Coupling SWAT+ and SWMM Models to Quantify Streamflow in Mixed Rural–Urban Watersheds

Abstract

The fast-paced urbanisation has resulted in urban–rural transitional zones known as ‘sprawl’ that exhibit both urban and rural characteristics. This transformation significantly impacts water balance, leading to changes in surface runoff patterns in these mixed urban rural watersheds. The contrasting runoff generation patterns in urban and rural land uses pose significant challenges for accurately simulating runoff in these urban rural watersheds. To address these mixed watersheds, we developed an Integrated Environmental Modelling (IEM) approach to effectively represent complex hydrological behaviours in both urban and rural areas by integrating two widely used models: the Storm Water Management Model (SWMM) and the Soil and Water Assessment Tool Plus (SWAT+). The integrated model provides a means for coupling models, data storage, analysis, visualisation, and decision support systems. This framework links PYSWMM and PYSWAT+, offering automatic calibration of the integrated model in Python, which improves runoff simulation by considering distinct runoff generation patterns for each land use and the unique characteristics of each subcatchment. This study aims to improve our understanding of how runoff generation and streamflow behaviour vary across urban and rural areas by evaluating the performance of a coupled SWMM-SWAT+ model. A case study conducted in the urban–rural watershed of Stroubles Creek, Virginia, USA, demonstrated that the integrated model outperformed both standalone models in predicting streamflow during dry and wet periods. This study successfully shows that the SWMM-SWAT+ integration can effectively represent hydrological processes in mixed urban–rural watersheds, achieving higher accuracy than standalone models. The integrated model achieved higher accuracy than the standalone models, improving NSE from 0.51 (SWAT+) and 0.64 (SWMM) to 0.85, reducing RSR from 0.70 (SWAT+) and 0.60 (SWMM) to 0.39, and lowering PBIAS to 16.8%, compared to 37.87% (SWAT+) and −70.37% (SWMM), demonstrating its superior performance in streamflow simulation. These findings provide a practical and adaptable modelling framework for improving stormwater management, flood risk assessment, and sustainable water resources planning in transitional watersheds experiencing both urban and rural influences.