Simulation of pre-monsoon temperature over Bangladesh using dynamical downscaling with high Resolution wrf-arw model

Abstract

In the present study, the Weather Research and Forecast (WRF-ARW V3.5.1) model have been used to simulate the station wise pre-monsoon temperature during 2010 – 2014 over Bangladesh. The initial and boundary conditions are drawn from the global operational analysis and forecast products of National Center for Environmental Prediction (NCEP-GFS) available for the public at 1°×1o resolution. The model was configured in single domain, 6 km horizontal grid spacing with 161×183 grids in the east-west and north-south directions and 30 vertical levels. For the simulation of pre-monsoon temperature WSM6-class graupel scheme and Kain-Fritsch (KF) cumulus parameterization (CP) scheme has been used. Initially the model has run for 107 days for long term prediction starting with the initial condition of 0000 UTC of 17 February up to 0000 UTC of 1 June for the period 2010-2014. The model has also run for 72 hours with everyday 0000 UTC initial conditions for 94 days for the prediction of 24, 48 and 72 hours lead time temperature in the pre-monsoon season of 2014. In this research, 2m level temperature have been simulated at 3 hourly interval, then daily and monthly average temperature data have been determined for 24, 48, 72 hour and 107 days during the studied period. This data has been compared with the observed temperature at 33 meteorological stations of BMD. The simulated monthly average temperature of March, April and May 2014 for 24, 48 and 72 hours prediction has been found maximum in the western region whereas observed temperature has been found maximum in the southwestern region of the country. The distribution pattern of observed and model simulated monthly average temperature for 24, 48 and 72 hours prediction are almost similar all over except southeastern region of the country. The difference of observed and simulated temperature lies within 1-2oC all over except southeastern region. The observed temperature is found to be minimum in the northeastern region but 107 days predicted temperature simulated minimum in the southeastern region and the difference lies within 5-6oC in the southeastern region. The correlation coefficients (CC) between observed and simulated temperature have found maximum for 24 hour lead time prediction in the month of March. As the time of the season progresses and also the time of prediction have increased, the CC has decreased all over and this decrease is significant in the southeastern region of the country. 24-72 hour lead time predicted temperatures are in good agreement with the observed temperature.