Assessment of SWAT Model Applicability and Performance for Predicting Surface Runoff in an Ungauged Watershed in the Philippines

Abstract

The Soil and Water Assessment Tool (SWAT) model, owing to its robustness and wide applicability, was tested in Mabacan watershed in Laguna, Philippines, with the objective of developing a model for predicting surface runoff or streamflow in ungauged watersheds under conditions of limited hydrologic data. Model parameterization was performed based on the physical characteristics of the watershed and guided by sensitivity analysis. SWAT-predicted monthly surface runoff values for a 7-year period were then compared with manually computed values obtained using water balance and SCS-Curve Number methods. Agreement between runoff values was evaluated using the coefficient of determination (R2) and Nash-Sutcliffe model efficiency (NSE). Correlation analysis between SWAT-simulated monthly runoff and observed monthly rainfall was also performed to further assess the plausibility of simulation results. Sensitivity analysis showed that that the curve number (CN2), soil evaporation compensation factor (ESCO), and soil depth (SOL_Z) are the most sensitive parameters with sensitivity index (I) of 2.92, 0.08, and 0.04, respectively. Comparison between SWAT-predicted surface runoff with those obtained using water balance and SCS-CN methods yielded R2 values of 0.95 and 0.94, and NSE values of 0.70 and 0.63, respectively. The SWAT-simulated runoff series also proved to be well-correlated with observed rainfall series (R2=0.82). Results of this study suggest that the SWAT model is highly applicable for predicting surface runoff or streamflow in ungauged watersheds in the Philippines even with limited hydrologic data and that familiarity with the physical characteristics of the watershed area is essential in enhancing model parameterization and performance.