Using stream flow and chemistry data to estimate catchment scale groundwater and nitrate fluxes

Abstract

Groundwater is the dominant flow path carrying land surface recharge, including dissolved contaminants, to surface waters draining a catchment. The dominance of the groundwater pathway poses a challenge to management of water quality in agricultural catchments, because groundwater quantity and quality are difficult and expensive to monitor, and groundwater assimilative capacity for nitrate is generally unknown. On the other hand, rainfall and evapotranspiration as inputs, and stream flow and nitrate concentration as outputs, can be recorded relatively easily, especially if inexpensive in-stream nitrate sensors can be developed.

The eigenmodel approach has previously been used to estimate the land surface area and groundwater discharge contributing to stream flow in a small hill catchment. We extended this approach to explain seasonal patterns of nitrate and silica concentrations observed in the Toenepi Stream, which drains a lowland dairying catchment near Morrinsville, and to estimate the water and nitrate fluxes driving these observations.

The resulting model (“StreamGEM”) was calibrated for the four-year period 1 April 2007 to 31 March 2011, and cross-validated using data from the period 1 April 1995 to 31 March 1997. Estimated discharge, nitrate concentration (as nitrate-N) and nitrate load from near-surface, fast groundwater, and slow groundwater flowpaths were then calculated. On an annual basis, stream flow was dominated by discharge from fast, shallow groundwater. In summer however, slow, deeper groundwater dominated both flow and chemistry.

The total catchment input load (at the bottom of the root zone) was estimated to be 40 kg N ha⁻¹ y⁻¹ nitrate nitrogen (NO₃-N). Nitrate attenuation in the groundwater components accounted for 20 kg N ha⁻¹ y⁻¹ of this, with the remaining 50% being discharged to the stream. At the catchment scale, nitrate assimilation appears to occur dominantly in the shallower flow near the redox boundary, despite the strongly reduced conditions and much lower nitrate concentrations found in the deeper groundwater.

The ability to estimate catchment water and nitrate fluxes from weather and in-stream data offers an inexpensive and potentially widely applicable tool for improved management of New Zealand’s land and water resources. Current research focuses on ascertaining in which type of catchments StreamGEM can be applied successfully.