Using factor analysis and end-member mixing techniques to infer sources of runoff generation

Factor analysis and (two-component) end member mixing analysis (EMMA) were applied to high resolution stream chemistry data from three catchments to infer sources of iron- and manganese-rich runoff in the catchment area and to evaluate their temporal and spatial influence on the stream water quality. Two Fe and Mn sources were identified: (a) an organic soilwater source, associated with Fe and Mn accumulation in the organic-rich upper soil horizons and (b) a deep soilwater/groundwater source arising from reduced metal mobilisation in the deeper soils. While (a) provides the dominant Fe and Mn inputs during storm events, (b) becomes important just before peak flow when riparian groundwater is displaced into the stream, giving rise to high Fe and Mn concentrations. The timing of such high Fe and Mn loadings in the stream runoff has important implications for the water supply management of the catchment.