2007 GSA Denver Annual Meeting (28–31 October 2007)

Paper No. 1
Presentation Time: 1:30 PM

SIMULATING PLAYA LAKES ON THE TEXAS HIGH PLAINS IN SWAT: ONE ASPECT OF A COMPREHENSIVE MODEL LINKAGE AIMED AT THE MANAGEMENT OF WATER QUALITY IN LAKE WHITNEY, TEXAS


PROCHNOW, Shane J., FILSTRUP, C.T., SCHREINER, Matt and BYARS, B.W., Center for Applied Geographic and Spatial Research, Baylor University, Waco, TX 76798, Shane_J_Prochnow@Baylor.edu

Playa lakes on the Texas High Plains pose a unique problem because they add a level of complexity beyond the processes normally accounted for in watershed models. Playa lakes develop at the center of internally draining catchments, concentrating and storing surface flow. Most of this stored water is then either evaporated or infiltrates into aquifers. The playa lake phenomenon is inherently difficult to simulate on a watershed scale since most GIS models are unable to reliably detect these subtle playa depressions even when the finest resolution digital elevation models (30 m) are used as a basis for stream network delineation. Moreover, recent studies on the High Plains have demonstrated a disconnection between surface water infiltration from playas and groundwater contribution to flow in nearby streams. Our study illustrates this issue by comparing the base flow index between measured stream flow and a series of scenarios simulated in the Soil and Water Assessment Tool (SWAT) that incorporate representations of playa processes. The 20,214 km2 catchment at the USGS gauging station on the South Fork of the Brazos River near Aspermont, Texas is used for this comparison analysis. A scenario using SWAT defaults for the study catchment results in nearly two orders of magnitude greater surface flow than that observed for the same time interval. However, a scenario that accounts for unlimited internal drainage of surface flow for the portion of the watershed where playas occur approximated the same amount of surface flow as the observed data from the gauging station. Our study also finds that playa processes are a more critical calibration parameter for surface flow than overland curve number and soil infiltration parameters. Our study illustrates the importance of representing the playa process for comprehensive watershed models.