devices and infiltration swales can be designed to reduce the concentration of these metals through various physical and bio-chemical processes; best practices can achieve significant reductions of the metals of concern. Below I discuss analyses to determine concentrations of copper, cad-mium, iron, lead, nickel, zinc, and aluminum, relating the findings to typical road runoff concentrations, reductions through bioretention devices similar to that installed in Silver Valley, and whether or not these reductions help meet CCME metals guidelines. 3.6.10.1 Copper Both dissolved and total concentrations of copper were measured in composited samples of both road runoff and BMP effluent. After laboratory analyses, a curious pattern was discovered among the data: that perhaps because of elevated copper concentrations in samples measured previous to mine, which lingered in the tubing of the ICP, the first ten samples of each batch were elevated, before copper concentrations washed through and appeared to drop to a steady state, disallowing confidence in representativeness of the data. As such, all measurements of copper were not used, and no conclusions can be made about copper concentrations in runoff or BMP effluent. 3.6.10.2 Cadmium Maximum concentration levels for the protection of aquatic life are a function of hardness, and of the few measured values of cadmium in the samples that were above the detection limit of 2 µg/L, none exceeded the CCME guideline that was just slightly higher than the detection limit. Table 3.30 lists the measured values (and identifies the meas-urements below detection limit). No statistical analyses were performed on this data, but the two road runoff sam-ples with cadmium concentrations above the detection limit are similar to that published in the literature. Guideline concentrations calculated for the fairly soft water were generally found to be below the detection limit, so just be-cause a sample’s concentration is below the detection limit does not necessarily indicate whether it exceeds the 135  short-term guidelines. Since the majority of the samples were below the detection limit, we can make no conclusions as to the BMP’s impact on cadmium removal. Table 3.30 Cadmium concentrations in Silver Valley runoff & treated effluent samples    Dissolved Total HardnessCCMEGuideline (acute)Exceeds? Dissolved Total HardnessCCMEGuideline (acute)Exceeds?10-Feb < D.L. 5.2 27.6 1.3 yes11-Feb < D.L. 2.5 4.3 0.3 yes < D.L. 3.2 48.3 2.1 yes14-Feb < D.L. < D.L. 5.0 0.3 ? < D.L. < D.L. 46.2 2.1 no15-Feb < D.L. < D.L. 4.2 0.3 ? < D.L. 2.1 33.8 1.6 yes17-Feb < D.L. < D.L. 3.8 0.3 ? < D.L. < D.L. 28.9 1.4 ?25-Feb < D.L. 2.5 69.6 2.9 no < D.L. 2.0 51.6 2.3 no02-Mar < D.L. < D.L. 5.3 0.3 ? < D.L. < D.L. 115.1 4.4 no04-Mar < D.L. < D.L. 1.5 0.1 ? < D.L. < D.L. 56.6 2.4 no14-Mar < D.L. < D.L. 9.5 0.6 ? < D.L. < D.L. 14.7 0.8 ?28-Mar < D.L. < D.L. 32.0 1.5 ? < D.L. < D.L. 12.3 0.7 ?29-Mar < D.L. < D.L. 11.4 0.6 ?29-Mar < D.L. < D.L. 8.3 0.5 ? < D.L. < D.L. 13.0 0.7 ?30-Mar < D.L. < D.L. 5.6 0.4 ? < D.L. < D.L. 12.8 0.7 ?03-Apr < D.L. < D.L. 5.7 0.4 ? < D.L. < D.L. 12.6 0.7 ?16-Apr < D.L. < D.L. 4.9 0.3 ? < D.L. < D.L. 11.7 0.7 ?17-Apr < D.L. < D.L. 3.3 0.2 ? < D.L. < D.L. 11.5 0.6 ?23-Apr < D.L. < D.L. 4.8 0.3 ? < D.L. < D.L. 9.8 0.6 ?24-Apr < D.L. < D.L. 3.5 0.2 ? < D.L. < D.L. 7.5 0.5 ?08-May < D.L. < D.L. 0.7 0.1 ? < D.L. < D.L. 11.6 0.7 ?26-May < D.L. < D.L. 6.4 0.4 ? < D.L. < D.L. 15.9 0.8 ?28-May < D.L. < D.L. 4.6 0.3 ? < D.L. < D.L. 7.3 0.4 ?19-Jun < D.L. < D.L. 9.8 0.6 ?All units in mg/L. Detection Limit = 0.003 mg/L All units in µg/L. Detection Limit = 2 µg/LCadmiumEffluentRunoffEvent136  3.6.10.3 Iron Dissolved and total iron concentrations measured road runoff and treated effluent samples are listed in Table 3.31 for each of the 22 storms sampled. CCME’s (2014) Guidelines for the Protection of Aquatic Life set maximum iron concentration at 0.3 mg/L for freshwater environments. All samples of direct road runoff exceeded this guideline, and runoff from the rain-on-snow event on 25 February exceeded this guideline by 56 times. If urban runoff at these iron concentrations were discharged into an urban creek (which, in this case, it is not), and making up most of the flow, it would be at levels acutely toxic to the aquatic life. Effluent from the BMP, however, exceeded the guideline only 36% of the time, and with much lower magnitude of exceedance. See Figure 3.39 below for this comparison. Table 3.31 Iron concentrations in Silver Valley runoff & treated effluent samples  Event Dissolved TotalCCMEGuidelineExceeds? Dissolved TotalCCMEGuidelineExceeds?10-Feb < D.L. 0.421 0.300 yes11-Feb < D.L. 1.070 0.300 yes < D.L. 0.121 0.300 no14-Feb < D.L. 1.517 0.300 yes 0.053 0.138 0.300 no15-Feb < D.L. 1.124 0.300 yes < D.L. 0.122 0.300 no17-Feb < D.L. 0.915 0.300 yes 0.055 0.858 0.300 yes25-Feb < D.L. 17.059 0.300 yes < D.L. 0.079 0.300 no02-Mar < D.L. 1.863 0.300 yes < D.L. 0.151 0.300 no04-Mar < D.L. 0.961 0.300 yes < D.L. 0.237 0.300 no14-Mar < D.L. 1.753 0.300 yes 0.225 0.287 0.300 no28-Mar < D.L. 9.396 0.300 yes 0.069 0.271 0.300 no29-Mar 0.068 0.200 0.300 no29-Mar < D.L. 0.911 0.300 yes 0.086 0.312 0.300 yes30-Mar < D.L. 0.367 0.300 yes 0.073 0.224 0.300 no03-Apr < D.L. 0.529 0.300 yes 0.069 0.203 0.300 no16-Apr < D.L. 1.007 0.300 yes 0.100 0.339 0.300 yes17-Apr < D.L. 0.674 0.300 yes 0.069 0.330 0.300 yes23-Apr < D.L. 1.096 0.300 yes 0.069 0.185 0.300 no24-Apr < D.L. 0.407 0.300 yes 0.061 0.158 0.300 no08-May < D.L. 0.845 0.300 yes 0.068 0.473 0.300 yes26-May < D.L. 0.765 0.300 yes 0.144 2.461 0.300 yes28-May < D.L. 2.481 0.300 yes 0.064 0.255 0.300 no19-Jun 0.179 0.338 0.300 yesAll units in mg/L. Detection Limit = 0.001 mg/L All units in mg/L. Detection Limit = 0.05 mg/LEffluentIronRunoff137   Figure 3.39 Total iron concentrations in runoff and treated effluent, compared to CCME guidelines Mean road runoff total iron concentration was 2.36 mg/L (median 1.01 mg/L, range 0.37-17.06 mg/L), similar to the concentration and (wide) range reported in the literature. With mean effluent concentration of 0.37 mg/L (median 0.25 mg/L, range 0.08-2.46 mg/L), an 82% reduction is observed, statistically significant at a 90% confidence level (plotted in Figure 3.40 below), is similar to total iron removal rates reported by Geosyntec & Wright (2012) of 43% in bioswales and up to 100% in bioretention devices. Dissolved iron fractions made up less than 2% of the total in road runoff, but increased to 29% in BMP effluent, indicating solubilization of iron through the BMP, perhaps a function of the reduced pH in the bioswale. Significant total iron reductions through the BMP brought concentra-tions down to below guideline maximum value of 0.3 mg/L. Fully 100% of road runoff samples exceeded the guide-line; only one-third of the effluent samples exceeded, and those only by small margins; compare concentrations to the dashed line in Figure 3.39 above. 138   Figure 3.40 Calculated mean total iron concentration and 90% CI for runoff and BMP effluent 3.6.10.4 Lead Measurements were made of both dissolved and total lead in all stormwater samples. All values were very low, and close to the ICP’s detection limit of 0.01 mg/L. Given uncertainties discovered among laboratory blanks, this detec-tion limit was raised slightly to 0.03 mg/L, above which only one sample exceeded. Since guideline maximum lead concentration is a function of hardness, and the runoff is fairly soft, guideline concentrations fell fully two orders of magnitude below the detection limits. Thus, even values identified as below the detection limit cannot be determined to be below the CCME guideline maximum concentration. No attempt can be made to quantify reductions of lead through the infiltration swale. 3.6.10.5 Nickel Concentrations of nickel were low in the stormwater sampled. Thirteen of the 19 runoff samples and 19 of the 22 BMP effluent samples were below detection limit of 0.01 mg/L (raised slightly from 0.007 mg/L driven by uncer-tainty discovered among laboratory blanks). CCME guidelines are the lower value between 0.025 mg/L and a calcu-lation that is a function of hardness. Listed in Table 3.32, five of 19 road runoff samples exceeded the maximum concentration; two of 22 treated samples exceeded the maximum concentrations. 139  Table 3.32 Nickel concentrations in Silver Valley runoff & treated effluent samples  Since a high proportion of the samples were below the detection limit, I excluded them from statistical calculations. Among the few samples compared, a statistically weak 29% reduction in nickel concentration is achieved from mean total nickel concentration in runoff of 0.051 mg/L (median 0.037 mg/L, range 0.013-0.143 mg/L) to mean nickel concentration in effluent of 0.036 mg/L (median 0.036 mg/L, range 0.026 – 0.046 mg/L). Though difficult to make comparisons with so little usable data, this BMP shows somewhat lower nickel reductions than the 66% (or so) predicted in the literature. Perhaps with a longer period of observation (say, over an entire year) and more storms, a different picture would emerge. 3.6.10.6 Zinc Nearly all measurements of zinc in both road runoff and BMP effluent were below the detection limit. Given the relative position of the remaining few samples’ concentrations to the detection limit, it is impossible to make conclu-sions about road runoff or effluent zinc concentrations, let alone reductions through the infiltration swale, other than that concentrations were generally low. Macdonald (2003) had reported a mean zinc concentration of 73 µg/L (very close to the detection limit of 70 µg/L); KWL (2010) reported zinc concentrations also below this detection limit, so Dissolved Total HardnessCCMEGuidelineExceeds? Dissolved Total HardnessCCMEGuidelineExceeds?10-Feb < D.L. < D.L. 27.6 0.025 no11-Feb < D.L. < D.L. 4.3 0.009 no < D.L. < D.L. 48.3 0.025 no14-Feb < D.L. < D.L. 5.0 0.010 no < D.L. < D.L. 46.2 0.025 no15-Feb < D.L. < D.L. 4.2 0.009 no < D.L. < D.L. 33.8 0.025 no17-Feb < D.L. < D.L. 3.8 0.008 no < D.L. 0.046 28.9 0.025 yes25-Feb < D.L. 0.013 69.6 0.025 no < D.L. < D.L. 51.6 0.025 no02-Mar < D.L. < D.L. 5.3 0.010 no < D.L. < D.L. 115.1 0.025 no04-Mar < D.L. < D.L. 1.5 0.004 no < D.L. < D.L. 56.6 0.025 no14-Mar < D.L. 0.143 9.5 0.016 yes < D.L. < D.L. 14.7 0.022 no28-Mar < D.L. < D.L. 32.0 0.025 no < D.L. < D.L. 12.3 0.019 no29-Mar < D.L. < D.L. 11.4 0.018 no29-Mar < D.L. < D.L. 8.3 0.014 no < D.L. < D.L. 13.0 0.020 no30-Mar < D.L. < D.L. 5.6 0.011 no < D.L. < D.L. 12.8 0.020 no03-Apr < D.L. < D.L. 5.7 0.011 no < D.L. < D.L. 12.6 0.020 no16-Apr < D.L. 0.059 4.9 0.010 yes < D.L. < D.L. 11.7 0.019 no17-Apr < D.L. < D.L. 3.3 0.007 no < D.L. < D.L. 11.5 0.018 no23-Apr < D.L. < D.L. 4.8 0.009 no < D.L. < D.L. 9.8 0.016 no24-Apr < D.L. < D.L. 3.5 0.008 no < D.L. < D.L. 7.5 0.013 no08-May < D.L. 0.034 0.7 0.002 yes < D.L. < D.L. 11.6 0.019 no26-May < D.L. 0.016 6.4 0.012 yes < D.L. 0.026 15.9 0.024 yes28-May < D.L. 0.040 4.6 0.009 yes < D.L. < D.L. 7.3 0.013 no19-Jun < D.L. < D.L. 9.8 0.016 noAll units in mg/L. Detection Limit = 0.01 mg/L All units in mg/L. Detection Limit = 0.001 mg/LNickelEffluentEventRunoff140  the low concentrations are not surprising. Guideline maximum concentration is set at 30 µg/L, somewhat below the laboratory’s detection limit. 