UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Using artificial tracers to observe timing of runoff from different landscape units in a small headwater… Bier, Anthony Friedrich 2008

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

Item Metadata

Download

Media
24-ubc_2008_fall_bier_anthony.pdf [ 811.12kB ]
Metadata
JSON: 24-1.0066629.json
JSON-LD: 24-1.0066629-ld.json
RDF/XML (Pretty): 24-1.0066629-rdf.xml
RDF/JSON: 24-1.0066629-rdf.json
Turtle: 24-1.0066629-turtle.txt
N-Triples: 24-1.0066629-rdf-ntriples.txt
Original Record: 24-1.0066629-source.json
Full Text
24-1.0066629-fulltext.txt
Citation
24-1.0066629.ris

Full Text

INVESTIGATING RUNOFF GENERATION PROCESSES IN A SMALL FORESTED HEADWATER CATCHMENT USING ARTIFICIAL TRACERS  by Anthony Friedrich Bier B.Sc.Hon. Trent Univerity, 2004 A THESIS SUBMITTED IN PARTIAL FULLFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in The Faculty of Graduate Studies (Geography) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) September 2008 ©Anthony Friedrich Bier, 2008  ABSTRACT Four artificial tracers were applied to a small headwater catchment in south western British Columbia to study runoff generated from topographically distinct landscape units. The seven hectare catchment is located in the University of British Columbia Malcolm Knapp Research Forest at low elevation (190-280 masl). A weir, multiple tipping bucket rain gauges and several piezometers were used to collect hydrological data. Three separate landscape units were identified based on topography, soil properties and proximity to the stream. The units included an area of shallow slope and deep soil, a riparian area along the intermittent stream channel and an area of very shallow soil with bedrock outcrops on a steep slope. Tracers used included rhodamine-WT, uranine, sodium chloride and potassium bromide. A suite of ion selective and fluorometric probes were used along with automated water sampling to monitor tracer breakthrough. The collected samples were analysed in the lab to validate the field measurements. Tracers were dissolved in solution and applied aerially with a backpack sprayer at the onset of forecasted precipitation events to facilitate rapid infiltration into the soil. The first application took place January 4th, 2006. Measurements were then taken continuously until March 20th, 2006, when a second round of tracers was applied to the landscape units. During the first measurement period, 532 mm of precipitation fell below the forest canopy over 75 days. During the second 78 day measurement period, 290 mm of rain fell. It was found that the overall wetness of the catchment affected travel times significantly. Large storms during the first, significantly wetter, application period exhibited similar lag times from peak event discharge to tracer arrival between the different landscape units. During small precipitation events and under dryer conditions, travel times were greatest  ii  in the area of shallow slope and deep soils. These lag times are indicative of longer pathways and perhaps the non-initiation of preferential flow below certain thresholds. In general, it was concluded that delineating catchments into groups of similar landscape units based on physical characteristics may be a promising new approach to explaining catchment runoff response.  iii  TABLE OF CONTENTS Abstract...................................................................................................................... ii Table of Contents ...................................................................................................... iv List of Tables ............................................................................................................ vi List of Figures .......................................................................................................... vii Acknowledgements ................................................................................................... ix 1.0 Introduction .......................................................................................................... 1 1.1 Terminology ................................................................................................... 2 1.2 The Development of Runoff Theory ............................................................... 3 1.2.1 Early research ....................................................................................... 3 1.2.2 Quantifying the process: defining the pathways of storm runoff ............. 5 1.2.3 Sources of runoff: changing mechanisms of subsurface stormflow ......... 7 1.2.4 Topographic controls on runoff .............................................................. 9 1.3 Current Directions in Subsurface Stormflow Research...................................10 1.3.1 Preferential flow as rapid infiltration .....................................................10 1.3.2 Rapid transmission of stormflow as preferential flow ............................12 1.3.3 From surface to bedrock as a control on subsurface flow .......................13 1.4 Tracers as a Tool for Studying Runoff Generation .........................................15 1.5 Motivation, Objectives and Hypothesis ..........................................................19 2.0 Methods ..............................................................................................................22 2.1 Site Description .............................................................................................22 2.1.1 Location................................................................................................22 2.1.2 Climate .................................................................................................23 2.1.3 Topography...........................................................................................25 2.1.4 Geology and soil ...................................................................................27 2.1.5 Vegetation.............................................................................................27 2.1.6 Hydrology.............................................................................................27 2.2 Data Collection – Instrumentation..................................................................28 2.2.1 Precipitation..........................................................................................28 2.2.2 Surface flow..........................................................................................29 2.2.3 Subsurface flow ....................................................................................31 iv  2.3 Tracer Experiments........................................................................................31 2.3.1 Site selection .........................................................................................31 2.3.2 Tracer application .................................................................................32 2.3.3 Tracer breakthrough..............................................................................33 3.0 Results ................................................................................................................35 3.1 Data Collection/Experimental Design ............................................................35 3.1.1 Measuring input – precipitation.............................................................35 3.1.2 Measuring output – discharge................................................................35 3.1.3 Subsurface measurements .....................................................................38 3.1.4 Performance and logistics of artificial tracer use....................................39 3.2 Experimental Results .....................................................................................45 3.2.1 Precipitation..........................................................................................45 3.2.2 Discharge ..............................................................................................47 3.2.3 Tracer breakthrough and recovery .........................................................48 3.2.4 Subsurface response ..............................................................................55 4.0 Discussion ...........................................................................................................59 4.1 Success Using Artificial Tracers ....................................................................59 4.2 Landscape Unit Response ..............................................................................61 5.0 Conclusions .........................................................................................................68 5.1 Review of Key Findings ................................................................................68 5.2 Recommendations for Further Research.........................................................69 Bibliography ............................................................................................................71  v  LIST OF TABLES Table 2.1  Measured inputs and outputs in the study catchment during .............. 28 the 2006 study period 29  Table 3.1.1  Manual stage measurements compared to Odyssey data ................... 37  Table 3.2.1  Measured monthly inputs and outputs in the study ............................ 47 catchment during the 2006 study period  Table 3.2.2  Summary of tracer mass recovery...................................................... 55  vi  LIST OF FIGURES Figure 2.1  Location of University of British Columbia......................................23 Malcolm Knapp Research Forest  Figure 2.2  1971-2000 Average monthly precipitation and temperature..............24 normals from Environment Canada UBC research forest administration climate station  Figure 2.3  1971-2000 average monthly temperature and precipitation at the .....25 UBC research forest administation climate station versus averages from 2006 study period  Figure 2.4  Map of study site with UTM coordinates. Numbers 1, 2, and 3 ........26 represent tracer applications areas R-CL, SS-Br and Sts-Ur, respectively.  Figure 2.5  Road cut bisecting upper and lower halves of the study catchment ...26 in Malcolm Knapp Research Forest, Fall 2006.  Figure 2.6  Main weir and probes at research catchment, MKRF .......................30  Figure 3.1.1  Comparison of measured average daily throughfall to ......................36 daily rainfall measured in the clearing (taken to be above canopy) from January 1st to April 10th 2006  Figure 3.1.2  Theoretical discharge curve compared to dilution gaugings .............38  Figure 3.1.3  Deviation of in situ Cyclops-measured uranine concentration...........41 from ISCO water sample concentration analysed using the GOT fluorometer  Figure 3.1.4  Uranine concentration from ISCO water samples determined...........41 using the GOT fluorometer versus the corresponding reading from the in situ Turner probe; corrected using a moving median filter  Figure 3.1.5  Difference between GOT fluorometer readings taken.......................42 from ISCO water samples and corresponding in situ Turner Probe readings in millivolts over time for uranine  Figure 3.1.6  Bromide probe millivolt reading from TempHion ion.......................43 selective probe for bromide over time  vii  Figure 3.1.7  ISCO water sample bromide concentration as determined ................44 using ion chromatography versus corresponding in situ TempHion probe reading for bromide  Figure 3.1.8  ISCO water sample chloride concentration determined.....................45 using ion chromatography versus corresponding in situ TempHion probe reading  Figure 3.2.1  Daily rainfall below canopy ............................................................ 46  Figure 3.2.2  Discharge from main weir in MKRF study catchment ......................48 during 2006 study period  Figure 3.2.3 A Discharge and tracer concentration over time for bromide (SS-Br) ..50 Figure 3.2.3 B Discharge and tracer concentration over time for uranine (StS-Ur) ..51 Figure 3.2.3 C Discharge and tracer concentration over time for chloride (R-Cl) ....52 Figure 3.2.4  Peak event discharge versus lag time to peak tracer concentration ...53  Figure 3.2.5  24 hour discharge volume surrounding even peak versus .................54 24 hour tracer mass flux (as a percent of total tracer applied) surrounding peak tracer concentration  Figure 3.2.6  Peak hydraulic potential (height above confining layer) ...................56 versus lag time from peak hydraulic potential to peak event stream discharge  Figure 3.2.7  24 hour tracer mass flux (as a percent of total tracer applied) ............57 surrounding concentration peak versus peak depth of water table below surface  Figure 3.2.8  24 hour tracer mass flux (as a percent of total tracer applied) ............58 surrounding concentration peak versus peak fractional depth of water table below surface  viii  ACKNOWLEDGMENTS Firstly, I would like to thank my supervisor Markus Weiler for granting me this opportunity and taking me on as his first graduate student at UBC. His gracious financial support and inspiration made this thesis possible. I must also thank Dan Moore for providing guidance and many excellent suggestions, especially in the editing stages. His assistance was greatly appreciated and dearly needed. Many others contributed to this thesis during the field work stage, most of all Joel Trubilowicz who worked concurrently on his own thesis and with whom I shared many hours of field work. Joel’s technical skills with the digital elevation data and creating the catchment Digital Elevation Model were very important. There were also several field assistants including Charlotte Argue, Cheryl So and Fabian Nippgen. I also extend my thanks to the hydrology community at UBC in both Geography and Forestry for their companionship and community. The many hours spent at UBC would have been very dull without all of you. Finally, I would also like to thank my parents for their continuous faith and moral support throughout this journey. A special thanks to my mother for without her example I would not be here. Lastly and perhaps most importantly, I must thank my loving partner Valley for seeing me through the several years it has taken to finalize this thesis and even assisting me on occasion in the field. Her patience (and impatience) has helped me see this thesis to its end.  ix  1.0 INTRODUCTION Subsurface stormflow is a basic mechanism of runoff generation on hillslopes and in small watersheds. Subsurface stormflow is particularly important in humid climates and forested watersheds. Studies have shown that as input and antecedent moisture increase so does the relative contribution of subsurface flow to the storm hydrograph. Subsurface flow, then, becomes increasingly important in moist areas and during large storms (Sidle et al. 2000). The importance of subsurface stormflow has been documented worldwide; examples include Dunne and Black’s (1970) study in Vermont, Eastern USA, Mosley’s (1979) and McGlynn et al.’s (2002) work in South Island, New Zealand, Rodhe et al.’s (1996) tracer study in Gardsjon, Sweden, Sidle et al.’s (2000) work in Japan, and De Vries and Chow’s (1978) study in Coastal British Columbia, Canada. Currently, hydrologists are still trying to define first order controls on runoff generation. Most studies aimed at identifying first order controls are concentrated at the hillslope scale. The key factors which determine the quantity, sources and timing of runoff are still being determined for a wide variety of landscape types. Runoff generation processes vary greatly from location to location, making it extremely difficult to generalize what controls runoff for more than one geographic area. Nonetheless, Kirchner et al. (2003) called for the development of a unifying theory which explains both physical and chemical phenomena in a single catchment as contradictory results are often obtained from quantitative flow observations and chemical analyses. A combined explanation of both physical and chemical phenomena in catchment hydrology is still lacking although significant advances have been made over the last century (Weiler et al. 2005). Understanding runoff generation is of interest to multiple disciplines because runoff  1  pathways and timing determine chemical composition of stream flow, stream flow quantity and slope stability (Weiler et al. 2005). Tracers have proved to be a valuable tool in studying runoff generation. Both naturally occurring and artificial tracers are used and have been for some time (e.g. Pinder and Jones 1969, Smart and Laidlaw 1977). I propose that artificial tracers will provide a valuable tool in examining and identifying runoff generation processes within this study. The following review progresses through the historical development of current runoff theory to the direction of current research, thus establishing a basis for the motivation behind this study. This chapter will conclude with the research objectives of this study.  