SWAT ungauged: Water quality modeling in the Upper Mississippi River Basin

• Published in: Journal of hydrology (Amsterdam) Vol. 584; p. 124601
• Main Authors: Qi, Junyu, Zhang, Xuesong, Yang, Qichuan, Srinivasan, R., Arnold, Jeffrey G., Li, Jia, Waldholf, Stephanie T., Cole, Jefferson
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.05.2020
• Subjects: algorithms; basins; climate; conservation practices; energy; hydrologic models; land use change; Mississippi River; model validation; Nitrate; nitrate nitrogen; pollution load; Sediment; sediments; soil organic matter; soil temperature; Stream flow; Total nitrogen; water quality; Watershed; watersheds; Mississippi River; Stream flow; Sediment; Nitrate; Total nitrogen; Watershed
• ISSN: 0022-1694; 1879-2707; 1879-2707
• DOI: 10.1016/j.jhydrol.2020.124601

•New soil energy and biogeochemistry algorithms are incorporated into SWAT.•Process-based algorithms outperform empirical methods in a large ungauged basin.•Analyzed contributions of energy and biogeochemistry algorithms to performance improvement.•Advancing process representation helps improve water quality modeling in ungauged basins. Improving model performance in ungauged basins has been a chronic challenge in watershed model application to understand and assess water quality impacts of agricultural conservation practices, land use change, and climate adaptation measures in large river basins. Here, we evaluate a modified version of SWAT2012 (referred to as SWAT-EC hereafter), which integrates an energy balanced soil temperature module (STM) and the CENTRUY-based soil organic matter algorithm, for simulating water quality parameters in the Upper Mississippi River Basin (UMRB), and compare it against the original SWAT2012. Model evaluation was performed for simulating streamflow, sediment, and nitrate-N (NO3-N) and total nitrogen (TN) loadings at three stations near the outlets of UMRB. The model comparison was conducted without parameter calibration in order to assess their performance under ungauged conditions. The results indicate that SWAT-EC outperformed SWAT2012 for stream flow and NO3-N and TN loading simulation on both monthly and annual scales. For sediment, SWAT-EC performed better than SWAT2012 on a monthly time step basis, but no noticeable improvement was found at the annual scale. In addition, the performance of the uncalibrated SWAT-EC was comparable to other calibrated SWAT models reported in previous publications with respect to sediment and NO3-N loadings. These findings highlight the importance of advancing process representation in physically-based models to improve model credibility, particularly in ungauged basins.