Effect of Rainfall Station Density, Distribution and Missing Values on SWAT Outputs in Tropical Region

Mou Leong Tan,Xiaoying Yang
DOI: https://doi.org/10.1016/j.jhydrol.2020.124660
IF: 6.4
2020-01-01
Journal of Hydrology
Abstract:Rainfall is one of the main inputs to the Soil and Water Assessment Tool (SWAT) model. Any error in the rainfall data could transfer to the model outputs and lead to inaccurate judgements and decisions. This study aims to evaluate the effect of spatial (station number and location effect) and temporal (randomly and selective missing values) rainfall scarcities on SWAT streamflow simulations. A total of 58 rainfall data scarcity scenarios was incorporated into a calibrated SWAT model to test the effect on monthly streamflow simulations of the Kelantan River Basin (KRB), Malaysia, during the period of 1980 to 1999. This study provides new insights of rainfall station distance, missing values in specific periods and different streamflow intensities elements on SWAT-based rainfall data scarcity (SWAT-Rainfall) assessment framework, with new evaluation indicators of extreme flows, trend and magnitude changes analysis. The results show that (1) a minimum of four stations is needed to achieve accurate monthly streamflow in the KRB; (2) distance of rainfall stations to the streamflow station has a considerable impact on SWAT modelling; (3) temporal rainfall scarcity has a large impact on general and extreme streamflow simulations compared to spatial rainfall scarcity; (4) missing rainfall values of more than 20% in rainfall data would significantly affect tropical streamflow simulation; (5) the impact of missing rainfall values during the high flow period is lower than those in the moderate and low flow periods; and (6) excessive missing value in rainfall data would lead to errors in trend analysis. A special caution need to be taken in the selection and pre-processing of rainfall data before their incorporation into SWAT. Future works include integration of satellite precipitation products in SWAT, improvement of SWAT weather generator and development of a SWAT-based rainfall data pre-processing scheme to reduce rainfall data uncertainty.
What problem does this paper attempt to address?