GSA Annual Meeting, November 5-8, 2001

Paper No. 0
Presentation Time: 1:30 PM-5:30 PM

GROUND AND SURFACE WATER CONTRIBUTION TO CHEMICAL MASS DISCHARGE IN A SEMI ARID DRYLAND AGRICULTURAL WATERSHED


VAN BIERSEL, Thomas P.1, ALLEN-KING, Richelle M.2, KELLER, C. Kent2 and SMITH, Jeffrey L.3, (1)Geology, Washington State University, Pullman, WA 99164, (2)Geology, Washington State Univ, Pullman, WA 99164, (3)USDA-ARS, 215 Johnson Hall, Washington State University, Pullman, WA 99164, tvanbiersel@wsu.edu

The goal of this research is to use environmental tracers to quantify the contributions of subsurface and surface runoff to predict the loading of non-point pollutants to rivers at multiple scales of study (field to basin). We use electrical conductivity to separate stream hydrographs and predict the mass discharge of an environmental tracer (dissolved silica) and an agrichemical (nitrate). The measured electrical conductivity is corrected for nitrate concentration. The study area is the Missouri Flat Creek watershed, a 14,400 ha semi-arid dryland agricultural setting located near Pullman, WA. N-fertilizer is generally applied to fields in the fall or spring in a liquid form (urea + ammonium nitrate). Ground and surface water samples are collected at approximately two-week intervals from an ephemeral stream and a tile drain located in actively farmed fields, and from seven stream-gaging stations. Surface water discharge is monitored continuously. A two-component (ground and surface water) hydrograph separation is performed using measured electrical conductivity end-members estimated from our field and basin study. Mass discharge predictions for silica and nitrate employ observed tile drain and ephemeral stream chemistry to represent the groundwater and surface runoff chemistry, respectively. The model provides a good fit between the predicted and measured dissolved silica and nitrate mass discharges. The match between predicted and observed silica discharge is better at the small watershed scale (underestimated by ~20% for 600 ha watershed) than the basin scale (overestimated by ~30%), where some urban runoff affects both the silica and electrical conductivity measurements. The analysis shows that the majority of annual nitrate discharge in Missouri Flat Creek is derived from the groundwater flowpath (>70%). Temporally, the nitrate predictions show that ground water supplies almost all of the observed nitrate (>80%) during the autumn and late spring.