Notice: Although this work was reviewed by USEPA and approved for publication, this scientific paper may not necessarily reflect official Agency policy. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.
Assessing SWAT's performance in the Kaskaskia River watershed as influenced by the number of calibration stations used
Article first published online: 20 NOV 2012
Copyright © 2012 John Wiley & Sons, Ltd.
Volume 28, Issue 3, pages 676–687, 30 January 2014
How to Cite
Chiang, L.-C., Yuan, Y., Mehaffey, M., Jackson, M. and Chaubey, I. (2014), Assessing SWAT's performance in the Kaskaskia River watershed as influenced by the number of calibration stations used. Hydrol. Process., 28: 676–687. doi: 10.1002/hyp.9589
- Issue published online: 7 JAN 2014
- Article first published online: 20 NOV 2012
- Accepted manuscript online: 11 OCT 2012 12:25PM EST
- Manuscript Accepted: 3 OCT 2012
- Manuscript Received: 17 DEC 2011
- Future Midwestern Landscapes study;
- single-site calibration;
- multi-site calibration;
- model validation
The Future Midwestern Landscapes (FML) project is part of the U.S. Environmental Protection Agency's Ecosystem Services Research Program. The goal of the FML project is to quantify changes in ecosystem services across the Midwestern region as a result of the growing demand for biofuels. Watershed models are an efficient way to quantify ecosystem services of water quality and quantity. By calibrating models, we can better capture watershed characteristics before they are applied to make predictions. The Kaskaskia River watershed in Illinois was selected to investigate the effectiveness of different calibration strategies (single-site and multi-site calibrations) for streamflow, total suspended sediment (TSS) and total nitrogen (TN) loadings using the Soil and Water Assessment Tool. Four USGS gauges were evaluated in this study. Single-site calibration was performed from a downstream site to an upstream site, and multi-site calibration was performed and fine-tuned based on the single-site calibration results. Generally, simulated streamflow and TSS were not much affected by different calibration strategies. However, when single-site calibration was performed at the most downstream site, the Nash–Sutcliffe efficiency (NSE) values for TN ranged between −0.09 and 0.53 at the other sites; and when single-site calibration was performed at the most upstream site, the NSE values ranged between −8.38 and −0.07 for the other sites. The NSE values for TN were improved to 0.5 – 0.59 for all four sites when multi-site calibration was performed. The results of the multi-site calibration and validation showed an improvement on model performance on TN and highlighted that multi-site calibrations are needed to assess the hydrological and water quality processes at various spatial scales. Copyright © 2012 John Wiley & Sons, Ltd.