Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Application of SWAT Model on Chapel Branch Creek Watershed with a Karst Topography, SC

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2010 Pittsburgh, Pennsylvania, June 20 - June 23, 2010  1009824.(doi:10.13031/2013.32002)
Authors:   Devendra Amatya, Manoj Jha, Amy Edwards, Thomas Williams, Daniel Hitchcock
Keywords:   Hydrologic Modeling, ArcViewSWAT, Streamflow, Water balance, Nutrient, Losing streams

SWAT, a widely used GIS-based basin scale model to characterize the hydrology and water quality for large, complex watersheds as a function of land use, soils, topography, management practices and their interactions with variability in climate, has not been fully tested on watersheds on karst topography. In this study, SWAT was used to evaluate monthly stream flow dynamics for a 1,555 ha mixed land use Chapel Branch Creek (CBC) watershed with karst features draining to Lake Marion near Santee, SC. In order to apply the SWAT model for simulating the hydrology, efforts were first made to calibrate the stream flows at three in-stream subwatersheds draining to the downstream flooded portion of the CBC at the edge of the Lake Marion using three years (2006-09) of climatic data across the watershed. The watershed was delineated into 31 smaller subwatersheds and 472 HRUs using the ArcView SWAT2003 model with DEMs for topography, SSURGO data base for soils, NAIP2005 imagery for land use, and other field observations. Out of the three main in-stream locations used for the flow calibration, the first one at the Cave Spring CS outlet (1089 ha) drained mostly agricultural and forested lands and a golf course besides an unknown subsurface area for groundwater discharging at this point. Land use draining the second one at SL-2 (550 ha) was comprised of mostly developed area of Town of Santee, highways and roads, and some agricultural and forested lands some of which had karst features characterizing the sinkholes and depressions and loosing streams. The third subwatershed drained an area of 63 ha containing a part of a golf course receiving effluents from a wastewater treatment plant within it. The model was able to capture the monthly streamflow at the cave spring (CS) outlet only after adding a known monthly baseflow as a point source at that location with a limited data. The monthly flow at the SL-2 location was over-predicted for the months with high and normal rainfall as expected as a result of loosing streams during lower lake levels. The model under-predicted the flows during high lake levels likely due to high conductivity and deep percolation coefficient used to lose flow to shallow groundwater and out of the system, respectively. The model validated at the main outlet of the watershed performing as a reservoir embayment was able to capture the measured monthly outflows during the wet periods. The measured monthly outflows of the flooded reservoir like embayment at the outlet were estimated by using a daily water balance approach with bathymetric survey and LIDAR data to obtain the volume of the embayment. This result showed the significance of the watershed karst features in the water balance with conduit and diffuse flow as a possible explanation for the missing flow from the SL-2 station that apparently appeared downstream at the cave spring as ground water flow, indicating a need to modify SWAT for accounting karst features for an accurate water balance. Secondly, the results also reflected on the effects of interactions of lake levels and voids/conduits of the karst features on the hydrology of the study watershed. These results may have large implication on estimates of nutrient and sediment loadings including TMDL development.

(Download PDF)    (Export to EndNotes)