Simulating sediment loading into the major reservoirs in Trinity River Basin

Abstract

The Upper Trinity River Basin supplies water to about one-fourth of Texas's population. The anticipated rapid growth of North Central Texas will certainly increase regional demands for high-quality drinking water. This has increased concerns that sediment and nutrient loads received by drinking water reservoirs are reducing and will continue to reduce reservoir volumes and water quality. The objectives of this study are to calibrate and validate the Soil and Water Assessment Tool (SWAT) model for streamflow and sediment to assess current rates and sources of sediment loadings to 12 major reservoirs in the Upper Trinity River Basin (in 7 eight-digit watersheds) and to use the calibrated model for assessing the effects of upland ponds. SWAT performed well for streamflow, as evidenced by r² values ranging from 0.55 to 0.95. Nash-Sutcliffe efficiency values ranged from 0.5 to 0.9 based on monthly streamflow comparisons between simulated and observed values for calibration, and r² values ranged from 0.58 to 0.95 for validation. SWAT simulated sediment loads reasonably well, as evidenced by the percentage of errors within 11%. Streamflow and sediment loading were quite diverse across the Trinity River Basin, resulting in a multitude of parameter adjustments during calibration. Long-term predictions indicate that the Richland-Chambers, Ray Hubbard, and Lavon watersheds have significant channel contribution to sediment loading reaching the reservoirs. Pond removal scenario analysis shows a 4% to 48% reduction in sediment loadings to reservoirs via pond detention of overland flow. This wide range is mainly due to the vastly different proportion of land area draining to ponds, the locations of ponds, and the basin's erosion (upland and channel) characteristics within each study watershed. The results indicate that in addition to implementing conservation practices such as ponds in upland areas, it is also necessary to have conservation practices in channels to further reduce erosion and subsequent loss to reservoirs. One limitation of this study is the lack of site-specific management information, and it is known that poor management practices at the field level can dramatically elevate sediment loads from an area. In this study, reasonable management operations were applied mainly at the county conservation district level. Opportunity exists for further data collection, including detailed data of field management and channel dimensions, which will allow the model to provide greater insight in identifying sensitive areas and reaches for stabilization and restoration. Opportunity also exists for further evaluation of the effects of optimizing pond size and placement to minimize reservoir sediment loading.