1.1 Terminology Basic terminology in runoff theory is generally well understood and does not need to be reviewed here. However, terminology referring to subsurface stormflow may often be confusing even within current literature. For example, subsurface stormflow may be described as interflow, translatory flow, lateral flow, subsurface runoff, soil water flow or transient flow. These terms will be discussed separately as they are introduced. Flow in the subsurface as a result of precipitation or snowmelt is hereafter referred to as subsurface stormflow. Subsurface stormflow is now recognized as the dominant mechanism for water delivery to the stream in humid environments and there are a multitude of terms used to describe this process. Stormflow itself is streamflow greater than baseflow that is generated by a precipitation event within the duration of the event. Typically, the duration is defined from the event onset to the time when streamflow returns to pre-event baseflow or another  2  event begins. Therefore, subsurface stormflow must be rapid enough to contribute to the stormflow portion of the hydrograph to qualify as stormflow and not simply subsurface flow. Subsurface flow can include both saturated and unsaturated components. It is generally accepted that saturated flow dominates storm runoff and is usually initiated when a saturated zone develops. A saturated zone can develop either at bedrock or a lower impermeable boundary, or within the soil profile between layers of contrasting permeability.  1.2 The Development of Runoff Theory 1.2.1 Early research Infiltration was the first process recognized as being significant to runoff generation during a precipitation event. In the early part of the 20th century, Robert Elymer Horton first described quantitatively the process of infiltration into the soil surface and introduced terminology still used by hydrologists today (Horton 1933). Following Horton, others recognized that surface runoff was not the dominant process responsible for increased stream discharge observed during precipitation events. Beginning with Hursh and Brater’s (1941) work at Coweeta (North Carolina), subsurface stormflow became recognized as a potentially important component of stormflow. Later, studies identified the concept of variable runoff source areas and the importance of subsurface flow as a contributor to event streamflow response (Betson 1964, Hewlett and Hibbert 1963, 1967). Shortly after these developments, “old” water (preevent water stored in the soil/groundwater) was identified as being a significant contributor to runoff (e.g. Pinder and Jones 1969). It is now widely accepted that old water constitutes the majority of stormflow in most cases (e.g. Pearce et al. 1986).  3  Horton (1933) defined infiltration as a result of the need to describe the physical process by which water moves into the soil; it is distinct from other terms sometimes used such as percolation or absorption. Horton defined infiltration capacity as “the maximum rate at which rain can be absorbed by a given soil at a given condition” (Horton 1933, pg. 453). Horton incorrectly concluded that runoff for an individual storm event was mainly or wholly surface runoff, and he attributed surface runoff to rainfall intensities that exceeded the infiltration capacity of the soil. This is widely known as Hortonian overland flow, or infiltration excess overland flow. However, Horton was not working in forested environments. The storm hydrograph response in a forested watershed was shown to consist of subsurface stormflow and channel precipitation by Hursh and Brater (1941). Already, Engler (1919) had recognized the importance of subsurface stormflow after making detailed measurements of infiltration and soil physical properties including porosity, water content, soil texture and hydraulic conductivity (in Weiler et al. 2005). Subsequently, soil depth, topography, and hydrologic characteristics associated with different elevations were shown to influence peak discharge (Hoover and Hursh 1943). The importance of unsaturated flow was first recognized by Hewlett and Hibbert (1963), who concluded that unsaturated flow could not be ignored in hydrograph analysis. Utilizing a concrete trough to observe unsaturated flow at the Coweeta experimental watershed, they coined the term “translatory flow” to describe unsaturated flow. This “translatory flow” was attributed it to the thickening of water films surrounding soil particles, which resulted in a pulse of water.  4  The variable source area concept was developed in the early 1960s and is largely attributed to Betson (1964). Measuring three independent variables (storm rainfall, runoff and soil moisture), Betson noticed that only certain areas of the watershed produced runoff and attributed the observed spatial phenomenon to variations in infiltration capacity; he referred to these as partial contributing areas. Later it was recognized that infiltration capacities are rarely, if ever, exceeded in forest soils and the variable source area concept was extended to incorporate the idea of the expansion and contraction of saturated areas that affect direct runoff of rain falling on those areas (Hewlett and Hibbert 1967). Subsurface stormflow was finally recognized as being an important contributor to event-based stream discharge. In addition, it was previously observed that preferential subsurface stormflow could occur in forest soils (i.e. water moving faster than the soil matrix should allow, typically through some form of soil pipe) (Whipkey 1965). Whipkey was the father of trench studies, in which trenches are commonly excavated along the base of a hillslope down to the impermeable layer and flow from the soil horizons is collected and measured.  1.2.2 Quantifying the process: defining the pathways of storm runoff Development of runoff theory proceeded rapidly through the 1960s and 1970s and the studies conducted by Dunne and Black (1970) set a precedent rarely exceeded for the following two decades. Dunne and Black used intensive instrumentation across various hillslope types to observe subsurface processes in a wet, mountainous area of Vermont. Three hillslopes consisting of well drained sandy loams over glacial till, with convex, concave and straight contours were instrumented with wells and piezometers to measure water table elevation and pressure potential. A nuclear depth probe was used to measure soil  5  moisture along a transect up the middle of each slope. A trench was excavated along the base of the hillslopes to measure runoff at various levels and weirs were installed above and below the reach of river channel running at the base of the study site. Subsurface stormflow was found to occur only during large events and saturation overland flow occurred in significant quantities only on the concave (hollow) hillslope. Overland flow occurring on the concave hillslope during large precipitation events was the only flow measured in large enough quantities to account for the measured stream flow. Anderson and Burt (1978) recognized the importance of the disproportionate contribution of hollows to stormflow in terms of their relative area within a catchment. Saturated wedge development in hollows was observed and shown to be the main contributor to event hydrographs (Anderson and Burt 1978). Other important contributions during this decade include Weyman’s (1973) study, which advocated the theory of a saturated wedge developing from riparian margins and moving upslope with increasing precipitation. Groundwater hydrologists, such as Alan Freeze, were developing theories on regional groundwater flow in the early 1970s (e.g. Freeze and Witherspoon 1967). Freeze (1972) suggested that the majority of event hydrograph response could be attributed to subsurface stormflow. Near the end of the 1970s a series of studies focused on searching for the mechanism that could explain this process (e.g. Sklash and Farvolden 1979). Until this time, subsurface flow was considered to be a function of measurable soil physical properties, namely hydraulic conductivity. However, measurements of soil hydraulic conductivity could not account for the rates of flow necessary to deliver water, via the subsurface, to the stream channel in order to affect the observed stream response. This  6  quandary is resolved by considering, separately, flow in the soil matrix described by Darcy’s Law (where flow is dependent on soil hydraulic conductivity) and preferential flow pathways such as soil pipes and macropores (e.g. Harr 1977). Studies such as that by Mosley (1979) showed that rates of subsurface flow could be large enough to account for the observed hydrograph response in a steep headwater catchment with very moist conditions (M8 catchment, Maimai, New Zealand). Large peak flow rates observed at concentrated locations of soil pit faces were found to coincide with stream hydrograph peaks and dye-tracing experiments were used to quantify the rate of water movement through the profile. Mosley’s dye experiments led him to conclude that the majority of the streamflow response was from the contribution of “event” or “new” water (water contributed by the current precipitation event). Significant debate over the source of water that affects the storm hydrograph response followed Mosley’s (1979) work.  1.2.3 Sources of runoff: changing mechanisms of subsurface stormflow Around the time Mosley was working in Maimai, a new method of examining the source of stormflow was conceived. Pinder and Jones (1969) were the first to use the two component mixing model to separate event water on the basis of chemical signatures by measuring various ions in rain water, storm discharge and stream baseflow. However, it would be almost 20 years before hydrochemical observations were combined with hydrometric observations. Pinder and Jones (1969) concluded that up to 42% of event stream flow might be old water in the Nova Scotia catchment studied. Later, Sklash and Farvolden (1979) measured tritium, oxygen-18 (δ18O), and deuterium isotope ratios across various watersheds and concluded that groundwater was the main contributor to the event hydrograph. The  7  process responsible for transferring old water to the stream in sufficient quantities to explain their observations was attributed to groundwater ridging near the riparian margins. It was thought that rapid conversion of the tension saturated capillary fringe to phreatic water (i.e. saturation occurred soon after the commencement of an event) was responsible for groundwater ridging. Following Mosley (1979), Pearce et al. (1986) and Sklash et al. (1986) published the results of studies in which they examined the relative concentrations of chloride, deuterium and δ18O and electrical conductivity in samples of rainfall, streamflow and soil water flowing from pit faces in the Maimai catchment, New Zealand. Generally, old (pre-event) water and new (event) water were thought to be mixing in the soil profile and discharging to the stream in a fairly uniform mixture in terms of isotopic and chemical composition (Pearce et al. 1996, Sklash et al. 1996). Groundwater ridging and saturated wedge development from the rapid conversion of tension saturated zones to positive potentials were cited as the mechanisms responsible for the delivery of stormflow. However, hydrometric data were not available to verify these claims. If conversion of tension saturated zones to positive potentials was occurring, rapid transmission of new water was not needed to explain stormflow. The majority of stormflow would be contributed by old water already stored in the soil and only a small amount of new water would be needed as input (Pearce et al. 1996, Sklash et al. 1996). To solve the old-water new-water dichotomy, a unification of hydrochemical and hydrometric measurements was necessary and McDonnell (1990) did just that in the same catchment (Maimai-M8) as studied by Mosley (1979), Pearce et al. (1986) and Sklash et al. (1986). Using the same soil pits excavated in the previous studies, a combination of isotope and chemical tracing, and an extensive tensiometer network, McDonnell (1990) observed  8  that water-tables arising at the soil bedrock interface were not maintained, but correlated well with throughflow rates. Soil piping (connection of macropores) was suggested to explain the rapid dissipation of the water table and pore water pressures (McDonnell 1990). To explain his observations McDonnell (1990) suggested that rapidly infiltrating new water perched at the impermeable layer and mixed with larger volumes of old water that subsequently drained as the saturated areas in the hillslope expanded. This would create continuous saturated areas conducive to rapid subsurface stormflow. The formation of these saturated areas is largely dependent on topography. McGlynn et al. (2002) provided a thorough review of the experiments to date at the Maimai research area. No definitive theory yet explains the details of subsurface stormflow and hydrologists are using various approaches to tackle this issue. The above review was focused on setting the background for discussion of current literature. Hydrochemical and hydrometric studies still have not conclusively determined how runoff occurs spatially and temporally in detail. In almost all cases the temporal variability of water composition is not well known. Implicit in the understanding of spatial and temporal variation in runoff is the knowledge of how first order controls vary in space. Topography provides a logical starting point for spatial study of runoff; among the first to examine runoff in relation to varying topography were Beven and Kirky (1979).  