Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sediment budget of gold and magnetite and their distribution in stream sediment in lower Harris Creek,… Hou, Zhihui 1998

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

Item Metadata

Download

Media
831-ubc_1998-27165X.pdf [ 9.72MB ]
Metadata
JSON: 831-1.0088673.json
JSON-LD: 831-1.0088673-ld.json
RDF/XML (Pretty): 831-1.0088673-rdf.xml
RDF/JSON: 831-1.0088673-rdf.json
Turtle: 831-1.0088673-turtle.txt
N-Triples: 831-1.0088673-rdf-ntriples.txt
Original Record: 831-1.0088673-source.json
Full Text
831-1.0088673-fulltext.txt
Citation
831-1.0088673.ris

Full Text

S E D I M E N T B U D G E T O F G O L D A N D M A G N E T I T E A N D T H E I R D I S T R I B U T I O N IN S T R E A M S E D I M E N T IN L O W E R H A R R I S C R E E K , S O U T H - C E N T R A L B R I T I S H C O L U M B I A , C A N A D A by ZHIHUIHOU B.Eng., Hebei College of Geology, 1982 M.Eng., Institute of Geophysical & Geochemical Exploration, P. R. China, 1985 A THESIS FOR SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Earth & Ocean Sciences We accept this thesis as confirming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA OCTOBER, 1997 ©ZhihuiHou, 1997 In p resent ing this thesis in partial fu l f i lment of the requ i rements for an advanced deg ree at the Univers ity of British C o l u m b i a , I agree that the Library shall make it freely available for reference and study. I further agree that pe rmi s s i on for extens ive c o p y i n g of this thesis for scholar ly pu rposes may be granted by the head of my depar tment or by his o r her representat ives. It is u n d e r s t o o d that c o p y i n g o r pub l i ca t i on of this thesis for f inancial gain shall not be a l l o w e d w i t hou t my wr i t ten permi s s ion . Depa r tment of Earf'ft <^"A PoSO* Sa««Ce.s The Univers i ty of British C o l u m b i a Vancouve r , C a n a d a Date Oet*L+r 3 d , 1 ^ 7 DE-6 (2/88) 11 ABSTRACT Although understanding the development of anomalous dispersion trains of gold in stream sediment is important to design and interpretation of exploration geochemical surveys, it has not been studied systematically with respect to variations in supply and transport of sediment. The problem is addressed by applying the sediment budget approach, with the following functional form: I(nput) - S(torage) = O(utput) Harris Creek, a gold-rich stream in south-central British Columbia was chosen as the study area. Primary sources of sediment are landslides in glacial deposits; secondary sources include bank erosion, erosion of sediment accumulated behind log jams, and sediment from tributary streams. Transport of bedload sediments within the Harris Creek drainage basin is strongly seasonal and dependent on the magnitude of the annual snowmelt flood. Field methods included collection of representative geochemical samples; direct field measurement of erosion; and collection of bedload samples using Helley-Smith samplers. Samples were analyzed in the laboratory to determine gold and magnetite. The data show that the supply of gold is discontinuous both spatially and temporally, with little or no gold being delivered to or transported out of Harris Creek during years with low or normal flood discharges (<20 m3/s). During years of large snow meltwater floods gold is transported for brief periods during very high discharge events. Coarse gold (<0.149>0.053 mm) is preferentially deposited in the voids at bar heads with exceptionally anomalous values at bar Ul heads at breaks in slope. Conversely, fine gold (<0.053 mm) is swept in suspension over bar heads to deposit at bar tails but also has peak values at breaks in slope. Estimates of input and output of sediment, magnetite and gold show that gold is less readily mobilized and transported out of Harris Creek than magnetite and sediment. Through time, this has resulted in development of substantially greater concentrations of gold (up to twenty times) in stream sediment in Harris Creek than in the glacial deposits that are the primary sources of sediment. With respect to the anomaly dilution model in exploration geochemistry, the combined effect of accumulation of gold at breaks in slope and dilution by material derived from landslides create an apparent cut-off for gold anomalies a short distance downstream from a landslide. iv TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vii List of Figures x Acknowledgment xvi CHAPTER 1. INTRODUCTION 1 1.1 Overview 1 1.2 Problem 7 1.3 Study Objectives 8 1.4 Study Approach and Techniques 9 CHAPTER 2 ENVIRONMENTAL SETTINGS AND MAJOR SEDIMENT SOURCES 12 2.1 Location and Access 12 2.2 Physiography 12 2.2.1 The Harris Creek basin 12 2.2.2 Geomorphology of Harris Creek 14 2.3 Bedrock Geology 15 2.4 Quaternary Geology 21 2.5 Climate 25 2.6 Geomorphic Processes and Sediment Sources 28 CHAPTER 3 STUDY METHODS AND DATA QUALITY EVALUATIONS 33 3.1 Area Selection 33 3.2 Bulk density 33 3.3 Direct Measurements of Erosion Rates 35 3.4 Discharge Measurements 41 3.5 Geochemical Sampling 44 3.5.1 Bedload sampling with Helley-Smith sampler 44 3.5.2 Sampling from landslides, bank and logjam 47 3.5.3 Stream sediment 48 3.6 Laboratory Methods 50 3.6.1 Wet and dry sieving 50 3.6.2 Heavy mineral separation 52 3.6.3 Combination and split 52 3.6.4 Chemical analysis 53 3.7 Data Accuracy and Precision 54 3.7.1 Accuracy 54 3.7.2 Analytical precision 54 CHAPTER 4 INPUT OF SEDIMENT, MAGNETITE AND GOLD: ESTIMATE WITH FIELD MEASUREMENTS 60 4.1 Calculation Formulas 60 4.2 Inputs of Sediment, Magnetite and Gold from McAuley Creek 62 V 4.3 Inputs of Sediment, Magnetite and Gold from Mosquito Creek 70 4.3.1 Inputs of sediment, magnetite and gold 73 4.4 Mass Input of Sediment, Magnetite and Gold from Landslides 84 4.4.1 Landslide #1 84 4.4.2 Landslide #7 85 4.4.3 Landslide #17 90 4.4.4 Landslide #28 90 4.5 Inputs of Sediment, Magnetite and Gold from Logjam #1 94 4.6 Inputs of Sediment, Magnetite and Gold from Unstable Banks 100 4.6.1 Section I 100 4.6.2 Section II 104 4.6.3 Section III 104 4.7 Summary 108 4.7.1 Concentration of magnetite and gold 108 4.7.2 Mass Inputs 112 CHAPTER 5 OUTPUTS OF SEDIMENT, MAGNETITE AND GOLD FROM HARRIS CREEK 117 5.1 Hydraulic Conditions 117 5.2 Critical Discharge 119 5.3 Sediment Transport 124 5.3.1 Transport rate 124 5.3.2 Output 129 5.4 Magnetite Output 129 5.4! Transport rate 129 5.4.2 Output 135 5.5 Output of Gold 135 5.5.1 Transport rate of gold in the -0.053 mm size fraction 135 5.5.2 Transport rate for the -0.149+0.053 mm gold 142 5.5.3 Gold Outputs 147 CHAPTER 6 BUDGETS FOR SEDIMENT, MAGNETITE AND GOLD, AND THEIR DISTRIBUTIONS ALONG LOWER HARRIS CREEK 148 6! Introduction 148 6.2 Sediment, Magnetite and gold Budget 149 6.2! Flow history 149 6.2.2 Sediment budget 151 6.2.3 Magnetite budget 151 6.2.4 Gold budget analysis 154 6.3 Distribution of Gold and Magnetite along Lower Harris Creek 158 6.3! Magnetite distribution 158 6.3.2 Distribution of gold 162 6.4 Landslides, Peak Gold Values and Stream Gradient 169 6.4.1 Further investigation of the problem 169 6.4.2 Landslides #1 and #7: anomalous gold sources or not? 170 6.4.3 Clay ball distribution 170 vi 6.4.4 Fingerprinting the effect of the clay balls 173 CHAPTER 7 DATA INTERPRETATION 178 7.1 Introduction 178 7.2 Sediment Sources: Effect of the Fraser Glaciation 179 7.3 Behavior of gold during floods 182 7.3.1 Sediment 182 7.3.2 Magnetite 187 7.3.3 Gold 187 7.4 Behavior of Gold between Floods 191 7.4! Sediment and magnetite 191 7.4.2 Gold 192 7.5 "Knick Point" and Gold Anomalies: Long Time Evolution 195 CHAPTER 8. CONCLUSIONS 199 REFERENCES 203 APPENDIX 209 vii List of tables Table 2.1 Exposure area proportions (%) of different rock types in Harris Creek catchment 2 Table 2.2 Mean temperature and precipitation at Lumby Sigalet Station 26 Table 3.1 Gold analysis (ppb) of Reference Till Sample Till-1 by AF-AAS and AQ-ICP 55 Table 3.2 Duplicate analyses of gold in -149+53 pm fractions with FA-AAS and AQ-ICP (Au in ppb) 56 Table 3.3 Duplicate results for 24 trace elements in -53 pm size fraction of sediment 59 Table 4.1 Concentrations of gold and magnetite in McAuley Creek and Mosquito Creek 63 Table 4.2 Sample parameters with Helley-Smith sampler on McAuley Creek in 1992 and 1993 66 Table 4.3 Showing concentrations of gold and magnetite in bedload samples from McAuley Creek 69 Table 4.4 Sampling parameters with Helley-Smith sampler on Mosquito Creek 74 Table 4.5 Statistical summaries for the transport rates of magnetite in -0.149+0.105 mm, -0.105+0.075 mm, and -0.075+0.053 mm, and gold in -0.053 mm on Mosquito Creek in 1993 78 Table 4.6 Mass input of gold and magnetite from Mosquito Creek in 1993 80 Table 4.7 Gold concentrations of bedload samples from Mosquito Creek in 1993 81 Table 4.8 Summaries of physical and chemical parameters of Landslide #1 87 Table 4.9 Inputs of sediment, magnetite and gold from Landslide #1 88 Table 4.10 Summaries of physical and chemical parameters of Landslide #7 91 Table 4.11 Inputs of sediment, magnetite and gold from Landslide #7 92 2 Summarie  of physical and chemical parameters of Landslide #28 93 viii Table 4. 13 Inputs of sediment, magnetite and gold from Landslide #28 95 Table 4! 4 Summaries of physical and chemical parameters of Log jam #1 97 Table 4! 5 Inputs of sediment, magnetite and gold from Logjam #1 99 Table 4!6 Summaries of physical and chemical parameters of Station 1 102 Table 4.17 Inputs of sediment, magnetite and gold from Station 1 103 Table 4.18 Summaries of physical and chemical parameters of Station 2 105 Table 4!9 Inputs of sediment, magnetite and gold from Station 2 106 Table 4.20 Summaries of physical and chemical parameters from Station 3 107 Table 4.21 Inputs of sediment, magnetite and gold from Station 3 109 Table 5.1 Sampling time and discharge on Harris Creek Station in 1993 125 Table 5.2 Regression equations for rates of sediment and magnetite of bedload samples versus discharge on Harris Creek in 1993 128 Table 5.3 Output of stream sediment from the Harris Creek basin 130 Table 5.4 Output of magnetite in the three size fractions from Harris Creek 136 Table 5.5 Concentrations of magnetite (%) and gold in bedload samples on Harris Creek 137 Table 5.6 Showing weights of -0.053 mm fraction and estimated gold concentrations in bedload samples on Harris Creek 141 Table 5.7 weight of -0.149+0.053 mm fraction and estimated gold concentrations in the Helley-Smith samples on Harris Creek 145 Table 6.1 Summary of flow conditions of Harris Creek and 18-year records of WSC station 08LC005 located 6 km downstream from this study site 150 Table 6.2 Showing the difference of transport of sediment, magnetite and gold in the Harris Creek basin 157 Table 6.3 Concentrations of magnetite and gold in heavy mineral concentrates along lower Harris Creek 161 Table 6.4 Geometric mean concentration ratios for gold and magnetite in Harris Creek Table 6.5 Gold content of the four landslides in the Harris Creek basin Table 6.6 Chemical elements selected for fingerprinting the influence of Landslide #1 X List of Figures Fig. 1.1 Definition of parameters used in Equation (i) 3 Fig. 2.1 The physiographic zonation of Harris Creek catchment 13 Fig. 2.2 The longitudinal profile showing the locations of four active landslides, relative elevation and stream slope variation along lower Harris Creek 16 Plate 2.1 Showing particle arrangement at a bar head on Harris Creek 17 Plate 2.2 Showing particle arrangement at a bar tail on Harris Creek 18 Fig. 2.3 Textural characteristics of stream sediments from high and low energy environments in Harris Creek 19 Fig. 2.4 Simplified geological map of Harris Creek Basin from Jones (1959) 20 Fig. 2.5 Quaternary geological map of the source zone in Harris Creek 23 Fig. 2.6 Showing the relation between the water equivalent depth of snow and maximum daily discharge in the Harris Creek basin 27 Fig. 2.7 Showing spatial locations of local and episodic processes in the source zone of the Harris Creek basin 29 Fig. 2.8 Locations of laterally unstable reaches along lower Harris Creek 32 Fig. 3.1 Relative locations of study sites in the Harris Creek basin 34 Fig. 3.2 Sketch map showing locations of bulk density samples, erosion pins and geochemical samples on Landslide #1 36 Fig. 3.3 Sketch map showing locations of bulk density samples, erosion pins and geochemical samples on Landslide #7 37 Fig. 3.4 Sketch map showing locations of bulk density samples, erosion pins and geochemical samples on Landslide #17 38 Fig. 3.5 Sketch map showing locations of bulk density samples, erosion pins and geochemical samples on Landslide #28 39 Plate 3.1 Logjam #1 on lower Harris Creek 40 Fig. 3.6 Showing relation between discharge and staff gauge on McAuley Creek 42 xi Fig. 3.7 Showing relation between discharge and staff gauge on Mosquito Creek 43 Fig. 3.8 Helley-Smith sampler 45 Plate 3.2 Showing the portable Helley-Smith sampler with a metal rod 46 Fig. 3.9 Stream sediment sites along Harris Creek 49 Fig. 3.10 Sample preparation flowchart 51 Fig. 3.11 Duplicate analysis of -0.053 mm sediments for gold by AQ-ICP 57 Fig. 4.1 Stream hydrograph for McAuley Creek in 1993 64 Fig. 4.2 Showing the relation between transport rate of stream sediment and discharge on McAuley Creek 67 Fig. 4.3 Showing population distribution in one composited Helley-Smith sample on McAuley Creek 71 Fig. 4.4 Stream hydrograph for Mosquito Creek 72 Fig. 4.5 Showing the relation of stream sediment transport rate with discharge on Mosquito Creek in 1993 75 Fig. 4.6a Showing the relationship between concentration of magnetite in the three size fractions and discharge on Mosquito Creek in 1993 77 Fig 4.6b Showing the relationship between transport rates of magnetite in the three size fractions and discharge on Mosquito Creek in 1993 77 Fig. 4.7 Relation between the transport rate of the -0.053 mm gold and discharge using Helley-Smith sampler on Mosquito Creek 82 Fig. 4.8 Showing the relation between transport rate of the -0.149+0.053 mm gold and discharge 83 Fig. 4.9 Textural characterization of sediments from the four active landslides 86 Plate 4.1 Showing Landslide #7 undercut by the May 1993 flood 89 Fig. 4.10 Textural characterization of sediments from Log jam #1 and three stations along Harris Creek 96 xii Fig. 4.11 Variation of magnetite concentrations in the three size fractions in a vertical profile of Log jam #1 98 Fig. 4.12 Variation of gold concentrations in the two size fractions in a vertical profile on Log jam #1 98 Fig. 4.13. Laterally unstable reaches and monitor stations along lower Harris Creek 101 Fig. 4.14 Magnetite concentrations of different sources 110 Fig. 4.15 Gold concentrations of in different sources 111 Fig. 4.16 Temporal and spatial variation of total sediment inputs from different sources in the Harris Creek basin 113 Fig. 4.17 Temporal and spatial variations of magnetite inputs from different sources 114 Fig. 4.18 Temporal and spatial variations of gold inputs f  iff t  to Harris Creek 6 Fig. 5.1 Stream hydrograph for Harris Creek 118 Fig. 5.2 Showing population distribution in one bedload sample at discharge of 5.3 m3/s on Harris Creek 120 Fig. 5.3 Showing population distribution in one bedload sample at discharge of 11.1 m3/s on Harris Creek 121 Fig. 5.4 Showing population distribution in one bedload sample at discharge of 26.1 m3/s on Harris Creek 122 Fig. 5.5 The relation between the proportion of the framework population and discharge for the composited Helley-Smith samples on Harris Creek 123 Fig. 5.6 The relationship between transport rate of stream sediment and discharge on Harris Creek in 1993 127 Fig. 5.7 Variation in concentrations of magnetite with discharge 131 Fig. 5.8 Transport rates of -0.149+0.105 mm magnetite versus discharge 132 Fig. 5.9 Transport rates of -0.105+0.075 mm magnetite versus discharge 133 xiii Fig. 5.10 Transport rates of -0.075+0.053 mm magnetite versus discharge 134 Fig. 5.11 Distribution pattern of gold values in the -0.053 mm Helley-Smith sample fraction 138 Fig 5.12 The distribution of gold and Poisson distribution calculation results 140 Fig 5.13 Transport rate of the -0.053 mm gold with Helley-Smith sampler 143 Fig 5.14 Estimated numbers of gold particles in -0.149+0.053 mm size fraction 144 Fig 5.15 Transport rate of 0.149+0.053 mm gold versus discharge on Harris Creek 146 Fig. 6.1 Sediment budget (in metric tones) for stream sediment from July 28, 1990 to April 17, 1992 152 Fig. 6.2 Sediment budget (in metric tonnes) for stream sediment from April 18, 1992 to April 30, 1993 152 Fig. 6.3 Sediment budget (in metric tonnes) for stream sediment from May 1 to May 31, 1993 152 Fig. 6.4 Sediment budget (in kg) for the -0.149+0.053 mm magnetite from July 28, 1990 to April 17, 1992 153 Fig. 6.5 Sediment budget (in kg) for the -0.149+0.053 mm magnetite from April 18, 1992 to April 30 1993 153 Fig. 6.6 Sediment budget (in kg) for the -0.149+0.053 mm magnetite from May 1 to May 31,1993 153 Fig. 6.7 Budget (g) for the -0.149+0.053 mm gold from July 28, 1991 to April 17, 1992 155 Fig. 6.8 Budget (g) for the -0.149+0.053 mm gold from April 17, 1992 to Aril 30, 1993 155 Fig. 6.9 Budget (g) for the -0.149+0.053 mm gold from May 1 to May 31, 1993 155 Fig. 6.10 Budget (g) for the -0.053 mm gold from July 28, 1991 to April 17, 1992 156 Fig. 6.11 Budget (g) for the -0.053 mm gold  April 18, 1 92 t  A 30 3 xiv Fig. 6.12 Budget (g) for the -0.053 mm gold from May 1 to May 31, 1993 156 Fig. 6 13 Downstream distribution of magnetite in stream sediments from high energy environments 159 Fig. 6.14 Downstream distribution of magnetite in stream sediment from low energy environments 160 Fig. 6.15 Downstream distribution of gold in -0.149+0.053 mm non-magnetic heavy mineral concentrates (NMHMC) from high energy environments, and its relationship with gold sources, landslides and stream gradients 163 Fig. 6.16 Downstream distribution of gold in -0.149+0.053 mm non-magnetic heavy mineral concentrates (NMHMC) from low energy environments, and its relationship with gold sources, landslides and stream gradients 164 Fig. 6.17 Downstream distribution of gold in -0.053 mm stream sediments from high energy environments, and its relationship with gold sources, landslides and stream gradients 165 Fig. 6.18 Downstream distribution of gold in -0.053 mm stream sediments from low energy environments, and its relationship with gold sources, landslides and stream gradients 166 Fig. 6.19 Downstream distribution of gold in -0.053 mm stream sediment in the detailed study reach 171 Fig. 6.20 Downstream distribution of the seven elements around Landslide #1 175 Fig. 6. 21 Distribution pattern of the relative concentrations of the four elements to the stream sediments in high and low energy sites downstream from Landslide #1 177 Fig. 7.1 Possible gold sources for lower Harris Creek and their transfer routes 180 Fig. 7.2 Transport rates of sediments coarser than medium sand versus discharge on Harris Creek in 1993 183 Fig. 7.3 Transport rates of sediments finer than medium sand versus discharge on Harris Creek in 1993 184 Fig. 7.4 Transport rate of pa ticles of 0 1+164 mm on Harris Creek in 1993 185 XV Fig. 7.5 Textural variation of sediment in various stage of flow conditions with Helley-Smith sampler on Harris Creek 186 Fig. 7.6 Variation of concentrations of gold against flow discharge 189 Fig. 7.7 Schematic diagram showing sedimentological behavior of gold particles in different environments along lower Harris Creek 190 Fig. 7.8 Downstream profile of gold in stream sediment in lower Harris Creek in the year with extremely high flood 194 Fig. 7.9 Stream profiles (A) Concave, (B) Concavo-convex, and (C) Interrupted by Knick points 196 xvi Acknowledgment Primary thanks go to Dr. W. K. Fletcher, my supervisor, for his full support, solid advice, and patient encouragement during my entire study period at UBC. I am grateful to Dr. Michael Church and Dr. June Ryder who helped me in selecting suitable study locations and sampling methods, and provided me with guidance and encouragement as well as critical comments on ideas and concepts. Comments and suggestions by Drs. W. K. Fletcher, M . Church, J. Ryder and A. J. Sinclair greatly improved my thesis. Thanks are due to Scott Babakaieff, T. Nelson and Brett Eaton for their excellent and enthusiastic assistance in the field. The study was made possible by a Graduate Fellowship from The University of British Columbia. Thanks are also due to many people in the Department of Geological Sciences at UBC who helped me with this project and my study here, especially L. A. Groat and B. Cranston. 1 C H A P T E R 1 INTRODUCTION 1.1. Overview Application of geochemistry to mineral exploration emphasizes the practical aspects of prospecting - how and where to recognize geochemical patterns related to mineral occurrences, or how and where to detect geochemical anomalies caused by mineralization (Garrett, 1983). To satisfy this objective, exploration geochemists have used geochemical surveys or, in other words, employed methods based on systematic measurement of one or more chemical properties of a naturally occurring material. The chemical property measured is most commonly the trace content of some element or group of elements; the naturally occurring material may be rock, soil, gossan, glacial drift, vegetation, stream or lake sediment, water, or vapor (Rose et al., 1979). Widespread acceptance of stream sediment geochemical surveys as one of the most important low-cost exploration tools is based on the premise that stream sediment composition is representative of the geochemistry of the catchment basin upstream of the sample site. If an anomalous dispersion train is followed upstream, the point of maximum metal values is known as the cut-off. This is usually considered to be close to the source of the anomaly and thus is the starting point for follow-up. As a guide to design and interpretation of stream sediment surveys, Hawkes (1976) described an anomaly dilution model: M e m A m = A a ( M e a " M e b ) + A m M e b The parameters used in the above formula are illustrated graphically in Fig. can be rewritten as follows: (1.1) 1.1. This model 2 Aa(Ml :ea-Meb) = AmCMem-Meb) (1.2) If the drainage area is large relative to the mineralized area, then the term AJVfeb can be neglected (Rose et al., 1979), so that, The term AmMem is a measure of the size (Am) and grade (Mem) of the deposit and is a constant for a given deposit. The quantity (Mea-Met,) is a measure of the strength of the anomaly above background, and A a is the drainage area above a sampling site. Equation (1.3), which generates a smooth hyperbolic decay curve for the anomalous dispersion train downbasin from the cut-off point (Fig. 1.1), is based on several assumptions: uniform rate of erosion; uniform geochemical background; no feedback between water and sediment; no sampling error; a single anomalous source; and no contamination. The parameters used in Equation (1.1) can be divided into two categories: (a) chemical parameters, such as Me a, Met, and Me m , that require samples to be collected and analyzed and (b) natural or geomorphic parameters, such as A a and An, that are estimated or calculated based on geomorphic patterns. Almost as early as the first application of geochemical surveys to mineral exploration, exploration geochemists realized that the data quality of the chemical parameters was very important for detection of geochemical anomalies caused by mineral occurrences, and many studies have been completed. In contrast, the geomorphic parameters have not received sufficient consideration. Aa(Mea-Meb) = A m Me m (1.3) Outline o f mineralised area A m Downstream direction Fig. 1.1. Definition of parameters used in Equation (1). 4 To further explore the importance of the data quality of the chemical parameters, the overall variability (S )^ in a data set is simplified theoretically into the following form on the basis of the classic discussions of Krumbein and Slack (1956), Miesch (1976), Garrett (1983), and Nichol (1983): = S^ ntr + S2 s mpi + S2 a n aj (1.4) Where: S^ ntr " natural fluctuation of the element in the sample medium; S s^mpl - sampling variance and S^ anal - variance due to sample preparation and analysis. Equation (1.4) shows that it is critical for the compositional changes to be stable in order to identify significant features in a set of geochemical data. By stable it is meant that the variability due to sampling (S^ s mpi) and analytical sources (S^ a n a j) is significantly smaller than the natural fluctuation (S2ntr) (Garrett, 1983). For Au and other weathering-resistant heavy minerals, the sampling variance (S2 Smpl) constitutes variance due to local segregation of the target material. In stream deposits, this could include lithological inhomogeneity of grains; depositional stratigraphy; lateral and vertical sorting processes promoted by the transporting medium. From the viewpoint of survey design, this sampling variance could be looked upon as a problem of "subgrid" structure at a scale finer than the practical sampling density allows. Therefore, in geochemical survey practice, the concern is to so structure the exploration design that it is possible to determine if any trends revealed by the surveys are largely caused by the natural fluctuation of S2 n t r, or could be due to the accumulated effects of sampling and analytical variability (Garrett, 1979). Probably the single most widely used technique that has been employed to aid sampling design and the a posterior evaluation of the variability from regional 5 (S^ntr )> sampling (S^ s mpi) and analytical sources (S^anaj) is Analysis of Variance (ANOVA). For example Fletcher (1981) used ANOVA to evaluate the variance between sample batches analyzed; Garrett and Goss (1978) applied it to evaluate sampling and analytical variation in regional geochemical surveys; Howarth and Lowenstein (1971) adopted it to interpret the sampling variability of stream sediments; Miesch (1976) employed ANOVA to study methods of sampling and laboratory analysis; and Sinclair et al. (1977) applied ANOVA to investigate the effect of analytical precision on the recognition of real variation (S2ntr). These studies have resulted in remarkable successes in geochemical exploration for metal deposits such as Cu, Zn and Ag. However, for metals, such as Au, Sn and W, that usually occur as major constituents of rare heavy minerals, the sampling and analytical variances become a serious problem, which has received much attention but has not been solved yet. Clifton et al. (1969) indicated that, based on the binomial distribution, to obtain the relative precision of about ±50% at the 95% confidence limits, a sample containing 20 gold particles was the minimum sample size adequate for gold analysis. A similar result can be predicted, more simply, from the Poisson distribution that relative error of rare grains (RE%) is approximated by (Fletcher, 1990): RE(%) = ±200/Vn (where n is the average number of gold particles in a sample of a given size). Harris (1982) demonstrated that considerable subsampling imprecision exists due to the non-homogeneous distribution of gold within the individual portions taken for analysis. Day and Fletcher (1986) found that, to ensure an adequate number (20) of gold particles, it was necessary to collect 20 kg of minus 5 mm field samples to provide enough sediment for gold determination in the minus 0.053 mm fraction, while up to 300 kg of minus 5 mm field samples were needed to ensure an adequate number (20) of gold particles in heavy mineral concentrates sized to minus 0.1 mm. 6 Recently, Fletcher and his colleagues have shown that even where samples are large enough to be representative of bulk sediments, there are still considerable spatial and temporal variances in concentrations of the heavy minerals in stream sediment. Their studies in Malaysia (Fletcher et al., 1987; Fletcher and Loh, 1996a, b), in Canada (Day & Fletcher, 1986, 1989, and 1990; Fletcher, 1990; Fletcher & Day, 1989; Fletcher & Wolcott, 1991; and Saxby & Fletcher, 1986a, b), and in Thailand (Paopongsawan and Fletcher 1993) have shown that the change of hydraulic conditions (including flow velocity, bed roughness, stream width, depth and discharge, sediment texture, and channel gradient) can cause preferential accumulation of heavy minerals at specific sites, for example, coarse gold in the high energy environment at bar heads. These results are consistent with the theoretical model of heavy mineral transport by streams (Day and Fletcher, 1991; Fletcher and Loh, in press). On the basis of their studies, to use those elements (such as Sn, W and Au) as the indicator elements (which are defined as those that form the main components of economic interest in a given type of mineral deposit (Butt and Smith, 1980)) in a stream sediment geochemical survey, local hydraulic variability must be minimized. Fletcher et al. (1987) suggested that the ratio of Sn/magnetite be used to suppress the Sn variations resulting from local hydraulic effects because the hydraulic behaviors of cassiterite (S.G. = 6.8 - 7.1) and magnetite (S.G. = 5.18) are similar, and magnetite is ubiquitous whereas cassiterite is associated with the primary mineralization. Recently, Fletcher et al. (1992 ) put forward a new concept, "the transport equivalence of particles", as particles, that despite differences in their physical properties, have the same average net transport velocity and are transported and deposited together. Thus their relative concentrations should not change in response to changes in hydraulic-sedimentological conditions, and can be used to correct effects of hydraulic conditions on heavy mineral distributions. Fletcher and Loh (1996) have applied this concept to minimize the effect of local hydraulic factors on the downstream distribution of Sn in Malaysia, thereby highlighting the natural fluctuations caused by Sn mineralization. 