Sensitivity of WRF Model for Simulation of 2014 Massive Flood Over Kashmir Region: A Case of Very Heavy Precipitation

Abstract

The present study simulates the devastating floods in Kashmir that caused widespread damage in the valley from September 2-6 2014. The study used NCEP-NCAR FNL data for the initialization and simulation of the WRF ARW model. Statistical analysis of temperature over four places namely Anantnag, Srinagar, Pulwama, and Baramulla taking RMSE and MBIAS at 24, 48, 72, and 96 hours was also done against observed ECMWF-ERA5 temperature data. Further analysis of RMSE and MBIAS showed a minimum value at 48, 72, and 96 h indicating the improvement of prediction after 6 hours. Rainfall amount was under-predicted by the model with a time lag of 4 h while temperature time series over four districts were significantly closer to observation. Furthermore, the Model was able to capture the strong vertical velocities along with sufficient moisture content up to 600 hPa at the time of observed rainfall.

Keywords

• rmse, mbias, wrf model, vertical velocity