3.6.10.7 Aluminum Also for each event, dissolved and total aluminum concentrations in samples of road runoff and BMP effluent were measured. Results are tabulated in Table 3.33 below. As with other metals, it was necessary to raise the detection limit (in this case from 0.018 mg/L to 0.1 mg/L) given uncertainty observed among laboratory blanks at lower con-centrations). This precluded examination of most dissolved aluminum concentrations in runoff samples, but did not affect total aluminum concentrations that were well above this limit.  CCME guidelines for maximum aluminum concentration in aquatic environments is pH-dependent: if below pH of 6.5, the guideline is 0.005 mg/L; if above, it is 0.1 mg/L. As few values of pH were measured in the field, the guideline concentration was determined by using average pH (8.00 in runoff, 6.85 in effluent, both of which are above the critical value of 6.5). Regardless of this estimation, all samples exceeded even the least conservative guideline of 0.1 mg/L, as in Table 3.33 below. CCME guideline values in bold are based on estimated values of pH. Table 3.33 Aluminum concentrations in Silver Valley runoff & treated effluent samples  Dissolved Total pHCCMEGuidelineExceeds? Dissolved Total pHCCMEGuidelineExceeds?10-Feb < D.L. 0.365 0.100 yes11-Feb < D.L. 1.257 0.100 yes < D.L. 0.208 0.100 yes14-Feb < D.L. 1.772 0.100 yes < D.L. 0.307 0.100 yes15-Feb < D.L. 5.748 0.100 yes < D.L. 0.169 0.100 yes17-Feb < D.L. 1.032 7.20 0.100 yes 0.128 4.312 6.37 0.005 yes25-Feb < D.L. 20.146 0.100 yes < D.L. 0.184 0.100 yes02-Mar < D.L. 2.040 8.40 0.100 yes < D.L. 0.227 6.19 0.005 yes04-Mar < D.L. 1.134 8.80 0.100 yes < D.L. 0.215 6.67 0.100 yes14-Mar < D.L. 2.017 0.100 yes 0.444 0.503 7.58 0.100 yes28-Mar < D.L. 11.151 0.100 yes 0.172 0.499 7.52 0.100 yes29-Mar 0.178 0.394 6.91 0.100 yes29-Mar < D.L. 0.984 0.100 yes 0.190 0.494 7.01 0.100 yes30-Mar < D.L. 0.417 0.100 yes 0.195 0.413 6.79 0.100 yes03-Apr < D.L. 0.570 8.36 0.100 yes 0.160 0.409 7.08 0.100 yes16-Ap < D.L. 1.140 7.28 0.100 yes 0.242 0.606 6.71 0.100 yes17-Apr < D.L. 0.870 8.01 0.100 yes 0.174 0.655 6.65 0.100 yes23-A r < D.L. 1.362 0.100 yes 0.168 0.367 7.01 0.100 yes24-Apr 0.117 0.694 0.100 yes 0.172 0.382 6.51 0.100 yes08-May < D.L. 0.581 0.100 yes 0.177 0.662 7.05 0.100 yes26-May < D.L. 0.774 0.100 yes 0.315 1.362 7.01 0.100 yes28-May < D.L. 1.580 0.100 yes 0.196 0.473 6.49 0.005 yes19-Jun 0.330 0.539 0.100 yesAll units in mg/L. Detection Limit = 0.02 mg/L All units in mg/L. Detection Limit = 0.1 mg/LRunoff EffluentAluminumEvent141  Observable in Figure 3.41 below is the great spread of aluminum concentrations in direct road runoff, strongly skewing concentrations. Mean total aluminum concentration is 2.91 mg/L (median 1.14 mg/L, range 0.42-20.15 mg/L). In the treated effluent, total aluminum concentration varies considerably less: a mean of 0.63 mg/L (median 0.41 mg/L, range 0.17-4.31 mg/L).  Figure 3.41 Total aluminum concentrations in runoff and treated effluent (log scale), and CCME guideline Mean concentrations of total aluminum are compared in Figure 3.42, with error bars indicating the 90% confidence limits: a significant aluminum concentration reduction of 79% through the infiltration swale, though outlet concen-tration is still well above guideline maximum concentration (the dashed line in Figure 3.41 above).  Figure 3.42 Calculated mean total aluminum concentrations & 90% CI for runoff and BMP effluent 142  Urban runoff aluminum concentrations are typically high like this: Tuccillo (2006) reported a range of 0.1-7.1 mg/L; Macdonald (2003) reported a mean of 0.6 mg/L, and KWL (2010) reported a mean of 0.8mg/L in dry season runoff. The high reduction of aluminum concentration exceeds the mean reduction of 17% reported by Glass and Bissouma (2005). 3.6.11 Temperature Using the same measurement technique as that for pH – in fact, the same instrument – and the near impossibility of obtaining a temperature reading at time of sample collection (given logistical complexity afforded by the site), only six measurements were taken of direct road runoff temperature, and sixteen of infiltration swale effluent. Table 3.34 below lists these measurements; the gradual increase of BMP effluent temperature from mid-February to late May is observable in Figure 3.43. Insufficient paired data points were collected to draw conclusions about the benefits that the BMP offers for temperature buffering in warm weather. Table 3.34 Runoff and treated effluent temperature in Silver Valley  Date Runoff Effluent10-Feb11-Feb14-Feb15-Feb17-Feb 4.8 5.225-Feb02-Mar 2.1 4.704-Mar 6.6 5.414-Mar 8.428-Mar 8.329-Mar 8.629-Mar 8.430-Mar 8.103-Apr 9.2 8.716-Apr 10.9 9.817-Apr 14.3 10.123-Apr 10.524-Apr 10.208-May 11.726-May 13.628-May 13.519-JunUnits are degrees Celsius.Temperature143   Figure 3.43 Runoff and treated effluent temperatures over spring 2014 in Silver Valley 3.6.12 Performance evaluation Water quality of Foreman Drive runoff is fairly similar to that expected in the literature, with some exceptions. The range of mean concentrations from the literature, and the mean Foreman Drive runoff concentrations, are listed in Table 3.35. TSS, TDS, TP, Cl-, hardness, and Fe concentrations all fall within the range of mean values reported. Since traffic density has a direct relationship to particle loading, it would be no surprise to see some of the concen-trations in the stormwater runoff from periurban Foreman Drive to be on the low end of the range. With only a cou-ple of notable examples (see Table 3.35), this is what we observe here. TKN, NOX, NH3, TOC, alkalinity, and Al concentrations in Foreman Drive stormwater runoff were all slightly lower than the typical range of mean values in the literature, as are TOC, alkalinity, and aluminum. Only runoff pH and nickel exceeded the range of mean values reported in the literature. Quantification of the water quality benefits of the infiltration bioswale along Foreman Drive is also summarized in Table 3.35.  Trends include a strong reduction in TSS but increased TDS, a general decrease in N & P concentra-tions, increased hardness, alkalinity, and chloride, 1.2 unit drop in pH, statistically weak copper increase, statisti-cally weak but notable reductions in lead and nickel, and strong reductions in iron and aluminum. No reductions were noted in the very low cadmium and zinc concentrations in the urban runoff. 144  Table 3.35 Summary water quality benefits of the Foreman Drive infiltration swale  It was outlined in section 3.4.3 that the principal mechanisms of water quality treatment through bioswales are set-tling/sedimentation, filtration, adsorption, infiltration, phytoremediation, physical plant resistance, volatilization, thermal tempering, and solubilization (Davis et al. 2003, Hinman 2012, Chuan et al. 1996, Chen & Lin 2001). To-gether they form complex cycles within the bioswale soils, microorganisms, and vegetation that can be summarized as a time- and chemistry-dependent mass balance. Chemical conditions in the soil and street runoff determine solu-bility of metals, so those originally adsorbed to particles may solubilize and wash out. Uptake of bioavailable nutri-ents and metals by plants is subtracted from the bioswale load and stored in plant biomass in the short-term, but is permanently removed if it is cropped or harvested (Gottschal 2005). Through volatilization, some substances toxic in the environment are instead released to the atmosphere. But the main mechanism is filtration: the straining out of particles suspended in the runoff, and on its own does little to remove dissolved substances. The inevitable build-up Parameter UnitsMean (med.) in RunoffMean (med.)  in Effluent TrendSignif.at 90%?TSS mg/L 44 to 809 37% to 90% 83.9 (35.5) 5.3 (1.6) 92% ↓ yesTDS mg/L 19 to 120 ? 36.1 (5.7) 103.0 (43.0) 185% ↑ noTKN mg/L 1.73 to 2.5 36% to 80% 1.2 (0.8) - 38-63% ↓ noNOXmg/L 0.237 to 13.6 -194% to 23% 0.196 (0.127) 0.356 (0.375) 81% ↑ yesNH3mg/L 0.18 to 0.8 54% to 86% 0.153 (0.146) 90% ↓ DL - noTP µg/L 140 to 620 -1000% to 83% 182 (86) 84 (49) 63-70% ↓ noPO4µg/L 100 to 210 ? - - - -TOC mg/L 11.9 to 33.7 ~70% 6.7 (3.7) 3.2 (3.0) 53% ↓ noCl-mg/L 0.55 to 75.3 bleed-through 33.2 (1.2) 84.4 (19.9) 154% ↑ noHardness mg/L 8.2 to 80.3 ? 9.6 (4.9) 26 (12.9) 170% ↑ yesAlkalinity mg/L 28 to 38.2 ? 5.0 (5.0) 17.9 (18.0) 258% ↑ yespH 7 to 7.9 ? 8.01 (8.19) 6.85 (6.85) 15% ↓ yesCu µg/L 5.85 to 65.9 -50% to 97% - - - -Cd µg/L 0.035 to 2.3 5% to 43% - - - -Fe µg/L 61 to 88000 -100% to 43% 2,360 (1,010) 370 (250) 84% ↓ yesPb µg/L 0.3 to 278 21% to 98% - - - -Ni µg/L 12 to 22.6 ~66% 51 (37) 36 (36) 29% ↓ noZn µg/L 6.5 to 275 30% to 100% - - - -Al µg/L 26.5 to 1500 ~17% 2,910 (1,140) 625 (172) 79% ↓ yesRange of Means (Literature)Reductions (Literature)145  of pollutants in the bioswale can be interrupted by regular maintenance, including removal of accumulated sedi-ments, and cropping of above-ground biomass to physically remove both organic and inorganic pollutants, and to renew the bioswale’s treatment capacity. Since the removal of pollutants from urban runoff is essentially a mass-balance problem, with sediments, metals, nutrients, and chemicals accumulating in the bioswale matrix or being taken up in biological processes, there is una-voidably an increase in the concentration and mass loading of these contaminants within the bioswale. Eventually however, the capacity of the system to continue filtering, holding, absorbing, and adsorbing the various contami-nants will be overwhelmed, and a long-term steady-state will be reached when leach-through is equivalent to that received. This leads to the need to determine an appropriately progressive maintenance schedule. Further, long-term research projects on the full life-cycle of a bioswale could help to determine what and how fre-quently to rejuvenate the removal capacity of the swales. Future researchers might ask, how frequently should accu-mulated sediments be removed? How frequently should the soils be removed and replaced to renew the ion ex-change capacity of the matrix? What should be done with the removed soils? What is the heavy metal concentra-tion/content in plant clippings annually harvested from bioswales, and what should be done with the clippings? As observed in the results of this research during a single season several years after installation, there was variability in the removal rate of each parameter. How does that change over time for each parameter? Does suspended sediment removal remain high, while the removal of iron diminishes over time?  Academic research or long-term monitoring of a site can eventually contribute to the development of curves of pol-lutant removal efficiencies over time, and, for a given construction standard and traffic loading, the ecological bene-fit of renewal can be weighed against the effort’s economic costs. Several years after construction, this mature bioswale is successfully achieving concentration reductions of TSS (92%), TKN (38-63%), TP (63-70%), TOC (53%), Fe (84%), and Al (79%), though only that for TSS, Fe, and Al are statistically significant reductions. Conversely, street runoff is measured to have an increased concentration of TDS (185%), NOX (81%), and Cl- (154%). Increases of hardness (170%) and alkalinity (258%) are beneficial to the runoff water quality. Overall, it is apparent that this BMP is still providing water quality treatment to a degree that is protecting the receiving water body, especially as only the first facility in a treatment train. Water quality measurements had never been performed on the progressive infiltration swale designed and installed in the Silver Valley neighborhood. In addition to the value of determining the water quality benefits of this specific LID-BMP, these results provide an additional unique data point to the body of knowledge around the treatment per-formance of this type of device. That the site is mature, nine years after construction, the water quality data found in this study can be placed near the perimeter of that offered in the literature. That design details are known, including depth of rock trench and depth of the growing/filtering media, affords an indispensable point of reference for using this data to predict the pollutant reduction through similarly-designed infiltration swales in the Lower Fraser Valley. 