7 1.2. Problems It is becoming more and more evident that both the geomorphological and geochemical assumptions of the Hawkes' model oversimplify conditions. Geomorphologically, Church and Slaymaker (1989) showed that secondary mobilization of Quaternary sediments along river valleys dominates over primary denudation of the land surface for all areas up to 3xl04 km2 in British Columbia, Canada. Bjorklund et al. (1994) found that true sheet erosion is observed only during occasional heavy rainstorms and generally the major part of the stream sediments originates from limited sources within the drainage basins. They also found that a point sediment source was supplying sediment to the stream where a stream undercut an adjacent slope (Fig. 9-10 of Bjorklund et al. (1994)). Bogen (1992) reported that mass movement triggers sediment movement into stream channels in Norway. Mansikkaniemi (1975) found that land use provides important point sources for active channels. Ryder (1991), and Ryder and Fletcher (1991) found that the supply of sediment to Harris Creek was discontinuous both spatially and temporally in the Harris Creek basin in south British Columbia, Canada. Terrain mapping (Ryder, 1991, and Ryder and Fletcher, 1991) found that, at present, four of the fourteen landslides, three of the sixteen tributaries (on 1:50,000 topographic map), and bank erosion and a big logjam along lower Harris Creek were the major sediment sources. Local episodic processes rather than uniform runoff control the sediment supply. Geochemically, studies of tin in stream sediment in the Sungei Petal, Malaysia show that (1) the samples from bar heads invariably contain greater concentrations of tin, and (2) rather than decaying in a regular fashion with distance from the source, the tin anomaly exhibits erratic, anomalous spikes for several kilometers downstream from the mineralization. Studies of Sn in Pelly River, Yukon Territory (Saxby and Fletcher, 1986b), and Au in the Huai Hin Laep, Thailand (Paopongsawan and Fletcher, 1993) have also shown that both elements increase in 8 concentrations downstream from their sources. In the Huai Hin Laep, decreased stream width and increased stream velocity and bed roughness, all favor concentration of Au on the bed of the stream. Studies in Harris Creek indicate that the hydraulic conditions on bar heads are different from those on bar tails. The preferential accumulation of Au and magnetite on bar heads tends to counteract downstream dilution so that Au values increase downstream along a 5 km study reach (Day and Fletcher, 1989 and 1991). Concentrations of heavy minerals on a stream bed can also vary seasonally; for example, Fletcher and Day (1989) found that the highest concentrations of Au are present at bar head sites in Harris Creek shortly after spring snow melt floods. Subsequent studies (Fletcher and Wolcott, 1991), using pit traps inserted in a gravel bar, showed that Au is transported in significant quantities only during periods of high discharge when the cobble-gravel pavement is disturbed. 1.3. Study Objectives This thesis examines (1) how the combination of sediment sources, episodic processes, and local hydraulic conditions affects the downstream dilution pattern of geochemical anomalies (of gold and magnetite) in stream sediment; and (2) whether or not the conventional anomaly dilution model (Fig. 1.1) can be applied to regions like Harris Creek. The main task of the study is to describe and explain observed spatial and temporal variation in the budget of gold and magnetite. This study is somewhat atypical in contemporary studies in exploration geochemistry, which have tended to focus on sampling density, sample size, size fractions to be analyzed, analytical method, and determination of threshold, etc. The work is also distinguished from many previous studies by the scope and resolution of the field monitoring and sampling program. The detailed programs are as follows: (1) investigation of the distribution of active sediment sources and contents of gold and magnetite; 9 (2) measurement of magnitude and frequency of magnetite and gold inputs from the active sources; (3) measurement of transport of magnetite and gold in Harris Creek; (4) measurement of magnitude and frequency of outputs of magnetite and gold from Harris Creek; (5) investigation of longitudinal distributions of gold and magnetite; and finally (6) discussion of the applicability of some assumptions made in the traditional Hawkes' dilution model. 1.4. Study Approach and Techniques To attain these objectives, the sediment budget approach will be introduced and will be used as a basis for the establishment of study methods. The sediment budget is fundamental to geomorphology (cf. Sutherland and Bryan, 1991), but it is quite new to exploration geochemistry. Therefore, it is necessary to start from the very basic concept. The sediment budget is based on input from sediment sources, changes in volume (or mass) of sediment stored within the drainage basin, and sediment output from the drainage basin. Generally, it has the following functional form (Roberts and Church, 1986): I - AS = O defined as follows: I is the mass or volume of sediment input from the sediment sources during the specified period At; AS is the change of the mass or volume of sediment stored in the channel reach in At; and O is the mass or volume of sediment output from the channel reach in At. Lehre (1982) recognized three important components of fluvial sediment budgets: (1) identification of erosional processes and understanding of their controls and interrelationships; (2) measurement of magnitude and frequency of sediment mobilization by each process, and (3) identification of 10 sediment storage elements and quantification of the volume (mass), residence time, and changes in storage of sediment in each element. In practice, a catchment must be broke down into reservoirs (or control volumes), to quantify the input and output of sediment between reservoirs, and to understand the relative importance of continuous versus episodic sediment transfer processes in the catchment. The studies in the Harris Creek basin (Ryder and Fletcher, 1991) have investigated the distribution of point sediment sources. Here I will use the sediment budget concept to specifically examine the response of lower Harris Creek to the point sediment sources and episodic processes. Examples of application of the sediment budget approach to clastic sediment routing are numerous in geomorphology - those by Jackli (1957), Rapp (1960), Lehre (1982), Leopold et al. (1966), Roberts and Church (1986), Kelsey (1980), Swanson et al. (1982) and Trimble (1983) are perhaps the best known. They have identified the major sources, pathways, and storage of clastic sediment within basins or landform units and determined process rates. Rapp also identified the importance of dissolved solids. The sediment budget approach as traditionally applied has not met with universal acceptance because of at least two perceived limitations (Slaymaker 1993): (1) it is claimed that it is a merely descriptive or accounting methodology and (2) the large error bands associated with measurements and calculations have encouraged a tendency to close the budget artificially. Keeping these uncertainties in mind, this study attempts to demonstrate the applicability of the sediment budget concepts to understanding processes relevant to exploration geochemistry. The methods used in this program are: an explicit recognition of the time scale of integration of the budget; a recognition and quantification of spatial and temporal variation of input amounts, and of major transport processes. The budget results presented in this thesis should be understood to represent broad general trends in space and a "snapshot" in time representing contemporary 11 conditions because of the short time scale of the study, and limited number of observation sites. The techniques used to collect field and laboratory data are: i) Terrain mapping techniques (Resources Inventory Committee, 1996) for Quaternary geology and current active processes; ii) Direct field measurements of sediment erosion and input of magnetite and gold from the active sources with erosion pins (Wolman, 1959; Twidale, 1964; and Hill, 1973) and the Helley-Smith sampler (Emmett, 1980) in combination with discharge measurements. iii) Direct field measurement of output of magnetite and gold from the lower Harris Creek with the Helley-Smith sampler. iv) Geochemical sampling methods described by Fletcher (1990). More details of data collection and analysis to meet the above objectives will be described in the relevant chapters. CHAPTER 2 12 ENVIRONMENTAL SETTINGS AND MAJOR SEDIMENT SOURCES 2.1. Location and Access The Harris Creek basin (82L/2 in the 1:50,000 national topographic map series) is bounded by longitudes 118° 45' and 119° 00' west and by latitudes 50° 01* and 50° 13' north in southern British Columbia, Canada (Fig. 2.1). Access to the Harris Creek area is by gravel roads. Because of logging, grazing and outdoor recreation, road conditions within the Harris Creek basin are quite good, with most of the roads leading to clear-cut areas and lakes. For stream sediment geochemical surveys, however, climbing and hiking are indispensable, and field work is often very strenuous. 2.2. Physiography 2.2.1. The Harris Creek basin Harris Creek, a tributary of Shuswap River, is a cobble-gravel-bed stream. The Harris Creek basin, with sixteen tributaries (on the 1:50,000 topographic map), occupies a catchment with an area of about 225 km ,^ of which 60% is located on the dissected plateau of the Okanagan Highland between 1,500 and 2,000 meters above sea level. Based on natural landscape and terrain mapping results, the physiographic zonation of the Harris Creek basin is summarized in Fig. 2.1. The highland surface above the 1,500-meter contour line has a gently undulating topography embellished with isolated hummocky hills. Slopes are Fig. 2.1. The physiographic zonation of Harris Creek catchment. (The Transfer Boundary is along the 5,000 feet contour line; contour interval is 500 feet. 1 foot = 0.3048 m) 14 typically short and gentle, with gradients ranging from 0 to 8°. First-, and second-order streams, with a channel gradients from 1° to 5°, originate on the highland surface. Lakes and swamps (figured areas in Fig. 2.1, also see 1:50,000 topographic map (82 L/2)) trap any sediment that is mobilized from the upstream area. Consequently, sediment supply from the highland surface to Harris Creek is negligible (Ryder and Fletcher, 1991). From the point of view of mass transfer which is defined as the movement of sediment without regard for mechanism (Church, 1983), sediments from the highland do not move across the plateau edge to Harris Creek. The highland surface is, therefore, categorized as the Zero Transfer Zone. The zero transfer boundary is arbitrarily placed along the 1,500 meter contour. From the plateau edge all the way down to the main valley floor of Harris Creek, slopes are long, straight and steep with mean gradients varying from 011° up to 42°. The topography of the main valley floor is gentle. Banks upstream of Mosquito Creek consist of poorly sorted gravel sediments and are coupled locally with colluvial slopes and till. The floodplain, which is developed downstream from Mosquito Creek, has two layers, a fine-grained overbank deposits overlying in-channel sandy-gravels, overbank deposits varying from 0.5 to 1.5 m tend to become thicker downstream. Terrain mapping results (Ryder, 1991, and Ryder and Fletcher, 1991) indicated that all the contemporary sediment sources are located in this area which is, therefore, classified as the Source Zone to Harris Creek (Fig. 2.1). This accounts for about 40% of the whole catchment basin. Even within the Source Zone, the sediment sources are not distributed uniformly, but mainly located on the north side of the Harris Creek valley. 2.2.2. Geomorphology of Harris Creek Harris Creek, a fourth order stream based on Horton (1945) analysis at a scale of 1:50,000, flows along the main valley floor. The channel from Landslide #28 downstream to Landslide #7 is steep with a gradient of about 0.039 (2.2°), and then becomes gentle (1.3°) from Landslide #7 downstream (Fig. 2.2). The stream bed slope below Landslide #1 decreases first (0.6°), and then 15 increases slightly (1!°) . Breaks-in-slope occur just below two active landslides: Landslides #1 and #7. Bars are common in-channel sediment storage sites along lower Harris Creek. Field observations indicate that ten of the eleven bars studied have two distinctive depositional environments: high energy environment (HEE) at bar heads and low energy environment (LEE) at bar tails. The textural characteristics of stream sediments from the HEE and LEE are shown in Plates 2.1 and 2.2, and Fig. 2.3. The HEE is represented by poorly sorted gravel: voids between coarse particles are filled by fine particles. The HEE also has an obvious armored surface (Plate 2.1). Day and Fletcher (1991) have shown that the texture of the HEE sediment has two populations: the cobble-gravel framework and interstitial sand and fine gravels; and Dj (the diameter that provides the best separation between the two populations) is between 2.38 and 11.63 mm. Sediment from the LEE consists of poorly sorted fine gravels and sands (Plate 2.2 and Fig. 2.3). Studies at a gravel bar on lower Harris Creek indicated that the flow velocity was higher at the bar head than that at the bar tail (Fletcher and Wolcott, 1991). 2.3. Bedrock Geology The Harris Creek basin, situated in the Vernon map area (Map 1059A), was geologically mapped by Jones in 1959. The upper reaches and waterheads are underlain by Precambrian Monashee Group, which contains gneiss, mica-sillimanite-garnet schist, quartzite, marble and phyllite, and Miocene basaltic lava with well-developed columnar jointing (Mathews, 1988). Opposite banks of the middle and lower reaches have contrasting geology with andesitic volcanics, limestone and argillites of the Carboniferous Cache Creek Group on the north bank, and Jurassic granite and granodiorite on the south bank (Fig. 2.4). The valley of Harris Creek forms a strong northwest trending regional lineament that is thought to trace a strike slip fault (Jones, 1959). 17 Plate 2.1. Showing particle arrangement at a bar head on Harris Creek c Plate 2.2. Showing particle arrangement at a bar tail on Harris Creek 19 0.05 0 ! 0.5 1 10 50 100 500 Grain size (mm) Fig. 2.3. Characterization of subsurface stream sediments from high and low environments in Harris Creek (Samples were collected from HZ-26 in Fig. 3.9) Fig. 2.4. Simplified geological map o f Harris Creek Basin from Jones (1959) 21 In the entire catchment, the order of areal proportions of the three major rock types is: Monashee gneiss (42%) > Basaltic lava (35%) > Granite (8%) (Table 2.1). However, in the Source Zone, the areal proportion of Monashee gneiss decreases while the areal proportions of basaltic lava and granite are higher, so that the importance of the rock types becomes: Basaltic lava (43%) > Monashee gneiss (28%) > Granite (16%). Nelson (1993) has shown that local bedrock geology has a strong control on lithological composition and texture of glacial till that covers at least 80% of the total area. 2.4 Quaternary Geology Quaternary geology of the Harris Creek catchment was mapped by Ryder (1991), and Ryder and Fletcher (1991). Restruction of Quaternary history indicates that the fundamental geomorphic frame of the Harris Creek basin had taken shape before it was occupied by the Fraser Glaciation ice from 20 ka to 11 ka ago. At the onset of the Fraser Glaciation, ice advancing upstream first impounded a lake in the Harris Creek basin and then over-rode the sediments deposited in the lake (Ryder, 1991; Ryder and Fletcher, 1991). The lake was subsequently re-established as the ice sheet melted and retreated. The Fraser Glaciation thus produced variable 3-D complexes of till, and glaciolacustrine and glaciofluvial sediments (Fig. 2.5, also see Ryder et al., 1993). Till mantles all but the steepest slopes and is thickest on gentle slopes and along the valleysides of lower Harris Creek. The lithology and texture of the till are variable, being strongly influenced by the proximity and type of the underlying bedrock (Ryder and Fletcher, 1991; and Nelson, 1993). Along the main valley a lower unit of glaciolacustrine sediment at least 50 m thick underlies the till and extends down below the stream bed. An upper unit of glaciolacustrine sediment overlies the till and forms a constructional terrace along lower Harris Creek that is particularly well developed near Vidler and McAuley Creek. Glaciofluvial sediments occur in Table 2.1. Exposure area proportions (%) of different rock types in Harris Creek catchment* Rock type Proportion in entire catchment Proportion in source zone Monashee gneiss 42.00 28.30 Andesitic volcanics 2.80 2.80 Limestone 5.90 8.00 Granite 7.60 15.80 Basaltic lava 35.30 43.00 * The estimation was made from Jones' (1959) geological map Fig. 2.5. Quaternary geological map of the source zone in Harris Creek (simplified from Ryder 1990) (The map legend is listed on the next page) 24 Map Legend of Fig. 2.3 Texture. Surficial material-Process Surface expression Texture of Surficial Materials g b c r s clay silt sand gravel boulders rubble Materials C Colluvium F Fluvial sediment FG Glaciofluvial sediment LG Glaciolacustrine sediment MTill R Bedrock U Undifferentiated Quaternary sediments Surface Expression a moderate slope b blanket ffan h hummocky j gentle slopes m rolling topography p plain r ridge s steep slopes t terrace u undulating topography v veneer w mantle of variable thickness -EG Glacial meltwater channels -F Failing (active slow mass movement) -I Irregularly sinuos channel -R Rapid mass movement -V Gully erosion Geological processes 25 several parts of the study area. Well defined terraces along lower Mosquito Creek and the upper part of lower Harris Creek are underlain by 1 to 4 m of outwash sands and pebbly gravels that rest on glaciolacustrine sediment. Fluvial sediments underlie the modern flood plain and the active channels of Harris Creek. Mappable Colluvium is restricted to steep slopes immediately below rock cliffs where significant accumulations of bedrock-derived material exist, and where steepness indicate. 2.5 Climate The Harris Creek basin is located on the Interior Plateau with the continental nature of the thermal climate. The basic pattern is characterized by strong climatic contrast. Temperature and humidity are controlled partly by modified maritime air masses which have lost much of their moisture during their passage across the Coast Mountains. However, incursions of continental Arctic air in winter and continental warm air in summer produce extreme and variable climatic conditions (Ryder, 1989). Table 2.2 shows monthly temperature and precipitation at Sigalet Road Observation Station, Lumby, 6.4 km northeast of the Harris Creek Basin. Annual precipitation is about 834 mm (Environment Canada, 1982), varying from 25 mm to 147 mm monthly. Snow is usually greater than rainfall, and snow pack accumulates everywhere. Snowmelt occurs between late May and early June when temperature rises. The relation between daily maximum discharge and the water equivalent depth of snow pack is shown in Fig. 2.6. Field observations suggest that, beside the snow accumulation, a sudden rise of temperature in several consecutive days and rain storms trend to produce high flows. H 3 ft so 3 I 3 3 M "O (T> -1 2 £' o' 3 Vi TJ H N- 3^  to cn 0> CO 3 e. 05 3' 85 O 3 3' ^ 3' a g o 3 3 tn p KJ e 5.' •a o •6° &' o' 3 & OS 3 © O 3 13 o 05 3 I Es 5' 3 Ul U> 00 00 4». O S © N O N O N 00 to o KJ U) U) r H r © N O N ~> —i O N © O N O N O N *• ~ *• O N O O N O N O N O N O N -J o O N 4*. 4 * * H O N O N U i 4*-O N H - u i 4 * KJ >— U> N O 4 ^ Ui U) K) as 3 * 5 U> - J bo bo © © KJ KJ © © as H 3 TU 3 c 3 • as U> 3 0 <3 © 8-Ul 3» CIS NO 3, © o SJ* •*• ® ~ £ o hf g I KJ © £ ^ 2 g 2 3 KJ «_ © ^ H © 0 © <» N O O N •  I— H - K) O N © KJ H - © KJ © © © O N UJ © Ul © N O 4>> © O N U) © KJ > U> 3 © ES to 00 -NO 00 © © O N U) © © KJ © O N J -O N >— © © I I 8" z I U n> o I 27 500 600 700 800 900 Water equivalent snow depth (nun) Fig. 2.6. Showing the relation between the water equivalent depth of snow and maximum daily discharge for 17 years in the Harris Creek basin (see Environment Canada, 1982) 28 2.6 Geomorphic Processes and Sediment Sources On the basis of the terrain mapping and sediment source investigation (Ryder, 1991; and Ryder and Fletcher, 1991), episodic processes dominate over the surface processes (which include rain splash, sheet erosion, frost disturbance and wind transport) and creep processes (which include slow mass deformation, wet-dry and freeze-thaw induced settlement) in the Harris Creek basin (Fig. 2.7). The episodic geomorphic processes include gullying, rockfall, rockslide, debris slide, local bank erosion, and intermittent fluvial processes. In brief, gullying involves the loosening and removal of surface particles by surface runoff and debris flow and results in the development of gullies (Ryder and Howes, 1986) which are mainly restricted to valley sides along lower Harris Creek. Rockslide occurs where rock slides dowslope. Rockfall occurs on steep surfaces where weathering (chiefly freezing and thawing) looses and triggers releasing sections of rocks. Rockfalls are mainly distributed along cliffs at the plateau edges, especially in the basaltic areas where columnar jointing is well developed (Mathews, 1988). Landslides are important in the Harris Creek basin. For the purpose of this study, the collective term "landslide" used in this thesis includes slumps, block glides and debris slides and also contains rockslide. Landsliding is a movement of materials on slopes under the influence of gravity, water saturation and fluvial undercutting.. An active rockslide has been identified on the west slope of McAuley Creek. Bank erosion, or bank retreat occurs when grains or assemblages are removed from the bank face by the flow (Thorne, 1978). Erosion consists of two distinct events - fluvial undercutting and mechanical failure of cantilevers in the top sediment. Along lower Harris Creek, bank erosion occurs on laterally unstable reaches between the Research Site and Landslide #28. Intermittent fluvial processes include erosion, sorting, transportation and deposition by running water, which normally depend upon grain size and specific gravity, and are conditioned by 29 Fig. 2.7. Showing spatial locations of local and episodic processes in the source zone of Harris Creek basin (DSR stands for detailed study reach. The significance of colluvium and gullying lies in their coupling tributaries with hill slopes. Modified from Ryder and Fletcher, 1991). 30 fluid density, gravity acceleration, slope of the water surface and hydraulic radius (Slingerland and Smith, 1986). Fine and light materials move in suspension (the sediment held in the water by its turbulence) relatively frequently and rapidly through the channel system. Coarse and heavy materials move as bedload (the sediment supported by intermittent contact with the unmoving bed (Emmett, 1980). Data from pit traps installed in a bar on Harris Creek indicate that gold and magnetite in bedload are only transported for a brief period when increased discharge, caused by snowmelt, disrupts the cobble-gravel pavement (Fletcher and Wolcott, 1991). More detailed discussions on the fluvial processes will be described in the relevant chapters. Local and episodic processes of mass movement are shown in Fig 2.7 and Fig. 2.8. Based on the results of sediment source investigation: (1) at present only four landslides: Landslides #1, #7, #17 and #28 along lower Harris Creek are supplying sediment to the trunk stream; other landslides are either temporarily stable or debouch into inactive channels. (2) of sixteen tributaries, only McAuley Creek, Mosquito Creek and Vidler Creek are major contributors of lower Harris Creek; others are either small or decoupled from the trunk stream by the floodplain; (3) bank erosion along laterally unstable reaches and a big log jam in lower Harris Creek are producing sediment loads. Sediment sources are classified into primary and secondary sources based on whether the sediments have been reworked by current fluvial processes. The primary sources are those that have not been reworked by fluvial processes and are usually located on hill slopes. Landslides #1, #7, #17 and #28 are primary sources. Secondary sources of sediments are the previously deposited alluvial sediments in the channels, alluvial fans and floodplain, and behind logjams. 31 Textures and compositions of the secondary sources are to a large extent unrelated to the proximity and type of the underlying bedrock. McAuley Creek, Mosquito Creek, log jams and alluvial banks along lower Harris Creek are all secondary sources (Fig 2.8). 32 33 C H A P T E R 3 STUDY M E T H O D S AND D A T A Q U A L I T Y E V A L U A T I O N S 3.1 Area Selection The Harris Creek catchment was selected as the study area on the basis of previous studies by Fletcher and his colleagues of fluvial processes and transport of gold and magnetite (Fletcher and Day, 1989; Fletcher 1990; and Fletcher and Wolcott, 1991). Ryder's (1991), and Ryder and Fletcher's (1991) sediment source investigation provide the essential information on the characteristics and locations of geomorphic processes and sediment sources. Other reasons that Harris Creek was chosen for the study are that: (1) access is good; (2) it is in reasonably natural condition with some selective logging activities; and (3) it is typical of many gravel bed streams in British Columbia [Notebook of Geog. 405: Fluvial Geomorphology, 1991]. Study locations and measurement sites in the Harris Creek basin are shown in Fig. 3.1. 3.2. Bulk Density A bulk density sampler was constructed with the help of the technicians of the Department of Geological Sciences, UBC. This tool consists of a phleger corer with screwed cap, cutting head and a 3.49 cm diameter capped internal transparent plastic tube which is 29 cm long and marked at 1 cm intervals. To take a bulk density sample, the plastic tube was placed into the phleger corer, and the cap was screwed on tightly to ensure immobility of the internal plastic tube. Material was taken by hammering the sampler into ground after two to five cm of surface material was scraped off. The sample height was read from the plastic tube's height division, then its volume was calculated with the formula Vs = rcxr2xh (where Vs is material volume, r is the internal radius of the plastic tube which is 1.75 cm, and h is the sample height). Ten bulk density Fig. 3.1. Locations of study sites in the Harris Creek Basin (L #1, #7 #17 and #28 are Landslides #1, #7, #17 and #28). 35 samples were taken along the toe of Landslide #1 (Fig. 3.2), nine from Landslide #7 (Fig. 3.3), nine from Landslide #17 (Fig. 3.4), and seven from Landslide #28 (Fig. 3.5). Bulk density samples were also collected from stream banks at Stations 1, 2 and 3 (Fig. 3.1), and Log Jam #1 (Plate 3.1). In the laboratory, the samples were dried in an oven at 110 °C for twenty four hours to drive off moisture, weighed immediately after, and then bagged. The bulk density (D) of the dry material was determined by D = Ws/Vs, where Ws is the dry weight of the material, and Vs is the same as above. 3.3. Direct Measurements of Erosion Rates Erosion pins, consisting of 45-centimeter stakes, 75-centimeter rebars and 15-centimeter nails, were used to estimate erosion rates. On the landslides, the erosion pins were driven into the active toe above the stream bed and the surface of the slide block (Figs. 3.2 to 3.5). Distances between the erosion pins and the toe wall, and distances from the erosion pins to a reference line were measured and recorded. Along the laterally unstable reaches, three stations were established (Fig. 3.1). At each station and Log Jam #1, erosion pins were inserted into the inactive floodplain along a line approximately parallel to the stream bank (Plate 3.1). Bank position was determined by measuring perpendicular distances from the bank edge to the line of erosion pins. Measurements were taken on July 28, 1991; May 17, 1992; April 29, 1993; and May 30, 1993. The measurement method is simple, but rather laborious. Although it is not suggested that the methods adopted are ideal, or that the results are precise, they allow analysis of annual variations in erosion and some assessments of the relative importance of erosion processes. 36 37 38 39 c c _o '55 s w o 1/3 a. CO G 3 CQ '55 "a. o e O o CO CU co . £ C X) co" -a o 6 <u JC o o <u 00 o co c .2 CO 00 O t N S, ^  ITS C CO C 'S, e 43 O '55 E 1 Plate 3.1. Logjam #1 on lower Harris Creek 41 3.4. Discharge Measurements Stilling wells and staff gauges were installed in June 1992 on McAuley Creek and Mosquito Creek (Fig. 3.1). The Mosquito Creek stilling well was located on the left bank about 150 m upstream from its confluence with Harris Creek. The staff gauge was bolted against the right bank and inserted into a relatively quiescent water pool. McAuley Creek stilling well was on the right bank approximately 100 m upstream from its outlet to Harris Creek. The staff gauge was bolted against the right bank and a tree. Rating curves between discharges and staff gauge recordings were established from flow measurements made with an Ott current meter (impeller A-13329), a measuring tape and a stop watch. On Mosquito Creek measurements were made at five points 0.4 m apart across the stream section. On McAuley Creek measurements were made at seven points 0.5 m apart across the section. For both creeks, measurements were made in the summers of 1992 and 1993. Measurements of discharge versus the staff gauge for both creeks are shown in Appendix Tables 1 and 2 and Figs. 3.6 and 3.7. Mosquito Creek has two curves: one for the flows of 1992 and before 6:00 p.m. of May 13, 1993, the other for the flow after 6:00 p.m. of May 13, 1993. The reason for this is that the water pool in which the staff gauge was bolted was buried by sediment, and aggradation changed the base level. The staff gauge and stilling well on Harris Creek established by Fletcher's and Church's research groups at the Research Site were used to measure flow conditions. Detailed summaries of the rating curves and relation between staff readings and stilling well recordings have been given by Church (1993). Recording curves of the stilling wells were transferred into digital data using a GIS application program called ROOT, and the digital data were then interpreted into hourly discharge data using a computer program WATERFOR (which was provided by Church (Pers. comm., 1993)). 42 ., 0.19 0.24 Staff gauge G (m) 1992 1993 H • Fig. 3.6. Showing relation between discharge and staff gauge on McAuley Creek 43 cy 2 i | 1 j i j | 1 0.5 Curve 1 —yS^. t=>s<; 0.2 0 ! • :::::::::::::::::::::::i:::::::::::::::::::::;:::::::::::::::::::J::::::::::::::^ 0.56 0.61 0.66 0.81 0.86 0.91 1992 0.71 0.76 G ( m ) Before flood After flood in 1993 in 1993 • O Curve 1: lgQ = 5.853G - 4.304 (n = 26, r 2 = 0.98) Curve 2: lgQ = 4.847G - 3.992 (n = 11, r 2 = 0.98) Fig. 3.7 Showing relation between discharge and staff gauge on Mosquito Creek. 