1.2.4 Topographic controls on runoff Recognizing that topography controls runoff to a large extent, Beven and Kirky (1979) proposed a simple index to predict runoff that was incorporated into a predictive model. TOPMODEL was the product of Beven and Kirkby’s (1979) first attempt to construct a  9  model explaining runoff and is based on the assumption that elevation gradients are the dominant influence on total hydraulic potential in steep terrain. The simple index ln(a/tanβ), where a is the upslope contributing area per unit contour width and tanβ is the local slope, was the basis for predicting areas of preferred saturation. Therefore, points with a large contributing area and low local gradient would have larger values indicating a greater likelihood of saturation and are thus locations of high flow contributions (Beven and Kirkby 1979, Bonell 1998). The expanding use of digital elevation/terrain models made TOPMODEL an attractive tool for runoff forecasting. TOPMODEL was first applied to Crimple Creek in the United Kingdom and was found to perform well enough to merit further development (Beven and Kirkby 1979).  1.3 Current Directions in Subsurface Stormflow Research 1.3.1 Preferential flow as rapid infiltration Preferential flow has been shown to be important for both flow initiation and rapid lateral transport of water downslope (Mosley 1979, McDonnell 1990). Preferential infiltration has been identified as being significant enough for rapid development of saturated areas and water tables in more permeable soils (e.g. De Vries and Chow 1978, McDonnell 1990, Weiler and Naef 2003). In order for rapid streamflow response to be facilitated by subsurface stormflow, water must infiltrate and move downslope at rates greater than estimates based on soil matrix properties would predict. De Vries and Chow (1978) conducted a plot study in coastal British Columbia, which demonstrated that the removal of the humus layer retarded infiltration rates and suggested the top layer was responsible for rapid delivery of water to root channels and macropores. Preferential flow can occur both via macropores and soil  10  pipes and in areas of higher permeability in the soil, including highly permeable layers (Bonell 1998, McGlynn 2002). Preferential flow via macropores and soil pipes was first emphasized by Mosley (1979). Although much of Mosley’s original theory has been discredited, his work was responsible for facilitating years of debate and groundbreaking research within the area of subsurface stormflow. The rates at which Mosley (1979) observed flow emerging from the soil pit faces were indeed significant, but his attribution of the flow to new water fuelled subsequent debate in the literature. It was not until McDonnell (1990) performed a hydrometric-hydrochemical study at Maimai that a reasonable alternative theory, supported by extensive field measurements, was proposed. To explain the observed old water chemical signatures both in the stream and emerging at the soil pit faces, McDonnell (1990) suggested that rapidly infiltrating water was perching at the impermeable interface and mixing with large quantities of old soil water. Of particular interest is that matric potentials near the bottom of the soil profile responded almost instantaneously to large precipitation events. McDonnell (1990) suggested that the matric hydraulic conductivity was low (5-10 mm h-1) and that ponding of water on the surface occurred readily and flowed via soil cracks and macropores to depth. It is also likely that water was infiltrating the organic layer and flowing laterally to macropores and cracks along the mineral layer (McDonnell 1990). The observed tensiometer response was disproportionate to water table perching; it may be that it takes time for water to “back-up” into the soil matrix and produce a response. McDonnell (1990) suggested that the capillary effect may be discontinuous due to cracks and therefore the higher moisture conditions in the matrix effectively “prime” the soil for incoming new water to flow preferentially in macropores and cracks. It has also been recognized that the  11  permeability of macropore and crack walls may be lower than that of the soil matrix, which would allow for rapid unimpeded flow once water fills these conduits (Calver and Cammeraat 1993; McDonnell 1990). Research has also shown that rapid changes in fluid pressure head and water content could be incorrectly attributed to macropore flow. Torres et al. (1998) proposed that a pressure wave passing through the unsaturated zone might be responsible for the rapid hydrometric response observed lower in the soil profile. Using sprinklers to bring an Oregon Coast Range headwater catchment to steady state, it was observed that pulses or spikes in the water application generated a response that travelled too fast to be attributed to simple advection or plug-flow (Torres et al. 1998).  1.3.2 Rapid transmission of stormflow as preferential flow Sklash et al. (1986) and Pearce et al. (1986) showed that Mosley’s (1979) attribution of preferential flow to new water was not possible in the context of water age; McDonnell (1990) sought an explanation for this dichotomy. As mentioned previously, there could be no doubt that preferred flow was occurring because flow velocities were much higher from the pit faces than those estimated based on soil physical properties. McGlynn et al. (2002) pointed out that soil pipes were not continuous beyond about 25 cm at Maimai, which casts doubt on McDonnell’s (1990) findings. Soil pipes are generally found along the soil bedrock interface in steep forested catchments, though their location may vary (McDonnell 1990). Rates of pipeflow are largely determined by their diameter and it is recognized that there are certain precipitation thresholds that must be exceeded before pipeflow will dominate subsurface flow (Weiler and McDonnell 2007). In Bonell’s (1998) review of runoff  12  generation he states “reconciling their [soil pipes] hydrochemistry coupled with the need for more sophisticated hydrometric studies to address the pipeflow issue, stands out as one of the principal research challenges connected with storm hydrograph separations.”  1.3.3 From surface to bedrock as a control on subsurface flow Beven and Kirkby’s TOPMODEL opened the door for subsequent investigations into the influence of topography that soon lead below the surface. TOPMODEL did not account for subsurface variation. Depth to bedrock or a confining layer may vary from a couple meters below the surface to actually being the surface in a single catchment and this can have dramatic implications for flow routing. Woods and Rowe (1996) tested the TOPMODEL theory and found that bedrock topography was responsible for flow routing as saturated areas developed in hollows and converged as ribbons of concentrated flow. To do this they excavated a trench in the Maimai M8 catchment and measured flow rate and quantity in a series of troughs along a trench face coupled with tensiometer and piezometer measurements. Flow volumes were highly variable and not well predicted by the TOPMODEL structure (Woods and Rowe 1996). In a followup study, Woods et al. (1997) developed a new topographic index by modifying TOPMODEL to account for time varying source areas related to antecedent moisture conditions. Woods et al. (1997) concluded that variability in runoff depends on both topography and soil moisture conditions. Subsequently, Freer et al. (1997) re-evaluated the data of Woods and Rowe (1996) and determined that bedrock topography in place of topography in the TOPMODEL structure significantly improved predictions. Freer et al.  13  (1997) also noted that antecedent soil moisture conditions were a significant control on the occurrence of saturation. Variation in moisture content in the unsaturated zone is dependent on a myriad of factors including soil physical properties, soil depth and topography. Weiler and McDonnell (2004) used a model for the Maimai and Panola experimental hillslopes to demonstrate that soil depth has a significant effect on the volume of flow produced. This was accomplished by varying overall soil depth and the variation in soil depth. Soil depth may also be an important controlling factor on the accumulation of subsurface flow, especially for smaller events (McDonnell et al. 1996, McDonnell 1997). Deeper soils have a greater storage capacity and may significantly extend the length of time to saturation, an effect that will be amplified during small events (Weiler and McDonnell 2004). The evaluations of TOPMODEL conducted by Moore and Thompson (1996) in the UBC Malcolm Knapp Research Forest, Southwestern British Columbia are of great importance to this study. Moore and Thompson (1996) concluded that “estimated location effects were only weakly related to the topographic index ln (a/tanβ), possibly as a result of significant spatial variability in soil transmissivity (assumed negligible in many applications of TOPMODEL) and/or errors in specifying ln (a/tanβ) from a DEM of the surface topography”. This conclusion, coupled with the suggestion that the deviations between water table response and the linear model used may be due to “non-steady state flow conditions”, provides enough justification for a closer examination of how storm runoff in the subsurface is behaving. Mainly, how is stormflow produced in this catchment and how does it vary? Indeed, Hutchinson and Moore (2000) set out to do just that.  14  Utilizing troughs to collect discharge from the soil profile of a hillslope and a network of piezometers, Hutchinson and Moore (2000) examined how throughflow related to the topographies of the surface and the basal till/confining layer. The troughs were oriented such that the hillslope was divided up into units so they could be compared. They found that during the lowest flows the throughflow (subsurface flow) distribution was correlated well with upslope contributing area as calculated from the basal till layer topography. However, at the highest flows, subsurface flow was more closely related to the contributing area of the surface topography. In other words, the saturated area, or water table, shifted from being parallel to the confining layer to being parallel to the surface. Moreover, they observed macropores which can deliver enough discharge to negate the topography as a control on subsurface flow. It is suggested that macropores can route water laterally (at some angle other than 90 degrees to the contours of the water table, confining layer and surface), which questions the validity of models which use topographic controls to predict subsurface flow (Hutchinson and Moore 2000). Topographic models usually assume a quasi steady-state for throughflow, which Hutchinson and Moore (2000) did not find appropriate in that geographic location.  1.4 Tracers as a Tool for Studying Runoff Generation Hydrologists have used tracers to study water movement for several decades and there are a number of different tracers available, some being better suited to specific applications. There are two basic kinds of tracer, those considered natural and those which are artificial. Natural tracers are ones that can be found in the natural environment such as oxygen isotope 18 (18O), tritium, or weathered materials like silicates. These can be measured from water  15  samples taken from soil, precipitation, groundwater and the stream. Artificial tracers are applied to the system; this includes various kinds of dyes and anions like chloride. Chloride is naturally occurring, but it is often applied in much larger quantities to override the natural background concentrations. Natural and artificial tracers both have advantages and disadvantages and neither is necessarily better; even within these two types one tracer may be completely inappropriate where another is very useful. For example, rhodamine WT is not a very useful soil water tracer in lab applications while Lissamine FF is (Trudgill 1987). The first tracers to be used in runoff generation studies were naturally occurring. By measuring the relative concentrations in the different sources (soil water, precipitation, stream) researchers could separate the storm hydrograph into different component sources (e.g. Pinder and Jones 1969). McGuire et al. (2005) used 18O to measure the residence time of water falling on 8 different subcatchments in the H.J. Andrews experimental forest, coastal range Oregon. This was done by measuring the ratio of 18O to 16O in the various input sources to determine the source of the stormflow and how long it has been in the catchment. By accounting for variations in isotope ratios with changes in elevation along with values for snowpack melt water, they were able to determine residence times for the catchments and compare how it varied across scale. Interestingly, residence time was not dependent on scale but was more closely related to simple topographic measures such as median flow path length and gradient. Other types of natural tracers include measurements of silica content and alkalinity. Soulsby et al. (2004) used alkalinity and silica content measurements to determine the main sources of runoff in the Scottish Highlands. Natural tracers lend themselves well to  16  catchment scale and larger studies while artificial tracers are convenient for hillslope and plot scale applications. The main drawback of natural tracers is the uncertainty associated with characterising the sources. Chemical signatures are variable in both space and time. The chemical signature of soil moisture varies spatially and temporally depending on the length of time that moisture has resided in the soil. In turn, residence time of soil water is dependent on the length of time since the last precipitation event and the size of that event. This dependence makes it difficult to account for the soil water signature. Additionally, the influence of interception and the spatial variation in chemical composition of precipitation at the catchment scale has received little attention to date. Most hydrochemical studies have focused on the hillslope scale and often use only a single rain gauge to determine the chemical signature of rainwater. As McGuire et al. (2005) pointed out, the isotope signature of rainfall varies with elevation. Deposition of minerals and soil physical properties can vary over small spatial scales, which will alter the chemical signature of water flowing through those