Assessing the Impact of Spatial Resolution on Land Surface Model Based on Hydrologic Simulations

Abstract

Though land surface models (LSMs) are originally developed for representing water fluxes, carbon fluxes and energy fluxes between land and atmosphere, recently LSMs are being used for hydrological simulation because it has some positive traits in comparison to conventional hydrological models. In this study, the Joint UK Land Environment Simulator (JULES), a land surface scheme of the Met Office Unified Model (UM), is implemented to study the effect of different spatial resolutions on streamflow simulation at the Krishna River basin (catchment area 2,60,000 km²), India. The meteorological datasets used here are WFDEI (WATCH-Forcing-Data-ERA-Interim) global data at the resolution of 1° × 1° and 2° × 2°. The simulation is run for 2001–2008 period with 3 years (2001–2003) of spin-up with 50 spin-up cycle and further simulation of years from 2004–2008. To assess the performance of the stream flow simulation, appropriate statistical parameters such as mean error (ME), root-mean-square-error (RMSE), percentage BIAS (PBIAS) are used to see error statistics. The study results indicate that spatial resolution of routing, driving and ancillary data has a significant effect on model output at the river basin scale. Though we have implemented the simulations in coarser resolutions, such studies on hydrological fluxes with a change of spatial resolution are important to know associated uncertainties with driving data and routing data resolution selection, this, in turn, would also help researchers to make meaningful management decisions to deal with future water security issues.