44 3.5 Geochemical Sampling As already stated in Chapter 1, conventional geochemical method with sampling of about 500 g of material, and analysis of a 10 to 20 g subsample of the minus 0.177 mm (80 mesh ASTM) fraction can not produce stable results for Au because the variance from sampling and analysis could far exceed the natural fluctuation. Day and Fletcher (1986), on the basis of studies of representative sample size and optimum size fraction for gold determination from five streams in British Columbia (including Harris Creek), found that 20-kg of minus 5 mm field samples can provide a representative sediment for gold determination in the minus 0.053 mm fraction whereas up to 300-kg of minus 5 mm field sample is needed to provide a representative sediment for gold determination in coarse size factions. According to their study results and also in consideration of time taken to process samples in laboratory, instead of collecting many small samples, a few large samples are collected from landslides, banks, logjam and stream sediment. 3.5.1 Bedload sampling with Helley-Smith sampler Depending on discharge, a standard Helley-Smith bedload sampler (7.62 x 7.62-cm orifice with an aluminum frame (Emmett 1980)) (Fig. 3.8) and portable Helley-Smith sampler (7.62 x 7.62-cm orifice with a metal rod instead of the aluminum frame (Plate 3.2) were used to collect bedload samples. Although the sampler was designed to take bedload sediment, it can not be asserted unequivocally that only the bedload sediment is collected, since the possibility exists for "low-flying" suspended load to be carried into and trapped by the sampler. Therefore, for this study, the practical definition of bedload is taken to be the sediment moving on or near the stream bed. By recognition of this practical definition, the Helley-Smith sampler is used to measure the transport of sediment, magnetite and gold. The procedure recommended by Church (personal communication 1992) was adapted in the field. The sampler was deployed by hand if stream water was low so that the nozzle was facing and orthogonal to the flow direction. The base of the 45 Fig. 3.8. Helley-Smith sampler (Emmett, 1976) and its possible position under water. Plate 3.2. Showing the portable Helley-Smith sampler with a metal rod. 47 sampler was laid firmly in contact with the stream bed to minimize potential sampling errors due to scooping of bed material during sampling. On McAuley Creek, the samples were taken from points 0.5 m apart across the stream. Sampling time was 15 to 30 minutes at each point. On Mosquito Creek, samples were taken from points 0.4 m apart across the stream. Sampling time was 2 to 40 minutes at each point, depending on the flow condition and the availability of stream sediment. Because the sample from each point was small, samples from the all points, for each run, were pooled to form a single composite sample. On Harris Creek, samples were collected, depending on flow conditions, near or from the road bridge about 300 m upstream from the Research Site. When flow was low, the sampler was deployed by wading across the stream. When flow was too high, the bridge was used to lower and retrieve the sampler. A plastic fin was tied to the bottom of the sampling bag in order to stretch out the sampling bag during the sampling period. Samples were made from points spaced 1 m apart across the stream section. Sampling time was 2 to 30 minutes at each point. Samples from all points were pooled to form a single composite sample for each run. Weights of the bulk samples vary from 0.4 kg to 34 kg. 3.5.2 Sampling from landslides, bank and logjam Samples were collected along the active toes of the landslides. At each sampling site, 2 to 5 cm of rain washed-splashed surface material was scraped off and fresh material from near vertical faces was collected. Four samples were taken from Landslide #1, four samples from Landslide #7, two samples from Landslide #17, and three samples from Landslide #28 (Figs. 3.2 to 3.5). Samples weighed from 16 to 30 kg. It is, however, necessary to point out that, because of the heterogeneity of glacial deposits, the sampling representativeness may not be easily justified from the few samples from a landslides even though they are large in size. 48 For laterally unstable reaches, one sample was collected from the floodplain and one from the sandy gravel at Stations 1 and 2, respectively. At Station 3 two samples were collected after 2 to 5 cm of surface material was scraped off (Fig. 3.1). Each sample weighed 15 kg to 20 kg. Samples from landslides and bank erosion stations were not sieved in the field because most of them contain enough fine (sandy) material to produce representative subsamples. On sediment trapped behind Log Jam #1, samples were collected from the vertical profile after the surface organic litter was scraped off. The sampling interval was 0.2 m, and four samples were collected. To obtain sufficient fine material for analysis, the sediment was wet sieved through the 2 mm screen into a plastic pail in the field. When the pail was full of sediment and water, overflow was inevitable, and very fine sediment would flow out in suspension. To catch this fine sediment, the pail was put in a galvanized catch-basin. After completion of sampling, the basin was allowed to settle about one hour, then the water was siphoned out, and the fine sediment was washed back into the pail. Each sample consists of 2 pails of minus 2 mm sediments weighing about 50 kg. In order to obtain this size, about 280 kg sediment needs field-sieving. All of the +2 mm sediment was weighed and hand-sieved through a 16 mm field sieve. 10 to 20 kg of the -16 to +2 mm sediment was quartered, bagged and returned to the laboratory. The +16 mm material was hand-tested through a metal template with apertures of half phi intervals from 16 to 128 mm: each size fraction was weighed, noted and finally discarded. 3.5.3 Stream sediment Bars are the common sediment storage sites in the stream channel along lower Harris Creek. As mentioned before, most bars have two distinctive environments: high energy environment (HEE) at bar heads, and low energy environment (LEE) at bar tails. On each bar (Fig. 3.9), two samples were taken, one from the HEE, and the other from the LEE with one exception that at 49 Fig. 3.9. Stream sediment sample sites along Harris Creek (Small circles are sampling sites. The sample above the line is from HEE, and the sample under the line is from LEE. Other legends see Fig. 2.8). 50 91-HZ-56 site no LEE sample was obtained because the bar does not have a discernible LEE. Sampling procedure for HEE material was the same as that for the log jam sediment. Two 23-1 pails (equivalent to about 50 kg) of the -2 mm sediment were sieved from about 350 kg of stream sediment. Two 23-1 pails (equivalent to about 45 kg) of LEE sediment were collected directly from the LEE without sieving because most of the sediment in the LEE is finer than 2 mm. 3.6. Laboratory Methods 3.6.1. Wet and dry sieving Although wet sieving was extremely time-consuming (for example, the total sieving time for this program was about two years) since the samples were very large (up to about 50 kg), this technique was still used because fine sediments bind together forming clay balls which can only be easily disaggregated by wet sieving. This also provides the clean size fractions required for preparing heavy mineral concentrates. Each sample was wet-sieved into seven fractions (Fig. 3.10) between the +0.425 mm and -0.053 mm using a peristaltic pump and recirculating water system (Day, 1988) that catches the -0.053 mm material in a bucket. Clay balls were broken down with the fingers. For samples from LEE, only one bucket of sediment was sieved. After sieving the size fractions coarser than 0.053 mm were dried in an oven, weighed, and then bagged. Flocculant (Cat-Floe made by Calgon Corp. USA) was added to the suspended -0.053 mm fraction and allowed to settle until the water became clear: the water was then decanted off, and the sediment was transferred into drying trays, dried at 50°C, weighed, and bagged. Between samples, the sieves were thoroughly cleaned using a toothbrush, and visually checked to detect screen breakage. The +0.425 mm fractions for samples from landslides, the log jam and Helley-Smith samples, and the -16 mm +2 mm splits from the field samples were dry-sieved into half phi a Q •a I V) C N p 1 e E A a C N o <o ( N A A O O a S C N C N f—1 C N V V fa '/-> O E A E A r-A A ir> A O o E E E A A o v"i A v-l O 1—1 V V <7 b 3 •a 3 .8 a 13 a. . o •a o a .8 00 o - t - i a - s o. CL, IC 3 i • < 8> '•a a < •8 with a ^ 13 3 I 2 I " o •\a-a. B O E B O C >. i-o to 8 •a 1 '•a_ 'a ••a ! c Q--a 1 - t -B s •a •I o o 3 < u < < CJ T3 ^ f g C O e £ .5 -5 e £ -a c o 1 -2 g D. o "75 w 2 S ^ o D-X 52 fractions for sedimentary texture analysis (Fig. 3.10). All sieving was done manually, not by the Rotap automatic siever, because the disk-, blade-, and roller-shaped gravels and pebbles rest flat on the sieve screen and have to be put through the openings by hand. 3.6.2. Heavy mineral separation For the samples from the landslides, bank erosion stations, log jam, and stream sediments along lower Harris Creek, heavy minerals were separated from the -0.149 +0.105 mm, -0.105 +0.075 mm, and -0.075+0.053 mm size fractions (Fig. 3.10) using methylene iodide (CH2I2), a heavy liquid with a specific gravity of about 3.3. The detailed operations were described by Day (1988). The Helley-Smith samples did not provide sufficient material for heavy mineral separation, and the fractions were close to or less than the weight of 30 grams required for determination of Au. Magnetic minerals were separated using a piston magnet from the three fractions of heavy minerals (Fig. 3.10). The separation procedure was repeated carefully several times in order to release the non-magnetic particles trapped by the magnetic particles. The magnetic and non-magnetic heavy mineral fractions were then weighed and vialed. The magnetic heavy mineral fractions will be called magnetite in this thesis although some magnetite may be included in rock fragments or other mineral particles. Magnetic fractions were directly separated from the three size fractions of the Helley-Smith samples without prior heavy mineral separation. These magnetic fractions will be also called magnetite in this thesis, although the magnetic fraction produced without heavy liquid separation is not strictly equivalent to the magnetic fraction separated with the heavy liquid, because the former contains composite particles that can be extracted with the piston magnet but will not sink in the heavy liquid. However, comparison of this magnetic fraction with the magnetite separated with heavy liquid showed no visible difference. 53 3.6.3 Combination and split The non-magnetic heavy mineral concentrates of each of the three fractions were considerably less than the 30 grams needed for FA-AAS (Fire Assay-Atomic Absorption Spectrometry) analysis. They were therefore, pooled together to form a composite sample (Fig. 3.10) with a size fraction between the -0.149 mm and +0.053 mm, and the entire sample analyzed. For the few composite samples larger than 30 grams, a 30-gram split was taken using the Jones riffle splitter. The same procedure was applied to the non-magnetic fractions of the Helley-Smith samples. Because samples collected from the Harris Creek basin contain abundant -0.053 mm sediment, it is necessary to split this fraction for analysis. After being disaggregated by a heavy stainless steel roller, 30-gram of the -0.053 mm sediment was split off with the Jones riffle splitter from the original samples. After completion of splitting of each sample, the splitter and containers were thoroughly cleaned using a high-pressure airgun and paper towel. 3.6.5 Chemical analysis Gold Gold in the non-magnetic heavy mineral concentrates was determined by routine fire assay -atomic absorption spectrometry (FA - AAS) with a detection limit of 5 ppb. Gold in the -0.053 mm sediment and in the -0.149+0.053 mm composite fraction of Helley-Smith samples was determined using an aqua regia digestion - preconcentration with aqua regia - ion exchange column - inductively coupled plasma method (AQ- ICP) with a detection limit of 0.1 ppb (Fletcher et al., 1995). Other Elements 54 Splits of the -0.053 mm fractions of samples from Landslide #1 and the Detailed Study Reach (DSR in Fig. 2.8) were decomposed by evaporation to dryness with mixed hydrofluoric -nitric - perchloric acids, followed by resolubilization of the residue in hydrochloric acid and analysis by inductively coupled plasma spectrometry for Ag, Al, Ba, Be, Bi, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P, Pb, Sr, Ti, V, W and Zn. All analytical work was done by a commercial laboratory in Vancouver. 3.7 Data Accuracy and Precision 3.7! Accuracy Canadian Certified Reference Material Project Reference Till Sample Till-1 was submitted with the samples analyzed by FA-AAS and AQ-ICP, respectively. Due to the expense, the standard sample was submitted to be analyzed only once for each analytical method. Results are shown in Table 3.1. The analytical value of the reference sample by FA-AAS is very close to the reference value of 14 ppb; But the analytical value by AQ-ICP is 8.2 ppb i.e. lower than the recommended value of 14 ppb, probably as a result of incomplete dissolution of encapsulated gold. 3.7.3 Analytical precision Analysis of duplicate splits of gold in the -0.149+0.053 mm non magnetic heavy mineral concentrates and Helley-Smith samples are listed in Table 3.2. The results for the non-magnetic heavy mineral concentrates by FA-AAS gave poor reproducibility, as was also found by Day in Harris Creek in 1988. He attributed the poor reproducibility to the "nugget effect". The duplicate analyses of the -0.149+0.053 mm Helley-Smith sediment fraction for gold by AQ-ICP also show the poor reproducibility. Three of the five duplicates have results below the detection limit of 0.1 ppb. Table 3.1. Gold analysis(ppb) of Reference Till Sample Till-1 by AF-AAS and AQ-ICP Standard Reference value Till-1 14 ± 4 * Analytical method n result(ppb) FA-AAS 1 15 AQ-ICP 1 8.2 * details see Lynch(1990) n = the number of reference samples Table 3.2. Duplicate analyses of gold in -150+53 um fractions with FA-AAS and AQ-ICP (Au in ppb) FA-AAS -150+53 um non-magnetic heavy mineral concentrates Sample Original Duplicate 91-HZ-42 760 2070 91-HZ-34 <5* 145 91-HZ-41 5 5 91-HZ-73 105 405 91-HZ-74 20 15 AQ-ICP 150+53 um Helley-Smith stream sediment Sample Original Duplicate 93-HZ-25 6.1 <0! 93-HZ-26 <0.1* <0! 93-HZ-28 <0.1 <0! 93-HZ-42 <0! 0.2 93-HZ-52 0.7 0.6 * < means below the detection limit. 57 Fig. 3.11. Duplicate analyses of -0.053 mm sediments for gold by AQ-ICP. 58 Duplicate analyses of gold in the -0.053 mm size fraction of different samples are shown in Fig. 3.11. For samples from the high energy environment (HEE), low energy environment (LEE), bank erosion, log jam, and landslides, the precision is close to or within ±50%. For Helley-Smith samples, however, duplicate results display large variability. Duplicate analyses for Ag, Al, Ba, Be, Bi, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P, Pb, Sr, Ti, V, W and Zn in the -0.053 mm size fraction are listed in Table 3.4. Compared to gold, these trace elements have very good reproducibility. The duplicate results strongly suggest that, for gold in the -0.149+0.053 mm size fraction, even collecting of 50-kg of minus 2 mm sediment from 350 kg stream sediment, sieving into different size fractions and preparation of heavy mineral concentrates are still not good enough to produce reproducible results. By use of the minus 0.053 mm size fraction of sediment, 20-kg of minus 2 mm sample will be sufficient to provide satisfactory subsamples for both gold and other trace element determinations. 59 Table 3.3 Duplicate results for 24 trace elements in -53 um size fraction of sediments 91-HZ-10 93-HZ-15 Original Duplicate Original Duplicate Ag (ppm) <0.2 <0.2 <0.2 <0.2 Al (%) 7.59 7.85 8.27 7.92 Ba (ppm) 820 860 990 950 Be (ppm) <0.5 <0.5 <0.5 <0.5 Bi (ppm) <2 <2 2 2 Ca (%) 2.49 2.59 2.62 2.50 Cd (ppm) <0.5 <0.5 <0.5 <0.5 Co (ppm) 16 16 20 19 Cr (ppm) 87 92 101 97 Cu (ppm) 91 63 84 82 Fe (%) 4.31 4.49 5.71 5.45 K (%) 1.75 1.82 1.81 1.72 Mg (%) 1.69 1.76 1.59 1.53 Mn (ppm) 740 770 1350 1295 Mo (ppm) 1 2 1 2 Na (%) 1.66 1.70 1.87 1.80 Ni (ppm) 36 37 39 36 P (ppm) 870 920 1730 1660 Pb (ppm) 14 10 18 20 Sr (ppm) 330 343 437 419 Ti (ppm) 0.46 0.47 0.78 0.75 V (ppm) 140 147 172 164 W (ppm) <10 <10 <10 <10 Zn (ppm) 100 104 128 124 Note: 91-HZ-10 is from Landslide #1 93-HZ-15 is stream sediment 60 CHAPTER 4 INPUT OF SEDIMENT, MAGNETITE AND GOLD: ESTIMATE WITH FDZLD MEASUREMENTS 4.1 Calculation Formulas As stated in Chapter 2, four landslides, two tributaries, one log jam and laterally unstable channel banks along lower Harris Creek are the main active sediment sources (Figs. 2.7 and 2.8). In order to investigate relative contributions of these sources and to evaluate the importance of the geomorphic processes by which magnetite and gold are introduced to lower Harris Creek, the following equations are employed to calculate mass inputs from these different sediment sources. (1) Mass input of sediment, magnetite and gold from tributaries: I = 27(Q)xTxW (4.1) where I = total mass input from a tributary during the monitoring period, which is expressed as metric tons for sediment, kilograms for magnetite and grams for gold; J{Q) = the average transport rate per unit width and is expressed as a function of discharge (Q) based on the Helley-Smith data and flow records; T is the time interval during which the discharge (Q) lasts; and W is the width of channel. Z is sum of the mass input over the range of discharges. (2) Mass input of sediment from landslides, the logjam and bank erosion: I = D x A , x R x T (4.2) where I = total mass input expressed as metric tons; D = bulk density of source sediment measured from the active area using the bulk density sampler; A a = the active area of sediment 61 source as estimated from the field survey; R = linear erosion rate measured on site (see Chapter 3 for the methods); and T = time interval with the dimension of year or day, depending on the field season and measurement arrangement. (3) Mass input of magnetite and gold from landslides, the logjam and bank erosion: I = F x ( D x A a x R x T ) x C (4.3) where I, D , A a , R and T are the same as for Equation (4.2). For magnetite F = the percentage of the -0.149+0.105 mm, -0.105+0.075 mm or -0.075+0.053 mm size fraction in the sample. For gold F = the percentage of the -0.149+0.053 mm and the -0.053 mm size fraction. C = concentration of gold (ppb) and magnetite (%) in the specific size fraction. Reliability of results obviously depends upon the precision and constraints of the field survey and measurements, and quality of the analytical data. The temporal resolution of the results is constrained by the timing of the field season and schedule. Most surveys were conducted in the late spring and early summer in each of three years from 1991 to 1993 during or around the time during which the annual snowmelt flood occurs. Considering that transportation and deposition of sediments in Harris Creek are associated with flood events (Fletcher and Wolcott, 1991), results can be used to evaluate the relations of the input and output of gold and magnetite with flow conditions in the Harris Creek basin. Spatial scale considerations provide a way of interpreting and aggregating the budget results, and spatial disequilibrium can be defined by the difference between mass input and output. Therefore, the budget results presented below can provide a framework for examining fundamental processes in the basin. The sediment budget methodology, in combination with consideration of fluvial process, may lead to a better understanding of the downstream dispersion of gold and magnetite. 62 4.2 Inputs of Sediment, Magnetite and Gold from McAuley Creek Introduction to McAuley Creek McAuley Creek, the largest tributary in the study area, is on the south side of lower Harris Creek, 7.7 km upstream from the Research Site (Fig. 2.7). It is a gravel-bed stream with a catchment area of 36 km ,^ of which 15.2 km^ is located in the Zero Transfer Zone, and 20.8 km^ in the Source Zone. Rockslides and other colluvial slopes connect the stream with slopes around the headwater area. Along lower McAuley Creek the stream flows through the floodplain which is covered by trees and bushes. Like Harris Creek, most bars along lower McAuley Creek have two distinct environments: a high energy environment at bar heads and a low energy environment at bar tails. Geochemical results for sites HZ-43 and HZ-44 clearly show the preferential accumulation of coarse gold and magnetite in the high energy environment at bar-head gravels; but fine (-0.075+0.053 mm) magnetite and fine (-0.053 mm) gold do not show this trend (Table 4.1). Au concentration, however, is appreciably lower than in Mosquito Creek. Field panning found pristine gold particles in the stream sediment (Hou and Fletcher, 1991). Hydraulic Conditions Discharge measurements, made in the 1993 season, are summarized in Fig. 4.1. From late April to early May, McAuley Creek flows were low with discharges of about 0.5 m-Vs. The Spring snowmelt flood, with an instantaneous peak of 5.2 m-Vs, occurred on May 11 and ended on May 29. Several peaks occurred in June and July. 63 m o JO CD-S' 22 o O oo o o r-' (N r-& 2 •5 Q j * o w < s VD O o oo CO O rj-' O * p (N II C 3 cr 64 65 Bedload Sampling McAuley Creek was sampled with the Helley-Smith Sampler during 1992 and 1993 field seasons. Sampling techniques have been described in Chapter 3. Table 4.2 shows the sampling parameters. Mass Inputs of Stream Sediment Weights of different size fractions of the Helley-Smith samples are listed in Appendix Table 3. The transport rate of stream sediment is calculated with the following formula (Church, 1993, per. comm.): R = (W t/T)x( 1/0.0762) (4.4) where R is the average transport rate per unit width; W t is total weight in grams; T is total sampling time (expressed as seconds); 0.0762 is the nozzle width of the Helley-Smith sampler in meters. The transport rate increases in response to increasing stream flow (Q) (Fig. 4.2). When examined in detail by consideration of the rising and falling limbs, sediment seems to exhibit hysteresis with rates being greater on the rising limb than that on the falling limb. One data point on the rising limb is question-marked in Fig. 4.2 because it probably represents an anomalous flow condition due to sample compositing. Because there are only three points on the falling limb, and because the data are quite scattered on the rising limb, regression analysis for the rising limb is not significantly different from that for the falling limb, therefore, one regression analysis is used to describe the general effect of increasing discharge (Q) on the increase of the transport rate: R = 0.02Q4-8 4 (n = 10, r 2 = 0.81, AR = 0.43 to 2.34). where 4.84 is the slope coefficient, and 0.02 is the constant. An F-test indicates that this equation is highly significant at a = 0.05. AR is a 2-standard error range of the transport rate with 66 Table 4.2. Sampling parameters with Helley-Smith sampler on McAuley Creek in 1992 and 1993. Sampling Sampling Discharge Weight Sample No. Date stations time/station (mVs) Flood limb Composite (kg) 92-HZ-10 20-May-92 7 15* 1.10 R 0.14 92-HZ-ll 20-May-92 7 15 0.90 R 0.01 92-HZ-17 26-May-92 7 15 2.39 R 1.5 92-HZ-18 26-May-92 7 15 2.49 R 2.4 92-HZ-20 27-May-92 7 15 1.65 R 0.38 93-HZ-29 12-May-93 6 30 2.45 R 93-HZ-30 12-May-93 6 30 2.49 R 93-HZ-31 12-May-93 6 30 2.53 R 0.43 93-HZ-32 12-May-93 6 30 2.53 R 93-HZ-33 12-May-93 6 30 2.70 R 93-HZ-34 12-May-93 6 , 30 2.91 R ^ » 93-HZ-43 14-May-93 • 6 30 3.51 R HHP 6.6 93-HZ-44 14-May-93 6 30 3.55 R 93-HZ-45 15-May-93 6 30 3.28 R 93-HZ-46 15-May-93 6 30 3.34 F 7.4 93-HZ-47 15-May-93 6 30 3.40 F 6.0 93-HZ-51 16-May-93 6 30 2.51 F 1 III I I 1.9 93-HZ-53 17-May-93 6 40 2.15 F lllllllll lllllllll Note: R means that the sample was on the rising limb, and F means that the sample was on the falling limb. The samples with the same pattern were composited. * the sampling time is in minutes. 67 2 •c o t/J 0.0001 0.01 b 0.001 t-Discharge (mVs) Rising limb Falling limb O • Fig. 4.2. Showing the relation between transport rate of stream sediment and discharge on McAuley Creek. (The horizontal bar shows the discharge range of composited samples) (See text for the question mark) 68 confidence coefficient of 95%, which is estimated using a formula: ±to/2(n-2)CjV(l/n) simplified from Krumbein and Graybill (1965) (where ta/2(n-2) is two-tail student's t distribution, a is standard deviation and equal to [l/(n-2)]x[SSy2 - (SSXY)2/(SSX2)], and n is sample size). Use of it^n-2)<W(l/n) to represent a confidence interval for error estimates is justified by the following reasons: the above regression equation of transport rate against discharge is an empirical formula, the quality of which depends on the sample size (n) and the scatter of measurements (a). The statistical error of the slope coefficient is specifically ignored because no underlying theoretical result is supposed (Church, per. comm., 1997). Stream sediment input from McAuley Creek to lower Harris Creek is then estimated using Equation (4.1): I = IffQJxTxW = {E0.02Q484x3600x4.85} x IO"6 where 3600 (seconds) is the duration time of a specific discharge (Q) from the hydrograph, 4.85 (m) is the channel width, and 10'6 is the factor to convert grams to metric tons. The calculations estimate that, during the entire record period of 3,220 hours from late April to early September in 1993, McAuley Creek supplied 40 tons (with a range of 17.0 to 93.6 tonnes) of stream sediments to lower Harris Creek, of which 98% was mobilized when the flow discharge was equal to or higher than 1.0 nvVs. The snowmelt flood in May lasted 443 hours (about 14% of the total record hours) and mobilized about 69.0% of the sediment. Floods from June to July lasted 593 hours and transported 29.0% of the sediment. Flow with discharge less than 1.0 nvVs lasted 2184* hours (68% of the total hours), but introduced only 2.0% of the sediment into Harris Creek. Chemical analyses found that gold in the Helley-Smith samples was below the detection limit. Also, when separating magnetite with the piston magnet, no magnetic particles were obtained (Table 4.3) from the samples. To account for absence of gold and magnetite in bedload Table 4.3, Showing concentrations of gold and magnetite in bedload samples from M c A u l e y Creek Gold Magnetite Sample No . (ppb) (g) 93-HZ-29-34 <0.1/<0.2* no 93-HZ-43-45 <0.1/<0.2 no 93-HZ-46 <0.1/<0.1 no 93-HZ-47 <0.1/<0.2 no 93-HZ-51-53 <0.3/<0.2 no no = no magnetic particles were separated from the bulk sample. * the number at top is for gold in the -0.149+0.053 mm size fraction, and the number at bottom is for gold in the -0.053 mm size fraction. 70 samples from McAuley Creek, cumulative probability plots are used to identify the presence of bedload populations that correspond to the gravel framework and sand matrix (see Chapter 5 for details). The result (Fig. 4.3) shows that the material trapped in the Helley-Smith sampler is a single population without the bimodal distribution that appears when the framework is mobilized. This suggests that the flows in McAuley Creek did not reach to the critical discharge to disturb and mobilize the pavement gravels and hence release heavy minerals. 4.3 Inputs of Sediment, Magnetite and Gold from Mosquito Creek Introduction Mosquito Creek, 11.4 km upstream from the Research Site (Fig. 2.7), is located on the north side of lower Harris Creek. This creek has a coarse-sand bed with a catchment of 28 km2, of which 8.0 km 2 is in the Zero Transfer Zone, and 20 km2 is within the Source Zone. In this small catchment, rock falls, landslides and other colluvium couple the creek with slopes. The stream descends from a steep V-shaped valley and flows into Harris Creek. Along Mosquito Creek large organic debris often redirects flows and forms small waterfalls. Gold and magnetite in Mosquito Creek stream sediment were investigated in 1991. Results indicated that Mosquito Creek contains high concentrations of gold not only in the -0.149+0.053 mm size fraction, but also in the -0.053 mm size fraction (Table 4.1). Both coarse and fine gold show preferential accumulations in high energy environments. Field panning in 1991 found pristine gold particles in the stream sediment. Hydraulic Conditions Discharge measurements made in 1993 are shown in Fig. 4.4. In 1991 and 1992, Mosquito Creek had very low flow, and no measurable samples were collected with the Helley-Smith sampler. In 1993, the creek departed from its normal stage and experienced a high flow: in late F i g . 4.3. Showing population distribution in one composi ted H e l l e y - S m i t h sample on M c A u l e y Creek 9 9B 95 OO 80 70 60 50 40 30 20 10 5 . 2 5 10 20 30 40 SO 60 70 80 80 95 98 . Cumulative % coarser 72 \0 m Tt ro fN •—' O (s/ftu) aSreipsifj 73 April the creek was at low flow stage with discharge of about 0.3 m-Vs, that rose sharply on May 10 and reached an instantaneous peak of about 5.5 nvVs on May 13. After the spring freshet of May 24, flows quickly returned to low stage with a discharge of about 0.5 nvVs. Several very small peaks occurred in June and July. Bedload Sampling Sampling data using the Helly-Smith sampler are listed in Table 4.4. A consecutive series of twenty six samples were collected between April 30 and May 17 in 1993. However, due to their small sizes, samples taken from the same flood limb and having similar discharges were pooled to obtain seven composite samples for gold and magnetite analysis. 4.3.1 Inputs of sediment, magnetite and gold Sediment Weights of different size fractions of Helley-Smith samples from Mosquito Creek in 1993 are listed in Appendix Table 3. No results for 1991 and 1992 are reported here, because the gauge was tampered with and level changed by disturbance. The transport rate of sediment is calculated using the formula (4.4). The relation between the transport rate and discharge is shown in Fig. 4.5 and is expressed as the following regression equation: R = 2.82Q3 0 5 (n = 7, r 2 = 0.96, AR = 0.90 to 1.11) where definitions for the parameters are those given previously. An F-test shows that this equation is highly significant at a = 0.05. The sediment input from Mosquito Creek to Harris Creek is then calculated using the following formula: I =If(Q)xTxW - {E2.82Q3 05x3600xW}/10"6 J»S O e C O E CJ c-CO , O . I 00 _c E re oo op •sy £ 1^ J O s T 3 O E J S •3 60 ea JS 1> 6 00 d 1 s Q. e CO O O CO c a CO 00 _C "5. 