146  3.7 Conclusion This study was designed to determine whether the infiltration swale along Foreman Drive, designed and installed to the best practices of the day for both infiltration and treatment of road runoff, is removing contaminants of concern according to the levels predicted in the literature. It was hypothesized that runoff water quality is significantly im-proved by treatment through the infiltration swale, evidenced in lower concentrations of the urban pollutants meas-ured. It can be concluded that this hypothesis is correct. Experimentation has demonstrated a statistically significant 92% reduction in TSS, an important water quality parameter. Concentrations of the heavy metals Fe, Ni, and Al (the only metals for which reliable comparisons could be made in this study) all had lower concentrations in the effluent than in the road runoff. Very low alkalinity and hardness were increased through the bioswale, reducing the discharged water’s sensitivity to acid inputs and decreasing the toxicity of metals to aquatic life. The pH was reduced through the bioswale, but only to just below neutral, and still within healthy limits. During the winter/spring sampling, TDS and Cl- concentrations were found to be generally higher in BMP effluent than in road runoff (which does not pose a problem to aquatic life in this case), both of which indicating a similar time lag as concentration spikes made their way through the bioswale but remained at elevated concentrations. A full year of water quality sampling, coupled with flow monitoring, would be necessary to determine exact loadings and whether there are seasonal periods of build-up and wash-out of dissolved molecules. Finally, concentrations of NOX in runoff were found to increase through the bioswale, which indicates insufficient denitrification (either lack of saturated zone or insufficient resi-dence time in the bioswale). Compared to the published performance values of similar bioswales, this mature bioswale in an active community is performing as it was designed. As a stand-alone BMP, it would almost be treating the stormwater runoff from the residential street to guideline concentrations. As part of the recommended treatment train of best management prac-tices, this bioswale makes a significant contribution to the water quality treatment of urban runoff.  147  Chapter 4: Conclusions & recommendations This research effort into the hydrologic and water quality benefits of low impact development best management practices sought to fill a local knowledge gap regarding the performance of the latest generation of these designs, installed in some areas throughout entire neighborhoods. Two unique research projects were performed. 4.1 Source controls for hydrologic response The first was a desktop research effort (supported in the field with ground-truthing missions and the installation of rain gauges) using high-resolution historical rainfall and urban runoff monitoring data enabled comparison of peak flows and cumulative event runoff volumes between a catchment with a progressive suite of on-lot source controls, and a control catchment built with conventional urban drainage infrastructure (and no intentional source controls). A literature review summarized the impacts of urbanization to the hydrology of a watershed, and subsequent impacts to aquatic life. To ensure that the paired study could appropriately be made, considerable effort was made to com-pare catchment area, contributions to total impervious area, rainfall, soils and hydraulic conductivity, time of pipe flow to outlet, and other catchment parameters. Most parameters demonstrated that the catchments were well paired, but the distance apart (and subsequent difference in timing of rainfall sometimes) precluded use of the historical rain gauge data to allow a comparison of time to peak flow between the two catchments. This research effort had two principal objectives: to determine whether the suite of lot-only source controls (discon-necting roofs, or 31% of the catchment area, from the drainage network and infiltrating precipitation on-site) (a) re-duces peak flows in proportion to the impervious area reduction, and (b) reduces runoff volumes in proportion to the impervious area reduction. Through a multiple-step process that involved identifying 128 unique rainfall events between 2011 and 2013, nor-malizing the runoff flow by catchment area, removing groundwater and interflow, and plotting storm-sourced runoff from the paired catchments for each of the storms, mean reductions of both peak flow and cumulative runoff volume were calculated. The mean peak flow reduction of 41% (± 3% at the 95% confidence level) was found to be statistically indistin-guishable from the 39% reduction in Routley neighborhood impervious area, attributed to the installation of the BMPs. This demonstrates a direct correlation between reduction of impervious area and reduction of peak flow. The mean cumulative runoff volume reduction of 30% (± 2% at the 95% confidence level) confirms a similar but slightly less pronounced relationship with the effective reduction of impervious area. Together, these indicate that the BMPs are working, and that the efforts made to disconnect roof surfaces from the drainage network are paying off.  Calculations of infiltration capacity of the soils and rock trenches, along with a sensitivity analysis on the site varia-bles, confirm that they should together infiltrate rainfall that falls on the lots up to the design storm of 45 mm. The reduced performance of the source controls at infiltrating rainfall may be attributable to subsequent re-connection of originally disconnected roof downspouts, sedimentation of the rock trenches, or compaction of the amended soil’s depth. That the more impervious Routley neighborhood, with its higher population density, is behaving hydrau-lically ‘less impervious’ than the Langley Meadows catchment, is a testimony to the BMPs’ effective disconnection 148  of the roof area from the storm sewer network. This suite of on-lot source controls, through the supposed disconnec-tion of 100% of the roofs in the Routley catchment, achieves a hydraulic benefit of 75% (total event runoff) and 100% (peak reduction) of this effort. 4.2 Source controls for improved runoff water quality The second research effort involved the installation of automated sampling equipment in the idyllic hamlet of Silver Valley, to measure road runoff water quality, and improvements through an eight year old infiltration swale. The central objective of this research effort was to determine to what degree the infiltration swale along Foreman Drive is removing contaminants, as compared to how it is expected to, as a BMP comparable with those described and re-ported on in the literature. It was hypothesized that the runoff water is significantly improved by treatment through the infiltration swale, and would be observable in reduced concentrations of the pollutants in the effluent. A review of the literature traced the sources and typical concentrations of several urban pollutants of concern, and outlined the expected treatment performance for the many parameters. To determine treatment performance, two autosamplers with external water level sensors/controllers and telemetry-equipped dataloggers, housed in secure boxes, served as a stand-alone sampling system that sent me a notification when rainfall caused enough runoff to trigger the sensor and begin sampling. The samples were composited by sam-pler – providing a single data point each for road runoff and treated effluent – for the several water quality parame-ters analyzed. Concentrations of sediments (TSS and TDS), nutrients (TKN, NOX, NH3, TP, and PO4), TOC, chlo-ride, hardness, alkalinity, and metals (Cu, Cd, Fe, Pb, Ni, Zn, and Al) were measured in the Environmental Labora-tory at UBC. Among the parameters for which a reduction was observed, relative strength of reductions were TSS > Fe > Al > TP > TOC > TKN > Ni > pH; underlined parameters expressed a statistically significant reduction at the 95% confi-dence level. The same analysis revealed that the concentration of several parameters increased in the effluent, with relative strength alkalinity > TDS > hardness > chloride > NOX; underlined parameters expressed a statistically sig-nificant increase. That alkalinity and hardness increased is considered beneficial; water quality guidelines do not regulate TDS itself; and the increase in chloride concentrations through the spring may be a time-lag expression of a straightforward chloride mass-balance that works itself out over the course of a year’s rainfall. NOX concentrations were found to significantly increase, but effluent levels were still below both acute and long term water quality guideline maximum concentrations. Table 3.35 lists a full summary of the mean (and median) road runoff and BMP effluent concentrations, trends and significance, and how they relate to meeting water quality guidelines. The 92% TSS reduction is excellent, achieving a mean effluent concentration of 5.3 mg/L and median of 1.6 mg/L. Measured concentrations of copper, cadmium, lead, and zinc were all below laboratory detection limits. But because copper and lead detection limits are above guideline concentrations, it is impossible to determine the relationship to the guidelines. Significantly increased water hardness through the BMP helps to mitigate the impact of some of the elevated heavy metals concentrations since toxicity of copper, cadmium, nickel, and lead decreases with increasing hardness. 149  From these water quality analyses of the Foreman Drive infiltration swale as a stand-alone facility, it can be con-cluded that it is largely performing as expected when compared with both the published literature on water quality treatment through bioretention devices, and with water quality guidelines designed to protect the aquatic life in An-derson Creek and the Fraser River. However, the bioswale is only the first facility in a treatment train of BMPs, and further tempering of both flow and water quality occur in the parkette and the detention pond before remaining run-off is discharged to the environment at a controlled rate. It is therefore recommended that municipalities encourage the installation of such robustly-constructed road runoff practices – as the first line of defense to infiltrate rainfall to groundwater, reduce hydrologic impacts to the receiving water bodies, and reduce the burden of contamination to the aquatic life in the streams. Finally, it is recommended to perform further, long-term research on the fate of contaminants in mature, in-situ bioswales, to determine the rate at which the pollutant removal capacity is reduced, to inform the development of progressive maintenance and renewal schedules to ensure that the water quality treatment feature continues to contribute aesthetically to the livability of the neighborhood.   