Improvements to runoffpredictions from a land surface model with a lateral flow scheme using remote sensing and in situ observations

Abstract

Like most land surface models (LSMs) coupled to regional climate models (RCMs), the original Common Land Model (CoLM) predicts runofffrom net water at each computational grid without explicit lateral flow (LF) schemes. This study has therefore proposed a CoLM+LF model incorporating a set of lateral surface and subsurface runoffcomputations controlled by topography into the existing terrestrial hydrologic processes in the CoLM to improve runoffpredictions in land surface parameterizations. This study has assessed the new CoLM+LF using Earth observations at the 30-km resolution targeted for mesoscale climate applications, especially for surface and subsurface runoffpredictions in the Nakdong River Watershed of Korea under study. Both the baseline CoLM and the new CoLM+LF are implemented in a standalone mode using the realistic surface boundary conditions (SBCs) and meteorological forcings constructed from remote sensing products and in situ observations, mainly by geoprocessing tools in a Geographic Information System (GIS) for the study domain. The performance of the CoLM and the CoLM+LF simulations are evaluated by the comparison of daily runoffresults from both models with observations during 2009 at the Jindong stream gauge station in the study watershed. The proposed CoLM+LF, which can simulate the effect of runofftravel time over a watershed by an explicit lateral flow scheme, more effectively captures seasonal variations in daily streamflow than the baseline CoLM.