6 ca 00 o Z _a> CO oo rf r t cn C O D i O i D i O i D i C i O i C t : C i o i C i Oi Oi Oi Qi Oi P i Oi Di Ii. UH CL, ^ -t rt cn cn <n cn r f r f r f cn cn cn cn cn cn cn rf' r t T f T t T t r t r t T f cn —i 00 N O cn cn N O N O cn cn cr, o o o o o' c 00 N O O N O O f-O , *-H r f C N r- cn o cn r f 00 00 N O N O v£l 00 00 „ _ _ _ ^ o o o o o o o O N m cn oo N O C N cn r— —i rf" rf cn oo cn o o o o o o o o C N C N c N C N C N r f r f c n o o o o o cn cn cn cn m o o o o o o o r f r f Tf r f Tf T f r f 22 22 o o cn t— t— NO NO NO t"~ I— 00 O O O O O O O O O O 00 00 00 00 00 00 00 00 O N O N O N O N N O K >» >» >» M< ca ca re cs < s s s's s s s O I i l i i ' I s s s co ca >>>>>-> ca ca ca ca ca ca co ca co cn N O N O c^  r- O O O O O N O N O N co ca N O r~ cn N O oc C N o — ' C N m r f c n N o r ^ o o o o o ' — O —' CN m r f <N CN CN CN CN 00 CN —« CN O N r f r f r f O CN cn cn ^ a s a s a a a a a a a a a a ^ a a a ? ^ a cn cn cn rn c*~. r^. m cn O N O N O N C N O ^ ^ C ^ O N cn cn cn cn cn C \ O ON ON O cn cn cn cn cn cn G \ O N O N O N O N O N O N cn cn cn cn ON ON ON ON cn cn ON ON 100 h 30 10 h 0.3 0.5 1 1.5 2 3 Discharge (m3/s) Fig. 4.5. Showing the relation of stream-sediment transport rate with discharge on Mosquito Creek in 1993. Note: the horizontal lines show the discharge ranges for the composited samples. The data with question mark is not used in the text. 76 where the explanation for the parameters is the same as for McAuley Creek. During the entire recording period of 3196 hours in 1993, Mosquito Creek delivered 332.7 tons (with range of 298.77 to 370.47 tons) of stream sediment into Harris Creek. The flow below 1.0 m-Vs lasted for 2931 hours, but only delivered 2.8% of the sediment. The snowmelt flood with discharges higher than 1.0 m3/s in May lasted 265 hours (which accounts for 8.3% of the entire recording time) and transported 323.5 tons of sediment, accounting for about 97% of the total. Magnetite Concentrations of magnetite in the -0.149+0.105 mm, -0.105+0.075 mm and -0.075+0.053 mm size fractions using Helley-Smith samples are shown in Fig. 4.6a. The relation of transport rates of magnetite in the three size fractions with discharge is shown in Fig. 4.6b. The relative abundance (weight percent) of magnetite versus discharge is quite erratic; no systematic trends or hysteresis are evident. The reason that transport rates of magnetite in the three size fractions have good relation to discharge (Fig. 4.6b) is because the transport rate is a product of concentration and discharge. Therefore, the transport rate exaggerates the real correlation of the independent quantities of concentration and discharge. Statistical summaries for the regression analysis, with an error interval (AR) for each equation, are listed in Table 4.5. F-test ratios show that all of the three equations are highly significant at a = 0.05. But slope coefficients are not significantly different at a = 0.05. That is, transport rates of magnetite in -0.149+0.105 mm, -0.105+0.075 mm and -0.075+0.053 mm fractions respond similarly to changing discharge. Similar results are obtained from lower Harris Creek (see Chapter 5), and were also obtained by Fletcher and Wolcott (1991). The mass input of magnetite, estimated with equation (4.1), is listed in Table 4.6. The spring snowmelt flood accounts for only 8.3% of the entire time but delivered 98% of the magnetite in the three size fractions. 77 g o c o CJ 2.8 2.3 1.8 1.3 0.8 - R F F • R • • F • R • F • O F R o • R F O • • R O F O R O F • F — R • • F O R F • i i • i 0.5 1 1.5 2 3 Discharge (mVs) -0.149+0.105 mm -0.105+0.075 mm -0.075+0.0^3 mm • O Fig. 4.6a. Showing the relationship between concentration of magnetite in the three size fractions and discharge on Mosquito Creek in 1993 OR. = Rising limb and F = Falling limb) 100 X I o s o a. 0.01 1.5 2 Discharge (mVs) -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm • ° Fig. 4.6b. Showing the relationship between transport rates of magnetite in the three size fractions and discharge on Mosquito Creek in 1993 78 Table 4.5. Statistical summaries for the transport rates of magnetite in -0.149+0.105 mm, 0. 105+0.075 mm, and -0.075+0.053 mm, and gold in -0.053 mm on Mosquito Creek in 1993. 1. Transport rate of magnetite in the -0.149+0.105 mm size fraction (in mg/sxm) R = 0.339xQ316 Observations (n) 6 R 2 0.68 F 8.44 Critical F (<x=0.05) 6.61 AR 0.86-1.16 R = 0.145xQ318 Observations(n) 6 R 2 0.71 F 9.71 Critical F (a=0.05) 6.61 AR 0.84- 1.19 3. Transoort rate of magnetite in -0.075+0.053 mm size fraction R = 0.079xQ330 Observations(n) 6 R 2 0.77 F 13.5 Critical F (a=0.05) 6.61 AR 0.84-1.19 4. Transoort rate of gold in -0.053 mm size fraction (in ug/sxm) R = 0.0000794Q5 0 2 r2 0.90 F 42.3 Critical F (a=0.05) 7.71 AR 0.72 - 1.38 5. Transoort rate of gold in -0.149+0.053 mm size fraction R = 0.0000741Q2 5 2 6 r2 0.93 F 329.5 Critical F (1,4) a=o.o5 7.71 AR 0.90-1.11 79 Gold Concentrations of gold in the -0.149+0.053 mm and -0.053 mm size fractions of the Helley-Smith samples are shown in Table 4.7 and Appendix Table 4. Gold values in the -0.053 mm size fraction are above the detection limit in six of the seven samples. At discharges between 0.5 and 5 m-Vs, the transport rate of gold in the -0.053 mm size fraction increases in response to increasing discharge (Fig. 4.7). Like sediment and magnetite, the effect of increasing discharge on the rates of transport of the -0.053 mm gold can be expressed using statistical regression equation. The statistical summary is listed in Table 4.5. An F-test indicates that the regression equations is highly significant at a = 0.05. Values of gold in the -0.149+0.053 mm size fraction are below the detection limit of 0.1 ppb in four of the seven samples (Table 4.7 and Fig. 4.8 a). According to the field notes, six of the seven samples were taken across the stream; sample 93-HZ-28 is not used here because this sample was not taken across the stream because of high flood. Gold values in the six samples are strongly skewed: three samples have no gold, two samples have very low gold values, and one sample contains 6! ppb gold. Calculation using computer program NUGGET indicates (Stanley, 1989) that only sample 93-HZ-19-25 has sufficient gold to result from contribution of a free grain of gold with a diameter of 0.053 mm and a thickness:diameter of about 1:10. (Hou and Fletcher, 1991). In the size fraction between 0.149 mm and 0.053 mm, 93-HZ-49 with a gold value of 0.8 ppb and 93-HZ-52 with a value of 0.7 ppb may reflect the gold concentration of gangue material. Ingamells and Pitard (1986) suggested that such strongly skewed analytical values with a preponderance of zero values can be described by a Poisson distribution. With one free gold grain in only one out of six samples, the observation of finding zero grains is 5/6 or about 83%, and the observation of finding one grain is 1/6 or 17% (inset of Fig. 4.8 a) in the total of six samples. Calculation, using Pn = e"z(z7n!) (where P„ is the probability that n grains will be found in a sample with the average number (z) of grains) (see Chapter 5 for details), indicates that the 80 32 22 o O CO cn o co cn o o + Os C O VO o 1 O vo co o co o VO CN o o oo o" o VO o co as as CD CD u. CJ o -t-» '3 CT c« O <D -*-» CD C Of. E •o c cd vO r f J D X) 03 H ca cn O o + cn r-o cn r-o o + cn o cn o o + as r f CD s <D 3 T3 O O K o 1 — <D Pi C o » c o o CO oi r f co O C N C N l vq cn C N r f r f r-co vo Ov 5 o CS CD c CD CD J S H C N o C N r f O as o CO r f C O ON C N vo C N CO s O qS ^ JS o CD o « 6 CD , C - G | a. or. J S Table 4.7. Gold concentrations o f bedload samples from Mosquito Creek in 1993 Sample N o . Discharge (mVs) Weight (g) of -0.149+0.053 mm Gold (ppb) -0.149+0.053 mm -0.053 mm 93-HZ-03-12 0.61 59.2 < 0 ! (30)* 0.6(30) 93-HZ-13-17 1.15 164.4 <0.1 (30) 2.1(30) 93-HZ-19-25 0.76 110.9 6 ! (30) 0.7(30) 93-HZ-28 1.69 237.3 < 0 ! (30) 2.7(30) 93-HZ-41-42 4.28 177.3 < 0 ! (30) 3(30) 93-HZ-49 1.78 25.63 0.8 (25.6) <0.2(30) 93-HZ-50-52 1.56 88.5 0.7 (30) 2.9(30) * the number is the parentheses is weight (grams) of the fraction analyzed. 82 0! F X i 3 ro u •c o 0.01 0.001 0.0001 0.00001 0.000001 1 1 ' — 1 — ' — L 1 1 1 1 0.5 1 1.5 2 3 5 Discharge (mVs) Fig. 4.7. Relation between the transport rate of the -0.053 mm gold and discharge using Helley-Smith sampler on Mosquito Creek. (The data in the square is not used and the explanation is given in the text) 0.0003 0.000 0.00003 -3 0.00001 2 I, 0.000003 0.000001 t-3E-7 1E-7 1 1.5 2 Discharge (mVs) 0.005 0.002 0.001 ? £ 0.0005 a. o •c O Q. 0.0002 § 0.0001 0.00005 0.00002 0.00001 1 1.5 2 3 5 Discharge (mVs) Fig. 4.8. Showing the relation between transport rate of the -0.149+0.053 mm gold and discharge, a. based on analytical results; b: according to the estimate based on the Poisson distribution (The arrows indicate that gold values are below detection limit at that discharge, and the number is weight (kg) of -2 mm sediments). 84 average number of grains that corresponds approximately to the above conditions is 0.2 (because Po = e' 0 2x(0.2°/0!)= 82%, and Pi = e"°'2x(0.27l!) = 16%)! The result thus suggests that the analytical values confirm a Poisson distribution with 0.2 grains of 0.053 mm in diameter which is equivalent to 1.6 ppb. Using this concentration, the transport rate is calculated using formula (4.1) and shown in Fig. 4.8.b. The effect of increasing discharge on the increase of transport of the gold in the -0149+0.053 mm size fraction is analyzed using the linear regression equation of the form: R = aQb, and the result is shown in Table 4.5. An F-test indicates that the equation of the transport rate of the gold with discharge is highly significant at a = 0.05. As described before, because transport rate is a product of concentration and flow, transport rate can increase with increasing discharge even though concentration does not do so. The mass input of gold in the two size fractions is estimated with the equation: I = £/(Q)xT xW. Results indicate that Mosquito Creek supplied 0.16 grams of the -0.149+0.053 mm gold, and 0.36 grams of the -0.053 mm gold to Harris Creek during the 3196 hours of recording period in 1993. Unlike sediment and magnetite, gold was transported only during the spring freshet which lasted for 265 hours, accounting for 8.3% of the entire time (Table 4.6). 4.4. Mass Input of Sediment, Magnetite and Gold from Landslides 4.4.1. Landslide #1 Basic Properties Landslide #1, one of the four active landslides, is located on the north bank of Harris Creek, 4.3 km upstream of the Research Site (Fig. 2.7). Landslide debris consists of about 90% glaciolacustrine sediment and 10% glacial till by volume. The field survey estimates that the total 85 volume of material in the slide scar is 9xl0 4 m-* which is being supplied gradually to Harris Creek through the landslide toe. The main processes to trigger the material being delivered to the creek are fluvial undercutting and mechanical collapse of the overhanging block. The size fraction distributions of the toe material are shown in Fig. 4.9A. It is obvious that, although Landslide #1 consists of a wide range of particle sizes, approximately 50% is silt and clay. The field survey determined that the active area (Aa) is: LxH = 25x2.1 m 2 where L is the length and H is the average height of the active toe. Physical and chemical parameters of the landslide toe are listed in Table 4.8. For landslides, log jam and bank erosion samples, weights of sediment fractions, magnetic and non-magnetic heavy mineral concentrates, and concentrations of gold in -100+270 and the -270 meshes are listed in Appendix Tables 5, 6 and 7. Inputs of sediment, magnetite and gold Input of sediment is calculated with Equation (4.2), and inputs of magnetite and gold are calculated using the Equation (4.3) on the basis of the data in Table 4.8. Results with calculation errors are listed in Table 4.9. Results show that, (1) this landslide contains very low gold concentrations; (2) the amount of input varies temporally with maximum inputs from May 1 to May 31 in 1993 coinciding with the May 1993 flood (Fig. 4.3 and also see Chapter 5). 4.4.2. Landslide #7 Basic Properties Landslide #7, located on the north bank of Harris Creek, 8.1 km upstream from the Research Site (Fig. 2.7), consists of about 80% till and 20% glaciofluvial sediment by volume. The volume of the material accumulated in the scar was estimated as 1.2x105 nP, and the material is being gradually supplied to Harris Creek through the active toe measuring 12x2 m 2 before May 1 in 1993. During the May 1993 spring freshet, Landslide #7 became very active (Plate 4.1) with movement of a large block of material into Harris Creek being triggered by fluvial 86 J3UIJ JU30J3(J J3UIJ JU30J9J 87 ro II C o N 00 OO l « N r t ro o ro O o rt c o o i— o Cu =0 rt 2 N O CN i o NO <i0 o ro CN 4 3 C L CL O O o x * — j r— < > od E Q oo 0) e 2 CO C 1) -a 3 CQ c cj E u L H 3 CO rt <L> s ro ON ro ON ON ~ - c ON 15" ^ »S rt E S 2 o tU ~ ON ro ON ON — ' ON oo" >> f> rt — 2 o. E < 2 2 tu ON rs ON ON — 1 ON oo" C N r--—' s 2 2 ° tu T3 O o -H rt rt i> o r t O +1 CN rt o oo o +1 r t C J 3 O tu a o *-a CJ C J oo o C J 4 3 SP & o •s .s e o tS eg C J N T3 CJ c o c u o a o o O T3 C J -a C L O to 6 0 CJ •3 <4-l o 13 u. o a C J • J a, "ol) CJ > rt cj 4 3 co c G J - 3 OO u, rt C J i> s 2 =• cu „ cj rt O 55 00 rt •o CJ rt & Cu cj rt •a C J a to rt 6 o a e •c o ro ON ro ON ON —« ON 2 CO E 2 CN ON ON ON ON O ro •"E < o oo ( N >, 3 £ o CO s o NO c o o E a O -H o ro ro NO ON ON — — ' — ' CN +1 .+1 +1 CO NO CO oo oo ON ^ (N o i cn O O -H o c „ rt o o +1 o -H o -H CN CN ro ro co O cn co NO o o o o -H CN CN <N CN o o o o -H rt cn cn ro O r-- cn — o o o o o + + + ON >/0 CO r t o — — o o o o cn CN O CO s ° o ° r- _ . 00 ^ rt 2 § | O o ° 5 22 §8. o ° T3 CJ rt g o rt cj J3 C n O 8 <+-! O E eg T3 cj T3 O u. CJ "rt 'C o •*-» rt E rt cj £ T3 rt e CO J O O s rt £ co CJ CJ rt o O H fc 89 Plate 4.1. Showing Landslide #7 undercut by the May 1993 flood. 90 undercutting. The field survey after the flood found that the active area (AJ of the toe had increased to 40x2 m 2. The physical and chemical parameters of the toe material are presented in Fig. 4.9B and Table 4.10. Inputs of Sediment, Magnetite and Gold Input rates of sediment, magnetite and gold are presented in Table 4.11. Although the monitoring program spanned three years, major inputs only occurred in May 1993, when large amounts of sediment and magnetite, and about 1 gram of -0.149+0.053 mm gold and 0.3 grams of -0.053 mm gold were added to Harris Creek. 4.4.3. Landslide #17 Basic Properties Landslide #17 is located on the north bank of Harris Creek, 1.2 km upstream from McAuley Creek (Fig. 2.7). It consists of till with locally interlayered glaciofluvial sediment. It is estimated that the volume of material accumulated in the scar is 1.9x10^  m3- Textural characteristics of toe material are summarized in Fig. 4.9C. The erosion pins driven into the toe area indicated that Landslide #17 was stable from 1991 to 1993 and even during the May 1993 flood, it did not supply measurable quantities of sediment, magnetite and gold to Harris Creek. 4.4.4 Landslide #28 Basic properties Landslide #28 is located on the north bank of Harris Creek, 11.4 km upstream of the Research Site (Fig. 2.7). This landslide consists of glaciofluvial sediment and till. The physical and chemical properties of its toe are shown in Table 4.12 and Fig. 4.9D. The volume of material 91 NO CN r o NO £ E I T ) o o + o r o cn oo o i n O o + ON ON CN r t c-N^  oo B e o o o CU o c oo rt oo o cn o cn rf c o •8 rt <u N 00 r o O Cu Cu O O r t > NO £ Q oo E o c 2 CO c 1) T3 ca cn ON r o ON ON rt ON rt rt-tu CN ON r o ON ON —' ON oo" o r o C u O U i t U ON cN ON ON — I O N oo" CN r-rt O t U >> rt T3 o r t </"> +1 t-NO NO rt u o NO +1 CN, rt o oo CN -H O — i CO CJ rt U H C O CO O t— PJ. 15 CJ | S rt 92 Table 4.11. Inputs of sediment, magnetite and gold from Landslide #7 Measurement interval Total From July 28, 1991 From M a y 18, 1992 From M a y 1, 1993 to M a y 17, 1992 to A p r i l 30, 1993 to M a y 31, 1993 3.4±0.63 Sediment (t) 3±0.4 1952.0±158 -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm 0.76+0.14 0.95+0.18 0.97+0.18 Magnetite (kg) 0.68+0.091 0.84±0.11 0.74+0.099 405 .4±32 .82 504.9+40.88 518.9+42.00 -0.149+0.053 mm 0.0015 -0.053 mm 0.0005 Gold (g) 0.0013 0.0004 0.8±0.065 0.27+0.022 93 cn o E E I r o c o o o + c o «~-o | < N | II 3 1 c ,o o I* cj N Iv5 c E CO o o Ox V o c o _ ' c o _ N r o CO 2 o c-CJ 0-rt JO a. CL 22 o O > 2 E CJ — ' ^ o Q oo E o 2 c r-o T J 3 CO > N cj 3 0 0 r o O N r o O N O N O N >, rt T3 - ~ ^  ro >. rt . -H *S CN CN O N r o O N O N —i O N oo" — 1 O rt 2 E o E o C H Cu-rt cj > N ••5 ~g Cc, O < O o O N C N O N O N oo . •CN r- g ^ t O N > N rt o +1 oo o cu oo _o rt H CL) (L) T3 o •c CJ a, rt •5 0 0 C •c 3 X ) <D > t> rt O c co rt 94 accumulated in the scar was estimated to be 6.5xl04 m ,^ and it was supplied to Harris Creek by fluvial undercutting and mechanical collapse of the active toe measuring 9x1.8 m 2. Inputs of Sediment, Magnetite and Gold Inputs of sediment, magnetite and gold are listed in Table 4.13. Landslide #28 contains a low concentration of gold. In the measurement periods from 1991 to 1993, May 1993 is the major delivery time for sediment and magnetite. Because of its low gold content, this landslide delivers very little gold to Harris Creek. 4.5 Inputs of Sediment, Magnetite and Gold from Logjam #1 Basic Properties Logjam #1, located 1.9 km upstream of the Research Site (Fig 2.8), has an approximately trapezoidal shape with a longitudinal length of about 60 meters. The volume of sediments stored behind the organic debris is estimated as 5x10^  m .^ Textural characteristics of the sediments (Fig. 4.1 OA) indicate that this log jam is a body of gravelly-pebble sediments locally interlayered with coarse sands. Its physical and chemical properties are presented in Table 4.14. Compared to the landslides in the previous section, the log jam sediments contain high values of gold and magnetite. The variations of gold and magnetite with increasing depth are irregular with no obvious trend down the profile (Figs. 4.11 and 4.12). The logjam supplies sediments to Harris Creek by fluvial undercutting and mechanical collapse of overhanging banks. The active section was 20x1.2 m 2 before April 30, 1993, but increased to 35x1.2 m 2 during the flood in May 1993. Inputs of Sediment, Magnetite and Gold Inputs of sediment, magnetite and gold from Logjam #1 are estimated and listed in Table 4.15. Although Logjam #1 contains higher concentrations of gold, input is quite low, even in the major supply period from May 1 to May 31 in 1993. 95 Table 4.13. Inputs of sediment, magnetite and gold from Landslide #28 Survey interval From July 28, 1991 to May 17, 1992 From May 18, 1992 to April 30, 1993 From May 1, 1993 to May 31, 1993 Total 1.14±0.12 Sediment (t) 0.0 39.4±6.6 -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm 0.4±0.045 0.43+0.045 0.43+0.045 Magnetite (kg) 0 0 0 13+2.17 14+2.33 13.9+2.32 -0.149+0.053 mm -0.053 mm 0.0001 Gold (g) 0 0 0.0035 / Note: / means that gold is below detection limit; 0 means that erosion rate is equal to 0. o - o o o o o o o O 00 NO rj" CN J3UIJ 1U30J3J o o o o o o o oo N O rr C N J 3 U I J 96 o o o o o o S o oo N O CN o J3UIJ }U90J9<i *3 0> o o o o o o o oo N O C N J 3 U I J J U 3 0 J 3 J 97 Table 4.14. Summaries of physical and chemical parameters of Log jam ff 1 Size fraction (n = 4) -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm -0.053 mm Percentages (F) 0.19 0.11 0.11 0.56 Magnetite (%) 6_30 3_4 1_9 -0.149+0.053 mm Gold (ppb) 79.1 (46.1-97) -0.053 mm 10.7(4.7-17.5) n Mean(g/cm3) SD (g/cm3) CV (%) Bulk density (D) 10 1.51 0.077 5.12 Survey Interval • From July 28, 1991 From May 18, 1992 From May 1, 1993 to May 17, 1992 to April 30, 1993 to May 31, 1993 Erosion rates (R) 37+1.4 cm/year 6.8+0.9 cm/year 6.2+1.2 cm/day 2 4 6 Concentration (%) ED -0.149+0.105 nun H -0.105+0.075 mm • -0.075+0.053 mm Fig. 4.11. Variation of magnetite concentrations in the three size fractions in a vertical profile of Log jam # 1 20 40 60 80 Concentration (ppb) • -0.149+0.053 mm • -0.053 mm Fig. 4.12. Variation of gold concentrations in the two size fractions in a vertical profile on Logjam #1 99 Table 4.15. Inputs of sediment, magnetite and gold from Log jam U1 Survey Interval From July 28, 1991 From M a y 18, 1992 From M a y 1, 1993 to May 17, 1992 to Apr i l 30, 1993 to M a y 31, 1993 Sediment (t) Total 10.86+0.41 2.14+0.28 121.9+23.6 Magnetite (kg) -0.149+0.105 mm 2.2+0.1 0.5+0! 23.5+4.55 -0.105+0.075 mm 0.7+0.03 0.1+0.01 7.4+1.43 -0.075+0.053 mm 0.4+0.02 0.07+0.01 4.1+0.79 Gold (g) -0.149+0.053 mm 0.0062 0.0013 0.064+0.012 -0.053 mm 0.001 0.0002 0.011±0.002 100 4.6 Inputs of Sediment, Magnetite and Gold from Unstable Banks Introduction Bank erosion, one of the important sources of sediments to Harris Creek, is confined mainly to the flood plain along lower Harris Creek (Fig. 4.13). Terrain mapping (Ryder and Fletcher, 1991) and the field surveys show that the unstable reaches are about 5! km in length and account for about 40% of the total length of lower Harris Creek. Based on the longitudinal survey, the unstable banks are divided into three sections. Their basic properties will be explained using the results obtained from monitoring stations which were established on the corresponding sections. 4.6! Section I Basic Properties The laterally unstable reaches in Section I are located between the Research Site and Landslide #1 and are about 2.5 km long (Fig. 4.13). The banks in this section expose two layers of overbank deposits. The upper layer is about 0.3 m thick and dark brown. The lower layer is about 0.5 m thick and light brown. Physical and chemical properties obtained from Station 1 are presented in Fig 4. 10B and Table 4.16. The two layers are similar in that both of them consist mainly of -2 mm material. However, the former contains 55% of the -0.053 mm sediment, whereas the latter contains only about 15% of the -0.053 mm material. The active area (Aa) of the laterally unstable bank in this section is 2500x0.3 m 2 for the upper layer, and 2500x0.5 m 2 for the lower layer. Inputs of Sediment, Magnetite and Gold The mass inputs of sediment from the laterally unstable bank in Section I are calculated with the Equation (4.2), and the inputs of magnetite and gold are calculated using the Equation (4.3). Results are shown in Table 4.17. Inputs vary from year to year, but sediment and magnetite were 102 li cl 3 N 00 r t vq i n r t r t r o CN r t ON O m o y o 00 >^  j_v> £3 — — C i _ <-O o o " s- -« a. o a. D -J y i - >-2 D J O N N O O N O r o O , rt 2 8. £ O QH O o D J O N CJ s O N O "ob °. ^ o a c>0 6 -1 C NO Q to a T3 3 OQ rt rt - ~g r o ON r o ON ON ON I—I — 1 ay r o ay E S Fro o •»-> CN ON r o ON O N O N oo •—< o >> rt •s pri s < p o UH U H ON CN O N O N rt O N 1—* oo CN _>> *3 • — 1 i — , s s o o U i tu T3 r t CN O -H N O rt t o o rt <o >, 1 o r t r o +1 NO CO ts U i c o to O u w C O DH X 1). , <u IS U i i i2 oo r t CD I H <D I oo 103 Table 4.17. Inputs of sediment, magnetite and gold from Section 1 Survey interval From July 28, 1991 From M a y 18, 1992 From M a y 1, 1993 t o M a y l 7 , 1992 to A p r i l 30, 1993 to M a y 31, 1993 Sediment (t) Total 293+52.4 0 1120.96+168 Magnetite (kg) -0.149+0.105 mm 445.3±94.6 0 1609.7±241.5 -0.105+0.075 mm 408.8+86.9 0 1483.3+222.5 -0.075+0.053 mm 181.7+38.6 0 663.6+99.5 Go ld (g) -0.149+0.053 mm 0.14+0.03 0 0.52+0.08 -0.053 mm 0.12+0.03 0 0.86+0.13 104 introduced to Harris Creek mainly in May 1993. In the same period, the unstable bank in Section I supplied approximately 0.5 grams of -0.149+0.053 mm gold and 1 gram of -0.053 mm gold to Harris Creek. 4.6.2 Section II Basic Properties Laterally unstable reaches in Section II are about 1.1 km long between Landslide #1 and McAuley Creek (Fig. 4.13). In this section, banks have a two-layer structure: the upper layer is a fine, brown overbank deposit; the lower layer is grey sandy gravels (Fig. 4.10C). Physical and chemical properties obtained from Station 2 are presented in Table 4.18. The active areas (A,) in Section II are: 1100x0.7 m 2 for the overbank sediment, and 1100x0.5 m 2 for the sandy gravels. Inputs of Sediment, Magnetite and Gold Results in Table 4.19 show that the greatest inputs of sediment, magnetite and gold took place from May 1 to May 31 in 1993. 4.6.3 Section III Basic Properties Section III, located between Station 3 and Landslide #28 (Fig. 4.13), consists of 1.5 km of laterally unstable reaches. Within this section, the floodplain is not well developed and banks consist largely of sandy gravels but locally include glacial till and colluvial slopes (Fig. 4.10D). Physical and chemical properties from Station 3 are shown in Table 4.20. Table 4.18. Summaries of physical and chemical parameters of Station 2 Size fraction (n = 1) -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm -0.053 mm Percentage* Upper layer (F) 12.2 7.2 4.7 11.2 Lower layer (F) 1.5 0.9 0.7 2.7 Magnetite (%) Upper layer 3.6 5 5.5 Lower layer 3_40 4_1 3_5 -0.149+0.053 mm -0.053 mm Gold (ppb) Upper layer 0.62 Lower layer 13.10 ' Bulk density (D) n Mean (g/cm3) Upper layer 5 1.07 Lower layer* 10 1.51 Survey interval From July 28, 1991 From May 18, 1992 From May 1, 1993 to May 17, 1992 to April 30, 1993 to May 31, 1993 Erosion rates (R) 9±1.1 cm/year 0 cm/year 1.5±0.34 cm/day SD (g/cm3) C V (%) 0.1 9.47 0.077 5.12 Table 4.19. Inputs of sediment, magnetite and gold from Section 2 Survey interval From July 28, 1991 From M a y 18, 1992 From M a y 1, 1993 to May 17, 1992 to A p r i l 30, 1993 to M a y 31, 1993 Total 120.6±14.7 -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm 274.5+33.6 222.2+27.2 158.4+19.4 -0.149+0.053 mm -0.053 mm 0.033 0.031 Sediment (t) 0 769.3+174.4 Magnetite (kg) 0 0 0 1653.6+374.8 1338.8+303.5 954.5+216.4 Gold (g) 0 0 0.21+0.05 0.19+0.04 Table 4.20. Summaries of physical and chemical parameters from Station 3 Size fraction -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm -0.053 mm Percentages (F)* 3 2.7 2.6 19 Magnetite (%) 3.50 4.6 4 -0.149+0.053 mm -0.053 mm Gold (ppb) 5.30 0.9 n Mean (g/cm 3) S D (g/cm 3) C V (%) Bulk density (D) 3 1.06 0.093 8.74 Survey interval From July 28, 1991 From M a y 18, 1992 From M a y 1, 1993 to M a y 17, 1992 to A p r i l 30, 1993 to M a y 31, 1993 Erosion rates (R) 4.8+0.7 cm/year 0 cm/year 1.3+0.2 cm/day * See Table 4.8 for the explanation 108 Inputs of Sediment, Magnetite and Gold Calculated inputs of sediment, magnetite and gold are shown in Table 4.21. The important input period is in May 1993. 4.7 Summary 4.7.1 Concentration of magnetite and gold Magnetite As stated in Chapter 2, sediment sources in the Harris Creek basin are classified into primary sources and secondary sources. Concentrations of magnetite in the two types of sources are very different (Fig 4.14). In the primary sediments magnetite is rather uniform and averaged about 0.7% in each of the three size fractions. In contrast, concentrations of magnetite in the secondary sources are higher and more variable, concentrations averaged about 3.1% in the -0.149+0.105 mm size fraction, 2.5% in the -0.105+0.075 mm size fraction and 1.9% in the -0.075+0.053 mm size fraction. Gold Although gold concentrations are much more variable than magnetite, the difference between that in the primary sources and in secondary sources is even more obvious (Fig. 4.15). In the primary sources, the concentrations of gold are low and averaged 2.2 ppb in -0.149+0.053 mm, and 0.7 ppb in -0.053 mm fraction. Gold concentrations in the secondary sources are much higher. For example, in Mosquito Creek gold concentration in stream sediment is 477 ppb in the -0.149+0.0523 mm size fraction, and 42 ppb in the -0.053 mm size fraction (Fig. 4.15). On average, Au in the secondary source sediments is about 50 times higher in the -0.149+0.053 mm size fraction and about 27 times higher in the -0.053 mm size fraction than that in the corresponding size fraction of the primary sources. 109 Table 4.21. Inputs of sediment, magnetite and gold from Section 3 Survey interval From July 28, 1991 From M a y 18, 1992 From M a y 1, 1993 to May 17, 1992 to A p r i l 30, 1993 to M a y 31, 1993 Sediment (t) Total 86.5+12.6 0 897.1+138 Magnetite (kg) -0.149+0.105 mm 23.9+3.48 0 234.4+36.1 -0.105+0.075 mm 19.2+2.80 0 187.5+28.5 -0.075+0.053 mm 10.4+1.52 0 101.6+15.63 Gold (g) -0.149+0.053 mm 0 0 0.2+0.03 -0.053 mm 0 0 0.07+0.01 110 7 6 1 5 ca c 4 8 c Prirnary source -.itimgll *f *f S? ^ Sediment source -0.105+0.075 mm • -0.075+0.053 mm M -0.149+0.105 mm Fig. 4.14. Magnetite concentrations of different sources I l l 600 Primary source 500 400 CL. a c o ra l l 300 8 c | 200 S 100 0 <r <r <r y rfb t l Sediment source H -0.149+0.053 mm • -0.053 mm Secondary source # J Fig. 4.15. Gold concentrations in different sources 112 4.7.2 Mass Inputs Sediment Fig. 4.16 shows temporal and spatial variations of sediment inputs in the Harris Creek basin. In the first monitoring period (July 28, 1991 to April 17, 1992), relative contributions of different sources are as follows: Section I (56.7%) > Section II (23.4%) > Section III (16.6%) > Log Jam #1 (2.1%) > Landslide #7 (0.7%) > Landslide #1 (0.3%) > Landslide #28 (0.2%). It is immediately clear that sediment contributed from the secondary sources accounts for 98.8%, of which 96.7% is contributed from bank erosion of the laterally unstable banks along lower Harris Creek. In the second period (April 17, 1992 to April 30, 1993), total input was as small as 6.2 tons, of which 17.6% is from Landslide #1, 48.1% is from Landslide #7, and 34.3% is from Log jam #1. In this period, the banks did not make contributions. In the third period (May 1993), total input amount is very large (Fig. 4.16), of which the relative contribution from the different sources is: Landslide #7 (36.6%)> Section I (21%) > Section III (16.8%) > Section II (14.4) > Mosquito Creek (6.1%) > Log Jam #1 (2.3%) > Landslide #1 (1.3) > McAuley Creek (0.8) > Landslide #28 (0.7%). Based on sediment source classification, 38.6% is from the primary sources; and 61.4% is contributed from the secondary sources, of which 52.2% is from erosion of the laterally unstable banks. Magnetite Temporal and spatial variations of inputs Of magnetite from different sources are summarized in Fig. 4.17. It can be seen from the difference of scales on Y-axes in Fig 4.17 that, almost all the magnetite is supplied from bank erosion along the laterally unstable reaches of ro ON ON O ro Cu < O 4—1 C N ON ON Cu < CQ saojnos Xjrpuoo?s saojnos /CJEUIUJ wo ro (i) jndui 4> 7 ro wo C N wo — uo O C N — ' O •V<j>, C J o o c C J E -3 o CO C J > 3 Cu C C C 01 4 ) 1) <U 2 CO <4-< ' C o g o ^ • r t OJ ea 4 3 > 13 CL, O 13 o 13 c '•fS O CO •5 c c o "S S3 > 2 OJ aj 4 3 OJ o 4J O G ' H cfa NO 0 0 C N ON ON Q. < O ON ON O O > 3 Full scale for B o wo ro saojnos XjepuoD?s saojnos /Lrciuuj 0 (i) sjnduj 4ti '3/ O O O O O O O O WO O WO O uo rO C N C N —' — •AO/ <L> O o co -*-» C C J E '•5 C J ro ON ON o sr1 E o Full scale for A ssojnos Areuiuj tt tt tt -fe, °7 3 A O O O O O O O O O O O O O O O wo o^ wo^  o^ o_^  in ro" ro" C N " C N " —<" - H " (l) smduj From July 28, 1991 to April 17, 1992 2,000 1,000 From May 1, 1993 to May 31, 1993 800 600 h 00 400 200 0 3. Secondary source CO a 8 C/l Ok 2,000 H 1,500 1,000 . ••c 500 ^ A N T 4" / Fig. 4.17. Temporal and spatial variations of magnetite inputs from different sources. (The first column from left is for -0.149+0.105 mm, the middle one is for -0.105+0.075 mm, and the one at right is for -0.075+0.053 mmsize fractioa Note the difference between primary and secondary sources. Note differences in scale between primary and secondary sources) 115 lower Harris Creek in 1991. In 1992 the total input is very low with 79.7% from Landslide #7 and 20.3% is from Logjam #1. Bank erosion did not make any contributions in this year. In the May 1993 flood, total supply of magnetite in the -0.149+0.053 mm size fraction is 9787.3 kg, of which 15.3% is from the primary sources and 84.7% from the secondary sources. Landslide #7 is the major contributor of the former, and bank erosion is the major contributor (84.1%) of the latter. Gold Variations of gold inputs from the different sources and in different monitoring periods are shown in Fig. 4.18. During the first period, the major contributor of gold is bank (Fig. 4.13). Log jam #1 and Landslides #7 and #1 supply a small amount of gold to Harris Creek. During the second period, banks along Harris Creek were stable. Landslide #1 becomes a main contributor for gold in the -0.053 mm fraction. Landslide #7 and Logjam #1 become major donors for gold in the -0.149+0.053 mm fraction. But the input amount was two orders of magnitude lower than in the first period (Fig. 4.18). The important input of gold into Harris Creek took place largely in May 1993: the major supplier of the -0.149+0.053 mm gold was Landslide #7, followed by the unstable bank - Section I and Section II (Fig. 4.18). The major supplier for -0.053 mm gold was Section I, followed by Mosquito Creek and Landslide #7. The relation between input and flow condition will be discussed in Chapter 6 in combination with output results. 