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.

SWAT Discharge Response to Spatial Rainfall Variability in a Kansas Watershed

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

Citation:  Transactions of the ASABE. 53(1): 65-74. (doi: 10.13031/2013.29503) @2010
Authors:   P. Tuppad, K. R. Douglas-Mankin, J. K. Koelliker, J. M. Shawn Hutchinson
Keywords:   NEXRAD, Resolution, Spatial rainfall variability, SWAT modeling

The spatial resolution of input rainfall data affects hydrologic model response and accuracy. The objective of this study was to evaluate the effects of spatial rainfall aggregation on predicted streamflow at five streamgage sites in the 6,316 km2 Smoky Hill River/Kanopolis Lake watershed in central Kansas. NEXRAD Stage III hourly rainfall estimates were accumulated for 24 h periods for 1995 through 2002. The original NEXRAD grid cells, approximately 4 4 km, were aggregated with incrementally coarser spatial scale resolutions of 8, 16, 32, 64, 128, and 256 km. These data along with interpolated ground-based raingage data were used to evaluate the influence of rainfall input type and spatial scale on streamflow modeled with the Soil and Water Assessment Tool (SWAT) model (2000 version). Interpolated ground-based rainfall data improved daily streamflow simulation compared to the SWAT default method of assigning subwatershed rainfall from the nearest raingage. Streamflow simulation improved further when NEXRAD-derived rainfall data were input for most of the aggregated grid-cell resolutions. However, the best model performance was not for the finest (4 km) grid resolution, but rather at resolutions ranging from 32 km to 128 km, depending on the location of streamflow measurement within the watershed. Greater variability in model performance was observed among the five streamgage sites within the watershed than among model runs using a range of aggregated grid-cell resolutions. These results indicate that greater rainfall spatial resolution for interpolated raingage data or, to a greater extent, aggregated NEXRAD precipitation data has the potential to improve SWAT simulation results compared to the typical use of nearest gage data.

(Download PDF)    (Export to EndNotes)