Runoff Parameterization Enhances Regional Climate Model Accuracy for the South Asian Summer Monsoon

The land surface processes that affect climate models across South Asia have been significantly influenced by the runoff parameterization. In this study, we investigated the sensitivity of summer monsoons to runoff parameterizations in a non-hydrostatic Regional Climate Model (RegCM) during 1982-2018 across South Asia, specifically over the Indian Summer Monsoon (ISM). The experiments comprised parameterizations, namely, the SIMTOP control (CTL) scheme and the Variable Infiltration Capacity (VIC) scheme. Both simulations were forced with an ERA Interim Reanalysis and evaluated with respect to the observation and reanalysis products. The evaluation of relevant hydrometeorological variables investigated the effect of the implementation of the VIC scheme in the simulation. The relevant parameters to the land surface processes have been evaluated with spatio-temporal variabilities with respect to the change in runoff parameterization. The results revealed that the VIC simulated rainfall was in better agreement with the gridded rainfall of the Indian Meteorological Department (IMD) than the CTL experiments. The temperature features were simulated relatively well in VIC compared to CTL, particularly over the monsoon core region. Similarly, the spatial characteristics of soil moisture, relative humidity, total cloud cover, surface net radiation, sensible heat flux, and latent heat flux also performed better with VIC than with the ECMWF Reanalysis fifth version (ERA5) data. Furthermore, extreme monsoon rainfall years (excess and deficit monsoons) were better simulated in the VIC-driven experiment. The interannual variations in rainfall, soil moisture, sensible heat flux, and latent heat flux further explain the dominance of the VIC simulations over CTL, which is consistent with the aforementioned findings. Furthermore, these simulations were validated with a statistical skill score called the Nash-Sutcliffe coefficient and mean absolute error. These statistical analyses further corroborated that the ISM simulations performed better with the VIC runoff parameterization than with the CTL simulations. Therefore, the utilization of VIC parameterization is suggested for simulating the regional climate characteristics over South Asia.