0.15 0.1 From July 28, 1991 lo April 17, 1992 O 0.05 0.0015 0.001 0.0005 From April 18, 1992 to April 30, 1993 From May 1, 1993 to May 31, 1993 y y y ^ Fig. 4.18. Temporal and spatial variations of gold inputs from the different sources to Harris Creek (The filled column is for -0.149+0.053 mm, and the open one is for -0.053 mm size fraction Note: different scale on Y-axis). 117 C H A P T E R 5 O U T P U T S O F S E D I M E N T , M A G N E T I T E A N D G O L D F R O M H A R R I S C R E E K 5.1 Hydraulic Conditions Using the methods presented in Chapter 3, flow conditions were measured and monitored near the Research Site in 1991, 1992 and 1993. Hourly discharge, prepared by digitizing with the GIS-ROOT program and interpreting with WATERFOR and staff gauge ratings, is summarized in Fig. 5.1. In 1991, several high-discharge events occurred in May and June with a flood in mid May. This flood, which resulted from snowmelt, had an instantaneous discharge of 20 rn^/s, and discharges greater than 10 m-Vs lasted for about seven days. A few small peaks occurred in June and July, but discharges greater than 10 mVs were relatively brief. From July through September flow was low. In 1992, the snowmelt flood with an instantaneous discharge of about 15 m-Vs was delayed until late May and early June. Discharges higher than 10 m /^s lasted only for two days. After this relatively brief flood, low-flow conditions were recorded through the monitoring period (Fig. 5.1). Compared to 1991 and 1992 flows, the 1993 spring flood, which occurred on May 11 and lasted for about fifteen days (Fig. 5.1), was unusually large with an instantaneous discharge up to 45 m-Vs. After the freshet, flow quickly declined to its low stage of 3 to 4.5 m-Vs. There were several minor floods: one of them reached to 15 nvVs in July, which was higher than in 1992. 118 Date-1993 g. 5.1. Stream hydrograph for Harris Creek (Shaded area is Helley-Smith sampling period) Note: Check of WSC station (08LC005) 6 km downstream from the study site indicates that no high flow happened in July, 1992 119 It is clear that the hydrological regime of Harris Creek is dominated by the snowmelt flood that usually occurs in late spring or early summer. This matches the regional pattern of the hydrological regime in south central British Columbia, Canada (Environment Canada, 1987). The intensity and duration of the snowmelt flood varies from year to year, depending on the water equivalent depth of snow accumulation (Fig. 2.6) in winter and the change of temperature in the following spring. The remainder of the year is characterized by long periods of low flow, occasionally interrupted by simple, discrete flood events from June to October, and a freezing period from November to March. 5.2. Critical Discharge Harris Creek is a gravel-bed stream with a bimodal sediment population: a coarse population, corresponding to the gravel framework, and a fine population corresponding to the sandy matrix (Day and Fletcher, 1991). The critical discharge is defined here as the discharge under which the bimodal bed sediments start to be mobilized and can be recognized by the appearance of a bimodal size distribution in the Helley-Smith samples. However, it is not suggested that this approach is precise because the stream may erode and carry bimodal materials from local sediment sources, such as stream banks or logjam storage sites. The cumulative probability plot technique (Sinclair, 1976) is used to identify the occurrence of multiple populations and partition them. At discharge below 10 m /^s, sediment collected by the Helley-Smith sampler has a single-population with D50 of about 0.6 mm (Fig. 5.2). When discharge rises over 10 m-Vs, the population corresponding to the framework appears and increases with increasing discharge (Figs. 5.3). At a discharge of 26 m?/s the Helley-Smith sample shows a strongly bimodal grain size distribution with D50 of about 14.4 mm (Fig. 5.4). Fig. 5.5 shows the relation between proportions of the framework populations and flow conditions for all bedload samples. It is inferred that the critical discharge is close to 10 nvVs where the coarse F i g . 5.2. Showing population distribution in one bedload sample at discharge o f 5.3 m 3 /s on Harris Creek ( D 5 0 = 0.60 mm) 121 Fig. 5.4. Showing population distribution in one bedload sample at discharge of 26.1 m3/s on Harris Creek (A: 68%, D 5 0 = 24.3 mm; B: 32%, D 5 0 = 0.62 mm) 98 95 90 80 70 60 50 40 30 20 10 5 2 1 . i 1" I , 1 ;" - i4H-4444-Li—J-- T - l - -„ U . J f ILL!- L, r__| L — 44. - + 4 H + 4 - r J ~n—rrp-Yr-r r - i M i l I i i i 1111II 4 ... - i :SE^^ES: | : := :=^; ..... , . J 4 j . f 4j.-j ._ i ^! , 2__J 3 1 r - - - j - H H 4 3 | | : : : : : - : ^ : : : : ; .1 ... _ 1 ' I ! L 1 . i t i. -1 fillillllllMM 1 1 1 Sl?;|:|EEEEEtEE::.E:;„ r _1 ~i zr „ ™ * .+-4 4 - - 4 - . -J - _ H J 4 _ A L , _ ' , J M i - M 4 H ' T f + r - f c -j. r** • ~zi : p ^ - - ; m ^ J g : ^ : ± + : : • r ~._U.J 1 , • ; J - LU . iii. 1 . •H-4- -44ft-144 F^-h h - -! i ffi-4±J^::i::::::::::::i - ——:—r— 3U+j-u±y^ |_u y.-A I I B - i / — , — / t—r-1— 3c±:::::||~~::::::::::::: t " . " J — -----r---^ EpE^ : r -~r i . i _ l . HriiTr4t+ j - H i — - i t-t^-M- — ' 1 1 ' ' I ' • I • 1 i- I I I I ! I I I I I I I I I ! I [ 1 I I I I I 11 I I [ I I I I I I I I | | I I I I I I 2 5 10 20 30 40 50 60 70 80 90 95 98 9! Cumulative % coarser 123 Discharge (mVs) Fig. 5.5. The relation between the proportion of the framework population and discharge for the composited Helley-Smith samples on Harris Creek. 124 population starts to appear in the Helley-Smith sampler indicating coarse gravels being mobilized and transported. This inference is in good agreement with the suggestion of Fletcher and Day (1989), and Fletcher and Wolcott (1991) that the critical discharge is at about 10 m-Vs. 5.3 Sediment Output 5.3.1 Transport rate The Helley-Smith sampler was used to collect bedload sediment using the operational procedures described in Chapter 3. Numbers of samples, sampling time, average discharge, and river width during each sampling period are listed in Table 5.1. A total of 39 samples were collected from Harris Creek in 1993. Due to their small size, however, some had to be pooled into composite samples to provide enough material for gold analysis (Table 5.1). Fewer samples were collected on the rising limb because the diurnal rise occurred overnight. Weights of the Helley-Smith samples are listed in Appendix Table 8. The transport rate of sediment is calculated using formula (4.4). The transport rate (Fig. 5.6) against discharge increases with increasing discharge. When examined in detail, however, the sediment transport rate exhibits both considerable complexity and hysteresis with rates of transport for a given discharge appearing greater on falling limb than on the rising limb (also cf. Klingem and Emmett, 1987). Apparently, it is necessary to develop bedload rating curves to predict the transport for flow conditions on the rising and falling limbs. The regression equations for both the rising limb and falling limb are shown in Table 5.2. An F-test indicates that the two equations are not only statistically significant at a = 0.05 but are also significantly different from each other at a = 0.05. 125 Table 5.1. Sampling time and discharge on H a m s Creek Station in 1993 (to be continued on next page) Sample No . Date Sampling Discharge River width Flood time (h) (rnVs) (m) limb 93-HZ-18 7-May 3.05 5.32 13.60 R 93-HZ-26 11-May 4.07 8.61 14.00 R 93-HZ-27 11-May 4.03 9.50 14.00 R 93-HZ-35 13-May 1.83 24.45 14.30 R 93-HZ-36 13-May 2.17 22.86 14.30 R 93-HZ-37 13-May 1.17 22.75 14.30 R 93-HZ-38 13-May 2.33 23.54 14.30 R 93-HZ-39 13-May 2.00 24.54 14.30 R 93-HZ-40 13-May 0.88 26.06 14.30 R 93-HZ-48 16-May 1.25 23.72 14.50 R 93-HZ-54 17-May 0.87 17.69 14.00 F 93-HZ-55 17-May 0.97 18.53 14.00 F 93-HZ-56 17-May 1.08 20.00 13.90 F 93-HZ-57 18-May 1.00 16.46 13.90 F 93-HZ-58 18-May 0.75 15.94 14.00 F 93-HZ-59 18-May 1.08 15.67 14.00 F 93-HZ-60 18-May 1.00 15.40 14.00 F 93-HZ-61 18-May 0.83 15.63 14.00 F 93-HZ-62 18-May 0.78 15.95 14.00 F 93-HZ-63 20-May 1.42 17.40 14.50 F 93-HZ-64 20-May 1.33 17.80 14.50 F 93-HZ-65 21-May 1.25 17.43 14.80 F 93-HZ-66 21-May 1.22 17.18 14.70 F 93-HZ-67 21-May 1.67 16.92 14.50 F 93-HZ-68 21-May 1.33 16.63 14.60 F 93-HZ-69 22-May 1.42 14.49 14.00 F 93-HZ-70 22-May 2.50 13.93 14.00 F 93-HZ-71 23-May 5.92 10.97 14.00 F 93-H772 23-Mav 5.16 10.97 14.00 F Remarks W e i 8 h t (kg) 0.92 8.64 23.32 15.41 21.70 22.75 26.72 19.49 14.66 8.39 25.77 23.66 20.58 34.06 30.52 24.00 14.45 | 28.80 15.38 5.74 Tabic 5.1. Sampling time and discharge on Harris Creek Station in 1993 (continued) Sample No. Date Time Discharge River width Flood (hours) (nvVs) (m) limb 93-HZ-73 24-May 7.75 9.72 13.30 F 93-HZ-74 24-May 7.58 9.72 13.50 F 93-HZ-75 25-May 6.67 8.69 13.50 F 93-HZ-76 25-May 6.63 8.69 13.30 F 93-HZ-77 26-May 6.65 8.78 13.30 F 93-HZ-78 26-May 6.67 8.78 13.60 F 93-HZ-79 27-May 6.75 8.19 13.20 F 93-HZ-80 27-May 6.70 8.19 13.20 F 93-HZ-81 28-May 6.75 8.18 13.20 F 93-HZ-82 28-Mav 6.75 8.18 13.20 F (kg) 10.81 6.06 14.80 12.08 10.86 *R indicates that the sample was on the rising limb; and F indicates that the sample was on the falling limb. Samples with the same pattern were composited. 127 Fig. 5.6. The relationship between transport rate of stream sediment and discharge on Harris Creek in 1993 128 Table 5.2. Regression equations for rates of sediment and magnetite of bedload samples versus discharge on Harris Creek in 1993. Size fraction (mm) a b n rj F-ratio AR 1. Sediment 2. Magnetite IO 3' 1 5 4.1 24 0.86 138 0.72-1.38 (3.4-4.8)* -0.149+0.105 10"3-81 3.3 24 0.88 165.5 0.79-1.26 (2.9-3.7) -0.105+0.074 IO"4 1 3 3.3 24 0.88 154.6 0.83-1.20 (3.0-3.6) -0.074+0.053 IO"4 1 5 3.2 24 0.83 109.8 0.87-1.15 (2.6-3.6) F(l,6)o.05 = 5.99; F(l,14)0.o5 =4.60; F(l,22)0.o5 = 4.30 AR is 2-standard error ranges for transport rate equations with g/sxm for sediment and mg/sxm for magnetite; * the numbers in the parentheses are the range of the slope coefficient at a=0.05. 129 5.3.2 Output Sediment output from Harris Creek is calculated using Formula (4.1). Calculation results are shown in Table 5.3. The three years' studies show that major output of the sediment is controlled by the snowmelt flood. The higher the peak flood, and the longer the discharge above 10 m3/s lasts, the more sediment will be transported out. For example, the May 1993 flood which was 2 times higher than 1991 and 1992 and lasted two times longer than the 1991 flood and 8 times longer than the 1992 flood,, transported 4938.2 tons of stream sediment - 10 times more than 1991 and 100 times more than 1992. In 1993, flow with discharge below 5.0 m3/s lasted 1538 hours (62.2%), but delivered only 0.2% of the sediment. Flow with discharge between 5.0 and 10.0 m3/s lasted 582 hours (23.5%), and delivered 1% of the total. Flood with discharge equal to and above 10.0 m3/s lasted only 353 hours (14.3%), but transported 98.8% of the sediment, of which 91% was transported by the flood between May 11 and May 26 in 1993. 5.4 Magnetite Output 5.4.1 Transport Rate Weights of magnetite in the -100+140, -140+200 and -200+270 fractions of bedload samples are listed in Appendix Table 4. Concentrations of magnetite in the three size fractions are shown in Fig. 5.7. The concentration-discharge data suggest that concentrations of magnetite in the three fractions increase slightly with increasing discharge and seem to have three peaks within the measured flow conditions: one at discharge of 10.0 m3/s, one at 17.0 m3/s and one at 23.0 m3/s. Transport rates of magnetite in the -0.149+0.105 mm, -0.105+0.075 mm and -0.075+0.053 mm size fractions against discharge are shown in Figs. 5.8, 5.9 and 5.10, respectively. The rates of transport depend on stream discharge, but exhibit both considerable complexity and to some extent hysteresis with the rate of transport for a given discharge being greater on the rising limb than that on the falling limb, i.e., the reverse of the hysteresis of sediment transport (Fig. 5.6). 130 Table 5.3. Outputs of stream sediment from the Harris Creek basin Duration time Output Year (hours) (tons) Percentage 1991 The entire monitoring period 3380 431 (281-660) Flow below 5.0 m 3/s 2312 2.6 0.6 Flow between 5 and 10.0 nrVs 836 96.0 22.3 Flow higher than 10.0 m7s 232 332.4 77.1 1992 The entire monitoring period • 1723 41.2 (27-63) Flow below 5.0 nrVs 1614 4.4 10.6 Flow between 5 and 10.0 nrVs 77 8.2 19.8 Flow higher than 10.0 m 3/s 32 28.7 69.6 1993 The entire monitoring period 2473 4938.2 (3225.0-7561.3) Flow below 5.0 m 5/s 1538 3.9 0.1 Flow between 5 and 10.0 nrVs 582 55.4 1.1 Flow higher than 10.0 m 3/s 353 4878.9 98.8 Note numbers in parentheses are ranges of the output. o ro i n r o <N (o/0) 3 J I J 9 U § B I U JO U 0 I J C J 1 U 3 0 U 0 3 10 15 20 25 30 Discharge (mVs) Rising limb Falling limb Fig. 5.8. Transport rates of-0.149+0.105 mm magnetite versus discharge Shaded bar is the approximate location of the critical discharge Arrows indicate directions of rising and falling discharge 1 3 3 100 F 30 Discharge (m3/s) Rising limb Falling limb Fig. 5.9. Transport rafos of-0.105+0.075 mm magnetite versus discharge Shaded bar is the approximate location of the critical discharge Arrows indicate directions of rising and falling discharge 134 100 F a Discharge (mVs) Rising limb Falling limb Fig. 5.10. Transport rates of-0.075+0.053 mm magnetite versus discharge Shaded bar is the approximate location of the critical discharge Arrows indicate directions of rising and falling discharge 135 However, the hysteresis at discharge below 15.0 m3/s needs confirmation because it is largely based on one point at a discharge of 9.1 m3/s on the rising limb. Therefore, the data are analyzed without separation for the rising and falling limbs. The statistical results of the regression analysis are shown in Table 5.2. The transport rates vary with discharge in the form of R = aQD. F-test statistics shows that all three equations are highly significant at a = 0.05, but the slope coefficients for the three equations (Table 5.2) are not significantly different at the same confidence level of a = 0.05; that is, transport rates of magnetite in the three size fractions respond similarly to changing discharge. 5.4.2. Output Output is calculated using Formula (4.1). Calculation results are shown in Table 5.4. There is a similarity between the magnetite output and sediment output with the May 1993 flood being most important in terms of transport amount. This flood transported about 5,000 kg of the -0.149+0.053 mm magnetite. Flow with discharge below 5.0 m3/s accounted for 62.2% of the all monitoring time, but delivered only 0.5% of the magnetite. Flow with discharge between 5.0 and 10.0 m3/s lasted 582 hours (23.5%), and delivered 3% of the magnetite. Flow with discharge equal to and above 10.0 m3/s lasted only 353 hours (14.3%), but transported 4518.28 kg (96%) of the magnetite out of the Harris Creek basin. 5.5. Output of Gold 5.5.1. Transport rate of gold in the -0.053 mm size fraction Concentrations of gold in the -0.053 mm size fraction are above detection limit in fourteen of twenty five samples (Table 5.5 and Fig. 5.11). It can be seen from Fig. 5.11 that: (1) the values are strongly skewed with a preponderance of zero values; and (2) the gold concentrations are discrete and can be roughly grouped into intervals that increment with a factor of 0.9 ppb: thus the first group is about 0.9 including 1, 1.2 and 1.3; second group close to 1.8 including 1.5, 1.6 136 CD c cf CD E c o 3 3 O <L> c 6 0 3 OO in VD O N <N ^ CN £ £ 2 r f O 1 r j -•n CN co m r t O CN CN —< OO VD — ' OO co ro S M . ° H O * ^ » CN —' ro VD CN O CN vD CN 00 — ro ro ro ro oo cN ro CN g o a. s o c o o O N o vn o J O o o JO E o o' o J = r" >-» w p l i l i U -VD vo cN r t r-. '• ~ —< CN ro in ro <N wo CN >n r j -oo CN cN cn <N o CN ro i ^ O o vn <N m vt h I N CN r> co r-~ vo XJ O •c CL) a. oo c "C o 'S o E C N ON ON E o m I 2 CL) CL) J O J O E ") °- I o c —i O XJ o cn cl CL) J 3 toO J3 o j ~ VJ p l i l i , PH cn CN co CN r-~ r t CN CN r t O r--CN r t r t CO ON I OO oo o m ^— oo CO r t OO t r t O S NO CO •~ l NO r t 0 0 CO o OO NO CN O N co O N ON CO oo r-CO co oo C N o, r~ co oo cn r t cn cn ro CN — i xJ O •c CD CC toO s o o E o cn cn e <L) CL) O *—' •a o e 5 -a o I \ S CD CL) C - O J ^ w £ £ ^ 5 o CD J 3 6 0 IS o JZ; p t i l i l i 137 ro wo O o + wo r-o © wo r-o ©' + wo o wo o ON r t IT) o ON ON CO r t r t ON o r t WO WO ON CO WO ro wo ro r-ON ro NO ro CN .CO o r t O NO WO CN CN r t NO WO NO ro r t ON ON r t Tt ro NO rO CS —< — < ' CS CN - H cs CN CN CN CN CN ro CN CN CS CN CN CS CS CN CN o ro ro wo r t 00 o o © ON o CN ro o ON WO 00 r-00 © CO O CN o ro r t r t 00 r t ON NO C -CN CM ro ro ro f - r t © ON ro CN ro CN rt CN CN ro ro CS ro' CN ro CN ro ro ro r t ro ro ro ro ro r t ro ro ro r t O ro ro WO NO _ ON wo NO ON CN wo wo 00 rt r t r t CN WO r- CN O ON © © 00 O 00 ON r-- ON ON r t r t oo ON rt CS ro ro r t ro NO ON r t ro CN ro CO CN r t CN ro ro ro ro ro r t r t r t r t r t wo r t r t r t wo r t r t r t 60 O c o 1 rt ro wo • S t3 •B -3 2 - s <G to <L> -N O ' « '€ a rt 2 °" C T3 ro "o wo on I 2 12 IcS ro wo O ro wo o © + ON r t 6(2) 1.2(0) CN 1.2(0) •7(2) ' V Csi •5(2) CN —1 c_? V —< C_J V © CN —1 —1 © CN' o © V V © © © CN' CN' CN d V o o ^ CN' CST ~* © V s <=> ° O v V ON CO © —' ro ro r t ro r t ON CN CN ro O CN © v O -—' ~—" *—' s—' ^—' *—' o o o o o d v © o o o o V V V V o V o V o V O — i w x ^ ' V * o o o o • • 5 V V V O r t o © © © ^ © , o © © rt rt rt rt rt rt rt o V o V 0> 0> o V o V 0> 0> "8 "rt o CN o o o o o o o o o p o o o p CN d CO d CO d CO d CO d ro d ro d ro o o o o d d o o 00 o o O r-oo ON ON r t r t r t r» ON o O oo wo O O ro NO O WO d ro CN r t CN r t ro CN ro CN d CN NO CN NO CN d CO WO d CO r t CN r~i CN •3 1 O CN ON 00 oo ON NO ro NO r - rt rt O wo od 00 00 00 ON ON ON CN r t wo r t NO r t 00 ro o NO ON NO ro wo O o wo NO oo r t W 0 CN r--wo r t r t W 0 NO O d r t WO wo ND NO od d CN CN cs CS cs ro CN ro CN r t CN r t CN NO cs rt I"1 00 NO WO CN O OO 00_ rt" r-~ oo ON CN NO" CN rt ro —< o ON" NO O NO CN 00 NO WO ON —( W0 NO C -W0 00 NO^  l>" NO NO r t NO NO W0 NO r t wo wo NO r~ wo wo ro NO 00 ro ro oo wo r t ro ON O ro rj-O + ON tn . I* ° I I % rt l> 1-2 H 43 ro ON ro ro ON ON ro ro ON ON ro ro ON ON ro ro ro ON ON ON ro co ON ON ro ON ro ro ro ON ON ON co ro ON ON ro ro ON ON ro ro ON ON 12 10 d. o 6 138 12 10 Q. £ cd to O CD JO 6 3 6 h 4 h 0.9 1.8 2.7 Concentration intervals (ppb) Fig. 5.11. Distribution pattern of gold values in the -0.053 mm Helley-Smith sample fraction 139 and 1.7, etc. Ingamells and Pitard (1986) suggested that such data follow the Poisson distribution. Because the concentration increases with a factor of 0.9 ppb, 0.9 ppb is taken as representing the contribution of 1 grain of gold. The actual probability distribution of finding no gold, one gold grain, two gold grains, three and four gold grains from the data in Table 5.5 and Fig. 5.11 are shown in the upper part of Fig. 5.12. Based on the Poisson distribution (P„ = e"z(z7n!) where P„ is the probability that n grains will be found in a sample with the average number of grains (z)), the table in the lower part in Fig. 5.12 lists probability of finding 0, 1, 2, 3 and 4 grains if the average number is 0.5, 0.6, 0.7, 0.8, 0.81, 0.82, 0.83 and 0.84. By comparison to the number in the upper part of Fig. 5.12, when the average number of gold grain is 0.84, the estimates using the Poisson distribution are in reasonable agreement with the actual, discrete distribution up to finding 2 grains. The former is much lower than the latter for the probability of finding 2, 3 or 4 grains. The reason for this is probably because at discharge above 20.0 m3/s, more gold grains are released and transported. Taking 0.9 ppb equivalent to one grain is acceptable. Shape characteristics of visible gold particles in Harris Creek stream sediment were studied by Hou and Fletcher (1991), and Day and Fletcher (1986). Results suggest that more than 70% of the gold particles occurred in flakes with a thickness:diameter of about 1:10. The calculation using NUGGET (Stanley, 1989) shows that 0.9 ppb in 30 grams of -0.053 mm fraction is equivalent to about 1 particle with a thickness:diameter of about 1:10. Therefore, it is estimated that at discharge below 20.0 mVs, samples contain 0.84 grains or 0.74 ppb on average. At discharge above 20.0 m3/s, more gold grains are released and transported. It is estimated using the same method that the average gold concentration at discharge above 20.0 m3/s is 1.2 ppb. Using these estimates, the transport rate of gold in the -0.053 mm size fraction is calculated using formula (4.4) R = (Wt/T)x(l/0.076) (Wt is equal to the weight of-0.053 mm multiplied by gold concentration, T is the total sample time listed in Table 5.6). The transport rate against discharge 140 50 40 -30 h Gold grain z n=0 n=l n=2 n=3 n=4 0.5 6.0.65 30.33 7.58 1.26 0.158 0.6 54.88 32.93 9.88 1.98 0.296 0.7 49.66 34.76 12.17 2.84 0,497 0.8 44.93 35.95 14.38 3.83 0.767 0.81 44.49 36.03 14.59 3.94 0.798 0.82 44.04 36.12 14.81 4.05 0.830 0.83 43.60 36.19 15.02 4.16 0.862 0.84 43.17 36.26 15.23 4.26 0.896 Fig. 5.12. The distribution of gold particles and Poisson distribution calculation results Table 5.6 Showing weights of -0.053 mm fraction and estimated gold concentrations in bedload samples on Harris Creek Sample No. Discharge Weight (g) Gold concentration Sampling time (mVs) of -0.053 mm (ppb) (min.) 93-HZ-18 5.32 24.6 0.74 450 93-HZ-75,76 8.69 71.9 0.74 660 93-HZ-77.78 8.78 88.2 0.74 750 93-HZ-81.82 8.18 65.2 0.74 720 93-HZ-79,80 8.19 61.8 0.74 780 93-HZ-26,27 9.06 201 0.74 1200 93-HZ-73,74 9.72 79.6 0.74 650 93-HZ-71.72 10.97 52 0.74 540 93-HZ-69,70 14.21 55.3 0.74 169 93-HZ-59,60 15.54 59.3 0.74 40 93-HZ-61.62 15.74 58.7 0.74 40 93-HZ-57,58 16.46 48.2 0.74 40 93-HZ-67,68 16.78 43.8 0.74 91 93-HZ-65,66 17.31 25 0.74 52 93-HZ-63,64 17.60 47.1 0.74 52 93-HZ-54 17.69 16.4 0.74 20 93-HZ-55 18.53 38.9 0.74 20 93-HZ-56 20.00 33.8 0.74 26 93-HZ-37 22.75 4Z.9 1.2 26 93-HZ-36 22.86 45.5 1.2 25.5 93-HZ-38 23.54 43.8 1.2 26 93-HZ-48 23.72 23.7 1.2 28 93-HZ-35 24.45 65.9 1.2 41 93-HZ-39 24.54 39.2 1.2 20 93-HZ-40 26.06 37 1.2 10 142 is shown in Fig. 5.13a. For comparison, Fig. 5.13b shows the results directly based on the analytical data. Transport rate based on estimated gold concentration increases with increasing discharge. It is expressed as the following statistical regression equation which is highly significant at a = 0.05: R = 10"739Q3 0 2 (n = 18, r2 = 0.88 n = 23 and AR = 10±003) 5.5.2. Transport rate for the -0.149+0.053 mm gold Concentrations of gold in the -0.149+0.053 mm size fraction are listed in Table 5.5 and Appendix Table 4. In spite of the combination of the samples into larger composite samples, gold values are strongly skewed with twenty one out of the twenty five samples having no detectable gold. Only two samples, 93-HZ-55 and 93-HZ-57-58 (Table 5.5), are likely to contain one grain of free gold. Because 10.0 m3/s is the critical discharge at which the bed surface is mobilized, and because coarse gold starts to be released from the bed once the discharge exceeds the critical discharge (Fletcher and Wolcott, 1991), transport of gold is estimated only at discharge above 10.0 m3/s based on the Poisson distribution and absence of free gold from 84% (21/25) of the samples. The result (Fig. 5.14) indicates the average number of gold grains is 0.2, equivalent to 1.2 ppb. Based on the data in Table 5.7 the transport rate is calculated using formula (4.4) and shown in Fig. 5.15a. The transport rate based the original data is shown in Fig. 5.15b. It is evident from Fig. 15a and b that the gold transport rate based on the estimated gold concentrations shows a trend with discharge. Using statistical regression analysis, the equation of the transport rate of gold in the -0.149+0.053 mm size fraction is expressed as follows. R= io-708Q2 8 5 (r2 = 0.65, n= 18 and AR = 0.78 to 1.29) An F-test indicates that this equation is highly significant at a = 0.05. A 2-standard error range is given by AR (see Chapter 4). 143 0.002 0.001 « 0.0003 3. £ o.oooi a. 0.00001 a fl -critical discharge B • B The BB B • I* • -10 15 Discharge (mVs) 20 25 30 0.002 0.001 | 0.0005 0.0002 t: o 0.0001 t -0.00005 0.00002 b on rising limb • large • ical disci O • •• - The crit o on falling limb 0 f f -o • w u; 10 15 20 25 30 Discharge (mVs) Fig. 5.13. Transport rate of the -0.053 mm gold with Helley-Smith Sampler a: estimate based on Poisson distribution; b: original data Arrows means gold below the detection limit 144 100 80 h 60 £ 40 20 Gold grains n = 0 n = 1 0.1 0.904837 0.090484 0.2 0.818731 0.163746 0.3 0.740818 0.222245 0.4 0.67032 0.268128 0.5 0.606531 0.303265 Fig. 5.14. Estimated numbers of gold particles in -0.149+0.053 mm size fraction 145 Table 5.7 Weight of-0.149+0.053 mm fraction and estimated gold concentrations in the Helley-Smith samples on Harris Creek Sample No . Discharge (mVs) of Weight (g) -0.149+0.053 mm Estimated gold concentration (ppb) Sampling time (min.) 93-HZ-18 5.32 12.91 1.6 450 93-HZ-75,76 8.69 40.31 1.6 660 93-HZ-77,78 8.78 45.86 1.6 750 93-HZ-81.82 8.18 36.71 1.6 720 93-HZ-79,80 8.19 35.98 1.6 780 93-HZ-26,27 9.06 149.5 1.6 1200 93-HZ-73,74 9.72 46.65 1.6 650 93-HZ-71.72 10.97 35.71 1.6 540 93-HZ-69,70 14.21 33.67 1.6 169 93-HZ-59,60 15.54 32.71 1.6 40 93-HZ-61.62 15.74 33.32 1.6 40 93-HZ-57,58 16.46 27.81 1.6 40 93-HZ-67,68 16.78 24.66 1.6 91 93-HZ-65,66 17.31 14.72 1.6 52 93-HZ-63,64 17.60 23.66 1.6 52 93-HZ-54 17.69 11.98 1.6 ,20 93-HZ-55 18.53 23.19 1.6 20 93-HZ-56 20.00 20.44 1.6 26 93-HZ-37 22.75 26.74 1.6 26 93-HZ-36 22.86 26.97 1.6 25.5 93-HZ-38 23.54 30.7 1.6 26 93-HZ-48 23.72 15.58 1.6 28 93-HZ-35 24.45 39.28 1.6 41 93-HZ-39 24.54 24.63 16 20 93-HZ-40 26.06 22.5 1.6 10 146 0.002 0.00002 Discharge (rrf/s) a X co =J. o oo S ro cn + o cn C M O B f3 0.003 0.001 0.0003 0.0001 £ 0.00003 o ex | 0.00001 0.000003 M l 1 1 1 , UJUli IL 20 25 5 10 15 Discharge (mVs) Fig. 5.15. Transport rate of -0.149+0.053 mm gold versus discharge in Harris Cr a: estimate based on Poisson distributio; b: from original data (arrows indicate gold below detection limit) 30 147 5.5.3. Gold Outputs Calculation results suggest that no gold was transported out of Harris Creek in both 1991 and 1992. In 1993, it is estimated that Harris Creek output 0.13 grams of the -0.149+0.053 mm gold and about 0.18 grams of the -0.053 mm gold. All of the gold was transported by the spring freshet between May 11 and May 26 in 1993, in which more than 95% of the gold was transported by the flood discharge above 30.0 nvVs. In contrast, about 46% of magnetite and 56% of sediment was transported by the flood at discharge above 30.0 m3/s. This evidence strongly suggests that gold is transported mainly by a big flood. This general conclusion is good agreement with the previous studies in Harris Creek. Gravels which form the framework behave quite differently than the sand which forms the matrix. Gravels begin to move only during the higher flows near the snowmelt flood peak than those that move sand matrix (Church et al., 1991, and also see Chapter 7). Gold is transported only for brief periods when increased discharge caused by snowmelt floods, disrupts the framework (Fletcher and Wolcott, 1991). These results suggest that gold is not moved with a mobility equal to that of sandy material. C H A P T E R 6 148 B U D G E T S F O R S E D I M E N T , M A G N E T I T E A N D G O L D , A N D T H E I R D I S T R I B U T I O N S A L O N G L O W E R H A R R I S C R E E K 6.1 Introduction As stated in Chapter 1, sediment, magnetite and gold budgets for this study are constructed on the basis of the following relation: I - AS = O, or AS = I - O where I is the mass of sediment, magnetite or gold input to the system (lower Harris Creek) during a specific time period At; AS is change in the mass of sediment, magnetite or gold in storage sites in the system in At; and O is the mass of sediment, magnetite or gold output from the system in At. "Storage site" is used here collectively to include stream channels, bars, floodplain and log jams. If AS = 0, then I = O. In this case the output of sediment, magnetite or gold is the same as the amount being added from the source areas, and this status is called "steady state" (Trimble, 1983), or "equilibrium" (Church, 1983). If AS > 0, then I > O. This situation implies that input of sediment, magnetite or gold is more than its output, the excess material will deposit into storage sites, and the storage sites are aggrading. If AS < 0, then O > I. In this condition, sediment, magnetite or gold is being remobilized from storage sites, and degradation is occurring. So far, inputs of sediment, magnetite and gold with estimated errors have been discussed in Chapter 4 and outputs in Chapter 5. Using these results, behavior of gold and magnetite along Harris Creek can be evaluated. Because Helley-Smith samples are so small, the result for gold input and output can be evaluated only qualitatively. Therefore, the budget results presented here attempt to