areas. It would be nearly impossible, or at least very labour intensive, to account for such variations. Nonetheless, naturally occurring tracers continue to be used and do provide certain advantages over artificial tracers, mainly that they can be used on a larger scale and that they are ubiquitous. Artifical tracers overcome the uncertainty associated with characterizing naturally occurring tracers in that we can control when, where and how much to apply. Artificial tracers have been around for some time, having been used as early as the late 1960s (e.g. radioactive tracers used by Pilgrim 1966). Exploration of the utility of tracers in hydrological study continued through the 1970s (e.g. Pilgrim and Huff 1978, Smart and Laidlaw 1977). Pilgrim and Huff (1978) demonstrated the usefulness of artificial tracers for monitoring  17  water movement in the subsurface and observed irregular patterns of movement despite a uniform surface. More recently, dye experiments have been used to examine infiltration in greater detail (Weiler and Flühler 2004, Weiler and Naef 2003). Weiler and Flühler (2004) used simulated rainfall with brilliant blue dye followed by soil pit excavation and image analysis to examine infiltration. Extended vertically stained sections of macropore flow were observed which initiated close to the soil surface. This finding is in accordance with the conclusions of McDonnell (1990). Weiler and Naef (2003) concluded that although macropores make up a much smaller fraction of the total porosity (<1%) they account for the majority of saturated flow and preclude the use of Darcy’s law or Richard’s equation to predict flow rates. Artificial tracers have also been used at the hillslope scale, often being injected via piezometers at a specific depth (e.g. Talamba et al. 2000) or as a line application at the top of a hillslope or hillslope plot. There is now an opportunity to test artificial tracers at the small catchment scale. Application is the biggest limitation with the use of artificial tracers. It is either labour intensive or expensive in the case of sprinkler systems. Rodhe et al. (1996) and Lange et al. (1996) conducted studies in a catchment which has been covered below the canopy so the chemical signature of the input water could be controlled. Rodhe et al. (1996) used 18O ratios while Lange el al. (1996) used LiBr as their tracers. Of greatest interest are the results of Lange et al. (1996), who had low recovery and concluded that residence time was long enough to permit equilibrium exchange between the soil water and soil matrix. They believed that hydrochemical processes related to catchment runoff are underestimated because they are often based on soil column studies that do not account for lateral movement.  18  While artificial tracers have been used in hydrology for some time, their usefulness as a tool for studying runoff generation has not been extensively explored. Artificial tracers were chosen for this study in order to investigate their utility as a more meaningful tool at the small catchment scale.  1.5 Motivation, Objectives and Hypothesis Through the previous sections it has been shown that subsurface stormflow is the dominant runoff process in steep headwater catchments and a solid understanding of subsurface stormflow processes is still lacking. In order to fully understand how changes to the landscape may affect chemical composition of streamwater and the timing and volume of peak streamflows, we must first understand the dominant controls on subsurface stormflow. Weiler and McDonnell (2004) have shown that controls which may have been previously overlooked, such as drainable porosity, are significant for predicting subsurface stormflow. An experimental approach uniting hydrological modelling with field experiments is an effective way to explore first order controls on runoff dynamics (Weiler and McDonnell 2004). Future goals should include the explanation of the origin and timing of stormflow, which will lead to a better understanding of chemical interactions at the hillslope scale (Weiler et al. 2003). Thoughtful experiments coupling hydrochemical and hydrometric measurements with experimental modelling provide a promising tool for identifying first order controls (Weiler and McDonnell 2004). Once first order controls are established, transferability of results to larger scales will become more feasible. Recently, hydrologists have suggested that isolating definable catchment units rather than focusing on the methods of transport (e.g. preferential flow, overland flow) may be a  19  better way to map and thus predict runoff (e.g. Sidle et al. 2000, McGlynn et al. 2004). Sidle et al. (2000) presented a hydrogeomorphic conceptual model based on experiments in Hitachi Ohta Experimental Watershed, which explains that as a watershed becomes progressively wetter different areas or units are switched on as they become linked through nodes. The nodes may be features such as the base of a hollow where saturation is occurring but rapid transmission of water cannot take place until that area is linked to the stream via some other feature such as a macropore network. This concept provides a useful framework for continued study focusing on how these units behave and act together to explain storm runoff. McGlynn et al. (2004) also pointed out that modelling has outpaced field studies in recent years. McGuire et al. (2005) reported findings that suggest topographic features have a greater influence on residence time than catchment size in the Oregon coastal mountains. The aim of this research is to observe how transit time differs between different landscape units (hypothesized to be dominated by different runoff generation processes). The landscape units are areas of uniform physical features such as soil depth, slope and proximity to surface water. The climate and terrain of southwestern BC are similar to those studied by McGuire et al. (2005). To further investigate how physical characteristics may affect the timing and flowpath of runoff, three landscape units of contrasting topographic and soil characteristics were chosen for study. These were isolated using three different tracers applied aerially (in solution with a backpack sprayer) to measure the timing and contribution from these areas. This study was both an exercise in tracer performance as well as a study of the processes themselves. The hypothesis being tested is that a catchment can be delineated into distinctive landscape units based on the dominant runoff generation processes (DRPs) expected to occur  20  in that unit, which are based on physical characteristics (namely soil depth, slope and proximity to the surface water network). Furthermore it is hypothesised that each unit will deliver water to the stream on a unique time scale which is a function of the DRPs. The questions this study seeks to explore stem from a set of basic questions which, simply stated, are: (1)  Are the time lags between the peak stormflow and the peak contribution different from each area?  (2)  How do storm size and antecedent conditions affect the timing of the contributions from these areas?  (3)  How does subsurface response vary between these areas?  It is expected that the response from each landscape unit will be indicative of a different runoff process. Given our knowledge or runoff generation processes this study asks, what might that process be? We already have a good idea of what processes are at work from previous studies, particularly in BC and Oregon. Identifying and linking dominant runoff generation processes with a unique timing signature to different landscape units is the objective of this study.  21  2.0 METHODS 2.1 Site Description 2.1.1 Location The study site was located in the Malcolm Knapp Research Forest (MKRF) owned by the University of British Columbia. MKRF is located just outside the town of Maple Ridge approximately 50 km east of Vancouver between the Pitt and Fraser rivers immediately before the Pitt River joins the Fraser River (Figure 2.1). A study catchment 7 ha in area containing a small ephemeral headwater stream was chosen as the study site. Previous studies in the same watershed garnering vital survey data and useful hydrologic information (Thompson 1994), in addition to ease of access, were primary factors in choosing the site. An unmaintained forest road bisects the catchment into roughly two halves, an upper and a lower half. All tracer experiments discussed herein were conducted below the road during the winter and spring of 2006.  22  Figure 2.1 Location of University of British Columbia Malcolm Knapp Research Forest  2.1.2 Climate Thirty year climate normals taken from 1971-2000 at the UBC Research Forest entrance and administration building (Environment Canada station 1103332) are shown for average monthly precipitation, and temperature in Figure 2.2. The climate is typical of southwestern British Columbia, showing a distinct rainy season (winter months) with an average yearly temperature of 9.6C and all average monthly temperatures above zero. The climate normals show the average minimum temperature only falls below 0C in January (-0.5C) and December (-0.1C). This low elevation site is ideal for tracer experiments during the rainy season due to the mild temperatures, which result in little to no snowfall.  23  15 10 5 0  Temperature (oC)  20  Ja n Fe b M ar Ap ril M ay Ju n Ju ly Au g Se pt O ct No v De c  Precipitation (mm)  350 300 250 200 150 100 50 0  Month Precipitation"  Temperature  Figure 2.2 1971-2000 Average monthly precipitation and temperature normals from Environment Canada UBC research forest administration climate station  Figure 2.3 shows the 1971-2000 averages compared to the period of study. Maximum monthly precipitation typically occurs in November (323 mm). Total precipitation for January of the study period vastly exceeds that amount at 538 mm. The months following January all fell short of the monthly average temperatures. Temperature for the study period was similar to the 30 year normals, aside from January, which was 3 C warmer at 5.3 C than the normal at 2.3 C. Temperature data for February are missing.  24  20  500  15  400 300  10  200  5  100 0  0 Jan  Feb  Mar  April  May  Jun  Temperature (oC)  Precipitation (mm)  600  July  Month Precipitation 1971-2000  Precipitation 2006  Temperature 2006  Temperature 1971-2000  Figure 2.3 1971-2000 average monthly temperature and precipitation at the UBC research forest administation climate station versus averages from 2006 study period  2.1.3 Topography The catchment’s elevation ranges from 190 to 280 masl (meters above sea level) with the logging road bisecting the catchment roughly in the center at 220 masl (Figures 2.4 & 2.5). The average gradient of the lower catchment is approximately 17%. Figure 2.4 shows a contour map of the study site with the sites for tracer application and the observation points.  25  Figure 2.4 Map of study site with UTM coordinates. Numbers 1, 2, and 3 represent tracer applications areas R-CL, SS-Br and Sts-Ur, respectively.  Figure 2.5 Road cut bisecting upper and lower halves of the study catchment in Malcolm Knapp Research Forest, Fall 2006.  26  2.1.4 Geology and soil The granitic bedrock is overlain by thin Vashon Glacial Drift and coarse textured humoferric podzols that range in depth from 0 m, where bedrock is exposed, to 2 m. Soils exhibit a strongly structured B horizon with many roots, stones and aggregates (Tischer 1986, from Thompson 1994). Cheng (1988) measured hydraulic conductivities ranging from 3x10-5 to 9x10-5 m s-1 (from Thompson 1994). Thompson (1994) also measured saturated hydraulic conductivities of 9x10-5 m s-1 at 0.5 m depth and 3x10-3 m s-1 for the surface layers.  2.1.5 Vegetation Vegetation consists mainly of Thuja plicata (Western Redcedar) and Pseudotsuga menziesii (Douglas Fir) and Tsuga heterophylla (Western Hemlock). The area was logged and replanted in the 1920s; thus the current stand is 70-90 years of age. The average tree height was 29 to 37 m and the crown closure ranged from 76-85% as of 1994 (Thompson 1994). Underbrush is sparse except near the road where it flourishes during the spring and summer. There is substantial deadfall and decaying stumps on the forest floor as well as a fairly uniform covering of leaf litter.  2.1.6 Hydrology The small catchment has an ephemeral surface water channel that was present for approximately 11 months in 2005 and 10 months in 2006. The visible surface water network significantly expands in the winter months during large rainfall events. Due to varying soil depth and numerous bedrock outcrops, a variety of runoff process are present. These processes include overland flow from saturated and impermeable areas and both saturated  27  and unsaturated subsurface flow. Maximum measured discharge from the 7 ha catchment occurred in January 2006 at 34.6 L s-1. Table 2.1 summarizes the precipitation inputs and surface outputs during the study period showing substantial quantities are lost to evapotranspiration and/or deep seepage or groundwater contributions.  Table 2.1 Measured inputs and outputs in the study catchment during the 2006 study period Month  Jan  Feb  Mar  April  May  Jun  Total  Precipitation (mm)  285  225  200  160  131  103  1105  Discharge (mm)  292  66  51  46  44  30  529  Difference (mm)  -6  159  149  113  88  72  576  2.2 Data Collection – Instrumentation 2.2.1 Precipitation Precipitation data were collected using ten Rainwise tipping bucket rain gauges that utilise an internal data logger. Nine of the gauges were below the forest canopy, with a tenth placed in an adjacent clearing. The tops of the buckets were levelled at 1.3 m above the surface on fence posts driven into the ground. In addition, data are collected at the MKRF office, Environment Canada station 1103332 (HANEY UBC RF ADMIN). Gauges had to be monitored due to clogging of the funnels and curious bears. For the later part of the second tracer application data collection period, some of the Rainwise loggers were switched for remote loggers which were used for a number of other instruments at the tipping bucket sites in conjunction with a related project. The remote loggers are manufactured by Crossbow and were housed in waterproof Lexan boxes fixed to the tipping bucket fence posts. The tipping  28  buckets themselves were not changed. The Rainwise data loggers recorded tips at the time they occurred and these data were processed into rainfall intensity for daily and fifteen minute intervals.  