Estimation of watershed hydrologic processes in arid conditions with a modified watershed model

- Three new functions were proposed to refine a watershed model under arid conditions.
- Each of the added algorithms improve the model accuracy in low-flow months.
- These alternative algorithms have potential application in other watershed models.

SummaryWatershed models play an important role in modern water resource management, increasingly demanding a robust hydrologic data framework to estimate watershed hydrochemical processes. The Generalized Watershed Loading Function (GWLF), a typical watershed model with modest data requirements, has been applied to watershed-scale hydrochemical estimation worldwide. However, while it generally successfully estimates flows in humid regions, the model suffers from a weakness in hydrologic estimation during low-flow periods, which are projected to continue increasing with global climate change in many places. To address this issue, three algorithms describing functional responses of flows to saturated water storage, the segment function approach, linear function approach, and exponential function approach, have been proposed in this paper, integrated with a previous leakage mechanism for unsaturated water storage used in two earlier GWLF versions, and applied to a case study of Shuai Shui River watershed in China. Comparisons of this version, including new algorithms or algorithm linkages, with the earlier GWLF versions, show that all the new algorithms improve model accuracy in low-flow months; the linear function approach linking the leakage process has the best effect. This work refines the framework of GWLF model to address both humid and arid conditions that can be used as alternatives for future applications. These new functional dynamic responses should also have potential application in other similar watershed models.