The unsung role of SST in simulating mesoscale events: an evaluation of August 2018 extreme rainfall over Kerala using WRF model

In southern India’s State of Kerala during August 2018, a prolonged period of extreme rainfall resulted in devastating floods and landslides. In order to understand its influence on the modelling of this severe rainfall event, sea surface temperature (SST) from measurements and forecasts is updated at the lower boundary of a mesoscale model in this work. The one-way nested domains of the weather research and forecasting (WRF) model are used in this work. Optimum interpolation sea surface temperature (OISST) from the National Oceanic and Atmospheric Administration (NOAA) and coupled forecast system sea surface temperature (CFSSST) from the National Centers for Environmental Prediction (NCEP) are used to update SST during model integration. The control experiment, however, made use of skin temperature data from the final analysis (FNL) of the NCEP. When compared to observations, control simulations overpredict the patterns of rainfall. When OISST and CFSSST are used, the rainfall distribution could well be simulated fairly accurately. The overprediction shown in the control simulation may have been reduced as a result of the latent heat flux values being lower and closer to observation when OISST and CFSSST are used. When OISST and CFSSST are updated, the 850 hPa circulation fields are significantly improved. The work emphasizes the role of employing reliable time-varying lower boundary conditions, which can significantly affect the simulation of extreme precipitation events using a mesoscale model. The role of SST, as realized through the better representation of latent heat flux, might have resulted in the better simulation of heavy rainfall using WRF when forced with observed and forecasted SST.