150  Bibliography Alberta Environmental Protection. (1996). Draft Alberta water quality guideline for the protection of freshwater aquatic life: copper. Standards and Guidelines Branch, Environmental Assessment Division, Environmental Reg-ulatory Service. 129pp. APHA, AWWA, WEF. (2005). Standard methods for the examination of water and wastewater, 21st edition. Ameri-can Public Health Association. Washington, Dc. Arnold Jr, C. L., & Gibbons, C. J. (1996). Impervious surface coverage: the emergence of a key environmental indi-cator. Journal of the American planning Association, 62(2), 243-258. Arnold, C. L., Boison, P. J., & Patton, P. C. (1982). Sawmill Brook: an example of rapid geomorphic change related to urbanization. The Journal of Geology, 155-166. Baker, J. P., & Schofield, C. L. (1982). Aluminum toxicity to fish in acidic waters. In Long-Range Transport of Air-borne Pollutants (pp. 289-309). Springer Netherlands. Barton, B. A. (1977). Short‐term effects of highway construction on the limnology of a small stream in southern On-tario. Freshwater Biology, 7(2), 99-108. Battiata, J., Collins, K., Hirschman, D., & Hoffmann, G. (2010). The runoff reduction method. Journal of Contem-porary Water Research & Education,146(1), 11-21. BC Environment. (1992). Urban runoff quality control guidelines for the Province of British Columbia. Prepared by British Columbia Research Corporation for Municipal Waste Branch of Environmental Protection Division of B.C. Environment. 138pp. BC Environment. (2014). Water Quality Guidelines (Criteria) Reports. Environmental Protection Division, Ministry of the Environment, Government of British Columbia. Website accessed 07 Septemer 2014 at http://www.env.gov.bc.ca/wat/wq/wq_guidelines.html. Beck, H. J., & Birch, G. F. (2012). Spatial and temporal variance of metal and suspended solids relationships in ur-ban stormwater—implications for monitoring. Water, Air, & Soil Pollution, 223(3), 1005-1015. Bertrand-Krajewski, J. L., Chebbo, G., & Saget, A. (1998). Distribution of pollutant mass vs volume in stormwater discharges and the first flush phenomenon. Water Research, 32(8), 2341-2356. Bill C-38: An Act to implement certain provisions of the budget tabled in Parliament on March 29, 2012 and other measures. 1st Reading, 29 March, 2012, 41st Parliament, 1st Session, 2012. Ottawa: Public Works and Govern-ment Services Canada, 2012. Bill C-45: A second Act to implement certain provisions of the budget tabled in Parliament on March 29, 2012 and other measures. 1st Reading, 18 October, 2012, 41st Parliament, 1st Session, 2012. Ottawa: Public Works and Government Services Canada, 2012. 151  Blecken, G. T., Zinger, Y., Deletic, A., Fletcher, T. D., & Viklander, M. (2009). Impact of a submerged zone and a carbon source on heavy metal removal in stormwater biofilters. Ecological Engineering, 35(5), 769-778.  Blecken, G., Zinger, Y., Muthanna, T., Deletic, A., Fletcher, T., & Viklander, M. (2007). The influence of tempera-ture on nutrient treatment efficiency in stormwater biofilter systems. Water Science & Technology, 56(10), 83-91.   Booth, D. B. (1991). Urbanization and the natural drainage system--impacts, solutions, and prognoses. The North-west Environmental Journal. 7(1), 93-118.  Brabec, E., Schulte, S., & Richards, P. L. (2002). Impervious surfaces and water quality: a review of current litera-ture and its implications for watershed planning. Journal of planning literature, 16(4), 499-514.  Bradford, A., & Gharabaghi, B. (2004). Evolution of Ontario's stormwater management planning and design guid-ance. Water quality research journal of Canada, 39(4), 343-355.   Brombach, H., Weiss, G., & Fuchs, S. (2005). A new database on urban runoff pollution: comparison of separate and combined sewer systems. Water Science & Technology, 51(02), 119-128. Burton, G. A., & Pitt, R. E. (2002). Stormwater effects handbook. Lewis Publ., Boca Raton, FL. Canadian Council of Ministers of the Environment. 2009. Canada-wide strategy for the management of municipal wastewater effluent. Whitehorse, YT. Canadian Council of Ministers of the Environment. 2014. Canadian water quality guidelines for the protection of Aquatic Life. Accessed January 2014 online at http://ceqg-rcqe.ccme.ca/. Summary table available at http://st-ts.ccme.ca/. Carline, R. F., & McCullough, B. J. (2003). Effects of floods on brook trout populations in the Monongahela Na-tional Forest, West Virginia. Transactions of the American Fisheries Society, 132(5), 1014-1020. Center for Watershed Protection. (2010). New York State stormwater management design manual. Prepared by the Center for Watershed Protection for New York State Department of Environmental Conservation. Albany, NY. 642pp. CH2M Hill Canada Limited. (2002). Effectiveness of Stormwater Source Control. Prepared for The Greater Van-couver Regional District. Burnaby, BC. Available online at http://www.mapleridge.ca/-Docu-mentCenter/View/830. Chilibeck, B, and Sterling, M. 2001. Urban Stormwater Guidelines and Best Management Practices for Protection of Fish and Fish Habitat - Draft. Vancouver: Department of Fisheries and Oceans. 5 pages. Chilibeck, B., G. Chislett, and G. Norris. 1993. Land Development Guidelines for the Protection of Aquatic Habitat. Department of Fisheries and Oceans, Habitat Management Division, Ottawa, Canada. Chen, S. Y., & J.G. Lin. (2001). Bioleaching of heavy metals from sediment: significance of pH. Chemosphere, 44(5), 1093-1102. 152  Chow, V. T. (1964). Handbook of applied hydrology. McGraw-Hill, New York. Chuan, M. C., G.Y. Shu, & J.C. Liu. (1996). Solubility of heavy metals in a contaminated soil: effects of redox po-tential and pH. Water, Air, and Soil Pollution, 90(3-4), 543-556. City of Burnaby. (2014). City of Burnaby guide for developing near streams. 28pp. Available online at http://www.burnaby.ca/Assets/city+services/planning/Guide+for+Developing+Near+ Streams.pdf. Clark, M. & Acomb, G. (2008). Bioswales/vegetated swales. Florida field guide to low impact development. Uni-versity of Florida. 4pp. Available online at: http://buildgreen.ufl.edu/ Fact_sheet_Bioswales_Vege-tated_Swales.pdf. Crowe, B. (2014). Eliminating combined sewer overflows from Vancouver’s waterways [PDF copies of PowerPoint slides]. Retrieved from http://www.metrovancouver.org/region/breakfasts/ Presentations/BrianCrowe-CityofVancouver-EliminatingCSOs-April15breakfast.pdf. Davie, T. (2008). Fundamentals of hydrology: Fundamentals of physical geography. London and New York: Routledge. Davis, A. P., Shokouhian, M., Sharma, H., & Minami, C. (2001). Laboratory study of biological retention for urban stormwater management. Water Environment Research, 5-14. Debo, T. N., & Reese, A. (2002). Municipal stormwater management. CRC Press. DeGasperi, C. L., Berge, H. B., Whiting, K. R., Burkey, J. J., Cassin, J. L., & Fuerstenberg, R. R. (2009). Linking Hydrologic Alteration to Biological Impairment in Urbanizing Streams of the Puget Lowland, Washington, USA1.JAWRA Journal of the American Water Resources Association, 45(2), 512-533. Deletic, A. (1998). The first flush load of urban surface runoff. Water Research, 32(8), 2462-2470. Deletic, A. B., & Maksimovic, C. T. (1998). Evaluation of water quality factors in storm runoff from paved ar-eas. Journal of Environmental Engineering, 124(9), 869-879. Dietz, M. E., & Clausen, J. C. (2005). A field evaluation of rain garden flow and pollutant treatment. Water, Air, and Soil Pollution, 167(1-4), 123-138. Dinicola, R. S. (1990). Characterization and simulation of rainfall-runoff relations for headwater basins in western King and Snohomish Counties, Washington. Water-resources investigations report (USA). District of Maple Ridge. (2014). Silver Valley Area Plan, in Maple Ridge Official Community Plan Bylaw No. 7060-2014. Chapter 10, 57pp. Available online at: http://www.mapleridge.ca/ DocumentCenter/View/747. Ellis, J. B. (1986). Pollutional aspects of urban runoff. In Urban runoff pollution (pp. 1-38). Springer Berlin Heidel-berg. Elwood, J. W., & Waters, T. F. (1969). Effects of floods on food consumption and production rates of a stream brook trout population. Transactions of the American Fisheries Society, 98(2), 253-262. 153  Environmental Management Act. 2013. SBC 2003, c. 53. <http://canlii.ca/t/520xn> retrieved on 2014-07-02. Erickson, A. J., Weiss, P. T., & Gulliver, J. S. (2013). Optimizing Stormwater Treatment Practices. Springer. Fisheries Act. 2012. R.S.C., 1985, c. F-14. Parliament of Canada. Ottawa, ON. Valid 29 June 2012 through 24 No-vember 2013. Fisheries Act. 2013. R.S.C., 1985, c. F-14. Parliament of Canada. Ottawa, ON. Valid 25 November 2013 to 27 May 2014. Flint, K. R., & Davis, A. P. (2007). Pollutant mass flushing characterization of highway stormwater runoff from an ultra-urban area. Journal of Environmental Engineering, 133(6), 616-626. FlowWorks. (2014). FlowWorks History. Website http://www.flowworks.com/about/history accessed April 2014. Frederick, C. P. (2003). Water-quality characteristics of urban storm runoff at selected sites in East Baton Rouge Parish, Louisiana, February 1998 through April 2002. US Department of the Interior, US Geological Survey.  Frederick, C. P. (2003). Water-quality characteristics of urban storm runoff at selected sites in East Baton Rouge Parish, Louisiana, February 1998 through April 2002. US Department of the Interior, US Geological Survey. Galli, J. (1990). Thermal impacts associated with urbanization and stormwater management best management prac-tices. Department of Environmental Programs, Metropolitan Washington Council of Governments. Galloway, J. N., Aber, J. D., Erisman, J. W., Seitzinger, S. P., Howarth, R. W., Cowling, E. B., & Cosby, B. J. (2003). The nitrogen cascade. Bioscience, 53(4), 341-356. Geiger, W.F. (1984). Characteristics of combined sewer runoff. Proceedings of the 3rd International Conference on Urban Storm Drainage. Lausanne, Switzerland. 40-46. Geosyntec Consultants, Inc. and Wright Water Engineers, Inc. (2012a). International stormwater best management practices (BMP) database: pollutant category summary summary statistical addendum: TSS, bacteria, nutrients, and metals. 35 pp. Available online at: http://bmpdatabase.org/Docs/2012%20Water%20Quality%20Analy-sis%20Addendum/BMP%20Database%20Categorical_SummaryAddendumReport_Final.pdf. Geosyntec Consultants, Inc. and Wright Water Engineers, Inc. (2012b). International stormwater best management practices (BMP) database: user’s guide for BMP data entry spreadsheets. 105pp. Available online at: http://bmp-database.org/Docs/2010%20BMP%20Database%20 User's%20Guide.pdf. Gibb, A., H. Kelly, R. Horner, T. Scheuler, J. Simmler, and J. Knutson. (1999). Best Management Practices Guide for Stormwater. (Appendix H: Construction Site Erosion and Sediment Control Guide). Prepared for Greater Vancouver Sewerage and Drainage District. Accessed 03 July 2014 at http://www.metrovancou-ver.org/about/publications/Publications/BMPVol2a.pdf. Glass, C., & Bissouma, S. (2005). Evaluation of a parking lot bioretention cell for removal of stormwater pollu-tants. Transactions on Ecology and the Environment, 699-708. 