characterize only the most important processes and their spatial and temporal variation 149 in the system. The time scale used in sediment budgeting has a major influence on the conclusion drawn (Church, 1980) and, in this context, the results describe only the general pattern for time scales of single years or less. 6.2 Sediment, Magnetite and Gold Budget 6.2.1 Flow history An 18-year record of flows is available from a Water Survey of Canada (WSC) gauge (WSC station 08LC005) located 6 km downstream from the study area. This station controls a drainage area of 253 km2, which is slightly larger than the area (225 km2) this study includes. Data from this station will be used to approximately evaluate historic characteristics of flows in the study basin. Data in Table 6.1 shows that the hydrological regime is dominated by a snowmelt flood in the late spring or early summer. Because 1991, 1992 and 1993 were monitored for the purpose of study, therefore, peak floods of these three years are used as references to compare with the historic records. Results (column 4 of Table 6.1) indicate that 5 years' floods were below 1992's flood, or in other words, peak flood was not over 15.0 m3/s. 5 years' floods are equivalent to 1992's flood, that is, peak floods were equal or close to 15.0 m3/s; 8 years' floods are equivalent to 1991 's flood: their peak floods are equal or close to 20.0 m3/s. No flood in the past 18 years can match the 1993's flood. Based on the 18-year records, the mean annual flood of Harris Creek is 19.0 m3/s, therefore, the 1991 flood is very close to the mean annual flood. For convenience, in the following descriptions the years with floods equivalent to the 1992 record are called dry years, years with floods equivalent to the 1991 record are called normal years, and the year with the May 1993 flood record is called a wet year. 150 Table 6.1 Summary' of flow conditions of Harris Creek and 18-year records of W S C station 08LC005 located 6 km downstream from this study site Recording Peak flood Date of Comparison to reference Gauge station year (mVs) peak flood years (see text) 1993 45.0 11-May as a reference gauge in Harris Creek* 1992 15.0 24-May as a reference gauge in Harris Creek 1991 20.0 15-May as a reference gauge in Harris Creek 1983 17.8 28-May 1992 08LC005, drainage area 253 k m2 1982 23.1 2-Jul 1991 08LC005, drainage area 253 k m2 1981 16.1 6-Jun 1992 08LC005, drainage area 253 k m2 1980 11.8 27-May below 1992 08LC005 , drainage area 253 k m2 1979 22.8 26-May 1991 08LC005, drainage area 253 k m2 1978 16.3 6-Jun 1992 08LC005, drainage area 253 k m2 1977 11.0 7-Jun below 1992 08LC005 , drainage area 253 k m2 1976 26.5 15-Jun approximately 1991 08LC005 , drainage area 253 k m2 1975 24.9 3-Jun approximately 1991 08LC005 , drainage area 253 k m2 1974 17.7 27-May 1991 08LC005 , drainage area 253 k m2 1972 28.1 1-Jun approximately 1991 08LC005 , drainage area 253 k m2 1971 17.3 3-Jun 1992 08LC005 , drainage area 253 k m2 1970 13.4 4-Jun 1992 08LC005 , drainage area 253 k m2 1969 8.7 2-Jun below 1992 08LC005 , drainage area 253 k m2 1968 18.4 3-Jun 1991 08LC005 , drainage area 253 k m2 1967 17.8 3-Jun 1991 08LC005 , drainage area 253 k m2 1966 11.0 5-Jun below 1992 08LC005 , drainage area 253 k m2 1965 7.7 16-Jun below 1992 08LC005 , drainage area 253 k m2 * the drainage area of Harris Creek is 225 k m 2 151 6.2.2 Sediment budget Figs. 6.1, 6.2 and 6.3 represent the sediment budgets for lower Harris Creek. Although the data are insufficient for detailed description of the input and output, results demonstrate that: (1) in both normal and wet years, the yield of sediment from the basin is less than the amount of the sediment being mobilized from the sources. Therefore, lower Harris Creek receives excessive materials, and aggradation occurs; (2) in dry years both input and output are small; but output is larger than input (Fig. 6.2). In 1992, about 20 tons of sediment was mobilized from degrading of the stream bed. The sediment budget along Harris Creek varies from year to year. In dry years, banks are not eroded; landslides are more or less stable; tributaries do not transport significant sediment so that input amount is minor. During such years lower Harris Creek mobilizes sediments only from its own bed. But because it is only a few days that the peak flood exceeds the critical discharge of 10.0 m3/s, the output of sediment from the stream bed is small. In normal years, sediment input from the secondary sources account for 99% of the total input, of which 97% is contributed from laterally unstable banks. Of the input sediment 80% is transported out of Harris Creek and 20% is stored in channels. In wet years, 39% of the input sediment is contributed from landslides; and 61% is from the secondary sources, of which 52% comes from bank erosion of laterally unstable banks. In such years, much (92%) of the input sediment is transported through the system with only a small amount (8%) being stored along lower Harris Creek. 6.2.3 Magnetite budget The budget for magnetite in the combined -0.149+0.053 mm size fraction is shown in Figs. 6.4 to 6.6. Like the sediment, in both normal and wet years the output of magnetite from lower Harris Creek is smaller than the amount of the magnetite being input. Conversely in dry years, output exceeds input with magnetite being remobilized from the stream bed. These results suggest that, in the dry years, magnetite from the different sources is below the conveyance 152 153 154 capacity of the stream, and Harris Creek mobilizes 38 kg of magnetite from its bed through degrading. The amount of magnetite stored along Harris Creek thus varies from year to year and is related to flow conditions. In normal years, almost all the magnetite is supplied from bank erosion. However, only a small amount (25%) is carried through lower Harris Creek with a large amount (75%) being stored along lower Harris Creek. In wet years, 15% of the magnetite is from landslides and 85% from the secondary sources. Bank erosion contributes 84% of the latter. Floods carried 50% of the magnetite out of the Harris Creek basin with 50% of it being stored in the Harris Creek basin. In dry years, 80% of the input is contributed from landslides and 20% is from the log jam. 6.2.4 Gold budget analysis Budgets for gold in the -0.149+0.053 mm size fraction are presented in Figs. 6.7 to 6.9, and for gold in the -0.053 mm size fraction in Figs. 6.10 to 6.12. In 1991 and 1992, the output of gold in both the -0.149+0.053 mm and -0.053 mm size fractions was zero. This means all the gold added from the different sources was stored in the channel. But supply sources are different from dry years to normal years. In the dry years, Landslides #1 and 7, and Logjam #1 are the major suppliers. But in the normal years, bank erosion becomes important. In the wet years, which are quite rare in the Harris Creek basin, 93% of the coarse gold and 88% of the fine gold was stored along lower Harris Creek. It is clear by summarizing the budget results into Table 6.2 that: during normal years Harris Creek delivers 80% of the input sediment and 25% of the magnetite out of the basin; and in wet years, almost all the input sediment is transported out but only 50% of the magnetite is carried away. Harris Creek transports gold out of the basin during very high floods. The result suggests that even the May 1993 flood transported only 7% of the coarse gold and 12% of the fine gold 155 156 Table 6.2 Showing the difference of transport of sediment, magnetite and gold in the Harris Creek basin. Dray year Normal year Wet year -0.149+0.053 m m A u 0 0 7 -0.053 mm A u 0 0 12 Magnetite - 25 50 Sediment - 80 92 Note the difference of transport is expressed as 100 * output/input (%) - means that output > input, storage experiences degradation. 158 out of the basin. In brief, the storage order of sediment, magnetite and gold increases with the increasing specific gravity of the minerals, that is, storage of Au > storage of magnetite > storage of sediment. 6.3. Distribution of Gold and Magnetite along Lower Harris Creek Although there are several types of storage sites along lower Harris Creek, bars are most distinctive and most often used sites to collect geochemical samples. The longitudinal distribution of gold and magnetite in high and low energy environments from bar heads and tails is therefore discussed in the following sections. 6.3.1 Magnetite distribution Weights of magnetic and non-magnetic heavy mineral concentrates from high and low energy environments along Harris Creek are shown in Appendix Tables 9 and 10. Magnetite distribution is shown in Figs. 6.13 and 6.14, and Table 6.3. A t-test indicates that mean of magnetite concentration in the -0.149+0.105 mm size fraction from high energy environments is significantly higher than that from the low energy environments at a = 0.05, but magnetite concentrations in both the -0.105+0.075 mm and the -0.075+0.053 mm size fractions are not significantly different in the high and low energy environments (Table 6.3). Between the two environments, hydraulic effects can also be evaluated by the Geometric Mean Concentration Ratios (GMCRs) (Saxby and Fletcher, 1986b) which are expressed as following form: GMCR = exp([IlnCRi)] In) where CRi -{CpjjvlX, High Energy Environment^HM,X, Low Energy Environment)i. CHM,X,High Energy Environment *s heavy mineral concentration in size fraction X from the high energy environment, and Cf^yr x; Low Energy Environment's t n e same heavy mineral concentration in the same size fraction from the low energy environment at sampling bar i (i = 1, 2, n). Table 6.3. Concentrations of magnetite and gold in heavy mineral concentrates along the lower Harris Creek Magnetite (%) Number of samples Maxim. Mean Minim. In high energy environments -0.149+0.105 mm 11 8.48 5.03 2.99 -0.105+0.075 mm 11 6.73 3.90 2.56 -0.075+0.053 mm 11 6.44 3.18 1.90 In low energy environments -0.149+0.105 mm 10 7.85 2.96 0.83 -0.105+0.075 mm 10 8.42 3.85 1.82 -0.075+0.053 mm 10 5.74 3.37 1.99 t-test of the difference between HEE and LEE means |HEE mean-LEE mean| Critical value at 0.05 -0.149+0.105 mm 2.07 1.533 -0.105+0.075 mm 0.05 1.447 -0.075+0.053 mm 0.19 1.123 Gold (ppb) Number of samples Maxim. Mean Minim. In high energy environment -0.149+0.053 mm 11 6460 4120 2990 -0.053 mm 11 6.8 2.4 0.8 In low energy environment -0.149+0.053 mm 10 215.00 73.00 5.00 -0.053 mm 10 27.40 7.50 0.20 162 Results calculated with the above formula are listed in Table 6.4. A two-tail t-test of the equation indicates that for magnetite in the -0.149+0.105 mm size fraction the GMCR is significantly higher than 1; but that values for magnetite in the -0.105+0.075 mm and -0.075+0.053 mm fractions are not significantly different from 1.0 (Table 6.4)., That is, magnetite in the -0149+0.105 mm size fraction prefers to deposit in high energy environments, whereas magnetite in the latter two size fractions does not show this preference. Longitudinally, downstream distribution of magnetite in all the three size fractions from high energy environments shows similar patterns with relatively high concentrations above the confluence with Mosquito Creek, then the magnetite decreases to 3% to 4% downstream between km 3 and 10. Further downstream, where stream gradient decreases to 0.011, magnetite concentration seems to increase again slightly. Magnetite concentration in the same size fractions from low energy sites is approximately 3% except for a large increase in a single sample immediately upstream from Landslide #1 (Figs. 6.13 and 6.14). 6.3.2. Distribution of gold Analytical results of gold in non-magnetic fractions and in the -0.053 mm fraction of stream sediment from high and low energy environments along Harris Creek are shown in Appendix Tables 11 and 12. The distribution of gold along Harris Creek is shown in Figs. 6.15. to 6.18, and Tables 6.3 and 6.4. The mean of gold in non-magnetic heavy mineral concentrates (NMHMC) of the -0.149+0.053 mm size fraction is 4120 ppb at high energy environment sites versus only 73 ppb at low energy environment sites. Gold in this size fraction thus has a strong preferential accumulation in high energy environments. In contrast, the mean of gold in the -0.053 mm stream sediment is 2.4 ppb at high energy sites and 7.5 ppb at low energy sites. A t-test indicates that the GMCR value (0.83) is significantly less than 1.0, that is, fine gold preferentially accumulates at low energy sites. 163 (ui) 0^ 9 Jnojuoo SAOCJB jqSran NP r f <N —i o o d o d o (uidd) DJAjHjAiM ui ppo 165 166 Table 6.4. Geometric mean concentration ratios for gold and magnetite in Harris Creek -0.149+0.105 mm -0.105+0.075 mm -0.075+0.053 mm Magnetite 1 38 1.05 0.94 -0.149+0.053 mm -0.053 mm Gold 36.36 0-83 168 Longitudinal distribution of gold in -0.149+0.053 mm NMHMC from high energy environments is shown in Fig. 6.15. In order to test the downstream trend statistically, the lower and upper 95% confidence limits for each sample are calculated with the following formulas based on the Poisson Distribution (Zar, 1984): L l = (X2(l-a/2), v)/2, where Li = lower 1- a confidence limit, and v = IX. L 2 = 0c2(ct/2), v)/2, where L2 = upper 1 - a confidence limit, v = 2(X + .1) and X is the calculated number of gold particles. Based on the assumption of gold present as flakes with thickness:diameter ratio of 1:10 (Hou and Fletcher, 1991; and Day and Fletcher, 1986), the above formulas are used to calculate lower and upper 1-a confidence limits of gold particles. Then the gold particles are recalculated back to concentrations. The confidence intervals are shown as vertical dashed lines on each value in Figs. 6.15 to 6.18. For gold in the -0.149+0.053 mm size fraction from the high energy environment, because most samples contain more than 20 gold flakes in the non-magnetic heavy mineral concentrates, bands between lower and upper confidence limits (95%) are very narrow, and the downstream variations are statistically significant without any overlap. The gold anomaly exhibits erratic and anomalous spikes several kilometers downstream of Mosquito Creek (Fig. 6.15). First at a short distance downstream of Landslide #7, there is a peak value of about 3.8 ppm, which is coincident with a break-in-slope where the stream gradient decreases from 0.039 to 0.022. A second gold peak is just below Landslide #1 with a value of 6.5 ppm. This also coincides with a break-in-slope where the gradient decreases from 0.022 to 0.011. The thirteen kilometers long-profile clearly shows that concentrations of gold increase, rather than decrease downstream from Mosquito Creek - a gold source for Harris Creek. In contrast to gold from high energy environments, downstream variation of gold in the low energy environments are not significant although the overall trend seems to be for Au values to decrease downstream. 169 Gold in the -0.053 mm size fraction of stream sediment from high energy sites has a high value of about 3.5 ppb just below Mosquito Creek, then decreases to about 1.0 ppb as far as 10 km. Downstream from Landslide #1, gold concentrations increase to 7.0 ppb (Fig. 6.17). Distribution of gold in the -0.053 mm stream sediment from low energy sites (Fig. 6.18) is quite different to that in the high energy sites but is similar to distribution of gold in the -0.149+0.053 mm size fraction (Fig. 6.15) with peak values of 27.4 and 25.0 ppb at breaks-in-slope just below Landslides #7 and #1, respectively. Further downstream from both landslides concentrations of gold decline to less than 4.0 ppb over a distance of several hundred meters. 6.4. Landslides, Peak Gold Values and Stream Gradient 6.4.1 Further investigation of the problem As described above, gold and magnetite are not stored uniformly along lower Harris Creek, but are preferentially stored in very specific environments. Gold in the -0.149+0.053 mm size fraction is preferentially stored at bar heads, while gold in the -0.053 mm size fraction preferentially accumulates at bar tails. Both have peak values immediately downstream from Landslides #1 and #7 (Figs. 6.15 and 6.18) at breaks-in-slope of the gradient of Harris Creek. The results suggest a possible spatial relation between peak gold values in both the silt-clay size fraction and the f\TMHMC, and Landslides #1 and #7. Because both landslides are active (see Chapter 4) and because cut-offs for the gold anomalies appear to be present just downstream from the two landslides, it might be supposed that the landslides are the source of the gold. Alternative explanations are that the landslides are not the source of the gold and that either (i) the apparent association of two gold anomalies with the landslides is coincidental, or (ii) the association of the landslides with the gold anomalies is related to geomorphic processes in the Harris Creek catchment. To test these possibilities, detailed studies of what, if any, effect the 170 introduction of landslide material has on the composition of sediments in Harris Creek were undertaken in the detailed study reach from 100 m upstream to 2,500 m downstream of Landslide #1 (the Detailed Study Reach (DSR) in Fig. 2.7) in 1993. 6.4.2 Landslides #1 and #7: anomalous gold source or not? The 1993 studies in the DSR confirmed the 1991 data: that is, gold content of the -0.053 mm fraction of sediments from high energy sites is less than in the associated low energy sites and the peak concentration (8.9 ppb) in the latter occurs roughly 600 m downstream of Landslide #1 (Fig. 6.19). Beyond this point gold values at low energy sites fall to less than 2 ppb and are comparable to concentrations of gold at high energy sites. Gold content of the -0.053 mm fraction of the landslides ranges from 1.0 - 3.6 ppb for Landslide #1 with an average of 1.8 ppb, and <0.2 to 0.8 ppb with an average of 0.4 ppb for Landslide #7 (Table 6.5). These results together with the results in Chapter 4 suggest that Landslides #1 and #7 are not the source of the high gold values found just below them. What, then, is the link between the landslides and distribution of gold in Harris Creek? 6.4.3 Clay ball distribution Landslide #1 consists of about 90% glaciolacustrine sediments and 10% till by volume (Fig. 4.9A). The glaciolacustrine sediments are cohesive silts and clays. Field measurements show that the landslide supplied roughly 2 tons sediment from July 1991 to April 1992, 1.0 ton from May 1992 to April 1993, and 70 tons in the May 1993 flood period. Field observations in the DRS found clay balls, similar to the glaciolacustrine sediments, at high energy sites. Although it was difficult to remove the balls intact from the stream bed, field observations indicated that the number, size and angularity of the clay balls decreases over a distance of about 1000 m downstream from the landslide. They were not observed at all at low energy sites downstream, or from the high or low energy sites upstream from the landslide. Thus, 171 172 Table 6.5. Gold content of the four landslides in the Harris Creek basin Gold concentration (ppb) in -0.053 mm size fraction Landslide Mean Range #1(7) 1.8 1.0-3.6 #7 (3) 0.4 <0.2 - 0.8 #17(2) 0.2 <0.1-0.5 # 28 (3) <0_2 <0.2 a The number in the parentheses is the numbers of samples from that landslide. Fingerprinting properties Al (%) Mg (%) Ba (ppm) Na (%) P (ppm) Sr (ppm) Ti (%) Stream sediment Landslide #12 HEE (n=l) LEE (n=l) n = 3 7.5 7.8 8.5±0.37* 1.3 1.4 1.8±0.08 948 975 860±0.00 1.8 / 1.5±0.14 1880 1900 843±36.1 470 445 324±7.20 0.74 0.76 0.43±0.02 LEE = low energy environment and HEE = high energy environment * 95% confidence limits. 173 from the distribution of the clay balls, it is obvious that they are glaciolacustrine sediments derived from the toe of Landslide #1. The changes in the size, shape and number of the balls suggest that they are gradually disaggregated during their transport downstream. 6.4.4 Fingerprinting the effect of the clay balls Multi-element geochemical data have been used to fingerprint the effect of the input of material from Landslide #1 on the geochemical composition of the -0.053 mm sediments. Seven elements (Al, Mg, Ba, Na, P, Sr and Ti) of the twenty four determined were chosen on the basis that they statistically best differentiate the composition of the landslide from that of the sediment (Table 6.6). Of the seven elements, Al and Mg have higher concentrations at 95% confidence limits, and Ba, Na, P, Sr and Ti have lower concentrations at 95% confidence limits in the landslide than in the sediment upstream from the landslide. Distribution patterns of the seven elements are shown in Fig. 6.20. Geochemical patterns downstream from the landslide are of two distinctive types that depend on whether elemental concentrations in the active toe are higher or lower than in sediments upstream of the landslide. For the former (i.e., Al and Mg) concentrations at high energy sites increase immediately below the landslide but then decrease further downstream. Conversely, concentrations of these elements are initially unchanged at the associated low energy sites but then show a concomitant increase as concentrations at high energy sites decline. Elements with relatively low contents in the landslide (i.e., Ba, Na, P, Sr and Ti) show the exact opposite trends (Fig. 6.20). To further estimate the relative contributions of landslide material to the sediments, a simple mixing model modified from Peart and Walling (1986) has been used: C s = 100x{(Pd - PU)/(P, - Pu)} when Pj > P u; C s = 100x{ 1 - (Pd - P>)/(PU - Pj)} ( P u > Pf). 174 (o / o ) IUD.UCO iv (%) l u o i u c o SJAI 1 y -S 3 o -o o y 8 « c o o o c C3 O O O O O o K~> o m o >/o u o r f r f r o r o CN (Uldd) JU31UCO J g o o >/-> ° C N 65 c-o g CD is o J 53 u* 4) c A. on - '£ Q. r-c C3 OO X3 g g c o o c > 4> 2 c <+- 4) ° +3 c o 3 JO •c X 3 CD o rc! O o J 3 3 e Cu) 5 u. -—1 § -8 XJ <L) XJ c3 CL) •3 X 3 Q xj o w C N - J v O X3 . C c/) 3 OO O E id Si Cu) X3 XJ 176 where C s = Relative concentration (%); P d = value of selected property for stream sediment downstream from Landslide #1; Pj = value of selected property for the landslide; and P u = value of selected property of stream sediment upstream of the landslide. Although there are some differences between the estimates, the results of applying this model to Ba, P, Sr and Ti (Fig. 6. 21) show that the general trend for the contribution of the landslide material to the -0.053 mm fraction at high energy sites declines from 60 to 80% near the landslide to 30% or less over a distance of 1 to 1.5 km. Over the same distance the contribution of the landslide material to the low energy sediments increases from less than 10% to about 40%. However, it should be pointed out that the mixing model is based on several simplifying assumptions (e.g., no sediment-water interaction, and equal rates of transport for different mineralogical components of the sediments (Hou and Fletcher, 1996)) that may not be entirely valid. Departure from these assumptions probably accounts for the minor differences between the predictions based on different elements. The low Au content of the landslides (Table 6.5) and geochemical fingerprinting of material eroded from the Landslide #1 (Figs. 20 and 21) clearly show that the landslide #1 is not the principal source of fine gold and that landslide material will dilute Au concentrations in the -0.053 mm fraction of the stream sediment. However, because of the cohesive nature of the clay balls, this dilution becomes apparent only some distance downstream from a landslide. Thus the combined effect of accumulation of Au at breaks-in-slope, in response to the change in hydraulic conditions (Day and Fletcher, 1991), and dilution by material derived from the landslide (Hou and Fletcher, 1996), is to create peak Au values and an apparent cut-off for the Au anomaly a short distance downstream from a landslide. 177 178 C H A P T E R 7 D A T A INTERPRETATION 7.1 Introduction Hawkes (1976) described the following anomaly dilution model as a guideline to design and interpretation of stream sediment surveys: M e m A m = A a ( M e a " M e b ) + A m M e b The definitions of the parameters used in the above formula have been illustrated in Fig. 1.1. This model, which generates a smooth hyperbolic decay curve for the anomalous dispersion train downstream from the cut-off point, is based on several assumptions: uniform rate of erosion, uniform geochemical background; no feedback between water and sediment; no sampling error; a single anomalous source; and no contamination. Previous studies in Harris Creek focused on errors associated with field sampling and Au determination (Day and Fletcher, 1986); transport and depositional segregation of heavy minerals from light bulk sediments (Fletcher, 1990); hydraulic behavior of gold and other heavy minerals at the bar scale; relation between heavy mineral concentrations and gradient of the stream in a short reach; and a theoretical transport model of gold and other heavy minerals by streams (Day and Fletcher, 1991). Studies of Quaternary geology and mass wasting in Harris Creek (Ryder, 1991; and Ryder and Fletcher, 1991) have also laid a base for this study. This chapter attempts to show how combination of point sediment sources, episodic processes, and local hydraulic conditions affects the downstream dilution model of gold anomalies in stream sediment. To attain this objective, results have been presented in the previous chapters to show input from different sources, output from lower Harris Creek, and the downstream 179 distribution of Au and magnetite in stream sediment in the Harris Creek basin. The emphases here will focus on natural geomorphic conditions in the Harris Creek basin, and the evolution of Au anomalies in stream sediment with time. Applicability of some of the assumptions made in the traditional Hawkes dilution model will be examined, in particular the traditional assumption of uniform rate of erosion will be discussed further. 7.2. Sediment Sources: Effect of the Fraser Glaciation The Harris Creek basin, with an area of approximately 225 km ,^ is divided into two zones: the Zero Transfer Zone and Source Zone (Fig. 2.1). The former is located on the upland surface of the Okanagan Plateau which was uplifted during late Pliocene to early Pleistocene (Mathews, 1988). Because the highland surface is gentle, lakes and swamps on it trap any sediment that is mobilized from the upstream area (Ryder and Fletcher, 1991). Therefore, material in the Zero Transfer Zone can not cross the boundary and contribute to lower Harris Creek. But it is obvious that the Zero Transfer Zone only means no sediment transfer to or compositional link with lower Harris Creek; it by no means implies that the Zero Transfer Zone is a closed system and internally inactive. The Source Zone, with an area of about 68 km^ (which accounts for 40% of the whole area), flanks the Harris Creek valley (Fig. 2.2). All the sediment sources studied are located discontinuously within this zone (Fig. 2.8, and also see Ryder and Fletcher, 1991), and most of them flank Harris Creek. The sediment cascade in Fig. 7.1 provides a framework for analysis of possible gold transfer routes and storage sites in lower Harris Creek. This demonstrates that gold mobilization on valley sides does not necessarily coincide with gold supply to the stream channel. For example, gold moved by rockslides may reach a creek or it may be temporarily stored on lower valley sides and in fans. Gold that is transported by tributary creeks may accumulate on alluvial fans rather than being entrained by the trunk stream. The diagram also indicates that gold F i g . 7.1. Possible go ld sources for lower Harr is Creek and their transfer routes. (F rom Ryder and Fletcher, 1991) 181 supply from the upstream to the downstream reaches of lower Harris Creek is not a continuous process, but that channel material may be temporarily stored in the old floodplain or in channel bars for time periods ranging from one year to several millennia. This is in good agreement with the findings from studies of regional sediment yield in British Columbia (Slaymaker and McPherson, 1977; Roberts and Church, 1986; and Church et al., 1989) that secondary remobilization of Quaternary sediments along river valleys dominates over erosion of the headwater land surface. These fluvial deposits have been constructed from reworked glacial drift with general aggradation of the valley floor by streams and are not directly and genetically linked to underlying bedrock, but are easily accessible for erosion and transportation. Thus the glaciolacustrine, glaciofluvial and flood plain sediments decouple and isolate Harris Creek from the bedrock that underlies the valley slopes. It follows, therefore, that any Au mineralization in bedrock along the valley slopes and floor of lower Harris Creek is unlikely to contribute Au directly to the stream sediments of lower Harris Creek. Similarly, the budget results (Chapter 6) suggest that anomalous concentrations of gold found in the stream sediment are most likely largely derived from floodplain banks along lower Harris Creek. The relatively high gold concentration of this material results from reworking and recycling through the modern flood plain. The area between the plateau edge and the main stream is covered by glacial till and colluvial deposits (Fig. 2.5). Geochemical studies have shown that these materials are genetically coupled to local bedrock (Nelson, 1993). Presence of gold in the underlying bedrock would be expected to be reflected in the till. But, because of the decoupling effect of the floodplain and glaciolacustrine and glaciofluvial sediments, sediment derived from the till and colluvial material is supplied to lower Harris Creek mainly via the major tributaries (such as Beetle Creek, Vidler Creek, McAuley Creek and Mosquito Creek) rather than directly (Fig. 2.8). Small tributaries are less important to the transfer process because they are either too small or debouch sediment on to 182 old flood plain deposits or abandoned channels. The budget results and morphology of gold particles suggest that the coarse (-0.149+0.053 mm) gold found along Harris Creek is derived originally from upstream tributary sources where bedrock and surficial deposits are more closely coupled. Based on visual and SEM examination of gold grains, relatively pristine gold grains found in McAuley Creek and Mosquito Creek are most likely to have been derived from a nearby bedrock source (Hou and Fletcher, 1991). But because no measurable gold was transported in McAuley Creek due to low flow, Mosquito Creek is probably the only tributary to contribute gold to Lower Harris Creek during the study period. Gold delivery occurs only during infrequent high flows, which are major gold transport events. 