GPU Based Acceleration of WRF Model: A Review

The Weather Research and Forecasting model (WRF) is a simulating system developed for atmospheric weather prediction. WRF model is used for both operational as well as research purposes. The need for accurate weather and climate simulation to be carried out in shorter time is increasing day by day, which leads to the acceleration of existing Numerical Weather Prediction (NWP) system. This paper gives a brief review of implementation of WRF model on Graphical Processing Units (GPUs). The works done in last few years are compared, which shows that the GPU implementation is feasible and worthwhile for accelerating the performance of WRF model. The GPU implementation can be done in two ways i.e. using CUDA or by using OpenCL. The OpenCL implementation of WRF model gives the option of platform portability. The CUDA implementation of a single module of WRF i.e. WRF Single Moment-5 (WSM5) gives the speedup of 1.3x for the whole WRF model. OpenCL implementation is done for another modules named scalar advection module, which shows the speedup of 7x for the whole WRF model.