154  Gostomski, F. (1990). The toxicity of aluminum to aquatic species in the US. Environmental geochemistry and health, 12(1-2), 51-54. Green, T. R., Beavis, S. G., Dietrich, C. R., & Jakeman, A. J. (1999). Relating stream‐bank erosion to in‐stream transport of suspended sediment. Hydrological Processes, 13(5), 777-787. Greig, S. M., Sear, D. A., & Carling, P. A. (2005). The impact of fine sediment accumulation on the survival of in-cubating salmon progeny: implications for sediment management. Science of the Total Environment, 344(1), 241-258. Greig, S., Sear, D., & Carling, P. (2007). A field‐based assessment of oxygen supply to incubating Atlantic salmon (Salmo salar) embryos. Hydrological processes, 21(22), 3087-3100. Hall, K. J., & Anderson, B. C. (1988). The toxicity and chemical composition of urban stormwater runoff. Canadian Journal of Civil Engineering, 15(1), 98-106. Hall, K. J., Kiffney, P., Macdonald, R., McCallum, D., Larkin, G., Richardson, J., Schreir, H., Smith, J., Zandber-gen, P., Keen, P., Belzer, W., Brewer, R., Sekela, M., & Thomson, B. (1999). Non-point source contamination in the urban environment of Greater Vancouver: A case study of the Brunette River Watershed. Health of the Fra-ser River aquatic ecosystem, 2, 1998-11. Hammer, T. R. (1972). Stream channel enlargement due to urbanization. Water Resources Research, 8(6), 1530-1540. Hathaway, J. M., Tucker, R. S., Spooner, J. M., & Hunt, W. F. (2012). A traditional analysis of the first flush effect for nutrients in stormwater runoff from two small urban catchments. Water, Air, & Soil Pollution, 223(9), 5903-5915. Hatt, B. E., Fletcher, T. D., & Deletic, A. (2009). Hydrologic and pollutant removal performance of stormwater bio-filtration systems at the field scale.Journal of Hydrology, 365(3), 310-321. Heywood, M. J. T., & Walling, D. E. (2007). The sedimentation of salmonid spawning gravels in the Hampshire Avon catchment, UK: implications for the dissolved oxygen content of intragravel water and embryo sur-vival. Hydrological processes, 21(6), 770-788. Hinman, C. (2012). Low impact development technical guidance manual for Puget Sound. Washington State Uni-versity Extension and Puget Sound Partnership. 365pp. Available online at: http://www.psp.wa.gov/down-loads/LID/20121221_LIDmanual_FINAL_secure.pdf. Hirsch RM, Walker JF, Day JC, Kallio R. (1990). The influence of man on hydrologic systems. In Surface Water Hydrology (The Geology of America, Vol. O-1), ed. MG Wolman, HC Riggs, pp. 329–59. Boulder, CO: Geol. Soc. Am. Holman‐Dodds, J. K., Bradley, A. A., & Potter, K. W. (2003). Evaluation of hydrologic benefits of infiltration based urban storm water management1. 155  Hoopes, R. L. (1975). Flooding as the result of Hurricane Agnes, and its effect on a native brook trout population, in an infertile headwater stream in central Pennsylvania. Transactions of the American Fisheries Society, 104(1), 96-99. Izagirre, O., Serra, A., Guasch, H., & Elosegi, A. (2009). Effects of sediment deposition on periphytic biomass, pho-tosynthetic activity and algal community structure. Science of the total environment, 407(21), 5694-5700. Izonfuo, L. W. A., & Bariweni, A. P. (2001). The effect of urban runoff water and human activities on some phys-ico-chemical parameters of the Epie creek in the Niger Delta. Journal of Applied Sciences and Environmental Management, 5(1). Johnston, C. (2005). Template for integrated stormwater management planning 2005: terms of reference template, draft report. [Report prepared for the Greater Vancouver Regional District]. Retrieved from: http://www.metro-vancouver.org/about/publications/Publications/ISMP_Template-_2005v2.pdf. Johnston, D., A. Kang, and N. Volkow. (2014). Draft Eagle Creek integrated stormwater management plan. Report submitted to the City of Burnaby’s Environmental Committee 16 June 2014. Available online at: https://burn-aby.civicweb.net/Documents/DocumentList.aspx? ID=14885. Jones, J. I., Murphy, J. F., Collins, A. L., Sear, D. A., Naden, P. S., & Armitage, P. D. (2012). The impact of fine sediment on Macro‐Invertebrates.River Research and Applications, 28(8), 1055-1071. Jones, M. P., & Hunt, W. F. (2009). Bioretention impact on runoff temperature in trout sensitive waters. Journal of Environmental Engineering, 135(8): 577-585. Jones, T., C. Johnston, and C. Kipkie. (2002). Using annual hydrographs to determine effective impervious area. Unpublished paper from Kerr Wood Leidal Associates, Ltd. in Burnaby, BC.  Katopodis, C. and R. Gervais. 1991. Ichthyomechanics. Freshwater Institute. Central and Arctic Region. Department of Fisheries and Oceans Canada. Kaushal, S. S., Groffman, P. M., Likens, G. E., Belt, K. T., Stack, W. P., Kelly, V. R., Band, L. E., & Fisher, G. T. (2005). Increased salinization of fresh water in the northeastern United States. Proceedings of the National Acad-emy of Sciences of the United States of America, 102(38), 13517-13520. Kemp, P., Sear, D., Collins, A., Naden, P., & Jones, I. (2011). The impacts of fine sediment on riverine fish. Hydro-logical Processes, 25(11), 1800-1821. Kerr Wood Leidal, Lanarc Consultants, & Goya Nogan. (2012). Stormwater Source Control Design Guidelines - Final Report. Retrieved 01 July 2014 from: http://www.metrovancouver. org/ser-vices/wastewater/sources/Pages/StormwaterManagement.aspx Kerr Wood Leidal. (2010). Yorkson neighborhood water quality monitoring program report, final report December 2010. Prepared for the Township of Langley. 186pp. 156  Kerr Wood Leidal. (2013). 2011 Routley neighborhood area stormwater and environmental monitoring final report. Prepared for the Township of Langley. 138pp. Kincaid, D. W., & Findlay, S. E. (2009). Sources of elevated chloride in local streams: Groundwater and soils as potential reservoirs. Water, air, and soil pollution, 203(1-4), 335-342. King County. (2009). Surface Water Design Manual. King County, WA: King County Department of Natural Re-sources and Parks. Kinouchi, T., Yagi, H., & Miyamoto, M. (2007). Increase in stream temperature related to anthropogenic heat input from urban wastewater. Journal of Hydrology, 335(1), 78-88. LeBlanc, R. T., Brown, R. D., & FitzGibbon, J. E. (1997). Modeling the effects of land use change on the water tem-perature in unregulated urban streams.Journal of Environmental Management, 49(4), 445-469. Lee, H., Lau, S. L., Kayhanian, M., & Stenstrom, M. K. (2004). Seasonal first flush phenomenon of urban storm-water discharges. Water Research, 38(19), 4153-4163. Lee, J. G., & Heaney, J. P. (2003). Estimation of urban imperviousness and its impacts on storm water sys-tems. Journal of Water Resources Planning and Management, 129(5), 419-426. Lee, J. H., Bang, K. W., Ketchum Jr, L. H., Choe, J. S., & Yu, M. J. (2002). First flush analysis of urban storm run-off. Science of the Total Environment, 293(1), 163-175. Leopold, L. B. (1968). Hydrology for urban land planning: A guidebook on the hydrologic effects of urban land use. Li, H., & Davis, A. P. (2009). Water quality improvement through reductions of pollutant loads using bioretention. Journal of Environmental Engineering, 135(8), 567-576. Limburg, K. E., & Schmidt, R. E. (1990). Patterns of fish spawning in Hudson River tributaries: response to an ur-ban gradient?. Ecology, 1238-1245. Livingston, E. H., & McCarron, E. (1992). Stormwater management: A guide for Floridians. Tallahassee, FL: Flor-ida Department of Environmental Regulation. Luce, J. J., Steele, R., & Lapointe, M. F. (2010). A physically based statistical model of sand abrasion effects on pe-riphyton biomass. Ecological Modelling,221(2), 353-361. Macdonald, R. (2003). Estimated urban runoff character and contaminant loadings in the GVRD: analysis of urban stormwater quality data collected within the GVRD from 1993-1997. Report prepared for the Greater Vancouver Regional District. 123pp. May, C. W., Horner, R. R., Karr, J. R., Mar, B. W., & Welch, E. B. (1997). Effects of urbanization on small streams in the Puget Sound ecoregion. Watershed Protection Techniques, 2(4), 483-494. McLeod, S. M., Kells, J. A., & Putz, G. J. (2006). Urban runoff quality characterization and load estimation in Sas-katoon, Canada. Journal of environmental engineering, 132(11), 1470-1481.   157  Metro Vancouver (2010). Integrated liquid waste and resource management: A liquid waste management plan for the Greater Vancouver Sewerage & Drainage District and member municipalities. Vancouver, BC. Municipal Wastewater Regulation. 2012. B.C. Reg. 87/2012. <http://canlii.ca/t/51w3b> retrieved on 2014-07-02. Nagpal, N.K., L.W. Pommen, and L.G. Swain. (2006). A compendium of working water quality guidelines for Brit-ish Columbia. Science and Information Branch, Environmental Protection Division, Ministry of the Environ-ment, Government of British Columbia. Website accessed 07 September 2014 at http://www.env.gov.bc.ca/wat/wq/BCguidelines/working.html Nelson, K. C., & Palmer, M. A. (2007). Stream Temperature Surges under Urbanization and Climate Change: Data, Models, and Responses. Journal of the American Water Resources Association. 43(2): 440-452. Olley, J., Caitcheon, G., Hancock, G., and Wallbrink, P. (1996). The Sources of Phosphorus and Sediment in the Namoi River Basin: Progress Report. CSIRO Division of Water Resources, Canberra. Paul, M. J., & Meyer, J. L. (2001). Streams in the urban landscape. Annual Review of Ecology and Systematics, 333-365. Pluhowski, E. J. (1970). Urbanization and its effect on the temperature of the streams on Long Island, New York. USGS Professional Paper 627-D. 116pp. Poff, N. L., Allan, J. D., Bain, M. B., Karr, J. R., Prestegaard, K. L., Richter, B. D., Sparks, R. E., & Stromberg, J. C. (1997). The natural flow regime. BioScience, 769-784. Quinn, J. M., Davies-Colley, R. J., Hickey, C. W., Vickers, M. L., & Ryan, P. A. (1992). Effects of clay discharges on streams. Hydrobiologia, 248(3), 235-247. Reuss, J.O., & D.W. Johnson. (1985). Effect of soil processes on the acidification of water by acid deposition. Jour-nal of Environmental Quality,14(1), 26-31. Roesner, L. A., Bledsoe, B. P., & Brashear, R. W. (2001). Are best-management-practice criteria really environmen-tally friendly?. Journal of Water Resources Planning and Management, 127(3), 150-154. Ryan, P. A. (1991). Environmental effects of sediment on New Zealand streams: a review. New Zealand journal of marine and freshwater research, 25(2), 207-221. Sabin, L. D., Lim, J. H., Stolzenbach, K. D., & Schiff, K. C. (2005). Contribution of trace metals from atmospheric deposition to stormwater runoff in a small impervious urban catchment. Water research, 39(16), 3929-3937. Saget, A., Chebbo, G., & Bertrand-Krajewski, J. L. (1996). The first flush in sewer systems. Water Science and Technology, 33(9), 101-108. Sansalone, J. J., & Buchberger, S. G. (1997). Partitioning and first flush of metals in urban roadway storm water. Journal of Environmental engineering, 123(2), 134-143. 