2.2.2 Surface flow Streamflow was measured at three locations. The main weir at the bottom of the catchment was constructed using lumber, steel and bentonite clay (Figure 2.6). The weir is a 90 degree V-notch and was instrumented with an Odyssey water level recorder logging every 10 minutes and later a Trutrack water level record was added. This is also the site of the various tracer probes used to measure tracer concentration. The water level readings were calibrated by taking manual measurements of the height above the V-notch throughout the study period. Salt dilution gauging was also used to calibrate the weir and compare to the theoretical discharge based on the equation (Gordon et al. 1992): Q = 1.342H2.48  (1)  where Q is discharge (m3s-1), H is the height above the V-notch (m) and the coefficients are based on a 90 degree V-notch weir.  29  Figure 2.6 Main weir and probes at research catchment, MKRF  Two other V-notch weirs, constructed from plastic laundry tubs, measured flow entering the lower catchment at the lower side of the road (Figure 2.4). The western culvert drains the upper catchment while the eastern culvert drains water diverted from an adjacent catchment. Odyssey water level recorders were placed in PVC stand pipes connected to the tub drain. A tee joint was used with one side leading to the Odyssey and the other serving as a drain to clean out accumulated sediment. The V-notch was 60° and the weirs had plywood splash guards and wire mesh to capture debris coming through the culverts and prevent clogging. The tubs were cemented in place during low flow and no flow (dry) conditions.  30  2.2.3 Subsurface flow While subsurface flows were not measured directly, subsurface water table dynamics were measured using five piezometers. The piezometers were fitted with Odyssey data loggers housed in PVC pipes sunk down to the till or bedrock layer, depending on what was possible at a given location. The bottom of the pipes had a 10 cm section drilled with many holes, screened with landscaping fabric and capped on the bottom. These were installed in bore holes made with a hand auger. The piezometer was then placed in the hole and sand was used to fill the lower section, which was topped with bentonite clay to form a seal that would prevent inflow from the above soil horizons. Then the original substrate that was removed during the augering was used to fill the rest with a final seal of bentonite at the surface. Bentonite was used at the surface to prevent water from flowing vertically down the outside of the piezometer as the new disturbed soil would provide an ideal pathway.  2.3 Tracer Experiments 2.3.1 Site selection Three tracer application sites were chosen based on topography, soil depth and proximity to the stream network. These characteristics were to be as different as possible to increase the likelihood of distinct DRPs being responsible for runoff generation at each site. Figure 2.4 shows the three application sites. R-Cl (Riparian-Chloride) borders the stream channel and is 617 m2. SS-Br (Shallow slope-Bromide) is characterised by deeper soils and gentle slopes relative to the rest of the catchment and is 495 m2. The third site, StS-Ur (Steep SlopeUranine), is characterised by a very steep slope with large bedrock outcrops running perpendicular to the fall line which results in a terraced like terrain. StS-Ur is 496 m2. Each  31  of these sites was selected to represent commonly found combinations of physical characteristics in British Columbia’s watersheds.  2.3.2 Tracer application Three different tracers were selected for the experiments to allow the identification of the individual responses from each landscape unit concurrently. Two tracer applications took place, the first on January 4th and the second on March 20th, 2006. On January 4th, 58 g of rhodamine WT, 10 kg of KBr and 53 g of uranine were applied to sites 1 (R-Cl), 2 (SS-Br) and 3 (StS-Ur), respectively (Figure 2.4). Due to the low concentrations measured following the first application the amount of tracer was increased for the second application. The second application on March 20th, 2006, consisted of 27.3 kg of bromide, 30.5 kg of chloride and 59.89 g of uranine plus another 57.81 g of uranine applied on March 21st, 2006. Initially, two assumptions were employed to calculate the tracer mass required: 1) that the average flow would be 12 L/s (based on the previous four months of measured discharge) and 2) the bulk of the tracer could be expected to move through the system in one week (based on the small size and proximity to the stream it was thought that the tracer would move through the system rapidly). The masses were calculated to achieve average concentrations of 10 µg/L for dye and 10 mg/L for ion tracers. The first application had wet initial conditions and was followed by heavy rain while the second had dryer initial conditions followed by moderate amounts of rain.  32  2.3.3 Tracer breakthrough As with discharge, tracer breakthrough was measured at the main weir installed at the outlet of the catchment. Measurements consisted of fluorescence probes, ion selective probes and automated water samplers both for back-up in case of probe failure and for added precision. During the first tracer application the GOT Gotschy Optotechnik LLF-M Fiberoptic Fluorometer was used to detect uranine and rhodamine dye breakthrough. This is a stand alone unit with its own logger and software; it is easy to use and was powered with a 12V automobile battery. A Turner Designs Cyclops-7 fluorometer was also used to measure rhodamine during this period. The Turner Designs probes were less costly and could be used in conjunction with a non-proprietary data logger, which were significant advantages over the large GOT Fluorometer. Using the Turner Designs Cyclops probes also allowed more frequent data recording than the GOT fluorometer. With a dynamic range of 0.004 ppb to 1000 ppb (g/L) the Cyclops probe was ideal for the low concentrations associated with this type of experiment. Bromide and chloride concentrations were measured using Instrumentation Northwest Inc. TempHion-2 ion selective probes. The GOT fluorometer recorded data every 10 minutes while a Campbell CR1000 stored readings from the TempHion and Cyclops probes every minute. For the second application an ion selective probe for chloride was added as well as a Cyclops uranine probe. The fluorescent dye probes were shrouded in a drilled 2” white PVC pipe within a 6” white PVC pipe, which was drilled on the lower half to allow through flow of stream water. The ion selective probes were shrouded in drilled 2” PVC. All probes were placed as close as possible to the bottom center of the stream channel (Figure 2.5). The probes were removed after surface water ceased flowing in July 2006.  33  Two ISCO model 6712 automated water samplers were used to collect samples with intervals ranging from 2 hours to 8 hours. Each sampler contained 24 bottles. One sampler was programmed to begin sampling following the other. Samples had to be collected every 4 days during the most frequent sampling periods which immediately followed tracer applications. The samples were decanted into opaque high density polyethylene Nalgene bottles to prevent photodegradation of the fluorescent dyes. Once these samples were collected they could be analysed in the lab using the same instrumentation that was used in situ, but having the benefit of more control over variables like light and temperature. This procedure allowed for corrections to be applied, thus eliminating the error associated with variable field conditions. For the salt tracers a Metrohm model 861 Advanced Compact Ion Chromatographer (IC) was used to obtain precise measurements of the chloride and bromide concentrations. The relationship between the Cyclops measurement and the IC measured concentration was then used to correct the in situ data set.  34  3.0 RESULTS 3.1 Data Collection/Experimental Design Intensive data collection coupled with the tracer applications occurred between January 1st 2006 and June 30th 2006. Not all instruments or variables were continuously recorded during this time due to limited resources, instrument failure and natural interference. For example, it was common for the rain gauges to become clogged with falling litter, thus rendering blocks of data unreliable and making it necessary to discard them.  3.1.1. Measuring input - precipitation Data from all 9 below canopy gauges could not be used at all times due to clogging. Gauge 5 was so problematic it had to be eliminated altogether. The mean throughfall was calculated from as few as one gauge and as many as eight gauges for a single time interval. Data from the single gauge placed in an adjacent clearing were used as a proxy for above canopy precipitation. It was necessary to establish a relationship between below and above canopy precipitation from January 3rd to April 10th as data from the original loggers were problematic in the later part of April. Following this problematic period the loggers were independently replaced with new ones in conjunction with another study. Data collection with those loggers proved problematic and no usable data for the period of interest was obtained. Figure 3.1.1 shows the relationship between above canopy precipitation and throughfall described by P BC = 0.77(P AC ) – 0.56  (2)  35  where P BC is the throughfall and P AC is the above canopy precipitation for daily totals. Solving the equation for zero throughfall yields a threshold value for above canopy precipitation of 0.73 mm, below which no throughfall is measured. This equation was used on precipitation following April 10th to extend the below canopy rainfall record.  Above Canopy Precipitation (mm)  50 45 40 35  2  y = 0.77x - 0.56, R = 0.97  30 25 20 15 10 5 0 0  10  20  30  40  50  60  Below Canopy Precipitation (mm)  Figure 3.1.1 Comparison of measured average daily throughfall to daily rainfall measured in the clearing (taken to be above canopy) from January 1st to April 10th 2006  3.1.2 Measuring output – discharge Discharge was calculated based on the theoretical weir equation as mentioned in the methods section. Stage measured by the Odyssey probe had an approximately constant difference from manual staff gauge readings (Table 3.1.1). The mean difference was used to adjust the Odyssey data prior to analysis.  36  Table 3.1.1 Manual stage measurements compared to Odyssey data Date  Odyssey  Staff  Difference  Nov. 26  182  48  134  Dec. 8  182  47  135  Dec. 28  246  98  148  Jan. 3  210  76  134  Jan. 4  209.5  76  133.5  Jan. 6  259  126  133  Jan. 9  327.5  190  137.5  Jan. 13  340  208  132  Jan. 19  240  108  132  Jan. 26  222  89  133  Mean Difference  135.2  While the Odyssey sensor performed well for measuring stage, the streamflow data were verified through salt dilution gauging. Five dilution gaugings were performed, which yielded three useable results. Figure 3.1.2 shows the three dilution gauging results compared to the theoretical calculation. The theoretical discharge is clearly greater than that calculated by dilution gauging. The dilution gaugings were performed under less than ideal conditions, lending doubt to their validity. These conditions included a braided and turbulent channel containing significant amounts of woody debris and very low stream flow. With only three good measurements made during the summer, it was considered insufficient to adjust the entire theoretical discharge data set.  37  Height above V-notch (cm)  7 6 5 4 Dilution gauged measure  3  Theoretical value for dilutions Theoretical Relationship  2 1 0 0  0.2  0.4  0.6 0.8 Discharge (l/s)  1  1.2  1.4  Figure 3.1.2. Theoretical discharge curve compared to dilution gaugings  3.1.3 Subsurface measurements Piezometers were deployed prior to the selection of tracer application areas as the catchment was set up before the experiment took shape and because their spatial distribution was critical to a parallel study. The result was that there were two piezometers in each plot. One piezometer in ShS-Br exhibited problematic data, which appeared to be a result of the logger cutting in and out and those data were discarded. In StS-Ur the one piezometer was located on a rocky shelf and exhibited unique characteristics, so it could not be used in conjunction with other piezometers to compute an average response. In its place, a piezometer located outside of the plot area but at the base of the slope was used as an indicator of the slope's cumulative response in the convergence zone.  38  3.1.4 Performance and logistics of artificial tracer use Two types of probes were used in this experiment with mixed success. Performance varied greatly with the type and brand of probes, which were limited in different ways. This was expected based on differences in price as well as design. The GOT probe is more expensive than the Turner probes and had a built-in shroud over the sensor. The GOT probes also were a complete unit housed in a travel case with a proprietary logger requiring its own software. The unit was easy to program and download from. The data did not exhibit any drift over time. The Turner probes, on the other hand, are designed to be used with any stand alone data logger and the Campbell Scientific CR1000 worked well, although their use required more programming as the CR1000 is often performing an array of functions. Low cost was a big advantage of the Turner probes as was the ability to record at shorter time intervals due to the greater memory capacity of the CR1000. Data collected with both loggers had to be checked against the water samples taken using the ISCO automated sampler.  Rhodamine Rhodamine was one of the three tracers chosen for this study as it had been used successfully in the past for many hydrological studies. It was applied in the near stream zone (riparian area) because adsorption is a known disadvantage in the use of rhodamine and this zone has the shortest travel distances to the stream channel. However, there was no useable response measured with the GOT probe during the first application period and the water samples  39  yielded no measurable response. Rhodamine was replaced by chloride in the second application for this reason.  Uranine Uranine was used in the same location for both applications, characterised by steep slope, bedrock outcrops and shallow soils. The first application showed a measurable response with the GOT fluorometer, which was validated using the GOT in the lab on the collected ISCO water samples. The results were promising and the GOT was used in addition to the Turner Cyclops probe during the second application. The Turner probe tended to shift when the battery was changed and also drift as battery voltage dropped despite using the regulated 12V output of the CR1000 logger. Using the difference between the concentrations measured in the lab with the GOT and those measured with the Cyclops in situ, a moving median filter was used to calculate a correction for the Cyclops (Figure 3.1.3). However, even after the correction was applied there was no discernable relationship between the corrected readings and the ISCO samples (Figure 3.1.4). The uranine concentration was very low contributing to the lack of an observable response with the GOT fluorometer.  