Using StorAge Selection Functions to Improve Simulation of Groundwater Nitrate Lag Times in the SWAT Modeling Framework.
Abstract
Intensive agriculture and atmospheric deposition have dramatically increased the input of reactive nitrogen into many watersheds worldwide. Reactive nitrogen can leach as nitrate into groundwater, which is stored and eventually released over years to decades into surface waters, potentially degrading water quality. To simulate the fate and transport of groundwater nitrate, many researchers and practitioners use the Soil and Water Assessment Tool (SWAT) or an enhanced version of SWAT that accounts for topographically-driven variable source areas (TopoSWAT). Both SWAT and TopoSWAT effectively assume that nitrate in the groundwater reservoir is well-mixed, which is known to be a poor assumption at many sites. In this study, we describe modifications to TopoSWAT that (1) relax the assumption of groundwater well-mixedness, (2) more flexibly parameterize groundwater transport as a time-varying distribution of travel times using the recently developed theory of rank StorAge Selection (rSAS) functions, and (3) allow for groundwater age to be represented by position on the hillslope or hydrological distance from the stream. The approach conceptualizes the groundwater aquifer as a population of water parcels entering as recharge with a particular nitrate concentration, aging as they move through storage, and eventually exiting as baseflow. The rSAS function selects the distribution of parcel ages that exit as baseflow based on a parameterized probability distribution; this distribution can be adjusted to preferentially select different distributions of young and old parcels in storage so as to reproduce (in principle) any form of transport. The modified TopoSWAT model (TopoSWAT+rSAS) is tested at a small agricultural catchment in the Eastern Shore, MD with an extensive hydrologic and hydrochemical data record for calibration and evaluation. The results examine (1) the sensitivity of TopoSWAT+rSAS modeling of nitrate transport to assumptions about the distribution of travel times of the groundwater aquifer, (2) which travel times are most likely at our study site based on available data, and (3) how TopoSWAT+rSAS performs and can be applied to other catchments.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2017
- Bibcode:
- 2017AGUFM.U21B..08W
- Keywords:
-
- 0810 Post-secondary education;
- EDUCATION