158  Sawyer, J. (2008). Surface waters: Ammonium is not ammonia. Integrated Crop Management News, Iowa State University. Accessed online 08 July 2014 at http://www.extension.iastate.edu/ CropNews/2008/0421JohnSaw-yer.htm Schiff, K., Bay, S., & Stransky, C. (2002). Characterization of stormwater toxicants from an urban watershed to freshwater and marine organisms. Urban water, 4(3), 215-227. Seaburn, G. E. (1969). Effects of urban development on direct runoff to East Meadow Brook, Nassau County, Long Island, New York. US Government Printing Office. Sear, D. A., Frostick, L. B., Rollinson, G., & Lisle, T. E. (2008). The significance and mechanics of fine-sediment infiltration and accumulation in gravel spawning beds. In American Fisheries Society Symposium (Vol. 65, pp. 000-000). Servizi, J. A., & Martens, D. W. (1991). Effect of temperature, season, and fish size on acute lethality of suspended sediments to coho salmon (Oncorhynchus kisutch). Canadian Journal of Fisheries and Aquatic Sciences, 48(3), 493-497. SFE Global. 2008. Sanitary Sewer Flow Report. Prepared by SFE Global for Monterrey County. Available online at http://www.co.monterey.ca.us/planning/major/Moss%20Landing%20 Community%20Plan/Intrastructure_Sub-Committee/SFE_Sanitary_Sewer_Flow_Rpt.pdf. Shields Jr, F. D., Knight, S. S., & Cooper, C. M. (1994). Effects of channel incision on base flow stream habitats and fishes. Environmental Management,18(1), 43-57. Sickman, J. O., Zanoli, M. J., & Mann, H. L. (2007). Effects of urbanization on organic carbon loads in the Sacra-mento River, California. Water Resources Research, 43(11). Simon, A. 1989. Gradation processes and channel evolution in modified West Tennessee streams. U.S. Geological Survey Professional Paper 1470. 93pp. Smullen, J. T., Shallcross, A. L., & Cave, K. A. (1999). Updating the US nationwide urban runoff quality data base. Water Science and Technology,39(12), 9-16. Spellman, F. R. (2004). Mathematics Manual for water and wastewater Treatment plant Operators. CRC Press. Spinello, A. G., & Simmons, D. L. (1992). Base flow of 10 south-shore streams, Long Island, New York, 1976-85, and the effects of urbanization on base flow and flow duration (pp. 1-31). US Department of the Interior, US Ge-ological Survey. Steedman, R. J. (1988). Modification and assessment of an index of biotic integrity to quantify stream quality in southern Ontario. Canadian Journal of Fisheries and Aquatic Sciences, 45(3), 492-501. Stephens, K. A. (2002). Stormwater planning: A guidebook for British Columbia. British Columbia, Ministry of Water, Land and Air Protection. Stormwater BMP Databas 159  Strassler, E., Pritts, J., & Strellec, K. (1999). Preliminary data summary of urban storm water best management prac-tices. Environmental Protection Agency Report EPA-821-R-99-012. Taebi, A., & Droste, R. L. (2004). First flush pollution load of urban stormwater runoff. Journal of Environmental Engineering and Science, 3(4), 301-309. Taylor, G. D., Fletcher, T. D., Wong, T. H., Breen, P. F., & Duncan, H. P. (2005). Nitrogen composition in urban runoff—implications for stormwater management. Water Research, 39(10), 1982-1989. Teledyne ISCO. (2008). 2150 Area Velocity Flow Module and Sensor Installation and Operation Guide. Revision Y. Product specifications. 122pp. Available online at: http://www.johnmorris.com.au/files/product/attach-ments/5092/7875_manual_instr.pdf Tiefenthaler, L. L., Stein, E. D., & Schiff, K. C. (2008). Watershed and land use–based sources of trace metals in urban storm water. Environmental Toxicology and Chemistry, 27(2), 277-287. Tiefenthaler, L., Schiff, K., & Leecaster, M. (1999). Temporal variability patterns of stormwater concentrations in urban stormwater runoff. Southern California coastal water research project annual report, 28-44. Township of Langley. (2003). Storm Water Management Plan Certified Record Drawing. Municipal File No. 8-15-0108. Township of Langley. (2004). Individual Lot Infiltration Facility. Design drawings and notes. Municipal Project No. 08-22-0046. Received by email from Mr. Art Kastelein, engineer with the Township of Langley. Township of Langley. (2012). 2011 Census Data Release 1: Population and Dwelling Counts. Available online at http://www.tol.ca/Services-Contact/Document-Library/fid/385. Township of Langley. (2013). Open Data Catalogue. Online depository of geographic information mapping data. Accessed online July 2013 at http://www.tol.ca/Services-Contact/Open-Data/Open-Data-Catalogue. Trimble, S. W. (1997). Contribution of stream channel erosion to sediment yield from an urbanizing watershed. Sci-ence, 278(5342), 1442-1444. Trowsdale, S. A., & Simcock, R. (2011). Urban stormwater treatment using bioretention. Journal of Hydrol-ogy, 397(3), 167-174.   Tuccillo, M. E. (2006). Size fractionation of metals in runoff from residential and highway storm sewers. Science of the Total Environment, 355(1), 288-300. USEPA. (1983). Results of the nationwide urban runoff program, volume I- final report. NTIS PB84-185552. Wash-ington, DC: US Environmental Protection Agency. USEPA. (1986). Ambient aquatic life water quality criteria for aluminum. United States Environmental Protection Agency, Office of research and development, Environmental Research Laboratory. Duluth, Minnesota. 42pp. 160  USEPA. (1987). Water quality criteria for water 1986. EPA 440/5086-001. United States Environmental Protection Agency, Office of Water Regulations and Standards, Washington, DC. 477pp. USEPA. (1993). Guidance specifying management measures for sources of nonpoint pollution in coastal waters. United States Environmental Protection Agency, Office of Water, Washington, DC. EPA-840-B-92-002. USEPA. (1999a). Storm water technology fact sheet: Bioretention. United States Environmental Protection Agency, Office of Water. EPA 832-F-99-012. 8pp. Available online at: http://water.epa.gov/scitech/wastetech/up-load/2002_06_28_mtb_biortn.pdf. USEPA. (1999b). Storm water technology fact sheet: Vegetated Swales. United States Environmental Protection Agency, Office of Water. 832-F-99-006. 7pp. Available online at: http://water.epa.gov/scitech/wastetech/up-load/2002_06_28_mtb_vegswale.pdf. USEPA. (2014). National recommended water quality criteria. United States Environmental Protection Agency, Office of Water Regulations and Standards, Washington, DC. Website accessed 08 September 2014 at http://wa-ter.epa.gov/scitech/swguidance/standards/criteria/current/index.cfm#gold. Page last updated 01 August 2014. Viklander, M., & Marsalek, J. (2010). Controlling contaminants in urban stormwater: linking environmental science and policy. On the water front, 101-108. Walker, W. W. (1987). Phosphorus removal by urban runoff detention basins. Lake and reservoir manage-ment, 3(1), 314-326. Walsh, C. J. (2004). Protection of in-stream biota from urban impacts: minimise catchment imperviousness or im-prove drainage design?. Marine and Freshwater Research, 55(3), 317-326. Walton, B. M., Salling, M., Wyles, J., & Wolin, J. (2007). Biological integrity in urban streams: Toward resolving multiple dimensions of urbanization. Landscape and Urban Planning, 79(1), 110-123. Water Environment Federation. (1998). Urban Runoff Quality Management (No. 87). ASCE Publications. Whalen, P. J., & Cullum, M. G. (1988). An assessment of urban land use/stormwater runoff quality relationships and treatment efficiencies of selected stormwater management systems. Water Quality Division, Resource Planning Department, South Florida Water Management District.  Whitman, R. L., & Clark, W. J. (1982). Availability of dissolved oxygen in interstitial waters of a sandy creek. Hy-drobiologia. 92: 651-658. Springer Netherlands. Winer, R. (2000). National pollutant removal performance database for stormwater treatment practices. Ellicott City, MD: Center for Watershed Protection. Wolman, M. G. (1967). A cycle of sedimentation and erosion in urban river channels. Geografiska Annaler. Series A. Physical Geography, 385-395. 161  Yorke, T. H., & Herb, W. J. (1978). Effects of urbanization on streamflow and sediment transport in the Rock Creek and Anacostia River Basins, Montgomery County, Maryland, 1962-74. In Geological Survey professional pa-per (Vol. 1003). US Government Printing Office.  162  Appendices   163  Appendix A  Hydrologic data for 128 rainfall events in the Township of Langley 164   EVENT:RAINFALL: 2.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 35 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 54 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %27 March 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.25Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.5Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain165   EVENT:RAINFALL: 2.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 89 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %16 February 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.010.020.020.030.03Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.02Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.5Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain166  EVENT:RAINFALL: 2.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 79 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 30 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 31 %27 January 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.05Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.5Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain167   EVENT:RAINFALL: 2.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 56 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 12 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 9 %14 April 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.5Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain168   EVENT:RAINFALL: 2.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 56 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 26 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 16 %27 June 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.5Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain169   EVENT:RAINFALL: 2.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 68 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 15 %07 October 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.0Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain170   EVENT:RAINFALL: 2.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 52 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 10 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 6 %26 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.010.020.020.030.030.04Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.03Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.0Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain171   EVENT:RAINFALL: 2.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 41 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 11 %26 June 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.0Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain172   EVENT:RAINFALL: 2.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 35 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 15 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 6 %08 March 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Tue 0:00 Tue 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Tue 0:00 Tue 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.0Tue 0:00 Tue 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain173   EVENT:RAINFALL: 2.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 67 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %14 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.0Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain174   EVENT:RAINFALL: 3.