Comparison of two different approaches for sensitivity analysis in Heihe River basin (China)

Abstract

Distributed watershed models should pass through a careful sensitivity analysis and calibration procedure before they are utilized as a decision making aid in the planning and management of water resources. Although manual approaches are still frequently used for sensitivity and calibration, they are tedious, time consuming, and require experienced personnel. This paper describes two typical and effective automatic approaches for sensitivity analysis and calibration for the Soil and Water Assessment Tool (SWAT). These included Sequential Uncertainty Fitting (SUFI-2) algorithm and Shuffled Complex Evolution (SCE-UA) algorithm. The results show that: (1) The main factor that influences the simulated accuracy of the Heihe River basin runoff is the Soil Conservation Service (SCS) runoff curve parameters; (2) SWAT performed very well in the Heihe River basin. According to the observed runoff data from 2005 to 2013, the determination coefficient R² of the simulation and the efficiency coefficient (Ens) of the model was higher than 0.8; (3) Compared with Shuffled Complex Evolution, the SUFI-2 algorithm provides almost the same overall ranking of the sensitive parameters, but it is found to require less time with higher accuracy. The SUFI-2 provides a practical and flexible tool to attain reliable deterministic simulation and uncertainty analysis of SWAT, it can lead to a better understanding and to better estimated values and thus reduced uncertainty.