| dc.description.abstract | Integrating some publicly available data sets with the Soil and Water Assessment Tool (SWAT) can develop new approaches for water quality simulation. This study evaluated the individual and combined impacts of integrating Climate Forecast System Version 2 (CFSv2), Daily Surface Weather Data (Daymet), World Weather for Water Data Services (W3S), crop rotation layer based on multiple cropland data layers (CDLs), an evapotranspiration (ET) data set from Moderate Resolution Imaging Spectro-Radiometer (MODIS) products, and a snowmelt data set from the European Centre for Medium-Range Weather Forecasts (ECMWF) into SWAT on simulations of discharge, sediment and phosphorus at monthly and daily time steps for the Silver Creek watershed (105 km2) and Roy Creek watershed (11 km2), two subwatersheds of the Big Green Lake watershed, Wisconsin. The baseline SWAT utilized Global Historical Climatology Network-Daily (GHCN-D) for weather and a single-year CDL for land use. Integrating Daymet only outperformed other data sets in overall simulation accuracy of monthly and daily discharge, while maintaining accuracy in sediment and phosphorus simulation. Integrating crop rotation layer, ET, and snowmelt data sets improved monthly SWAT discharge simulation. The combined integration of Daymet, crop rotation, ET, and snowmelt data sets slightly improved or lowered the monthly SWAT accuracy of discharge, sediment, and phosphorus simulations compared with the integration of Daymet only while reducing uncertainty, showing mixed effects of integrating multiple data sets. Compared with MODIS ET data, uncalibrated SWAT underestimated summer ET peaks. The baseline calibration of monthly SWAT delayed snowmelt peaks. Results suggest that further improvement of crop rotation and snowmelt simulation are required to improve the daily SWAT model performance. | |