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 76 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 12 %29 February 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.04Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.030.03Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain175   EVENT:RAINFALL: 3.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 86 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 12 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 11 %02 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.04Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.030.03Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain176   EVENT:RAINFALL: 3.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 51 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 11 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 6 %14 February 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.010.020.020.030.030.04Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.030.03Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain177   EVENT:RAINFALL: 3.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 48 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 15 %19 March 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.10Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain178   EVENT:RAINFALL: 3.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 46 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 8 %14 June 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain179   EVENT:RAINFALL: 3.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 48 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 37 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 22 %09 March 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.120.14Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain180   EVENT:RAINFALL: 3.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 87 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 36 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 27 %25 December 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.10Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.5Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain181   EVENT:RAINFALL: 3.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 58 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %12 November 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain182   EVENT:RAINFALL: 3.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 38 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 31 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %24 April 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.20Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain183   EVENT:RAINFALL: 3.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 60 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 37 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 30 %25 January 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.15Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Wed 0:00 Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain184   EVENT:RAINFALL: 3.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 30 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 30 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %13 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Tue 0:00 Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain185   EVENT:RAINFALL: 3.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 89 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %19 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Mon 12:00 Tue 0:00 Tue 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Mon 12:00 Tue 0:00 Tue 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Mon 12:00 Tue 0:00 Tue 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain186   EVENT:RAINFALL: 3.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 44 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 25 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %16 February 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Sat 0:00 Sat 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Sat 0:00 Sat 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Sat 0:00 Sat 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain187   EVENT:RAINFALL: 3.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 77 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %26 February 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.00.51.01.52.02.53.03.54.0Tue 12:00 Wed 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain188   EVENT:RAINFALL: 4.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 44 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 9 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 4 %10 May 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.04Tue 12:00 Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.03Tue 12:00 Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Tue 12:00 Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain189   EVENT:RAINFALL: 4.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 38 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 25 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %27 May 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.15Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain190   EVENT:RAINFALL: 4.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 55 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 24 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %28 November 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.10Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain191   EVENT:RAINFALL: 4.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 78 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %18 March 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain192   EVENT:RAINFALL: 4.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 40 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 8 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 3 %21 May 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.010.020.020.030.030.04Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.030.03Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain193   EVENT:RAINFALL: 4.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 69 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 21 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %17 June 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain194   EVENT:RAINFALL: 4.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 87 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 22 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 20 %06 February 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain195   EVENT:RAINFALL: 4.4 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 29 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 55 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %12 December 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.20Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain196   EVENT:RAINFALL: 4.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 54 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 12 %29 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.10Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.10Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain197   EVENT:RAINFALL: 4.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 61 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 39 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 30 %22 December 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.20Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain198   EVENT:RAINFALL: 4.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 49 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 15 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %09 September 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.20Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain199   EVENT:RAINFALL: 4.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 50 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 20 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %26 October 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain200   EVENT:RAINFALL: 4.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 52 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 27 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %28 April 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.15Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.0Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain201   EVENT:RAINFALL: 5.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 86 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 9 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 10 %17 September 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain202   EVENT:RAINFALL: 5.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 55 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 29 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 21 %02 January 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.20Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain203   EVENT:RAINFALL: 5.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 57 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 7 %26 May 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Sun 0:00 Sun 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Sun 0:00 Sun 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Sun 0:00 Sun 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain204   EVENT:RAINFALL: 5.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 13 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 3 %21 February 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain205   EVENT:RAINFALL: 5.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 50 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 34 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 31 %23 June 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.100.200.300.40Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.100.200.300.40Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain206   EVENT:RAINFALL: 5.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 36 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 24 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 9 %31 July 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.10Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain207   EVENT:RAINFALL: 5.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 25 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 61 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 35 %30 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.100.200.300.400.50Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.100.200.300.400.50Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain208   EVENT:RAINFALL: 5.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 56 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 11 %07 June 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.0Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain209   EVENT:RAINFALL: 5.