40  Uranine Concentration Difference [ug/L]  1.5 1 0.5 0 -0.5 -1 -1.5 -2 60  80  100  120  140  160  Day Of Year 2006  Corrected Turner concentration [ug/L]  Figure 3.1.3 Deviation of in situ Cyclops-measured uranine concentration from ISCO water sample concentration analysed using the GOT fluorometer  1.5 1 0.5 0 -0.5 -1 0  0.5  1  1.5  GOT concentration [ug/L]  Figure 3.1.4 Uranine concentration from ISCO samples determined using the GOT fluorometer versus the corresponding reading from the in situ Turner probe; corrected using a moving median filter.  The GOT data exhibited a drifting deviation from the concentration of the ISCO samples as measured in the lab (Figure 3.1.5). The relationship between DOY (Day Of Year) and the deviation of the in situ reading from the ISCO sample was applied to the GOT readings and  41  showed much better correlation than the corrected Cyclops readings (Figure 3.1.5). This relationship was then applied to the in situ GOT uranine readings to correct the breakthrough concentrations. Unfortunately, the GOT fluorometer was removed from the field on May 3rd, 2006. Only the Turner-Cyclops probe remained in the field, which resulted in a shorter data set for uranine.  60  Voltage (mV)  50  y = 0.0019x 2 - 0.3592x + 40.32  40 30 20 10 0 0  50  100  150  200  Day of Year 2006  Figure 3.1.5 Difference between GOT fluorometer readings taken from ISCO water samples and corresponding in situ Turner Probe readings in millivolts over time for uranine  Bromide During the first experiment the ion selective probe showed a marked response with little noise (Figure 3.1.6). However, breakthrough concentrations were lower than expected due to the higher than expected discharge, prompting the use of a greater quantity in the second experiment.  42  Voltage (mV)  100 80 60 40 20 0 03  08  13  18  23  Day Of Year 2006 Figure 3.1.6 Bromide probe millivolt reading from TempHion ion selective probe for bromide over time  The ISCO samples were analysed in the lab using a Metrohm Ion Chromatograph. This allowed the establishment of a relationship between the assumed actual concentrations measured using the IC and the in situ field measurements with the TempHion probe (Figure 3.1.7). The IC allowed for precise measurements with a high degree of accuracy. The ion selective probe measurements could then be adjusted to fit the relationship between field concentrations and IC concentrations. On March 3rd there was a shift in the probe reading by about 10 millivolts and therefore Figure 3.1.7 shows two data sets, one before the shift and one following.  43  2.5  Sample Conc. [mg/L]  2  1.5  1  y = -0.6598Ln(x) + 2.7242  0.5 y = -0.5629Ln(x) + 2.3911  0 0  5  10  15  20  25  30  35  40  45  50  Probe Reading [mV] March 23-May 30  Jan. 6-Jan.14  Figure 3.1.7 ISCO water sample bromide concentration as determined using ion chromatography versus corresponding in situ TempHion probe reading for bromide  Chloride Chloride replaced rhodamine in the second tracer application. The same Instrumentation Northwest Inc. TempHion-2 probe was used in conjunction with the ion chromatograph to establish chloride breakthrough concentrations. Unlike the bromide measurements, the chloride probe data showed no clear relationship with the concentrations measured from the ISCO samples using the IC. As a result, several different relationships had to be fitted in order to adjust the probe data (Figure 3.1.8). This may have been a result of the unusual shifts associated with drifting battery voltage and the “jumps” which occurred due to battery changes.  44  8 y = 0.0059x2 - 1.34x + 78.47 Sample Conc. [mg/L] from IC  7 y = 0.0075x2 - 0.77x + 22.92  6 5 4 3 2  y = 0.014x2 - 2.58x + 120.26  1  y = 0.0031x2 - 0.23x + 5.76  0 20  40  60  80  100  120  Probe Reading [mV] April 14th to Apr. 17 14:00  March 23 19:00 - April 11 10:00  April 11 12:00 - April 14 6:00  April 17th 14:00 - May 5th  Figure 3.1.8 ISCO water sample chloride concentration determined using ion chromatography versus corresponding in situ TempHion probe reading  3.2 Experimental Results 3.2.1 Precipitation The total below canopy rainfall measured from January 1st to June 30th, 2006, was 813 mm. The total above canopy rainfall measured was 1157 mm, which translates into 334 mm or 29% of rainfall being intercepted. However, it should be mentioned that precipitation also fell as snow during this period, which melted quickly and did not accumulate. This could increase the interception rate as the canopy would be able to block more incoming snow than rain allowing more time for evaporation and sublimation to occur.  45  Maximum daily rainfall measured below the canopy was 44 mm, with 50 mm measured in the clearing (above canopy) on January 13th, 2006. The maximum daily rainfall measured in the clearing (above canopy) of 51 mm, with 41 mm measured below the canopy, was recorded on January 9th, 2006 (Figure 3.2.1). The difference in percentage throughfall on these two days is likely a result of which below canopy gauges were used in the calculation and the possibility that some precipitation fell as snow, though none was recorded at the research forest climate station. The highest precipitation intensity measured was 2.9 mm in 10 minutes on April 9th, 2006. During the first tracer experiment period 532 mm of rain fell below the canopy compared to 290mm for the second application period.  50 45  Precipitation (mm)  40 35 30 25 20 15 10 5  01/0 1/06 11/0 1/06 21/0 1/06 31/0 1/06 10/0 2/06 20/0 2/06 02/0 3/06 12/0 3/06 22/0 3/06 01/0 4/06 11/0 4/06 21/0 4/06 01/0 5/06 11/0 5/06 21/0 5/06 31/0 5/06 10/0 6/06 20/0 6/06 30/0 6/06 10/0 7/06 20/0 7/06 30/0 7/06  0  Date  Figure 3.2.1 Daily rainfall below canopy  Table 3.2.1 summarizes the monthly precipitation, throughfall and discharge values during the study period while Figure 2.3 shows the 1971-2000 normals compared to the period of study. There is some inconsistency between the climate station and the catchment, but the  46  discharge values agree well with the throughfall. Maximum monthly precipitation typically occurs in November (323 mm). Total precipitation for January of the study period vastly exceeded that amount at 538 mm. The months following January all fell short of the monthly average temperatures. Temperature for the study period was similar to the 30 year normals, aside from January, which was 3 C warmer at 5.3 C than the normal at 2.3 C. Temperature data for February are missing.  Table 3.2.1 Measured monthly inputs and outputs in the study catchment during the 2006 study period Month  Jan  Feb  Mar  April  May  Jun  Total  Precipitation (mm)  285  225  200  160  131  103  1105  Throughfall (mm)  402  58  95  100  94  46  795  Discharge (mm)  292  66  51  46  44  30  529  Difference T-D (mm)  110  -8  44  54  50  16  266  3.2.2 Discharge Figure 3.2.2 shows discharge from January 1st 2006 to June 31st 2006. January was the most eventful month with large discharge peaks that reflect the significant rainfall inputs. The largest events occurred during the first tracer application period. A maximum instantaneous discharge of 34.6 L s-1 was recorded on January 13th at 13:26 (Figure 3.2.2). From January 1st 2006 to June 30th 2006, 37.0 x 106 litres or 529 mm of water flowed past the lower weir, leaving 294 mm or 36.2% of the input presumably lost to evaporation, storage and deep groundwater contributions and subsurface flow around and below the main weir. Table 3.2.1  47  summarizes the monthly discharge and precipitation for the study period, showing substantial quantities are lost to evapotranspiration and/or deep seepage or groundwater contributions. 40 35 30  Q (L/s)  25 20 15 10 5  01 / 07 01 / 14 01 / 21 01 / 28 01 / 04 01 / 11 02 / 18 02 / 25 02 / 04 02 /0 11 3 / 18 03 / 25 03 / 01 03 / 08 04 /0 15 4 / 22 04 / 29 04 / 06 04 / 12 05 / 19 05 /0 26 5 / 02 05 / 09 06 / 16 06 / 23 06 / 30 06 /0 6  0  Day/Month in 2006 Figure 3.2.2 Discharge from main weir in MKRF study catchment during 2006 study period  3.2.3 Tracer breakthrough and recovery Unfortunately, the mass of rhodamine applied to the riparian plot was not enough to be detected for the first application. The variation of in situ concentration readings was within the margin of error to be considered reliable. However, bromide, uranine, and in the second application, chloride, bromide and uranine were detected in sufficient quantities to calculate tracer breakthrough.  48  Tracer breakthrough was dependent on both location of the application and the size of the event. The size of the event could be measured both as input, peak discharge and 24 hour discharge values. Figures 3.2.3 show tracer concentrations measured in the stream over time for bromide, uranine and chloride, respectively. Close examination of the discharge and concentration time series reveals a time lag between the peak storm flow and peak concentration. Figure 3.2.3 (A) shows a clear response from SS-Br with low levels of background noise, especially for the second application, where bromide concentrations are higher. There are several clearly definable peaks with sharp recessions following the pattern of discharge quite clearly. Interestingly, May and June show slow recessions after the major storms. Figure 3.2.3 (B) shows StS-Ur had significantly more background noise and fewer clearly definable peaks. However, there were still several peaks that could be used and they also followed the pattern of discharge more closely than SS-Br. In particular, the long recession observed in SS-Br was not observed in StS-Ur, partially due to the lack of a complete data set. Figure 3.2.3 (C) shows R-Cl to behave in a very flashy manner, with no recession and relatively steady base flow concentrations. It is difficult to observe the presence of a time lag in R-Cl. Pre-event concentrations return quickly in sharp contrast to a more gradual fall in concentration from SS-Br. To get a better idea of the difference between tracers and therefore application sites, it was necessary to measure this quantitatively. To do this the lag time from peak event discharge volume to peak tracer concentration was plotted against peak event size to compare between the response from the different landscape units.  49  40  2.5  A  Discharge Bromide  2  Q (L/s)  1.5  20  1  Concentration [mg/L]  30  10 0.5  0  0 1/3/06  2/12/06  3/24/06  5/3/06  6/12/06  Date (dd/mm/yy)  Figure 3.2.3 A Discharge and tracer concentration over time for bromide (SS-Br)  50  40  1.2  B  Discharge Uranine  30  Concentration [g/L]  Q (L/s)  0.8  20  0.4 10  0  0 1/3/06  2/12/06  3/24/06  5/3/06  6/12/06  Date (dd/mm/yy)  Figure 3.2.3 B Discharge and tracer concentration over time for uranine (StS-Ur).  51  40  8  Q (L/s)  30  6  20  4  10  2  0  0 1/3/06  2/12/06  3/24/06  5/3/06  Concentration [mg/L]  C  Discharge Chloride  6/12/06  Date (dd/mm/yy)  Figure 3.2.3 C Discharge and tracer concentration over time for chloride (R-Cl)  52  SS-Br clearly shows the greatest lag times. There are fewer data points for the comparison of R-Cl and SS-Ur; however, what is clear is that both R-Cl and StS-Ur show much smaller lag times than SS-Br. SS-Br lag time decreases with increased input, which results in larger discharge volume and discharge peaks. At some threshold value the lag to peak tracer concentration for all tracers/application areas becomes more or less equal (Figure 3.2.4). This threshold appears to be around 30-35 L/s.  50.00  Peak lag (hours)  40.00 30.00 20.00 10.00 0.00 -10.00 0  10  20  30  40  Discharge (L/s) ShS-Br Linear (ShS-Br)  StS-Ur Linear (StS-Ur)  R-Cl  Figure 3.2.4 Peak event discharge versus lag time to peak tracer concentration  When comparing overall 24 hour mass flux surrounding peak tracer breakthrough to 24 hour discharge surrounding the peak discharge, a distinct relationship emerges for each application site. For all tracers, mass flux increases as event discharge increases. However, the rate of increase is greatest for R-Cl followed by SS-Br and then StS-Ur (Figure 3.2.5). 53  Linear relationships have been plotted as a rough guide to help visualize the difference between the three locations (Figures 3.2.4 and 3.2.5).  25  24h Mass Flux (%)  20  15  10  5  0 0  500  1000  1500  2000  2500  3000  3500  3  24h discharge volume (m ) Bromide  Uranine  Chloride  Linear (Uranine)  Linear (Chloride)  Linear (Bromide)  Figure 3.2.5 24-hour discharge volume surrounding event peak versus 24 hour tracer mass flux (as a percent of total tracer applied) surrounding peak tracer concentration  Table 3.2.2 outlines the recovery of the tracers for both application periods. As of March 19th, 42.4% and 38.1% of the bromide and uranine were recovered, respectively. At the end of the second period (May 3rd – Ur & June 14th - Br & Cl), 31.3%, 19.2%, and 62.9% of bromide, uranine and chloride were recovered.  54  Table 3.2.2 Summary of tracer mass recovery Tracer  Period*  Mass Applied  Mass Recovered  % Recovered  Bromide  1  10 kg  4.24 kg  42.4  2  27.3 kg  8.55 kg  31.3  1  53 g  20.17 g  38.1  2**  117.7 g  22.59 g  19.2  1  N/A  N/A  N/A  2  30.5 g  19.19 g  62.9  Uranine  Chloride  * Period refers to the 1st (Jan. 3/06 - March 19/06) and 2nd (March 20/06 - June 1/06) application periods ** Period 2 for uranine ends May 3rd 2006  3.2.4 Subsurface response Each plot area exhibited a unique response of hydraulic head to rainfall. The piezometers are a measure of hydraulic potential showing the pressure head just above or at the impermeable layer (bedrock or till). The peak potential is also a good proxy for water table depth as there is no confining layer above the till/bedrock. The potential/water table height was examined as actual height (mm), depth below the surface and as a fractional depth of the overall soil depth at the measurement location. Of greatest interest is the timing of peak potential (maximum total pressure head) in relation to the peak event discharge. Figure 3.2.6 shows the lag of the peak hydraulic potential in each plot area versus the potential itself. In ShS-Br the peak hydraulic potential occurs prior to the peak event discharge, while StS-Ur occurs close to the peak discharge and R-Cl occurs after the peak event discharge (Figure 3.2.6).  55  Hydraulic potential is dependent on location. The lowest hydraulic potentials were observed in ShS-Br, which has a gentle slope and deeper soils followed by StS-Ur (steep slope and shallow soil) and R-Cl (the riparian area) (Figure 3.2.6). Also, greater hydraulic heads are associated with greater peak lag times.  20 ShS-Br  StS-Ur  R-Cl  Peak lag (hours)  15 10 5 0 -5 -10 0  200  400  600  800  1000  Hydraulic Head (mm)  Figure 3.2.6 Peak hydraulic potential (height above confining layer) versus lag time from peak hydraulic potential to peak event stream discharge  The peak depth of water below the surface (the distance from the soil surface to the water table) also shows some correlation with the 24 hour tracer mass flux surrounding the tracer peak concentration (Figures 3.2.6 and 3.2.7). The height of the water table below the surface may be an indicator of different transport zones being activated. Figure 3.2.7 shows that  56  greater mass fluxes are associated with higher (smaller depth below surface) water table peaks. Figure 3.2.8 shows the same pattern but the relationship appears more linear.  