Uncertainty of catchment nitrate flux estimates calculated using the “StreamGEM” approach with monthly monitoring data

Abstract

The dynamic catchment model “StreamGEM” was previously used to infer water and nitrate fluxes through the subsurface of the small (15 km2) Toenepi dairying catchment, based on long term (4 years) daily flow and monthly chemistry data collected at the catchment outlet. The approach allowed separation of stream flow into near-surface, fast- and slow-groundwater contributions, including estimation of nitrate attenuation and discharge along each flow path, as well as the uncertainty of these results (Woodward et al., 2013). In the current project, extension of the approach to analysis of datasets from three much larger catchments was explored. Additional complicating factors in upscaling included the presence of more diverse land use, topographical, meteorological and geological zones within each catchment, as well as the potentially greater importance of time lags between precipitation events and catchment response.