Catchment analysis presentation AT THE pacrim 2019 conference in auckland, new zealand
Dennis will presented a keynote address in a geochemistry session at the PACRIM 2019 conference April 3-5 in Auckland, New Zealand. This session was sponsored by the Association of Applied Geochemists (AAG). The title of his presentation was "A catchment analysis approach to the interpretation of regional stream sediment data: Examples from the Canadian Cordillera" . He also contributed to a short course “Exploration Geochemistry: Applying Fundamentals in the Field" that is also sponsored by the AAG.
introduction to stream sediment catchment analysis
The geochemistry of a stream sediment sample reflects the integrated effects of eroding lithological units within the catchment basin from which it has been derived, their relative contributions to the stream sediment load, the de-coupling of elements during weathering and hydrodynamic sorting, and deposition/precipitation of secondary minerals. These processes all have the potential to mask subtle responses from commodity and pathfinder elements derived from mineral occurrences within a catchment. In addition, the catchment area relative to the area of exposed mineralisation within the catchment determines the degree to which a geochemical response related to that mineralisation is diluted.
Several approaches have been developed to filter the masking effects of various processes affecting stream sediment geochemistry. These include the calculation of weighted average geochemical background for individual elements based on the relative proportions of each lithological unit exposed within the catchment, multiple regression analysis of stream sediment data given the areal extent of each lithological unit, levelling of geochemical data using the dominant lithological unit, and regression analysis of commodity or pathfinder elements against principal components that can be attributed to specific geochemical processes. Alternative approaches to the interpretation of regional stream sediment data include fractal analysis and the use of machine learning algorithms.
examples from the canadian cordillera
Over the past 10 to 15 years, both Geoscience BC and the Yukon Geological Survey have undertaken re-analysis of archived stream sediment samples collected by government regional stream sediment surveys through the Canadian cordillera. Samples were collected on a relatively consistent basis using standardised collection methods over a 30- to 50-year period. Re-analysis of the archived material has been undertaken following a modified aqua regia digestion in most cases, using a mixture of ICP-OES and ICP-MS instrumentation. These re-analyses were undertaken over a relatively short time period compared to the original analyses and provide a high-quality data set in which laboratory and analytical variability has been minimised. A generalised work flow for interpretation of the data is illustrated in Figure 1.
Interpretation of regional stream sediment data from British Columbia and Yukon emphasises the importance of understanding the processes affecting the geochemical data prior to selecting a specific interpretative approach. In southern British Columbia, regional stream sediment data from the QUEST South area underwent regression analysis to correct for possible scavenging of metals by secondary Fe hydroxides and calculated residuals levelled by dominant rock type in the catchments (Arne and Bluemel, 2011). On northern Vancouver Island the use of a weighted sums model (WSM), developed using commodity elements levelled for dominant catchment lithology, proved the best predictor of known porphyry Cu-Au occurrences (Arne and Brown, 2015; Figure 2). Mackie, Arne and Brown (2010) and Arne, Mackie and Pennimpede (2018) describe a methodology whereby regional stream sediment geochemical data from individual map sheets from the southern Yukon were levelled using the dominant lithological unit in the catchments, as well as using regression analysis of commodity and pathfinder elements against principal components in which those elements defined lithological or metal scavenging associations. Several methods were applied to data from north-western British Columbia by Arne et al. (2018) to remove lithological effects masking evidence for mineralisation prior to incorporating the data into a WSM for specific mineral deposit types. Most of the traditional approaches generated some minor improvement in the detection of known mineral occurrences provided a correction for dilution was also applied. The use of machine learning was also trialled, with the random forests algorithm generating plausible predictive models for a variety of porphyry Cu deposit types, as well as epithermal Ag-Au deposits.
These studies indicate that a variety of data processing approaches involving catchment analysis can be used to generate improved predictive models for mineral exploration, but that the performance of different interpretive approaches varies depending on the geological terrain and data set. This conclusion highlights the need for exploratory data analysis (EDA) of each data set to understand the main processes influencing the geochemical data rather than the routine application of a specific interpretive methodology in all circumstances.
Most of the case studies discussed were undertaken while the author was employed by CSA Global on projects funded by Geoscience BC and the Yukon Geological Survey. The support of both organisations is gratefully acknowledged. Catchment basins in British Columbia and Yukon were generated by the British Columbia and Yukon Geological Surveys, respectively. Colleagues who contributed to the work over many years include Britt Bluemel, Olivia Brown, Olywn Bruce, Yao Cui, Eric Grunsky, Kristen Kennedy, Rob Mackie and Chris Pennimpede. Tony Christie is thanked for his review and editing of this abstract.
Arne, D, Mackie, R and Pennimpede, C, 2018. Catchment analysis of re-analysed regional stream sediment geochemical data from the Yukon, Explore 179: 1-13.
Arne, D C, Mackie, R, Pennimpede, C, Grunsky, E and Bodnar, M, 2018. Integrated assessment of regional stream-sediment geochemistry for metallic deposits in northwestern British Columbia (parts of NTS 093, 094, 103, 104), Geoscience BC Report 2018-14, 96 p.
Arne, D and Brown, O, 2015. Catchment analysis applied to the interpretation of new stream sediment data, Northern Vancouver Island (NTS 102I and 92L), Geoscience BC Report 2015-4, 40 p.
Arne, DC and Bluemel, EB, 2011. Catchment analysis and interpretation of stream sediment data from QUEST South, British Columbia, Geoscience BC Report 2011-5, 25 p.
Mackie, RA, Arne, DC and Brown, O, 2015. Enhanced interpretation of regional stream sediment geochemistry from Yukon: catchment basin analysis and weighted sums modelling, Yukon Geological Survey Open File 2015-10, 11 p.