25  24h Mass Flux (%)  20  15  10  5  0 0  100  200  300  400  500  600  700  Depth Below Surface (mm) ShS-Br  StS-Ur  R-Cl  Figure 3.2.7 24 hour tracer mass flux (as a percent of total tracer applied) surrounding concentration peak versus peak depth of water table below surface  57  25.00  24h Mass Flux (%)  20.00  15.00  10.00  5.00  0.00 0.20  0.25  0.30  0.35  0.40  0.45  0.50  0.55  0.60  0.65  0.70  Fractional Depth Below Surface  ShS-Br  StS-Ur  R-Cl  Figure 3.2.8 24 hour tracer mass flux (as a percent of total tracer applied) surrounding concentration peak versus peak fractional depth of water table below surface  58  4.0 DISCUSSION 4.1 Success Using Artificial Tracers In general, it can be said that the artificial tracers themselves worked. That is, they travelled to the stream and provided a measurable signal. However, complications with equipment and general execution of the experiment impeded the overall performance of the tracers. It was estimated that the bulk of the tracer might move through the catchment in a matter of weeks and that the average discharge may be close to 12 L/s. Unfortunately, January 2006 was a particularly wet month with one storm having a peak flow nearly three times the expected average. As a result, tracer concentrations were low. For rhodamine the signal was not outside the margin of error and there was too much noise for that data to be used. In addition, rhodamine was applied to the riparian area to decrease its contact with woody debris as it has been known to be heavily absorbed by organic matter (Personal communication, Weiler 2006). However, there was a large amount of woody debris in the riparian plot and it could not be totally avoided. Roots below the surface as well as the organic layer itself may have taken up a substantial amount of the rhodamine, further compounding the poor performance of the tracer. Rhodamine has proved to be a good tracer for applications directly in streams (e.g. measuring discharge in large rivers and stream tracer tests), but the experience here has shown it to be a poor tracer for water moving through soil. The first application period yielded useable data from only two tracers. With the intensive field sampling regime and no time for extensive data analysis, the second set of tracers had to be applied. At a minimum it was discovered that the rhodamine signal was particularly poor and the decision was made to use chloride in its place. The quantities were also increased, but the rainfall inputs were smaller. This did provide a good trace, but the  59  GOT fluorometer was removed to be used at another site leaving only the Cyclops turner probes. The Cyclops probes had a lot of noise and drifting data, which were not used in the data analysis. May 3rd marked the end date of the uranine data collection. The second application saw the introduction of chloride as one of the artificial tracers and it worked well. In general, the salts were easier to work with than the dyes. They are considered to be sufficiently conservative and are easy to obtain and measure. Of course there was a background signal for chloride which had to be considered in the analysis but there was no detectable background of bromide. Another advantage of the salts is the ability to measure their concentrations from water samples with reasonable precision. The Ion Chromatograph was instrumental in providing the proper calibration for the probe data and allowed a high degree of confidence. In general the salt tracers were found to perform better than the dyes. Uranine may yet prove to be a useful tracer, but the results from this study are not detailed enough to determine the degree of its utility. Based on this study, potassium bromide would be the tracer of choice in a single tracer study. The cheap and readily available Turner probes performed well, making KBr the most user friendly tracer. In addition, the tracer is more stable compared to dyes, which have issues of absorption and photodegradation. Lange et al. (1996) also had success with using lithium bromide in Sweden, which was applied to a 1000 m2 catchment. There may well be other artificial tracers which prove to be useful for studying runoff generation. Naturally occurring tracers may also be useful in this type of study (e.g. deuterium) if they can be applied artificially in sufficient concentrations that they override the background concentration.  60  Despite the many limitations of the dye tracers some valuable data were collected and the analysis yielded some interesting results.  4.2 Landscape Unit Response A uniquely definable response was observed from each landscape unit. The units were a riparian zone with moderate slope (R-Cl), a zone with deep soils and gentler slope (ShS-Br) and a zone with a very steep gradient, shallow soils and bedrock outcrops (StS-Ur).  Spatial/Temporal Response Section 3.2.4 describes the water table response as characterized by piezometer measurements. It was seen that peak hydraulic head was greatest in R-Cl followed by StS-Ur and then ShS-Br. This finding is not surprising since the water table should be closest to the surface in the riparian area as it is a convergence zone. The piezometer characterizing StS-Ur was at the bottom of the slope so it is not surprising that the hydraulic head was greater there due to rapid delivery from a large steep contributing area. Lag times between the peak hydraulic head and peak stormflow were greatest in R-Cl followed again by StS-Ur and ShS-Br. What is interesting is that the peak in ShS-Br occurred before or close to the event peak. This finding may be explained by preferential pathways becoming connected and draining backed-up saturated areas which then contribute to the peak flow. Typically we might expect the greater storage capacity and depth to impede infiltration and for the peak hydraulic head to occur later as the profiles fills up, but this does not appear to be the case (Weiler and McDonnell 2004). Figure 3.2.6 shows that the peak hydraulic head is far lower for ShS-Br than for any of the other areas. This may be  61  representative of a well developed subsurface drainage network along bedrock troughs and through soil pipes. Interestingly, peak hydraulic head in StS-Ur seems to occur around the same time, showing it to be closely connected with the event peak. StS-Ur was located closer to the weir than ShS-Br, but no closer to the surface water network. It may be that the shallow soils and bedrock outcrops are dominated by more surface runoff, which would explain the similar timing of the peak hydraulic head to storm flow. R-Cl had hydraulic heads peaking on the falling limb of the hydrograph. In other words, the peaks lagged behind those of the stormflow, but usually within five hours. Sidle et al. (2000) proposed a new conceptual model which incorporated many runoff generation processes put forward over the past few decades (see Chapter 1). The concept of nodes which link up a network of varying processes was what made the model innovative. These nodes are like switches that are turned on and off when certain thresholds are met. There are two important characteristics of these nodes: the first is their location and the second is their threshold. Theoretically the location is dependent on physical characteristics like confining layer topography, location of macropores and soil physical properties. For example, a node may be a volume of soil matrix between two macropores or soil pipes, or perhaps a localised high spot in the confining layer behind which a saturated layer must reach sufficient height before the node can be “switched on”. The thresholds are likely to be controlled by a myriad of factors which could include both the physical characteristics of the node location and the area it controls as well as input intensity and amount combined with antecedent moisture conditions. It is also important to realise that while the nodes may be the “first order of controls” on the runoff dynamics within a catchment their behaviour cannot be  62  as simple as on or off. The relationship describing discharge from a subsurface node point may be described by a two phase non-linear relationship. The second phase would have a markedly steeper slope, and the threshold would be the point where the slope begins to increase rapidly. Figure 3.2.4 shows that peak tracer concentration lags from the various landscape units in the study catchment converged near events with peak discharges of 30 L/s. It would seem that at this threshold value of roughly 30 L/s, this catchment has all the nodes between the channel and the application sites turned on (or rather past their threshold). This results in the rapid delivery of water to the stream such that there is little difference between the units. This observation would not have been possible with steady state sprinkler experiments commonly associated with tracer studies (e.g. Lange et al. 1996). The results of this study agree nicely with the work of Weiler and McDonnell (2004) in which they setup a virtual hillslope experiment to study the effect of drainable porosity on runoff, among other factors. One of the most striking results is the visualization of flow for highly drainable soils at two different times during one experiment. The first time step, less than 20 hours from the start of the input (which is variable and ends just after 20 hours), shows high areas lower in the catchment with high relative flow (areas of steeper slope and shallow soil). This same area exhibits lower relative flow after 60 hours and the lower areas show higher flow. These different units are effectively being switched on and off with the meeting of certain thresholds of input (and time since input). This behaviour nicely depicts the concept of individual landscape units having definable responses which, when coupled together, form the overall response.  63  Residence Time The dynamic spatio-temporal response observed has important implications for residence time within a catchment such as the one used in this study. In climates with well defined rainy seasons, residence times will vary greatly from the wet to dry season (McGuire et al. 2006). As observed above, for storms below certain sizes the time to peak tracer concentration will vary greatly between different landscape units. The smaller the storm size the longer the time lag to peak discharge from SS-Br (Figure 3.2.4). This relationship translates into greater residence times for water falling on SS-Br, which becomes even more significant during the dry season; recall the extended tracer recession limbs during the dryer second study period for SS-Br as compared to StS-Ur and R-Cl (Figure 3.2.2). Table 3.2.2 shows that 42% bromide and 38% uranine were recovered over the first 75 day application period, in which 532 mm of precipitation fell, and dropped to 31% bromide and 19% uranine for the second 78 day period, in which 290 mm of precipitation was received (Figures 2.3 and 3.2.1). The differences in mass recovered exemplifies how dependent residence time is on initial conditions. It shows that residence time cannot be strictly defined, but is dependent on the spatial and temporal dynamics of the hydrological condition of a catchment, which can vary greatly in a short period of time. The variation in residence time between landscape units was not well defined in this study given the many unknown variables affecting recovery rates. However, in the second application period nearly 63% of the applied chloride (R-Cl) was recovered with only half that (31%) for bromide (SS-Br) and 19% for uranine (Sts-Ur). With the uncertainty surrounding the performance of uranine we will focus on the difference between R-Cl and SS-Br. R-Cl is the riparian landscape unit while SS-Br is the shallow slope, deep soil unit  64  some distance from the network. It is clear that it takes more time for water to travel from SS-Br to the stream than for R-Cl. Unfortunately, the lack of data for the riparian area in the first application period negates drawing further conclusions. In a sprinkler tracer experiment, Lange et al. (1996) recovered 14% of bromide applied within four days of the onset of the experiment, at which time bromide concentration reached background levels. This result, coupled with the findings here of recovery over much longer periods, implies a large degree of mixing between old and new water during an event. New water laced with tracer may mix with old water and diffuse back into the soil matrix becoming immobile until and event of sufficient size can remobilize it. The advantage of this study over other tracer studies such as those conducted in Gårdsjön, Sweden, by Rodhe et al. (1996) and Lange et al. (1996) is that the input is dynamic in time. However, the spatial variability cannot be controlled as in the steady state studies cited above. It is argued here that in order to be able to really determine how a catchment behaves under natural conditions, the conditions should not be controlled. Instead we should look for new ways to observe the processes naturally as artificial and controlled studies are already abundant. Controlled studies are important to give us new ideas, but they represent a mere snapshot of time in a single spatial unit. The spatial and temporal variability in physical characteristics, biota and input are so great that hydrological study must seek to integrate the variability, or rather integrate itself into the natural variability to be viable for the next generation of researchers.  65  Implications for land use The findings from this study suggest that we should be looking at landscapes more closely when trying to mitigate environmental damage due to logging, farming, development or any other interference we may impose on the landscape. The most immediately relevant topic is logging. A new method of patch cuts could be implemented around the landscape unit concept in which areas of lower hydrological sensitivity based on measures of proximity to surface networks, soil depth and slope are considerations. Designating a buffer around stream networks may not be enough when so much of how water is delivered to the stream is dependent on the physical features mentioned above. Buttle (2002) reviewed the need for cold temperature ground water recharge in Ontario lakes and streams for brook trout spawning. In this instance a traditional fixed width buffer zone would do little to protect against changes in ground water temperature and delivery. This study supports the need for a re-evaluation of how we harvest our forests and what kind of buffers we leave in place. This is not only a water quantity issue but also a water quality issue. Increased levels of nitrate leaching after harvesting may be mitigated by cutting in areas where water remains longer and has more chance to mix with the immobile fraction, such as in areas of shallow slope and deep soils. In addition, we may apply this knowledge and type of study to areas where agriculture takes place on a more limited land base. In Canada, we are lucky to have enough land that our agriculture largely takes place in flat valley bottoms, or in the open plains. However, in some parts of the globe, agriculture takes place on varied and steep terrain. Application of fertilizer, pesticides and herbicides should be subject to careful consideration; proximity to the stream channel is only one concern. It may be that a steep slope with  66  shallow soils far from any surface network has a well developed subsurface network that moves water rapidly to the stream channel. In such cases chemicals and fertilizers should be avoided. Tracers can be applied in a similar manner to how they have been in this study to agricultural settings to determine patterns in landscape physical properties and how water moves.  