Evaluating SWAT + model uncertainties for human and natural outcomes: Application in a Great Lakes agricultural watershed

Dataset · Mon May 20 00:00:00 EDT 2024

DOI:https://doi.org/10.57931/2352241· OSTI ID:2352241

Bunyon, Enock ^[1]; Hadjimichael, Antonia ^[2]; Son, Kyongho ^[3]; Chowdhury, Pranab Roy ^[3]; Deines, Jillian ^[3]; Hetland, Robert ^[3]

---

1. Pennsylvania State University; Pacific Northwest National Laboratory
2. Pennsylvania State University
3. Pacific Northwest National Laboratory

Nutrient exports from agricultural lands in the Great Lakes Region pose significant threats to water quality and ecological health through eutrophication, hypoxia, and harmful algal blooms. Climate change and agricultural adaptation practices complicate future nutrient loading due to intensified hydrologic cycles and land use decisions. Our research focuses on evaluating the Soil and Water Assessment Tool (SWAT) plus model parametric uncertainties for human and natural outcomes across different scales. These factors are integral to ensuring a balance between productive agricultural practices and maintaining the health of watershed hydrology. However, uncertainties in modeling such complex interactions pose significant challenges, limiting our ability to precisely determine critical factors that influence crop yield and soil moisture.     Our analysis employs Sobol global sensitivity analysis to evaluate first-order, second order, and total-order indices for SWAT crop growth parameter, ensuring comprehensive assessment of individual and interactive effects on model outputs. The objective is to identify the parameters that significantly affect model outputs for crop yield and soil moisture and improve our understanding of their interactions at the basin and hydrological response unit (HRU) scale. Our case study, the Portage River Watershed, which drains into Lake Erie, is chosen to better capture finer scale interactions crucial for predicting nutrient loading under future climate scenarios. This foundational work is aimed at setting the stage for the future development of an agent-based model (ABM). The ABM model would incorporate SWAT outputs to dynamically simulate decision-making processes.  

• Research Organization: MultiSector Dynamics - Living, Intuitive, Value-adding, Environment
• Sponsoring Organization: USDOE Office of Science (SC), Biological and Environmental Research (BER)
• OSTI ID: 2352241
• Country of Publication: United States
• Language: English