Impact of Land Surface Models on Simulation of Extreme Rainfall Events over Upper Catchments of The River Indus

Abstract

Water management in arid and sub-arid regions is critical and it could turn out to be precarious over the mountainous regions like Upper Indus basin due to related hazards; flash flooding, debris flow, land sliding, etc. The WRF-ARW (v5.3.1) modelling framework is coupled with Unified Noah and Noah-MP land surface models (LSM) with various physical schemes and domain setup. This setup helps to investigate the impact of LSM physics on rainfall simulation over the upper Indus basin and their possible links with landuse datasets. We use GIS platform to modify and rectify the landuse datasets to investigate the impact on rainfall simulations with different physical parameterizations in WRF modelling setup. In nested domain, all simulations coupled with NoahMP shows negative bias (i.e. overestimation) for rainfall over the mountain areas, while there is no significant impact on storm structure when the single domain is considered for both unified Noah and Noah-MP LSM. Furthermore, Noah-MP improves the rainfall simulations over the complex terrain, and there is no substantial impact in the plains of River Indus. Moreover, updated landuse and corrected elevation dataset also slightly improves the simulation qua lity and MODIS landuse shows marginally negative bias in urban areas as compared to default USGS landuse dataset.