Multi-variable sensitivity analysis, calibration, and validation of a field-scale swat model: Building stakeholder trust in hydrologic and water quality modeling

Authors: KARKI, RITESH - Auburn University; SRIVASTAVA, PUNEET - Auburn University; Bosch, David; KALIN, LATIF - Auburn University; LAMBA, JASMEET - Auburn University; Strickland, Timothy

Submitted to: American Society of Agricultural and Biological Engineers
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/13/2019
Publication Date: 5/12/2020
Citation: Karki, R., Srivastava, P., Bosch, D.D., Kalin, L., Lamba, J., Strickland, T.C. 2020. Multi-variable sensitivity analysis, calibration, and validation of a field-scale swat model: Building stakeholder trust in hydrologic and water quality modeling. American Society of Agricultural and Biological Engineers. 63(2):523-539. https://doi.org/10.13031/trans.13576.
DOI: https://doi.org/10.13031/trans.13576

Interpretive Summary: An understanding of the impacts of irrigation and nutrient best management practices on hydrology, water quality, and agricultural productivity is critical for optimization of land management. Models of these processes can expand our understanding of these systems. This study evaluated the ability of the SWAT model to simulate runoff, soil moisture, cotton and peanut yield, and nitrate transport in conventionally- and strip-tilled plots while also evaluating the differences in hydrological and nutrient simulation parameters in the two tillage practices. Modeling results showed that SWAT can adequately simulate runoff, soil moisture, cotton and peanut yield, and nitrate at the field scale. Sequential calibration of surface runoff, crop yield, soil moisture, and nitrate showed that crop yield can be an important consideration for improving SWAT model robustness in nutrient transport simulations. Soil moisture calibration did not have a significant effect on runoff simulations. The impacts of different management scenarios on crop yield, irrigation water use, and nutrient loss were also evaluated. The results demonstrated to stakeholders that the SWAT model can successfully quantify the impacts of different management scenarios on their farm fields.

Technical Abstract: Multi-variable calibration of a field scale Soil and Water Assessment (SWAT) model is critical for understanding the true impacts of irrigation and nutrient best management practices on hydrology, water quality, and agricultural productivity, and for building stakeholder trust for its eventual implementation at the watershed scale. This study evaluated the ability of the SWAT model to simulate runoff, soil moisture, cotton and peanut yield, and nitrate in conventionally- and strip-tilled plots while also evaluating the differences in hydrological and nutrient simulation parameters in the two tillage practices. Modeling results showed that SWAT can adequately simulate runoff, soil moisture, cotton and peanut yield, and nitrate at the field scale and that calibrated values for the Curve Number Of Operation (CNOP) were different for the conventionally- and strip-tilled plots and critical to runoff calibration. We found that it was also important to change the routing method from Variable Storage to Muskingum and adjust DIS_STREAM for runoff simulation, if the fields were to be simulated as a watershed rather than as an HRU. Sequential calibration of surface runoff, crop yield, soil moisture, and nitrate showed that crop yield can be an important consideration for improving SWAT model robustness in nutrient transport simulations, although soil moisture calibration did not have a significant effect on runoff simulations. We also evaluated the impacts of different management scenarios on crop yield, irrigation water use, and nutrient loss, and demonstrated to stakeholders that the SWAT model can successfully quantify the impacts of different management scenarios on their farm fields.