67  5.0 CONCLUSIONS 5.1 Review of Key Findings The goal of this study was to identify unique timing signatures from different landscape units characteristic of coastal British Columbia and the Pacific Northwest. It was also meant to explore the usefulness of artificial tracers as a tool for studies of water movement through soil on a small catchment scale. It was found that, in general, salts perform better than dyes as they do not have the drawbacks of absorption and photodegradation. Salts are also relatively inexpensive and easy to work with. In addition, more precise methods such as ion chromatography are available to measure ion concentrations in collected water samples. A variety of in situ instruments are available for both salts and dyes, but the more cost effective ones tested here (probes by Turner) performed better for salts than dyes. It is the recommendation of this study to use salts when possible in future studies of this nature. It is felt that the original hypothesis, that a catchment can be delineated into different landscape units with unique timing signatures for water delivery to the stream, is correct. These differences are based on the physical characteristics of the units and therefore the dominant runoff generation processes occurring within them. In general, shallower slopes and deeper soils are found to increase the time necessary for water to travel to the stream. It was observed that as a catchment becomes wetter (both from greater input and antecedent conditions) the time lag between units decreases; at some point preferential flow lines become “turned on,” allowing rapid delivery from all units. It is this dynamic response based on antecedent conditions, input and the landscape unit’s physical characteristics that needs to be better understood. If that can be done a better understanding and therefore description of  68  catchment response can be achieved. As discussed above this research has implications for how we manage land to better protect hydrological resources, the balance of which is crucial to ecosystem maintenance.  5.2 Recommendations for Further Research There are many avenues down which this study can be expanded upon. One of the most obvious is to take the techniques used here and apply them to other types of landscapes and climates as discussed above. This type of study can be conducted in both forested and agricultural land and in areas of differing topography and climate. In dryer areas it may be necessary to use salt tracers only as they are more conservative than dyes. More studies of this type would allow hydrologists to begin observing patterns in similar landscape units. Identified patterns could ultimately lead to the implementation of landscape unit models based on documented behaviours in field research. It may be possible to begin to measure the response quantitatively in relation to quantitatively measured physical characteristics of the landscape units. A more quantitative approach with a greater number of landscape units, perhaps in several catchments, would be an ideal expansion of this study. In such a study, measurements of average soil depth, slope, distance from catchment outlet and soil physical properties could be compared to the landscape unit response. This comparison would determine what factors are most important in controlling the response. As discussed in chapter 4, it may be possible to gain a better understanding of when thresholds are reached and what determines them. As with many hydrological studies, a denser network of instruments with more physical measurements would be ideal. Both soil moisture content and soil matric potential  69  could be measured. However, there are limitations to how much we can do currently, but field technology is growing rapidly to close the gap between hydrological modelling and field studies. Networks of inexpensive remote sensors are being tested currently and may provide the kind of intensive data collection desired by field hydrologists. It would be ideal to continue to work in this catchment now that its behaviour is better understood and the tracer quantities required to provide a good signal are known. With more time the measurement periods could be extended and the piezometer network expanded in addition to adding new types of instrumentation.  70  BIBLIOGRAPHY Anderson, M.G. and Burt, T.P., 1978. The role of topography in controlling throughflow generation. Earth Surfaces Processes and Landforms, 3: 331-334. Betson, R.P., 1964. What is watershed runoff? Journal of Geophysical Research, 69(8): 1541-1551. Beven, K.J., and Kirkby, M.J., 1979. A physically based, variable contributing area model of basin hydrology. Hydrological Sciences-Bulletin 24: 45-69. Bonell, M., 1998. Selected challenges in runoff generation research in forests from the hillslope to headwater drainage basin scale. Journal of the American Water Resources Association, 34(4): 765-786. Buttle, J.M. 2002. Rethinking the donut: the case for hydrologically relevant buffer zones. Hydrological Processes. 16: 3093-3096. Calver, A., and Cammeraat, L.H. 1993. Testing a physically based runoff model against field observations on a Luxembourg hillslope. Catena, 20: 273-288. Cheng, J.D. 1988. Subsurface stormflows in the highly permeable forested watersheds of southwestern British Columbia. Journal of Contaminant Hydrology, 3: 171-191. De Vries, J. and Chow T.L. 1978. Hydrologic behavior in a forested mountain soil in coastal British Columbia. Water Resources Research, 14(5): 935-942. Dunne, T. and Black, R.D., 1970. An experimental investigation of runoff production in permeable soils. Water Resources Research, 6(2): 478-490. Engler, A., 1919. Untersuchungen über den Einfluss des Waldes auf den Stand der Gewässer, 12. Kommissionsverlag von Beer & Cie, Zürich, 626 pp. in Wieler et al. 2005. Pinder G.F. and Jones, J.F., 1969. Determination of the Groundwater component of peak discharge from the chemistry of total runoff. Water Resources Research, 5: 438-445. Freer, J. McDonnel, J., Beven, K.J., Brammer, D., Burns, D., Hooper, R.P., and Kendal, C. 1997. Topographic controls on subsurface storm flow at the hillslope scale for two hydrologically distinct small catchments. Hydrological Processes, 11(9): 1347-1352. Freeze, A.R., 1972. Role of subsurface flow in generating surface runoff. 2.Upstream source areas. Water Resources Research, 8(5): 1272-1283.  71  Freeze, R. A., and Witherspoon, P.A., 1967. Theoretical analysis of regional groundwater flow. 2. Effect of water-table configuration and subsurface permeability variation: Water Resources Research, 3: 623-634. Gordon, N. D., McMahon, T. A. and Finlayson, B.L. 1992. Stream Hydrology: An Introduction for Ecologists. John Wiley & Sons Ltd. Chichester, England. Harr, R.D., 1977. Water flux in soil and subsoil on a steep forested slope. Journal of Hydrology, 33: 37-58. Hewlett, J.D. and Hibbert, A.R., 1963. Moisture and energy conditions within a sloping soil mass during drainage. Journal of Geophysical Research, 68: 1081-1087. Hewlett, J.D. and Hibbert, A.R., 1967. Factors affecting the response of small watersheds to precipitation in humid areas. In: W.E. Sopper and H.W. Lull (Editors), Forest Hydrology. Pergamon Press, New York, pp. 275-291. Hoover, M.D. and Hursh, C.R., 1943. Influence of topography and soil depth on runoff from forest land. Transactions of the American Geophysical Union, 2: 693-698. Horton, R.E. 1933. The role of infiltration in the hydrological cycle. Transactions of the American Geophysical Union, Fourteenth annual meeting, pp. 445-460. Hursh, C.R. and Brater, E.F., 1941. Separating storm-hydrographs from small drainage-areas into surface- and subsurface-flow. Transactions of the American Geophysical Union, 3: 863-871. Hutchinson, D.G. and Moore, R.D. 2000. Throughflow variability on a forested slope underlain by compacted glacial till. Hydrological Processes 14(10): 1751-1766. Kirchner, J.W., 2003. A double paradox in catchment hydrology and geochemistry. Hydrological Processes, 17: 871-874. Lange, H., Lischeid, G., Hoch, R., and Hauhs, M. 1996. Water flow paths and residence times in a small headwater catchment in Gårdsjön, Sweden, during steady state storm flow conditions. Water Resources Research. 32: 1689-1698. McDonnell, J.J., 1990. A rationale for old water discharge through macropores in a steep, humid catchment. Water Resources Research, 26(11): 2821-2832. McDonnell, J.J., 1997. Comment on "the changing spatial variability of subsurface flow across a hillside" by Ross Woods and Lindsay Rowe. Journal of Hydrology, New Zealand, 36(1): 97-100.  72  McDonnell, J.J. Freer, J., Hooper, R., Kendal, C., Burns, D., and Beven, K., 1996. New method developed for studying flow in hillslopes. EOS, Transactions of the American Geophysical Union, 77(47): 465. McGlynn, B.L., McDonnell, J.J. and Brammer, D.D., 2002. A review of the evolving perceptual model of hillslope flowpaths at the Maimai catchments, New Zealand. Journal of Hydrology, 257: 1-26. McGlynn, B.L., McDonnell, J.J., Seibert, J. and Kendall, C., 2004. Scale effects on headwater catchment runoff timing, flow sources, and groundwater-streamflow relations. Water Resources Research, 40: W07504. McGuire K.J., McDonnell, M., Weiler, M., Kendall, C., McGlynn, B. L., Welker, J. M., and Seibert, J. 2005. The role of topography on catchment-scale water residence time. Water Resources Research, 41: W05002. McGuire, K.J., Weiler, M., McDonnell, J.J. 2006. Integrating field experiments with modeling to infer water residence times, Advances in Water resources.doi: 10.1016/j.advwatres 2006.07.004 Moore, R.D. and Thompson, J.C. 1996. Are water table variations in a shallow forest soil consistent with the TOPMODEL concept? Water Resources Research, 32: 663669. Mosley, M.P., 1979. Streamflow generation in a forested watershed. Water Resources Research, 15: 795-806. Pearce, A.J., Stewart, M.K. and Sklash, M.G., 1986. Storm runoff generation in humid headwater catchments: 1. Where does the water come from? Water Resources Research, 22: 1263-1272. Pilgrim, D.H., 1966. Radioactive tracing of storm runoff on a small catchment. Journal of Hydrology, 4: 289-305; 306-326. Pilgrim, D.H., and Huff, D.D., 1978. A field evaluation of subsurface and surface runoff I. Tracer Studies. Journal of Hydrology, 38: 299-318. Pinder, G.F., and Jones, 1969. Determination of the groundwater component of peak discharge for the chemistry of total runoff. Water Resources Research, 5(2): 438445. Price, A.G. 1994. Measurement and variability of physical properties and soil water distribution in a forest podzol. Journal of Hydrology, 161: 347-364.  73  Rodhe, A., Nyberg, L., and Bishop, K. 1996. Transit times for water in a small till catchment from a step shift in the oxygen 18 content of the water input. Water Resources Research. 32: 3497-3511. Sidle, R.C., Tsuboyama, Y., Noguchi, S., Hosoda, I., Fujieda, M. and Shimizu T. 2000. Stormflow generation in steep forested headwaters: a linked hydrogeomorphic paradigm. Hydrological processes. 14: 369-385. Sklash, M.G., nad Farvolden, R.N., 1979. The role of groundwater in storm runoff. Journal of Hydrology, 43: 45-65. Sklash, M.G., Stewart, M.K. and Pearce, A.J., 1986. Storm runoff generation in humid headwater catchments: 2. a case study of hillslope and low-order stream response. Water Resources Research, 22(8): 1273-1282. Smart, P.L. and Laidlaw, I.M.S. 1977. An evaluation of some fluorescent dyes for water tracing. Water Resources Research. 13: 15-33. Soulsby, C., Rodgers, P. J., Petry, J., Hannah, D.M., Malcolm, I.A. and Dunn S.M. 2004. Using tracers to upscale flow path understanding in mesoscale mountainous catchments: Two examples from Scotland. Journal of Hydrology, 291:174-196. Talamba, D., Joerin, C. and Musy, A. 2000. Study of subsurface flow using environmental and artificial tracers: the Haute-Mentue case, Switzerland. IAHS Publication, 262: 559-264. Thompson, J.C. 1994. Behaviour and prediction of water table levels in shallow forest soils, Southwestern British Columbia. M.Sc. Thesis, Simon Fraser University, Burnaby, BC, Canada. Tischer, E. 1986. Hydrological behaviour of a forested mountain slope. M.Sc. Thesis, Univeristy of British Columbia, Vancouver, BC Canada. In Thompson 1994. Torres, R., Dietrich, W.E., Montgomery, D.R., Anderson, S.P. and Loague, K., 1998. Unsaturated zone processes and the hydrologic response of a steep, unchanneled catchment. Water Resources Research, 34(8): 1865-1879. Trudgill, S.T. 1987. Soil water dye tracing, with special reference to the use of rhodamine WT, Lissamine FF and amino G acid. Hydrological Processes, 1: 149-170. Weiler, M. and Flühler, H., 2004. Inferring flow types from dye patterns in macroporous soils. Geoderma, 120(1-2): 137-153. Weiler, M. and McDonnell, J., 2004. Virtual experiments: a new approach for improving process conceptualization in hillslope hydrology. Journal of Hydrology, 285(1-4): 318.  74  Weiler, M. and McDonnell, J., 2007. Conceptualizing lateral preferential flow and flow Networks and simulating the effects on gauged and ungauged hillslopes. Water Resources Research, 43, W03403, doi:10.1029/2006WR004867 Weiler, M., McDonnel, J., Tromp-van Meerveld, H.J., Uchida, T. 2005. Subsurface Stormflow. Encyclopedia of Hydrological Sciences.Wiley and Sons. Weiler, M., McGlynn, B.L., McGuire, K.J. and McDonnel, J.J., How does rainfall become runoff? A combined tracer and runoff transfer function approach. Water Resources research, 39(11): 1315, doi:10.1029/2003WR002331, 2003. Weiler, M. and Naef, F., 2003. An experimental tracer study of the role of macropores in infiltration in grassland soils. Hydrological Processes, 17(2): 477-493. Weyman, D.R., 1973. Measurements of the downslope flow of water in a soil. Journal of Hydrology, 20: 267-288. Whipkey, R.Z., 1965. Subsurface storm flow from forested slopes. International Association of Science, Hydrology Annual Bulletin, 2: 74-85. Woods, R. and Rowe, L., 1996. The changing spatial variability of subsurface flow across a hillside. Journal of Hydrology (NZ), 35(1): 51-86. Woods, R.A., Sivapalan, M. and Robinson J.S. 1997. Modelling the spacial variability of subsurface runoff using a topographic index. Water Resources Research, 31: 2097-2110.  75  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0066629/manifest

Comment

Related Items