A study on availability of ground observations and its impacts on bias correction of satellite precipitation products and hydrologic simulation efficiency

Abstract

Precipitation is a crucial input for hydrological models to achieve various purposes such as water resource management and flood forecasting. However, precipitation stations are usually limited in number and eccentric in location. Although satellite precipitation products (SPPs) have improved remarkably, their application to data-sparse areas for flow simulation has significant limitations due to insufficient ground observations for bias correction. This paper proposes an integrated methodology from precipitation to flow simulation in a data-scarce basin by studying the impact of a precipitation gauge network on average areal precipitation (AAP) estimation, SPPs correction, and resultant hydrological simulation efficiency. To this end, seventeen observation network patterns were configured from a data-sufficient basin, the Fuji River Basin (FRB), Japan. Four bias correction methods were conducted for six SPPs under seventeen configurations of gauge network. A total of 264 AAP and 346 discharge results were comprehensively evaluated in combination with various gauge network configurations, bias correction methods, and hydrological simulation schemes. We found that, on bias correction methods, the proposed dynamic use of a statistical bias correction method (SDBC) can be the most effective, significantly improving the accuracy of SPPs even with a small number of eccentrically set observation stations. This study also revealed that four precipitation stations in the FRB (around 900 km2/station) could work as well as networks with more than ten or twenty stations for AAP estimation and flow simulation at the outlet. In addition, it is confirmed that the hydrological model could be well-calibrated individually by different precipitation inputs to fit the precipitation pattern. Those findings provide a robust basis for the use of SPPs in many poorly gauged basins. Also, the results on the configuration of networks provide scientific suggestions for properly deploying ground observation networks in ungauged basins for the effective use of SPPs.