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 31 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 54 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %04 March 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.120.14Fri 12:00 Sat 0:00 Sat 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Fri 12:00 Sat 0:00 Sat 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Fri 12:00 Sat 0:00 Sat 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain210   EVENT:RAINFALL: 5.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 87 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 21 %02 November 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain211   EVENT:RAINFALL: 5.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 66 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 15 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 10 %09 February 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.05Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain212   EVENT:RAINFALL: 5.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 74 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %27 April 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.06Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.05Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain213   EVENT:RAINFALL: 5.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 58 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 16 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 11 %15 September 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.120.14Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain214   EVENT:RAINFALL: 6.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 52 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 25 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %27 October 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.250.30Sun 0:00 Sun 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.250.30Sun 0:00 Sun 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Sun 0:00 Sun 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain215   EVENT:RAINFALL: 6.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 39 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 10 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 8 %11 April 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.06Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Wed 12:00 Thu 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain216   EVENT:RAINFALL: 6.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 67 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %07 June 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.04Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.010.020.020.030.03Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain217   EVENT:RAINFALL: 6.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 60 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 28 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 22 %23 June 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.25Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain218   EVENT:RAINFALL: 6.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 46 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 38 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 26 %07 December 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.20Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain219   EVENT:RAINFALL: 6.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 123 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 16 %07 November 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain220   EVENT:RAINFALL: 6.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 44 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 21 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %22 May 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Wed 12:00 Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.10Wed 12:00 Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.0Wed 12:00 Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain221   EVENT:RAINFALL: 6.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 41 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 31 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %21 April 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain222   EVENT:RAINFALL: 6.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 62 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 29 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 28 %05 January 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.10Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain223   EVENT:RAINFALL: 7.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 38 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 27 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 16 %01 April 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.10Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain224   EVENT:RAINFALL: 7.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 62 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %28 February 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.06Tue 12:00 Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.05Tue 12:00 Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Tue 12:00 Wed 0:00 Wed 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain225   EVENT:RAINFALL: 7.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 77 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 21 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 19 %21 February 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain226   EVENT:RAINFALL: 7.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 54 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 14 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 8 %06 February 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.010.020.030.040.050.060.07Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.04Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Sun 0:00 Sun 12:00 Mon 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain227   EVENT:RAINFALL: 7.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 56 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 22 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %15 October 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.120.14Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.15Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.01.02.03.04.05.06.07.08.0Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain228   EVENT:RAINFALL: 7.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 49 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 50 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 28 %14 April 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.0Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain229   EVENT:RAINFALL: 7.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 28 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 18 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 7 %05 May 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.0Thu 0:00 Thu 12:00 Fri 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain230   EVENT:RAINFALL: 7.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 39 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 29 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 12 %01 June 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Wed 12:00 Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.25Wed 12:00 Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.0Wed 12:00 Thu 0:00 Thu 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain231   EVENT:RAINFALL: 8.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 70 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 17 %16 April 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.100.12Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.080.100.12Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.0Mon 0:00 Mon 12:00 Tue 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain232   EVENT:RAINFALL: 9.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 70 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 35 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 28 %31 March 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.080.10Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain233   EVENT:RAINFALL: 9.4 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 77 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 22 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 15 %23 June 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.250.30Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.25Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.0Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain234   EVENT:RAINFALL: 10.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 43 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 36 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 21 %28 January 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.25Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.20Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.012.0Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain235   EVENT:RAINFALL: 10.0 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 41 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 33 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 13 %02 October 20110%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.250.30Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.250.30Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.012.0Sun 12:00 Mon 0:00 Mon 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain236   EVENT:RAINFALL: 10.3 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 58 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 26 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %31 May 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.050.100.150.200.250.30Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.050.100.150.200.250.30Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.012.0Thu 12:00 Fri 0:00 Fri 12:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain237   EVENT:RAINFALL: 10.5 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 51 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 23 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 16 %15 December 20120%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.010.020.030.040.050.06Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.012.0Sat 0:00 Sat 12:00 Sun 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul R. Runoff Cumul-Rain238   EVENT:RAINFALL: 10.8 MM ROUTLEY PEAK RUNOFF / LANGLEY MEADOWS PEAK RUNOFF: 57 %LANGLEY MEADOWS CUMULATIVE RUNOFF / EVENT RAINFALL: 20 %ROUTLEY CUMULATIVE RUNOFF / EVENT RAINFALL: 11 %27 September 20130%20%40%60%80%100%0% 20% 40% 60% 80% 100%Percent of fullrainfall volumeOf time to full runoffCumul. Rain Cumul. LM Runoff Cumul. R Runoff0102030400.000.020.040.060.08Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Normalized Flow(mm/5min)LM Normalized Flow R Normalized Flow LM's GWI R's GWI0102030400.000.020.040.060.08Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Runoff(mm/5min)Langley Meadows Runoff Routley Runoff05101520253035400.02.04.06.08.010.012.0Fri 0:00 Fri 12:00 Sat 0:00Rainfall intensity (mm/hr)Cumulative rainfall& runoff (mm)Cumul LM Runoff Cumul