Preliminary Application of a Multi-Physical Ensemble Transform Kalman Filter in Precipitation Ensemble Prediction

The optimal initial perturbation method using the ensemble transform Kalman filter (ETKF) is a point of intense popular interest in ensemble prediction. However, problems remain with respect to short-term ensemble prediction, such as insufficient ensemble spread, too large a prediction error, and so on. In this study, multi-physical parameterizations and boundary perturbations were introduced into the initial ETKF, and a heavy rainfall event that occurred in Hainan Province during 30 September to 6 October 2010 was simulated, as an example, using the single-physical ETKF and multi-physical ETKF in WRF3.5. The main results were as follows: All ensemble schemes outperformed the contrast forecast, with the multi-physical ETKF found to be the best. The RMSE and ensemble spread were well improved. For the multi-physical ETKF, the improvement in the location of heavy rain was obvious. The results indicate that the introduction of a variety of physical processes in the initial perturbations for the ETKF could significantly amplify the ensemble spread and improve the ensemble forecast of each quantity. The application of the physical ETKF method may have great potential in precipitation ensemble prediction.