Enhancing ecohydrological simulation with improved dynamic vegetation growth module in SWAT

Abstract

The Soil and Water Assessment Tool (SWAT) model has been widely used for simulating ecohydrological processes such as streamflow and evapotranspiration (ET) under different land use and climate conditions. The growth of vegetation is a crucial link in the ecohydrological process. However, the model uses a simplified Environmental Policy Impact Climate (EPIC) model, which has limitations in simulation of dynamic vegetation growth. Therefore, this study aims to enhance SWAT by incorporating the forest aging data and optimizing the dormancy and forest vegetation growth modules. The application of this modified model with a case study shows that the accuracy of leaf area index (LAI) simulation has significantly improved (coefficient of determination (R2) increased by 0.16, the Nash-Sutcliffe efficiency (NSE) increased from −0.18 to 0.87, and the absolute value of percent bias (PBIAS) decreased by 50.99%), and with moderate improvement in streamflow simulation (NSE increased by 0.03, PBIAS decreased by 4.60%). Furthermore, optimizing the dynamic growth model of vegetation led to significant increases in LAI (151.58%) and ET (5.25%) values in the forest area, while streamflow and water yield decreased by 4.28% and 4.16%, respectively, across the watershed. The results indicate that dynamic vegetation growth led to increased ET, resulting in a decrease in streamflow and water yield. Overall, our results suggest that the modified SWAT model effectively simulated the process of forest growth from young forest to mature stages, significantly improving the accuracy of simulated hydrological processes. This methodology can be applied to other watersheds to simulate ecohydrological processes and can be valuable in management of ecosystems and water resources at watershed scale.