7.3 Behavior of gold during floods 7.3. J Sediment Sediment transport depends on discharge, and there is an anti-clockwise hysteresis (Fig. 5.6) with greater amounts of sediment being transported during falling limb than during rising flow conditions. At discharge between 5.0 and 10.0 m-Vs, transport is limited to accessible material <10 mm which the flow is able to transport (Church, et al., 1991). At the critical discharge of 10.0 m-^ /s, the pavement is disturbed and partly mobilized (Fig. 5.5, also see Fletcher and Wolcott, 1991); gravel-sized particles start to be mobilized and released from the bed. Under such conditions, the fine sand and silt are transported because of their small size; the medium sand is moved because of its great exposure to the flow, once it is on top of the bed layer and no longer hidden from the flow (Figs. 7.2. and 7.3, and also see Klingeman and Emmett, 1978). At discharge above 15.0 nrVs, large cobbles (-90.4+64 mm) start to move (Fig. 7.4). Conversely as the stream discharge recedes to 15.0 m /^s or below, big cobbles stop moving, and voids are formed between them. Particles <0.42 mm that fill in the interstices become trapped. At discharge below the critical threshold (10.0 m-Vs), coarse particles stop moving, and the pavement is reestablished, so the flow is able to carry only sediments < 10 mm in size (Fig. 7.5). (Sxui/S) 3]BJ yodsucjx (SxUi/3) 0]CJ yodsunji 185 10 X I o 0.1 0.01 '-L_l : l i p ; • • • -~& -L _ — 10 15 Discharge (mVs) -90.5+64 mm Fig. 7.4. Transport rate of particles of-91+64 mm on Harris Creek in 1993. 30 I 1 10 100 Grain size (mm) Fig. 7.5. Texrural variations of sediment in various stages of flow conditions with Helley-Smith sampler on Harris Creek. (The dotted lines are for discharges below 10 mVs. The solid lines are for discharges above 10 m7s). 187 7.3.2 Magnetite The transport rates of magnetite in the three size fractions respond similarly to changing discharge (Figs. 5.8 to 5.10 and Table 5.3). The reason why the magnetite in the three size fractions moves in similar way is not completely clear yet and seems likely because magnetite of the three size fractions is similarly "hidden" from the flow so that they have similar possibility to be transported by a flood. Along lower Harris Creek, the environment most favorable to trap heavy minerals is void structures, which are particularly developed at bar head sites (Fig. 2.3 and Plate 2.1). As flow falls, whether magnetite is trapped into the voids at bar head or deposited at bar tail sites is controlled by its grain size. Magnetite in the -0.149+0.105 mm size fraction, presumably because it travels lower in flow and tests the bed more often (Day and Fletcher, 1989), preferentially accumulates in the voids at bar heads - as shown by the GMCR ratio greater than 1 (Table 6.4). Magnetite in the -0.105+0.075 mm and -0.075+0.053 mm size fractions, with GMCRs equal to or less than 1, is not preferentially accumulated in the voids at bar heads. 7.3.3 Gold Interpretation of the behavior of gold is more problematic than magnetite because of the poor resolution of the data. Because no gold is detected in most Helley-Smith samples, the results can be interpreted as: (1) the gold is not transported and the few gold data reported are anomalous highs; (2) the gold is transported episodically, and the reason that most Helley-Smith samples do not have gold is because the gold is not released during the sampling intervals; (3) gold is systematically but infrequently transported, leading to the strongly skewed pattern, with a preponderance of zero values, such that results follows the Poisson distribution (Ingamells and Pitard, 1986). 188 Gold transport depends on discharge with the gold mainly being transported during (brief) periods of high discharge in the spring snowmelt flood (Section 4.3.4 and Section 5.5). The entrainment and movement of gold particles is complex and seems to be controlled by three categories of variables: sedimentological characteristics (texture and pavement), hydraulic conditions (discharge), and gold particles. Once discharge is around the critical discharge to disturb and break the pavement, both the coarse (-0.149+0.053 mm) and fine gold (-0.053 mm) are released from the bed. The fine gold seems to be immediately carried downstream in suspension since the concentration of the -0.053 mm gold collected using the Helley-Smith sampler decreases with the increasing discharge up to 20 m3/s (Fig. 7.6). For the coarse gold, the major release period seems to be at discharge above 15.0 m-Vs under which cobbles start to move (Fig. 7.6). The reason for the increase of the fine gold at discharge above 23.0 nvVs is not understood yet but possibly because of the complete activation of the bed or movement of boulders. The results of gold distribution along lower Harris Creek indicate that coarse gold is much more preferentially trapped than fine (-0.053 mm) gold at bar head sites (Table 6.3), and thereby has a greater GMCR ratio indicating its greater enrichment. That is, coarse gold moves in close contact with the bed and easily falls into the voids. Conversely, fine gold has a GMCR < 1 indicating that it is swept into suspension over bar heads and carried to bar tails or the overbank environment to be deposited in lower velocity regimes (Fig. 7.7). Because the framework cobbles are too large to be in settling or entrainment equivalence to the gold particles, Reid and Frostick (1985) refer to the process as "interstice entrapment". In summary, findings in lower Harris Creek are consistent with the traditional guidelines of placer miners and prospectors for finding accumulations of heavy minerals in bar head gravels, bedrock riffles, scour holes and, as described by Smith and Beukes (1983), in flow convergence zones. Pebble lags and constricted channel-ways are also recognized as controls on accumulation o C O o o o o o o uo CN o CN oo / > I / O U Q o C O CN (qdd) uonanuaoucQ CD 00 43 o T3 O c •a - a o 00 cn C o c5 u . +—' c CD O c o o o c o I vd ob 1 9 0 \ \ \ ^SKs \ Coarse gold particles Fine gold particles i i i i i i i i i i is characterised with high flow velocity and high roughness. Coarse gold particles are trapped behind or beneath cobbles and gravels (voids). is characterised with low flow velocity and low roughness. Fine gold particles are carried in suspension and then deposited at bar tail sites. F ig . 7.7. Schematic diagram showing sedimentological behaviour o f go ld particles in defferent environments along lower Harris Creek. 191 of gold in the Witwatersrand paleoplacers deposited in braided river systems (Smith and Minter 1980). The results are also consistent with theoretical models of transport and deposition of gold and other heavy minerals by streams (Day and Fletcher, 1991). These results also account for the concentration of gold and magnetite in the secondary sources being considerably higher than in the associated primary landslide sediment (Figs. 4.14 and 4.15). Based on the 18-year records (Table 6.1), Harris Creek has 44% of the years to experienced floods similar to 1991 flow condition, and 27% of the years to experience floods similar to 1992 flow condition. Under these conditions, bank erosion and landslides are important suppliers, the storage order is gold>magnetite>sediment. Therefore, Harris Creek resembles a conveyor belt: it is still redistributing sediment that was delivered to the valley by the Fraser Glaciation 10 kyr ago. During this process, Harris Creek transports more sediment out of the basin and leaves more gold behind so that gold is concentrated in the secondary sources. The important implication is that the bank material along Harris Creek is not a dilutant of but a concentrator of gold. When gold is delivered to Harris Creek, coarse gold is preferentially accumulated at the bar heads, and fine gold is preferentially accumulated at bar tails. The accumulation of the coarse gold at the bar head gravels tends to counteract downstream dilution so that gold values actually increase downstream (Fig. 6.15). Results reported in this thesis confirm that preferential accumulation sites occur at breaks in stream gradient throughout lower Harris Creek. 7.4 Behavior of Gold between Floods 7.4.1 Sediment and magnetite The results presented in Chapters 4 and 5 convincingly demonstrate that, although there are errors in the sediment and heavy mineral budgets (see Chapters 4, 5 and 6), input and output of sediment and magnetite vary considerably from year to year. The coincidence of input and output 192 with floods strongly supports the claim that flood events are responsible for almost the entire process. As stated previously, sediment and magnetite do not come from the headwaters area on the Okanagan Highland, but mainly come from sources flanking lower Harris Creek. The relative importance of different sources, and absolute amount of input and output strongly depend on the intensity and duration of discharge in lower Harris Creek. In a dry year, stream flow is not high enough and does not last long enough to penetrate and soak banks to cause instability. Banks therefore do not supply significant sediment and magnetite to lower Harris Creek in such a year. Log jams and landslides are the chief contributors although the absolute input amounts are very small (Fig. 6.2). Another feature of a dry year is that some sediment and magnetite will be mobilized and transported from the stream bed. In a normal or wet year, because the entire channel of lower Harris Creek is occupied by stream water, and because the high flow lasts for a relatively long time, banks are fully penetrated and soaked so that slumping and collapse cause bank erosion. The laterally unstable reaches then become the most important source for sediment and magnetite. In a wet year landslides are also important to sediment and magnetite budgets. 7.4.2 Gold Field panning and chemical analysis showed that there are most likely two tributaries that are close to bedrock sources in the Harris Creek basin: one is McAuley Creek and one is Mosquito Creek. The budget results indicated that flows in McAuley Creek during the study were not high enough to mobilize gold from the bed, whereas Mosquito Creek was found to transport gold to the trunk stream. Delivery of gold from Mosquito Creek to Harris Creek depends on the discharge in Mosquito Creek, while the downstream transportation of the gold depends on the discharge in lower Harris Creek. The output of gold also depends on the discharge in lower Harris Creek. Therefore, there are several possible scenarios for downstream distributions of gold in lower Harris Creek: 193 Scenario 1: If the discharge in Mosquito Creek is above 3.0 m-Vs, but the discharge in lower Harris Creek is less than the 10.0 mVs as in 1992 and five other years in the 18-year records, the coarse gold added from Mosquito Creek can not be transported down Harris Creek and will be locally stored in alluvial fans or gravel bars. A point Au anomaly or a short anomaly dispersion train would be expected to form just below Mosquito Creek. Scenario 2: If the discharge in Mosquito Creek is above 3.0 m-Vs, and flood flow in lower Harris Creek is average (> 10.0 m-Vs) and lasts long enough, the gold introduced from Mosquito Creek will continue to be transported downstream together with the fine gold supplied from bank erosion. As the flood falls, the coarse gold will preferentially accumulate at bar head sites and fine gold at bar tail sites. This preferential accumulation can counteract downstream dilution so that Au concentration can actually increase downstream (Figs. 6.15, 17 and 18). As discussed above, because coarse gold does not preferentially accumulate at bar tail sites, concentrations of this gold at these sites should more closely reflect the bulk composition of sediments being transported past the site. Therefore, concentrations of coarse gold at bar tail sites should decrease downstream. The downstream dilution corresponds roughly to the geochemical traditional models (Polikarpochin, 1971; and Hawkes, 1976). Au geochemistry in stream sediment in 1986 could be an example for this situation. In that year the spring snowmelt flood had a peak discharge of 37 m-Vs with 25 days > lOnvVs (Fletcher and Wolcott, 1991). Geochemical studies in a 5 km reach upstream of the Research Site found that concentration of coarse Au at bar tail sites shows significant downstream decrease with distance at a = 0.05. In contrast, coarse gold at bar head and fine gold at bar head and bar tail sites did not decrease downstream with distance at a - 0.05 (Fig 7. 8). 194 1,000 3 0 0 100 3 0 10 3 1 0.3 0.1 •••• m' ->s © 2,000 4 ,000 D i s t a n c e d o w n s t r e a m ( m ) B a r h e a d B a r t a i l • e 6,000 Concentrations of the coarse gold at bar tails = -0.015 xDi stance + 57.02 Concentrations of the coarse gold at bar heads = 204.28 Both equations are significant at 95% confidence interval. 1,000 r 3 0 0 : 100 -1 0.3 h 0.1 I 1 1 ' 1 ' 0 2 ,000 4 ,000 6 ,000 D i s t a n c e d o w n s t r e a m ( m ) B a r h e a d B a r t a i l • e Concentrations of fine gold at bar tails = 23.34 Concentrations of fine gold at bar heads = 20.03 Both equations are statistically significant at 95% confidence intervals. Fig. 7.8. Downstream profile of gold in stream sediment in lower Harris Creek in the year with extremely high flood. Modified after Day (1988). 195 Scenario 3 If discharge in Mosquito Creek is below 3.0 m-Vs, but the flow in the trunk channel is high (>10.0 m-Vs) for sufficient time, no gold is delivered from Mosquito Creek, but is supplied from unstable banks and landslides. The high flow mobilizes the bed pavement and releases gold from the stream bed, so that gold migrates downstream to a favorable region, such as breaks-in-slope of the stream gradient, where coarse gold is trapped in the voids at bar heads, and fine gold particles are deposited in bar tail sites. Because no gold is added from a gold source, concentration of coarse gold at bar tail sites will not show the downstream decaying pattern. The longitudinal profile of Au in the stream sediment in 1991 most probably describes this situation (Figs. 6.15 to 6.18): No cut-off point of Au anomalies occurred below Mosquito Creek. Coarse gold preferentially accumulates at bar heads, and fine gold at bar tails. Both displayed high values at breaks-in-slope of stream gradient. Concentrations of coarse gold in bar tails show no clear downstream dilution trend. 7.5 "Knick Point" and Gold Anomalies: Long Time Evolution The longitudinal profile of streams is typically concave upward and exponential in form (Fig. 7.9 A), even though some floodplains exhibit marked local convexity (Fig. 7.9B). The concavity, however, is usually not perfectly smooth in detail but is commonly interrupted by perturbations. These can be caused by reaches where the channel is floored by bedrock, or by local zones of erosion or deposition. Local filling may be initiated by an influx of bank material load that is too coarse or too great in volume to be transported on the preexisting gradient. Coarse or cohesive material, for example, may be introduced to a trunk channel from a local spot where active undercutting is occurring. The addition of the load requires deposition downstream until the local gradient is increased sufficiently to allow the bedload to be transported. In contrast, local scour is possible when there is an excess of stream competence and transport ability relative to available load (Butzer, 1976). 196 Fig. 7.9. Stream profiles: (A) Concave, (B) Concavo-convex, and (C) Interrupted by Knick points (from Butzer, 1976; Shadowed curves are original surface) 197 Few stream profiles are smoothly concave (or concavo-convex). Instead, there generally are inflection points, known as Knick points (Fig. 7.9C) which usually have lower downstream gradient and an oversteepened segment. As the Knick point migrates in a headward direction, erosion may begin upstream. Both flume and field data suggest that in channels formed of cohesionless material, pronounced knickpoints will be smoothed out after only a short distance of upstream migration (Brush and Wolman, 1960; Morisawa, 1964). Where the channel is composed of bedrock or cohesive sediments (such as silty and clay sediments that are very common in the Harris Creek basin), the knick point may retreat upstream for a considerable distance and can still be preserved. Study results in Harris Creek found that coarse gold preferentially accumulates at bar head sites and fine gold preferentially accumulates at tail sites of the same bars that are associated with breaks in stream gradient throughout lower Harris Creek (Figs. 6.15 and 6.18). Results confirm the field observations and theoretical models of transport and deposition of gold and other heavy minerals by streams (Day and Fletcher, 1989 and 1991). It is also apparent that two of the four active landslides, Landslides #1 and #7, coincide with major breaks in stream gradient where landslide activity is triggered by lateral instability and bank erosion of the stream. This association of channel instability and landslides with change in stream gradient may be coincidental. Alternatively, all three phenomena may be linked by geomorphic processes on either a local or catchment-wide scale by, for example, either (i) the knick-point, created by changes in stream gradient, retreating upstream (Hou and Fletcher, 1995 and 1996) or (ii) the excess energy of the stream, associated with the decrease in gradient at the knick-point, being partially dissipated by lateral migration of the channel and erosion of banks resulting in landsliding. In the case that the knickpoint is smoothed out after only a short distance of upstream migration, the cut-off points of the Au anomaly may also disappear. In the case that the knickpoint retreats for a considerable distance and is still preserved, the cut-off may be also preserved. As we know, a stream system always adjusts itself to a particular pattern of vegetation and climate, on the one hand, and of 198 lithology and structure, on the other. Changes in one or more of these variables will require major readjustments within the stream system, such that new knickpoints are formed, which are accompanied by formation of new cut-offs. Such processes keep going until significant external or internal changes (such as a new glacial period or volcanism) occur, or as described before, until the redistribution of the glacial sediment is dissipated. Ultimately, a new form of adjustment will start, with a different regime of erosion, transportation and deposition. Because of the regional perturbation of the Fraser Glaciation in the Harris Creek basin, it is believed that Harris Creek has to adjust itself to the particular pattern of source sediment distribution. Therefore, considering the long time-scale of the geomorphic processes involved in the evolution of a drainage basin after deglaciation, there are many possibilities to form knick points, but almost no possibility to prove or disprove such relationships in the short term. Nevertheless the association of two of four active landslides with breaks-of-slope in stream gradient suggests that the association may be more than a coincidence, and it could be a result of evolution of the Harris Creek basin. Whatever the relation between the changes in stream gradient, landslides and knick points, and whatever the processes responsible for it, the low gold content of the landslides (Table 6.3) and geochemical fingerprinting of material eroded from the Landslide #1 clearly show that the landslide will dilute Au concentrations in the -0.053 mm fraction of the stream sediments. However, because of the cohesive nature of the clay balls (Plate 6.1), this dilution only becomes apparent at low energy sites some distance downstream from the landslide. The combined effect of accumulation of gold at breaks-in-slope and dilution by material derived from the landslide, is therefore to create an apparent cut-off for the Au anomaly a short distance downstream of the landslide, and the cut-off may migrate upstream concomitant with migration of the knick point. 199 CHAPTER 8 CONCLUSIONS Based on the budgets of gold and magnetite, and their distributions along Harris Creek in British Columbia, my thesis comes to the following conclusions: Budget Results 1. The Harris Creek basin is an internally active system in which all the contemporary sediment sources discontinuously flank Harris Creek. This situation differs markedly from the relatively simple drainage basin model that assumes a uniform, continuous distribution of sediment sources throughout the catchment. The sediment sources are classified into primary and secondary sources based on whether they have been reworked by current fluvial processes. In the primary sources, magnetite is rather uniform and low in concentration. Concentrations of magnetite in the secondary sources are about twice as higher as in the primary sources. Similarly, gold in secondary sources is much higher. On average, concentrations of Au in the secondary sediments is about 5 times higher in the -0.149+0.053 mm size fraction and 20 times higher in the -0.053 mm size fraction than in the corresponding size fraction of the primary sediments. 2. The snowmelt flood that usually occurs in late spring is the most important sediment transport process and contributes input, storage and output of sediment, magnetite and gold. In years like 1992, stream water is not high enough and does not last long enough to penetrate and soak banks to cause instability, so that banks are not eroded; landslides are stable; and tributaries do not transport significant amount of sediment, magnetite and gold. In such years, input and output are both minor. In normal years like 1991, because the entire channel is occupied by stream water, banks are fully penetrated and soaked so that laterally unstable reaches become 200 the most important source for sediment, magnetite and gold. In a year like 1993, very high flood triggers landslides, therefore, landslides become the second most important contributor. Remobilization of sediment sources along the Harris Creek valley and Harris Creek banks dominates over the erosion of the headwater land surface of the Harris Creek basin. An implication of this analysis is that the headwaters area of the Harris Creek drainage does not contribute gold to Harris Creek. These observations show that the spatially discontinuous nature of sediment supply within the fluvial system conflicts with the conventional uniform erosion rate. 3. Supply of gold to Harris Creek is discontinuous both spatially and temporally. In both dry and normal years, gold is not delivered into and transported out Harris Creek. In a wet year, gold is transported only for brief periods during very high floods. The fact that more sediment is transported out, and more gold is left behind makes Harris Creek concentrate gold in fluvial sediments. Distributions of Gold and Magnetite in Stream Sediment along Harris Creek 1. Magnetite in the -0.149+0.105 mm size fraction preferentially accumulates in the high energy environment at the bar heads - as shown by a GMCR ratio greater than 1. Magnetite in the -0.105+0.075 mm and -0.075+0.053 mm size fractions, with GMCRs equal to or less than 1, is preferentially accumulated in the low energy environment at the bar tails and in overbank deposits. 2. Gold in the -0.149+0.053 mm size fraction, with the GMCR ratio much greater than 1, is preferentially accumulated in the voids at the bar heads. The gold shows anomalous values at bar heads which are coincidental with breaks-in-slope where the stream gradient decreases. The accumulation of the gold in the voids at the bar heads tends to counteract downstream dilution so that concentrations of gold do not generate a smooth hyperbolic decaying curve, they actually increase downstream along the thirteen km study reach. 201 3. Gold in the -0.053 mm size fraction, with a GMCR < 1, is swept into suspension over bar heads and carried to bar tails or the overbank environments to be deposited in lower velocity regime. The fine gold shows peak values at the bar tails which, similar to the gold in the -0.149+0.105 mm size fraction, are coincidental with the breaks-in-slope where the stream gradient decreases. 4. Landslides are important sediment sources but contain low concentration of gold that dilute Au concentration in the stream sediment. But because most landslides contain cohesive glaciolacustrine sediment, and the cohesive sediment is not easily broken down, therefore the dilution becomes apparent only some distance downstream from the landslide. The combined effect of accumulation of gold at breaks-in-slope and dilution by material derived from the landslide creates an apparent cut-off for the Au anomaly a short distance downstream of a landslide. An implication of these analyses is that the results obtained here do not match the assumptions presented in the traditional hyperbolic downstream decaying model (see Chapter 1), by which it is not meant that the traditional model is wrong, but that it is pointed out that understanding the fundamental processes in a drainage basin is the basis for successful and realistic application of a theoretic model. I do not claim to have achieved a new predictive model, nor did I set out to do so. But based on these results I got from Harris Creek, the interactions of episodic sediment sources and sedimentological processes can produce complex downstream distribution patterns for gold in stream sediment. Insofar as these interactions can create the appearance of Au anomalies and anomaly cut-offs, and higher concentrations of gold in stream sediment than in the material derived from till and other colluvial slopes, the effects of landslides, discontinuous transport of gold and changes in stream gradient should be taken into account when interpreting exploration geochemical data. Because both episodic landsliding and sedimentological 202 processes are common in areas like the Harris Creek basin, it is reasonable to expect that their influence on stream sediment geochemistry may be a correspondingly common occurrence. The relation of evolution of gold anomaly with basin-scale geomorphic processes is an obvious avenue for future research. 203 R E F E R E N C E S Bjorklund, A., Lehmuspelto, P., Ottesen, R. T. and Steefelt, A., 1994. Drainage geochemistry in glaciated terrain. In: Drainage Geochemistry edited by Hale, M. and Plant, J. A. Handbook of Exploration Geochemistry, Vol. 6 (Govett, G. J. S., editor). Elsevier, p. 307-340. Bogen, J., 1992. Environmental studies in Western Europe using overbank sediment, in: Erosion and Sediment transport Monitoring Programs in River Basins (Proceedings of the Oslo Symposium, August 1992) IAHS Pub. no. 210. p. 317-325. British Columbia Forest Service, 1972. Forest cover map 82-L-2-E. Forest Inventory Division, Ministry of Forests, British Columbia. Brush, L. M. and Wolman, M. G., 1960. Knickpoint behavior in non-cohesive material: a laboratory study. Geol. Soc. America Bull. 71: 59-74. Butt, C. R. M. and Smith, R. E., 1980. Conceptual models in exploration geochemistry -Australia: J. Geochem. Explor. 12(2/3) 89 - 365. Butzer, K. W., 1976, Geomorphology From The Earth. Harper & Row, Publishers. 461 pp. Church, M., 1980. Records of recent geomorphological events. In: R. A. Cullingford, D. A. Davidson, and J. Lewin, eds., Time Scales in Geomorphology, 13-30. Church, M., 1983. Concepts of sediment transfers and transport on the Queen Charlotte Island, British Columbia. Fish/Forestry Interaction Program. Ministry of Forests. 30 pp. Church, M., 1993. Harris Creek ratings. 15 pages, unpublished. Church, M., Kellerhais, R. and Day, T. J., 1989. Regional clastic sediment yield in British Columbia. Canadian Journal of Earth Sciences, 26: 31-45. Church, M. and Slaymaker, O., 1989. Disequilibrium of Holocene sediment yield in glaciated British Columbia. Nature, Vol. 337, No. 6202, pp. 452-454. Clifton, H. E., Hunter, R. E., Swanson, F. J. and Phillips, R. 1., 1969. Sample size and meaningful gold analysis: U. S. Geol. Surv., Professional Paper 625-C, 17 pp. Day, S., 1988. Sampling streams for gold in mineral exploration, south British Columbia. M. Sci. thesis. University of British Columbia, Vancouver, Canada. 233 p. Day, S. and Fletcher, W. K., 1986. Particle size and abundance of gold in selected stream sediments, Southern British Columbia, Canada. J. Geochem. Explor. 26: 203-214. Day, S. and Fletcher, W. K., 1989. Effects of valley and local channel morphology on the distribution of gold in stream sediments from Harris Creek, British Columbia. J. Geochem. Explor. 32, 1-16. Day, S. and Fletcher, W. K., 1991. Concentration of magnetite and gold at bar and reach scales in a gravel-bed stream, British Columbia. J. Sed. Petrol. 61, 871-882. Dietrich, W. E. and Dunne, T., 178. Sediment budget for a small catchment in mountainous terrain. Z. Geomorph. N F. Suppl. Bd. 191 - 206 204 Emmett, W. W., 1980. A field calibration of the sediment-trapping characteristics of the Helley-Smith bedload sampler. U. S. Geological Survey Professional Paper 1139, pp. 44. Environment Canada, 1982. Monthly record meteorological observations in Canada. Atmospheric Environment Service, vol. 67, 282 pages. Environment Canada, 1987. Surface water data British Columbia 1987. Inland waters Directorate. Water Survey of Canada, Ottawa, 316 pp. Fletcher, W. K., 1981. Analytical methods in geochemical prospecting. Elsevier, 255 pp. Fletcher, W. K., 1990. Dispersion and behavior of gold in stream sediments. Open File 1990-28, British Columbia Geological Survey, pp. 28. Fletcher, W. K., 1994. Behavior and exploration geochemistry of cassiterite and other heavy minerals in streams. Journal of Southeast Asian Earth Sciences, Vol. 10, No. 1/2, pp. 5-10. Fletcher, W. K., Church, M. and Wolcott, J., 1992. Fluvial transport equivalence of heavy minerals in the sand size range. Can. J. Earth Sci., 29, 2017-2021. Fletcher W. K. and Day, S., 1989. Behavior of gold and other heavy minerals in drainage sediments: some implications for exploration geochemical surveys. Trans. Inst. Min. Metall. 98, B130-B136. Fletcher, W. K., Dousset, P. E. and bin Ismail, Y., 1987. Elimination of hydraulic effects for cassiterite in a Malaysian stream. J. Geochem. Explor. 28, 385-408. Fletcher, W. K., Lahiri, R., Caughlin, B. L. and Block, H., 1995. Use of a sensitive analytical method and the silt-clay (<53 m) fraction of stream sediments in exploration for gold in northern Thailand. J. Geochem. Explor. 55: 301-307. Fletcher, W. K. and Loh, C. H., 1996a. Transport equivalence of cassiterite and its application to stream sediment surveys for heavy minerals. J. Geochem. Explor. 56, 47-57. Fletcher, W. K. and Loh, C. H., 1996b. Transport of cassiterite in a Malaysian stream: implication for geochemical exploration. J. Geochem. Explor. 57, 9 - 20. Fletcher, W. K. and Loh, C. H., 1997. Transport and deposition of cassiterite by a Malaysian stream, in press. Fletcher, W. K. and Wolcott, 1991. Transport of magnetite and gold in Harris Creek, British Columbia, and implications for exploration. J. Geochem. Explor. 41, 253-274. Garrett, R. G., 1979. Sampling considerations for regional geochemical surveys. In: Current Research, Part A. Geol. Surv. Can., Paper, 79-1A: 197-205. Garrett, R. G., 1983. Sampling methodology, in: Statistics and Data Analysis in Geochemical Prospecting, edited by R. J. Howarth. p. 83-110. Garrett, R. G. and Goss, T. I., 1978. The evaluation of sampling and analytical variation in regional geochemical surveys. In: Waterson J. R. and Theobald P. K. (editors), Geochemical Exploration 1978. Association of Exploration Geochemists, Rexdale, Ont. p. 371-383. 205 Harris, J. F., 1982. Sampling and analytical requirements for effective use of geochemistry in exploration for gold. In: Precious Metals in the Northern Cordillera, pp. 53 - 67; The Assoc. of Explor. Geochemists, Spec. Vol. 10. Hawkes, H. E., 1976. The downstream dilution of stream sediment anomalies. Journal of Geochemical Exploration Vol. 6, p. 345-358. Hill, A. R., 1973. Erosion of river banks composed of glacial till near Belfast, Northern Ireland. Z. Geomorph. N. F. 17 (4), 428-442. Horton, R. E., 1945. Erosional development of streams and their drainage basins: hydrophysical approach to quantitative morphology. Bull. Geol. Soc. Am., 56: 275-370. Hou, Z. and Fletcher, W. K., 1991. Distribution and morphological characteristics of visible gold in Harris Creek (82L/2). In Geological Fieldwork 1991, Paper 1992-1. B. C. Ministry of Energy, Mines and Petroleum Resources, pages 319-323. Hou, Z. and Fletcher, W. K., 1996. The relations between false gold anomalies, sedimentological processes and landslides in Harris Creek, British Columbia, Canada. Journal of Geoche. Explor. 57 21-30. Howarth, R. J. and Lowenstein, P. L., 1971. Sampling variability of stream sediments in broad-scale regional geochemcial reconnaissance. Trans. Inst. Min. Metall., Ser. B, 80: 363-372. Ingamells, C. O. and Pitard, F. F.,1986. Applied Geochemical Analysis. Wiley-Interscience publication. New York: Wiley. 733 pp. Jackli, H., 1957. Gegenwarts geologie des Bundnerischen Rheigebietes. Beitrage Zur Geologie der Schweiz, Geotech, Series No. 36. Jones, A. G , 1959. Vernon map-area, British Columbia. Geol. Surv. Can. Mem. 256. Kelley, C. C , and Spilsbury, R. W., 1949. Soil survey of the Okanagan and Similkameen valleys, British Columbia. B. C. Dept. Agric. in cooperation with Experimental Farms Service, Dominion Depart. Agric. 88 pages. Kelsey, H. M., 1980. A sediment budget and an analysis of geomorphic process in the Van Duzen River basin, north coastal California, 1941-1975: Summary. Geological Society of America Bulletin, Part I, Vol. 91, p. 190 - 195. Klingeman, P. C. and Emmett, W. W., 1987. Gravel bedload transport processes, in: Thorne, C. R., Bathurst J. C and Hey, R. D., eds., Sediment Transport in Gravel-Bed Rivers, John Wiley and Sons Ltd., Chichester, 141-179. Krumbein and Graybill, 1965. An introduction to statistical models in geology. Mcgraw-Hill Book Company. 475 p. Krumbein, W. C. and Slack, H. A., 1956. Statistical analysis of level radioactivity of Pennsylvanian black fissile shale in Illinois. Geol. Soc. Am. Bull. 67(6): 739-762. Lehre, A. K., 1982. Sediment mobilization and production from a small mountain catchment: Lone Tree Creek, Marin County, California, Ph. D. Thesis, University of California, Berkeley, 375 pp. 206 Leopold, L. B., Emmett, W. W., and Myrick, R W., 1966. Channel and hillslope processes in a semi-arid area, New Mexico. U. S. Geological Survey Professional Papers, 352-G. Mansikkaniemi, H., 1975. Monthly sedimentation in some reservoirs of hydroelectric stations in Finland. Fennia, 143: 38. Mathews, W. H., 1988. Neogene geology of the Okanagan Highland, British Columbia. Canadian Journal of Earth Sciences, 25, 725-731. Miesch, A. T., 1976. Geochemcial survey of Missouri - methods of sampling, laboratory analysis and statistical reduction of data. U. S. Geol. Surv. Prof. Paper, 954-A: 39 pp. Milhous, R. T. and Klingeman, P. C , 1973. Sediment transport system in a gravel-bottom stream. In: Hydraulic Engineering and the Environment, Proceedings of the 21st Annual Hydraulics Division Specialty Conference. American Society of Civil Engineers Boztana, Montana, pp. 293 -Morisawa, M. E., 1964. Development of drainage systems on an upraised lake floor. Am. Jour. Sci. 262: 340-354. Nelson, T. J, 1993. A geochemical comparison of tills overlying four different bedrock units at Harris Creek, British Columbia. 51 pp. B. Sc. Thesis of the University of British Columbia. Nichol, I., 1983. Geochemical exploration for gold: A special problem, in Applied Geochemistry edited by Thormton, I and Howarth R. J. p. 61-85. Notebook of Geog. 405 - Fluvial geomorphology, 1991. Department of Geography, The University of British Columbia. Paopongsawan, P. and Fletcher, W. K., 1993. Distribution and dispersion of gold in point bar and pavement sediments of the Huai Hin Laep, Loei, northern Thailand. J. Geochem. Explor. 47, 251-268. Peart, M. R. and Walling, D. E., 1986. Fingerprint sediment source: the example of a drainage basin in Devon, UK. In: R. F. Hadley (Editor), Drainage Basin Sediment Delivery. IAHS Publication No. 159, pp. 41-55. Phillips, J. D., 1991. Fluvial sediment budgets in the North Carlina Piedmont. Geomorphology, 4, 231 -241. Polikarpochikin, V. V. 1971. The quantitative estimation of ore-bearing areas from sample data of the drainage system. In: R. F. Boyle and J. J. McGerrigle (Editors), Geochemical Exploration, CIM Special Volume 11, pp. 41 - 55. Rapp, A., 1960. Recent development of mountain slopes in Karkevagge and surroundings. Geografiska Annaler, Vol. 42A, 65-200. Reid, I and Frostick, L. E., 1985. Role of settling, entrainment and dispersive equivalence and of interstice trapping in placer formation. Journal of geological Society London, Vol. 142, pp. 739 -' I A S ' Roberts, R. G. and Church, M., 1986. The sediment budget in severely disturbed watersheds, Queen Charlotte Ranges, British Columbia. Canadian Journal of Forest Research 16(5) 1092-1106. 207 Resources Inventory Committee, January 1996. Guidelines and Standards to Terrain Mapping in British Columbia. Surfical Geology Task Group, Earth Sciences Task Force, British Columbia. 216 pp. Rose, A. W., Hawkes, H. E. and J. S. Webb, 1979. Geochemistry in mineral exploration. Academic Press, pp. 657. Ryder, J., 1989. Climate, Chapter 1 in Quaternary Geology of Canada and Greenland, R. J. Fulton (ed.); Geological Survey of Canada, Geology of Canada, no. 1.p. 26 - 31. Ryder, JR., 1991. Quaternary geomorphology and Terrain map of the Harris Creek area, B. C. (Parts of NTS 82L/1&2 Scale: 1:20,000). Open File 1991-8. B. C. Mineral Resources Division, B. C. Ministry of Energy, Mines and Petroleum Resources. Ryder, J. R. and Fletcher, W. K., 1991. Exploration geochemistry - Sediment supply to Harris Creek (82L/2). B. C. Ministry of Energy, Mines and Petroleum Resources, Geological Fieldwork 1990, Paper 1991-1, papers 301-306. Ryder, J., Hou, Z. and Fletcher, W. K., 1993. Interpretive terrain maps and exploration geochemistry in glaciated Cordilleran regions. In: S. B. Romberger and D. I. Fletcher (editors), Integrated Methods in Exploration and Discovery (Conference Program and Extended Abstracts), Denver, Co, pp. AB101-AB102. Ryder, J., and Howes, D. E., 1986. Terrain Information: A User's guide to Terrain maps in British Columbia. B. C. Ministry of Environment and Parks. 16 p. Saxby, D. and Fletcher, W. K., 1986a. Behavior of scheelite in a Cordilleran stream. In: GEOEXPO '86: Exploration in the North American Cordillera, (edited by Elliott, I. L. and Smee, B. W.), Rexdale, Ont., Association of Exploration Geochemists, p. 177-183. Saxby, D. and Fletcher, W. K., 1986b. The geometric mean concentration ratio (GMCR) as an estimator of hydraulic effects in geochemical data for elements dispersed as heavy minerals. J. Geochem. Explor., 26: 223-230. Sinclar, A. J., 1976. Application of probability graphs in mineral exploration. The Association of Exploration Geochemists, Spec. Vol. 4, 95 pages. Sinclar, A. J., Fletcher, W. K., Bentz A., Price, B. A. and Wong, S. S., 1977. Minor elements in pyrites from some porphyry-type deposits, British Columbia. Trans. A.I.M.E., 262:94-100. Slaymaker, O., 1993. The sediment budget of the Lillooet River basin, British Columbia. Physical Geography, 14(3), pp. 304-320. Slaymaker, O. and Mcpherson, H. J., 1977. An overview of geomorphic processes in the Canadian Cordillera. Zeitschrift fur Geomorphologie, 21: 169-186. Slingerland, R and Smith N. D. 1986. Occurrence and formation of water-laid placers. Ann. Rev. Earth Planet. Sci. 14, 113 - 147. Smith, N. D. and Beukes, N. J., 1983. Bar to bank flow convergence zones: a contribution to the origin of alluvial placers. Econ. Geol. 78, 1342 - 1349. 208 Smith, N. D. and Minter, W. E. L., 1980. Sedimentological controls of gold and uranium in two Witwatersrand paleoplacers. Econ. Geol. 75, 1-14. Stanley, C. R., 1989. Nugget representativity calculator. MDRU, Dept. Geol. Sci. UBC, Vancouver, Canada. Sutherland, R. A. and Bryan, R. B., 1991. Sediment budgeting: a case study in the Katiorin drainage basin, Kenya. Earth Surface Processes and Landforms, Vol. 16, 383-398. Swanson, F. J., Janda, R. J. Dunne, T., and Swanston, D., eds., 1982. Sediment budgets and routing in forested drainage basins. U. S. Department of Agriculture, Forest Service, General Technical Report PNW-141. Thorne, C. R. 1978. Processes of bank erosion in river channels Ph. D. thesis, University of East Anglia, Norwich. 447 p. Trimble, S. W., 1983. A sediment budget for Coon Creek basin in the driftless area, Wisconsin, 1853-1977. American Journal of Science, Vol. 283, p. 454-474. Twidale, C. R., 1964. Erosion of an alluvial bank at Birdwood, South Australia. Z. f. Geomorph., N. F., 8: 189-211. Wolman, M. G., 1959. Factors influencing erosion of a cohesive river bank. Am. J. Sci., 257: 204-216. Zar, J. H., 1984. Biostatistical analysis. Prentice-Hall Inc., 718 pages. 209 APPENDIX Appendix Table 1. Mosquito Creek discharge measurements Date/time Discharge (mVs) Staff gauge (m) In 1992 May 10/11:30 0.296 0.640 May 12/09:55 0.240 0.620 May 13/08:50 0.203 0.615 May 15/09:20 0.180 0.605 May 17/08:50 0.172 0.600 May 20/09:15 0.140 0.588 May 21/08:49 0.147 0.590 May 22/09:05 0.136 0.588 May 23/08:50 0.124 0.585 May 24/08:45 0.108 0.582 May 25/08:45 0.088 0.579 May 26/10:10 0.240 0.625 May 29/18:24 0.190 0.610 In 1993 April 28/08:05 0.400 0.660 April 29/08:12 0.394 0.655 Apnl 30/08:09 0.660 0.700 May 01/08:20 0.710 0.715 May 02/08:15 0.671 0.710 May 03/08:11 0.642 0.705 May 04/08:10 0.616 0.700 May 05/10:03 0.910 0.725 May 05/15:15 1.010 0.745 May 06/08:09 1.190 vj.760 May 07/12:22 1.140 0.755 May 09/14:04 1.100 0.725 May 10/12:15 1.110 0.738 Appendix Tabic 1. Mosquito Creek discharge measurements Date/time Discharge (mVs) Staff gauge (m) In 1993 M a y 14/11:00 2.660 0.970 M a y 16/12:49 1.900 0.890 M a y 17/08:37 1.480 0.865 M a y 18/11:25 1.310 0.835 M a y 19/08:09 0.860 0.825 M a y 22/16:20 0.740 0.785 M a y 23/16:28 0.641 0.775 May 24/16:20 0.429 0.755 M a y 25/16:17 0.428 0.745 May 26/15:34 0.405 0.745 M a y 28/15:25 0.380 0.735 M a y 30/15:20 0.310 0.725 June 06/12:35 0.280 0.715 Appendix Table 2. 1992 and 1993 discharge rating curve measurements in McAuley Creek Date/time Discharge (m3/s) Staff gauge (m) In 1992 M a y 09/17:03 1.610 0.255 M a y 11/09:55 0.730 0.209 M a y 12/08:50 0.730 0.195 May 14/08:50 0.660 0.185 M a y 16/08:55 0.690 0.185 M a y 18/10:05 1.000 0.210 May 20/10:05 1.100 0.220 M a y 21/09:25 0.885 0.205 M a y 22/09:55 0.800 0.195 M a y 23/09:40 0.780 0.190 M a y 24/09:50 0.940 0.200 M a y 25/09:30 1.060 0.218 M a y 26/11:00 2.310 0.275 M a y 27/15:50 1.690 0.255 M a y 29/17:35 1.300 0.235 M a y 30/08:04 1.210 0.225 In 1993 A p r i l 28/09:18 0.460 0.145 Apr i l 29/08:42 0.464 0.150 M a y 01/14:06 0.810 0.175 May 04/13:37 0.675 0.165 M a y 05/08:05 0.818 0.185 M a y 06/13:00 1.120 0.215 May 12/09:30 2.620 0.265 M a y 18/12:12 2.440 0.275 M a y 19/08:47 2.890 0.280 Appendix Table 3. Weights of sediment fractions of Helley-Smith samples from Mosquito Creek and McAuley Creek (grams) Discharge Fraction in phi S A M P L E N O . (mVs) -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 Mosquito Creek 93-HZ-03 0.74 0.0 0.0 400.6 202.6 369.8 235.8 93-HZ-17 1.06 0.0 315.3 718.2 718.0 805.5 764.2 93-HZ-25 0.85 0.0 156.5 131.9 426.0 617.1 712.9 93-HZ-28 1.92 0.0 116.2 62.0 242.3 441.8 426.6 93-HZ-41.42 2.51 451.7 1180.5 1416.4 1483.0 1440.5 1271.9 93-HZ-49 1.72 0.0 0.0 48.7 140.5 116.1 117.2 93-HZ-50,52 1.61 0.0 0.0 124.8 339.2 367.8 481.5 McAuley Creek 93-HZ-34 2.14 0.0 0.0 0.0 0.0 0.0 0.0 93-HZ-43-45 3.40 0.0 0.0 0.0 137.6 131.2 258.9 93-HZ-46 3.28 0.0 351.8 513.1 268.2 442.6 327.8 93-HZ-47 3.53 0.0 0.0 292.7 627.2 628.9 554.7 93-HZ-51.53 2.46 0.0 0.0 0.0 47.4 78.3 46.9 92-HZ-10 1.10 0.0 0.0 0.0 0.0 0.0 0.0 92-HZ-17 2.39 0.0 65.9 0.0 91.5 89.5 83.5 92-HZ-18 2.49 0.0 0.^ 143.8 83.4 108.3 168.4 92-HZ-20 1.65 0.0 0.0 0.0 0.0 8.8 0.0 214 Appendix Table 3. Weights of sediment fractions of Helley-Smith samples from Mosquito Creek and McAuley Creek (grams) Fraction in phi -3.5 -3.0 -2.5 -1.8 -1.3 -0.8 -0.3 0.3 0.8 Mosquito Creek 272.9 357.1 178.4 140.7 167.5 186.1 198.5 171.6 231.4 868.6 1032.5 219.0 204.2 258.8 315.8 301.3 245.5 280.4 699.6 833.4 1064.6 894.0 1076.8 1194.6 1206.1 994.7 1003.5 490.4 596.2 216.5 153.1 174.2 187.2 185.0 156.6 204.9 1241.4 1309.3 1584.2 1068.3 1136.6 1222.7 1186.8 966.3 1101.7 189.1 272.1 417.1 352.1 382.0 367.4 325.4 257.5 272.4 548.1 829.6 1254.0 995.2 1055.3 1062.6 946.8 690.6 731.8 M c A u l e y Creek 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 297.9 382.0 488.9 483.0 668.0 950.0 615.1 853.5 617.8 362.6 458.3 643.1 528.5 644.7 717.3 750.4 493.5 360.0 417.8 374.9 422.1 315.5 388.1 470.7 430.2 342.2 277.6 72.9 71.8 107.0 125.3 154.4 250.8 301.8 205.1 217.5 0.8 4.1 8.8 5.2 6.1 7.1 9.2 8.2 9.5 69.3 92.0 138.2 104.5 126.5 129.8 133.5 99.2 79.4 145.3 160.7 252.6 199.1 227.2 209.8 180.8 134.9 104.8 28.1 36.8 41.6 31.0 33.9 33.5 30.2 20.4 14.8 Appendix Table 3. Weights of sediment fractions of Helley-Smith samples from Mosquito Creek and McAuley Creek (grams) Fraction in phi 1.3 2.3 2.8 3.3 3.8 4.3 Mosquito Creek 528.3 86.6 24.9 17.3 17.0 110.0 1235.0 186.0 91.0 44.3 29.1 239.7 1138.3 196.2 57.2 32.7 21.0 240.4 828.9 248.0 132.1 68.1 37.1 296.5 1399.9 258.2 92.8 53.4 31.1 250.4 330.5 48.4 11.0 8.3 6.4 59.0 871.1 131.4 43.7 27.0 17.8 206.6 M c A u l e y Creek 245.4 36.0 13.7 11.6 9.5 114.9 519.6 39.5 17.4 11.2 9.4 119.8 354.7 28.7 12.6 7.7 7.6 122.4 296.6 32.2 10.5 7.8 6.4 91.3 175.9 12.1 3.6 3.1 2.4 33.6 10.1 14.8 13.7 12.8 12.7 12.9 74.2 45.3 21.7 1.9 2.3 5.2 105.2 61.8 40.2 17.9 15.7 21.1 18.8 23.0 16.8 13.1 13.1 13.7 216 Appendix Table 4. Concentrations of gold (ppb) and magnetite (grams) in Helley-Smith samples from Mosquiot Creek, McAuley Creek and Harris Creek in 1993 Sampling Gold (ppb) in Gold (ppb) in Magnetite (grams) SAMPLE NO. Creek -0.149+0.053 mm -0.053 mm -100+140 -140+200 -200+270 93-HZ-03 Mosquito <0.1 0.6 0.6 0.3 0.15 93-HZ-17 Mosquito <0.1 2.1 2.34 0.72 0.49 93-HZ-25 Mosquito 6.1 0.7 1.18 0.5 0.25 93-HZ-28 Mosquito <0.1 2.7 3.41 1.27 0.63 93-HZ-41.42 Mosquito <0.1 3.0 2.49 1.1 0.59 93-HZ-49 Mosquito 0.8 <0.2 0.15 0.12 0.06 93-HZ-50.52 Mosquito 0.7 2.9 1.02 0.31 0.24 93-HZ-34 McAuley <0.1 0.0 NO NO NO 93-HZ-43-45 McAuley <0.1 0.0 NO NO NO 93-HZ-46 McAuley <0.1 0.0 NO NO NO 93-HZ-47 McAuley <0.1 0.0 NO NO NO 93-HZ-51.53 McAuley <0.3 0.0 NO NO NO 92-HZ-10 McAuley <0.1 0.0 NO NO NO 92-HZ-17 McAuley <0.1 0.0 NO NO NO 92-HZ-18 McAuley <0.1 0.0 NO NO NO 92-HZ-20 McAuley <0.1 0.0 NO NO NO 93-HZ-18 Harris <0.3 <0.1 0.23 0.11 0.08 93-HZ-26.27 Harris <0.1 0.9 3.13 1.24 0.76 93-HZ-35 Harris <0.1 2.4 0.88 0.34 0.23 93-HZ-36 Harris <0.1 1.3 0.56 0.26 0.16 93-HZ-37 Harris <0.1 0.9 0.56 0.25 0.15 93-HZ-38 Harris <0.1 <0.1 0.58 0.29 0.21 93-HZ-39 Harris <0.1 3.9 0.52 0.26 0.15 93-HZ-40 Harris <0.2 <0.2 0.41 0.22 0.16 93-HZ-48 Harris <0.2 2.3 0.34 0.2 0.13 93-HZ-54 Harris 0.9 0.9 0.24 0.16 0.12 93-HZ-55 Harris 4.7 <0.2 0.42 0.23 0.17 93-HZ-56 Harris <0.2 <0.2 0.43 0.22 0.14 93-HZ-57,58 Harris 4.0 <0.2 0.45 0.24 0.15 93-HZ-59.60 Harris <0.1 1.2 0.52 0.28 0.21 93-HZ-61.62 Harris <0.1 <0.2 0.51 0.34 0.22 93-HZ-63,64 Harris <0.2 1.0 0.46 0.24 0.16 93-HZ-65.66 Harris <0.2 <0.1 0.27 0.14 0.1 93-HZ-67,68 Harris <0.1 <0.2 0.48 0.23 0.18 93-HZ-69.70 Harris <0.1 1.5 0.53 0.31 0.22 93-HZ-71.72 Harris <0.1 2.1 0.59 0.33 0.18 93-HZ-73,74 Harris <0.1 1.7 0.63 0.49 0.25 93-HZ-75,76 Harris 0.6 1.6 0.59 0.28 0.17 93-HZ-77.78 Harris <0.1 <0.2 0.63 0.2 0.22 93-HZ-79,80 Harris <0.1 <0.2 0.43 0.22 0.19 83-HZ-8L82 Hams <0.1 1.2 0.48 0.25 0.13 Appendix Table 5. Weights (grants) of sediment fractions of the samples from landslides, log jam and unstable bank along lower Harris Creek Sampling Fraction in phi sites and No. -7.5 -7.0 -6.5 -6 -5.5 -5 -4.5 -4 Landslide #1 91-HZ-08 0.0 0.0 0.0 434.3 204.7 321.4 482.6 91-HZ-09 0.0 0.0 0.0 170.3 269.5 230.5 342.5 91-HZ-10 0.0 0.0 0.0 77.3 124.1 151.4 180.8 Landslide #7 91-HZ-15 0.0 0.0 0.0 3100.0 778.4 1025.4 954.0 91-HZ-16 0.0 0.0 514.8 383.1 881.3 737.7 639.6 91-HZ-17 0.0 319.3 673.8 369.6 836.0 655.9 791.7 Landslide # 17 91-HZ-23 0.0 0.0 989.0 930.3 1519.0 1597.3 1186.9 93-HZ-145 0.0 0.0 560.1 207.0 247.7 298.4 369.4 93-HZ-146 0.0 330.4 1401.3 983.2 754.2 741.3 663.1 Landslide #28 91-HZ-61 0.0 0.0 0.0 66.2 252.1 300.9 284.2 91-HZ-62 0.0 0.0 0.0 197.5 306.3 330.6 487.8 Log jam #1 92-HZ-05 0.0 5125.0 10825.0 6025.0 7450.0 6500.0 6550.0 25884.1 92-HZ-06 0.0 4850.0 6100.0 9600.0 8350.0 6075.0 5225.0 31360.2 92-HZ-07 0.0 1600.0 2800.0 6625.0 5450.0 3325.0 3350.0 6341.9 92-HZ-08 7025.0 9400.0 3275.0 8000.0 5250.0 4350.0 4550.0 46910.8 Bank erosion Station 1 91-HZ-72 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 91-HZ-73 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Bank erosion Station 2 91-HZ-74 0 0.0 0.0 0.0 0.0 0.0 198.0 5.8 91-HZ-75 0 0.0 1631.3 2811.6 4303.0 2436.3 2014.1 1592.3 Bank erosion Station 3 93-HZ-04 0 0.0 0.0 0.0 770.4 835.5 560.2 644.9 93-HZ-05 0 0.0 0.0 276.0 258.8 117.7 177.4 131.0 Appendix Table 5. Weights (grams) of sediment fractions o f the samples from landslides, logjam and bank erosion along lower Harris Creek Fraction in phi -3.5 -3 -2.5 -1.75 -1.25 -0.75 -0.25 0.25 0.75 561.9 570.2 640.0 286.6 370.5 382.2 198.4 228.4 250.2 878.4 909.3 1131.2 583.5 676.2 752.7 720.7 701.2 888.6 764.4 739.3 806.1 309.4 379.6 462.4 522.4 461.6 453.8 353.2 426.6 561.4 491.1 531.7 713.0 34568.3 40185.1 54435.0 27704.7 31700.9 41780.8 5283.8 5862.1 9856.6 38674.8 40407.3 49919.7 0.0 16.4 23.1 0.0 5.9 6.7 12.6 18.2 51.5 1155.2 1111.4 1311.9 616.9 708.0 1052.6 154.5 138.8 258.6 391.3 324.8 338.0 253.7 225.2 206.2 176.0 159.5 152.4 714.1 617.5 655.0 474.1 491.2 495.6 520.1 515.3 487.2 445.0 456.3 524.7 314.4 318.5 313.0 279.2 267.0 255.9 426.7 420.3 449.4 501.8 498.0 507.3 38753.5 35384.9 15923.2 32471.8 28077.2 16203.9 9041.8 8098.8 11822.2 33028.4 24781.4 13579.3 14.3 13.8 25.2 4.1 3.0 3.5 47.4 54.8 73.8 792.8 568.3 461.4 785.4 755.9 784.3 267.2 341.3 470.0 290.1 261.5 273.0 201.8 207.7 221.2 136.8 125.6 166.0 607.9 538.1 485.8 508.0 515.2 516.0 4 3 4 . 4 4 4 7 . 6 451.4 435.5 492.5 502.5 328.8 327.5 298.1 242.8 224.1 224.2 489.4 406.0 333.9 490.6 448.1 413.9 11929.0 8917.7 10481.4 14187.6 10378.1 8729.0 13383.9 12481.5 10733.8 12457.1 10256.1 10051.3 37.2 135.2 202.2 18.7 68.4 594.3 91.0 221.1 957.1 367.5 326.3 386.8 649.5 597.2 628.9 425.8 295.9 241.2 Appendix Table 5. Weights (grams) of sediment fractions of the samples from landslides, log jam and bank erosion along lower Harris Creek Fraction in phi 1.25 2.25 2.75 3.25 3.75 4.25 899.7 762.6 735 7 566.7 640.7 6896.7 727.5 448.2 516 6 486.8 606.4 10851.8 532.9 429.4 572 7 589.5 699.5 9775.4 1224.2 543.2 438.7 297.3 302.4 5832.9 1455.9 1066.7 1485.5 1178.5 1041.5 6952.7 1003.2 524.6 628.2 568.0 452.0 7285.2 1004.0 464.7 422.5 342.6 384.9 5582.9 560.4 245.3 262.2 232.7 292.7 3299.9 405.5 161.7 176.4 130.3 200.4 2340.8 918.9 514.5 514.2 461.0 505.0 4896.6 921.8 508.0 482.4 427.0 474.3 5938.8 3790.0 1282.3 660.5 408.1 319.1 1630.2 2970.7 950.4 576.1 375.1 313.2 2039.6 1990.6 375.2 169.2 96.2 72.0 598.1 4668.7 1287.2 688.0 396.6 279.6 1720.9 1337.7 1067.5 801.6 842.6 525.6 6272.7 4730.3 2316.2 1376.0 1012.3 434.1 1813.1 4980.0 3239.7 1837.5 1079.9 711.9 1688.0 1107.1 525.6 352.3 211.8 179.4 645.7 1173.6 487.0 348.2 285.8 237.5 2213.7 483.8 314.1 344.1 321.3 340.0 4295.9 220 on C L . C+-. o co O rt C M C o o c o O CD q E rt ro O N O N . q q 'co rt X) J 4 1 ) CD C H CJ C O •c Ui rt X o g) rt s r~ >n m O N 00 — ro CN •—' vo ro O N O — O O O N CN O N O - - Cvi 2 -, , O N r f O m ro O oo CN r f cn O N CN r f r f cn o O N ro CN O O oo ro O ro vd vd oo N O — ' o — i i n o r~ r f o CN 00 o ro CN ro O N O V D O N T t O O O - ^ CN O N r~- m —> r - <n ro — —• ro ON CN CN — ' O r f oo O N vq r - r f , — ' od r f ' ' — i <—i i n oo N O , — i cn ro CN oo ro rt V O CN — O N r— ro r f oo v o t -o - ° 92 r o _ • r f oo — ON O ( N O N V O O O C O O N O 00 r f r f N O —' ro ro ro —• CN vo r f ro vq O O oo ro O vd ,—* rt CN V O rt oo o o c c r rt o cj c en rt XJ C rt o c en rt c o c o o l —1 XJ C rt ° O £2 "55 j= p ,£P o o vo _o rt f— '-5 c o D. Q. < c E rt — CD c en rt E • c o Z O N rt CN ro — i ro cn r f o O N VO oo r f O N cn oo r f cn ro CN ro CN ro ro ro oo CN VO oo ro rt r - cn NO r f r f ' r f o r f r f — vd o ro O O o O ro r f r f CN CN CN O o r f o . „ vc r — r - — ^ ro vq ro — ^ CN CN ro ro V C rt o N O ro O O ro oo O N N O t—-H cn N O ro r f oo O N O CN cn r-' vd CN 00 rt CN © O O N r f t-~ o N O O N r f cn rt ro ro CN oo CN r f r f rt rt o r f cn r r + r - oc O cr, cr, O ro rN O N O N o r - cn O O N CN r f O N vo O N vq cn f—, t - O N r o CN r f O N r f cn O N CN ro O N vo r f r -oo N O O CN O N CN CN CN —• cn cn CN o i CN cn cn r-' ro cj -a on c c rt cy) o Z _ o Q. E rt C/) r - 00 oo r- « CN CN — — .—. % % 3fc o CJ o o Cu) CD CD CD CD CD CD e XJ XJ XJ XJ XJ XJ XJ XJ XJ XJ XJ rt rt rt "lo "co "w "co "co "co "w "co • — . • —i • — * ' rt T3 XJ XJ XJ XJ XJ XJ XJ XJ XJ XJ eo en en en c C c q g q q q q q O o o o rt rt rt rt rt rt rt rt rt _J _J _J 1—1 —1 q o GO cr, vC r-~ oc oo O N 0 O O O O O N N N N N N 1 _ I X I I i i i i i • r-i r-j r i r^l —• — ON 3-. O N O N O N ON o N O i i X I X X I X i O N O N O N O N cn vo r f r f rt rt q q c q c o o o o o *-*-< 1 cy) ry) cy) do ro r f cn r f cn r~- O o X • X 1 X 1 X ro rt X 1 ro O N O N O N O N O N 221 Appendix Table 7. Concentrations (ppb) of gold in landslides, bank material and log jam along lower Harris Creek Gold in Go ld in SampleNo. Sampling sites N M H M C -0.053 m m 2 92-HZ-05 Log jam #1 2710 11.4 92-HZ-06 Logjam #1 2430 4.7 92-HZ-07 Logjam #1 1370 17.5 92-HZ-08 Logjam #1 3330 8.8 91-HZ-08 Landslide #1 80 1.6 91-HZ-09 Landslide # 1 110 1.0 91-HZ-10 Landslide # 1 255 2.6 91-HZ-15 Landslide #7 1830 0.3 91-HZ-16 Landslide Ul 50 <0.2 91-HZ-17 Landslide Ul 320 0.8 91-HZ-23 Landslide #17 2410 0.5 93-HZ-145 Landslide #17 <5 <0.1 93-HZ-146 Landslide #17 <5 <0.1 91-HZ-61 Landslide #28 70 <0.2 91-HZ-62 Landslide #28 100 <0.2 91-HZ-72 Station I 305 0.9 91-HZ-73 Station I 105 6.9 91-HZ-74 Station II 20 4.9 91-HZ-75 Station II 650 <0.2 93-HZ-04 Station III 780 0.9 93-HZ-05 Station III <5 <0.1 1 Go ld in N M H M C = Gold in -0.149+0.053 mm non-magnetic heavy mineral concentrates. 2 Gold in -0.053 mm = Gold in -0.053 mm sediment. Appendix Table 8 Weights of sediment fractions of Hellcy -smith samples (grams) on Harris Creek. Discharge Size Fraction in (phi) SAMPLE NO. (mVs) -6.5 -6 -5.5 -5.0 -4.5 -4.0 93-HZ-26,27 9.05 0.0 0.0 0.0 0.0 67.6 75.4 93-HZ-35 24.11 0.0 1805.5 1574.2 2028.5 1854.7 1602.3 93-HZ-36 23.04 0.0 779.1 1084.3 1892.7 1676.4 1288.5 93-HZ-37 22.88 384.9 3800.8 2422.2 2037.5 1431.0 1342.1 93-HZ-38 23.32 474.9 2430.5 2964.5 1954.0 1765.3 1272.1 93-HZ-39 24.54 0.0 1473.7 1814.0 3009.1 2588.2 2175.7 93-HZ-18 5.31 0.0 0.0 0.0 0.0 0.0 0.0 93-HZ-40 26.1 0.0 2140.1 2916.5 2440.3 1737.3 1339.4 93-HZ-48 23.71 362.1 1384.1 1575.1 1165.5 1230.9 884.9 93-HZ-54 17.11 617.3 304.4 993.6 1297.9 896.7 667.0 93-HZ-55 17.7 0.0 1022.1 1518.9 2019.6 2322.0 1993.6 93-HZ-56 20 468.0 1969.0 2669.1 2427.5 1777.7 1526.7 93-HZ-57,58 16.62 378.5 3545.3 2010.9 1699.9 1227.2 1014.2 93-HZ-59,60 15.57 1953.0 3023.2 3780.6 3543.3 2519.7 2044.4 93-HZ-61.62 15.8 0.0 4539.4 2722.4 3020.4 2419.2 2042.9 93-HZ-63,64 17.58 0.0 2050.9 1773.1 1665.8 1680.6 1417.9 93-HZ-65,66 17.35 455.0 2237.3 1634.6 1086.1 883.9 699.6 93-HZ-67,68 16.78 1124.0 3855.7 2057.4 1911.3 1403.2 1451.0 93-HZ-69.70 14.41 0.0 559.9 427.6 375.2 698.1 816.1 93-HZ-71.72 11.11 0.0 0.0 331.8 242.0 242.6 158.8 93-HZ-73.74 9.85 0.0 407.1 75.9 72.6 212.4 277.9 93-HZ-75,76 8.52 0.0 0.0 0.0 0.0 17.5 36.8 93-HZ-77,78 8.78 0.0 0.0 85.9 195.8 323.9 219.6 93-HZ-79,80 8.19 0.0 155.3 150.3 77.9 174.0 143.6 83-HZ-81.82 8.26 0.0 0.0 0.0 176.5 133.6 142.0 Appendix Table 8 Weights of sediment fractions of Helley -smith samples (grams) on Harris Creek. Size Fraction in (phi) -3.5 -3.0 -2.5 -1.75 -1.25 -0.75 -0.25 0.25 0.75 59.3 125.9 246.4 233.6 453.0 922.4 1377.2 1366.4 1565.4 1579.2 1975.8 2408.2 1548.6 1515.4 1403.4 1205.2 875.2 958.4 1224.0 1134.8 1275.8 803.5 764.5 722.0 719.7 688.1 674.0 1131.9 1278.1 1484.0 958.7 977.6 1037.5 1038.1 885.8 831.0 1266.1 1336.3 1586.6 1040.1 1114.5 1312.6 1433.6 1143.7 956.7 2085.2 2112.7 2368.3 1477.8 1624.7 1758.8 1620.2 1095.5 875.2 0.0 0.0 28.5 12.2 22.7 60.2 112.0 153.8 228.5 1228.0 1287.3 1437.7 828.5 818.4 849.1 802.3 621.0 546.2 799.2 785.4 944.8 677.6 853.7 1042.8 1082.9 770.3 624.2 553.1 582.1 674.0 388.6 396.4 -831.0 521.4 472.5 437.9 1961.8 2056.1 2630.7 1812.0 1973.0 2087.5 1765.0 1124.4 862.0 1333.6 1462.0 1927.7 1351.7 1482.5 1509.7 1353.6 961.3 822.8 879.4 933.4 1260.5 963.4 1111.7 1324.7 1407.5 1064.2 916.7 1745.7 1917.7 2474.1 1746.3 2037.6 2402.2 1938.4 1177.0 873.0 1737.2 1818.6 2084.5 1492.7 1704.7 1935.4 1876.0 1207.7 966.8 1338.1 1426.6 2049.1 1442.7 1814.7 2142.7 2095.8 1345.9 1009.9 556.7 597.6 850.2 611.5 828.0 1025.4 1021.7 829.6 622.1 1319.0 1488.9 2168.0 1906.9 2038.0 2518.5 2271.8 1482.3 1062.0 915.3 1194.1 1831.7 1347.8 1376.3 1409.7 1451.7 1129.5 955.7 164.2 225.5 338.6 260.8 330.9 466.5 617.3 672.0 659.6 278.7 429.9 785.4 791.7 1092.4 1348.8 1433.8 1238.3 1128.0 29.4 66.2 164.6 198.9 440.6 768.0 1116.8 1118.4 1004.1 346.7 457.2 950.4 869.8 1306.8 2015.0 2500.0 2100.5 1816.0 230.6 317.2 619.2 593.5 1059.9 1667.6 2148.0 1873.7 1674.4 228.7 342.3 649.1 669.9 1027.7 1494.6 1704.9 1540.4 1387.6 Appendix Table 8 Size Fraction in (phi) 1.25 2.25 2.75 3.25 3.75 4.25 1571.9 229.9 77.7 40.1 31.7 201.0 798.0 77.7 18.9 11.2 9.2 65.9 552.7 57.7 12.9 7.8 6.2 45.5 535.8 49.2 12.9 7.6 6.2 43.9 572.9 55.3 13.1 8.7 8.9 43.8 527.8 45.9 10.6 7.7 6.4 39.2 235.2 26.5 6.1 3.6 2.5 24.6 400.8 39.5 9.3 7.1 6.1 37.0 398.3 35.9 6.4 4.8 4.4 23.7 357.7 27.3 4.9 3.6 3.5 16.4 513.8 44.4 10.1 6.6 6.5 38.9 521.4 43.6 8.3 6.0 6.2 33.8 705.0 63.6 11.4 8.8 7.6 48.2 725.6 68.0 13.7 10.1 8.9 59.3 790.6 68.7 12.9 11.1 9.4 58.7 638.0 38.6 9.5 7.3 6.9 47.1 440.7 30.9 6.1 4.4 4.2 25.0 629.7 46.4 10.7 7.4 6.6 43.8 734.1 71.5 14.3 9.8 9.5 55.3 852.0 87.8 14.9 11.4 9.4 52.0 996.3 112.1 22.1 14.8 9.8 79.6 880.2 110.0 20.1 12.0 8.2 71.9 1371.9 110.7 21.0 13.8 11.0 88.2 1004.8 94.5 15.2 11.0 9.8 61.8 1168.5 96.4 15.6 12.0 9.1 65.2 Appendix Table 9 A : Weights (grams) of magnetic heavy minerals in the three size fractions from high energy sites along Harris Creek Magnetic heavy minerals (grams) S A M P L E N O . -100+140 -140+200 -200+270 91-HZ-26 22.42 9.84 4.33 91-HZ-28 31.63 21.16 12.96 91-HZ-29 7.21 3.88 2.08 91-HZ-31 5.55 2.01 0.79 91-HZ-33 25.65 11.18 5.26 91-HZ-35 34.19 14.19 4.59 91-HZ-37 24.87 12.03 7.37 91-HZ-39 19.16 8.02 3.67 91-HZ-42 9.72 5.03 0.87 91-HZ-43 9.38 4.03 1.05 91-HZ-45 4.75 1.88 0.62 91-HZ-46 20.54 10.45 4.97 91-HZ-48 19.10 7.76 3.68 91-HZ-50 24.35 12.06 6.42 91-HZ-52 36.09 15.33 8.75 91-HZ-54 6.62 7.62 6.61 91-HZ-56 26.46 12.73 5.85 91-HZ-57 20.30 10.69 4.34 B: from low energy environment Magnetic heavy minerals (grams) Sample No. -100+140 -140+200 -200+270 91-HZ-27 13.66 7.46 2.9 91-HZ-32 13.09 10.93 6.47 91-HZ-34 34.34 31.64 19.63 91-HZ-36 6.92 5.44 5.57 91-HZ-38 24.32 13.67 8.83 91-HZ-40 24.76 13.43 10.04 91-HZ-41 7.44 5.28 4.73 91-HZ-44 1.02 0.61 0.43 91-HZ-47 26.22 19.09 13 91-HZ-49 9.45 5.29 3.14 91-HZ-51 10.52 9.53 7.53 91-HZ-53 35.54 17.59 8.42 91-HZ-55 3.35 3.73 4.45 91-HZ-58 50.45 22.74 7.73 Appendix Table 10 A : Weights of non-magnetic heavy mineral concentrates in the three size fractions from high energy sites along Harris Creek Non-magnetic minerals (grams) Sample No . -100+140 -140+200 200+270 91-HZ-26 18.84 6.79 4.32 91-HZ-28 22.62 16.05 5.84 91-HZ-29 10.12 6.90 3.47 91-HZ-30 21.42 3.83 1.53 91-HZ-31 6.86 0.72 0.28 91-HZ-33 28.68 11.43 3.38 91-HZ-35 23.56 10.92 1.43 91-HZ-37 20.68 6.83 4.33 91-HZ-39 12.24 4.46 1.52 91-HZ-42 60.20 30.22 1.93 91-HZ-43 11.05 4.29 0.86 91-HZ-45 5.80 2.95 0.33 91-HZ-46 14.48 4.49 2.01 91-HZ-48 12.35 3.66 1.28 91-HZ-50 14.67 5.12 3.40 91-HZ-52 16.93 6.82 4.55 91-HZ-54 9.54 5.00 2.50 91-HZ-56 14.06 6.95 2.16 91-HZ-57 12.31 5.23 1.76 B : from low energy sites along Harris Creek Non-magnetic minerals (grams) Sample No. -100+140 -140+200 -200+270 91-HZ-27 15.18 6.94 2.99 91-HZ-32 26.48 17.84 6.71 91-HZ-34 66.59 53.10 15.*2 91-HZ-36 16.80 8.77 5.53 91-HZ-38 25.24 11.91 4.72 91-HZ-40 24.42 12.68 5.38 91-HZ-41 87.93 50.82 33.63 91-HZ-44 1.96 0.94 0.39 91-HZ-47 27.54 15.97 7.20 91-HZ-49 9.49 3.88 1.25 91-HZ-51 19.85 7.98 4.64 91-HZ-53 26.02 8.34 3.59 91-HZ-55 7.27 2.65 2.22 91-HZ-58 35.82 15.50 3.43 Appendix Table 11 Concentrations of gold in the two size fractions from high energy sites along Harris Creek Gold (ppb) S A M P L E N O . N M H M C -270 sediment 91-HZ-26 4470 6.8 91-HZ-28 5770 2.2 91-HZ-29 240 0.9 91-HZ-31 28900 42.1 91-HZ-33 6140 0.9 91-HZ-35 4900 5.1 91-HZ-37 1270 1.5 91-HZ-39 1240 1.3 91-HZ-42 760 5.0 91-HZ-43 945 1.0 91-HZ-45 1080 38.0 91-HZ-46 3750 1.7 91-HZ-48 35 1.0 91-HZ-50 1440 3.6 91-HZ-52 640 0.8 91-HZ-54 2240 1.0 91-HZ-56 910 0.8 91-HZ-57 1930 1.6 Note: The sampling sites are shown in Fig. 3.9. 1 N M H M C = -0.149+0.053 mm non-magnetic heavy mineral concentrates. Appendix Table 12 Concentrations of gold in the two size fractions from low energy sites along Harris Creek Gold (ppb) SAMPLE NO. NMHMC1 -270 sediment 91-HZ-27 215 2.4 91-HZ-32 2010 1-4 91-HZ-34 <5 25 91-HZ-36 <5 0.5 . 91-HZ-38 80 2.1 91-HZ-40 20 0.3 91-HZ-41 <5 0.9 91-HZ-44 <5 0.8 91-HZ-47 45 27.4 91-HZ-49 135 4.4 91-HZ-51 75 0.2 91-HZ-53 125 1.3 91-HZ-55 10 0.4 91-HZ-58 25 11.4 Note: The sampling sites are shown in Fig. 3.9. 1 NMHMC = -0.149+0.053 mm non-magnetic heavy mineral concentrates 

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.831.1-0088673/manifest

Comment

Related Items