P8 Realtime WRF Ensemble-RTFDDA Forecast with Downscaling of Multiple Global Models for Electric Grid Applications

Cheng, William Y.Y., Yubao Liu, Gregory Roux, Linlin Pan, and Yuewei Liu, National Center for Atmospheric Research, Shuanglei Feng, Shuanlong Jin, Ju Hu and Chun Liu, State Grid Corporation of China, China

The WRF-based mesoscale multi-physics, multi-land surface perturbation and multi-scale (nested-grid) ensemble realtime four-dimensional data assimilation (E-RTFDDA) and forecasting system is currently being used to support electric power applications for the State Grid Corporation of China (SGCC). These applications include predicting icing on power lines, lightning threats on transmission grid, and wind power. Currently, there is no well-defined method to construct a mesoscale ensemble forecast system. Physics-based, observation and data assimilation perturbations, and stochastic perturbations are the main approaches for building mesoscale ensemble members. However, in general, they still do not generate enough spread in the mesoscale ensemble forecast. It is hypothesized that perturbations large-scale model forecasts that are used to drive mesoscale ensemble models (through the initial/boundary conditions) may have more important impact on the ensemble forecast performance. By using different available global models to initialize different mesoscale ensemble members, it may be possible to generate more desired spread in the ensemble forecast. In particular, it may be possible to generate the representative probability density function (PDF) in the mesoscale ensemble forecast with the fewest ensemble members by this method. This study investigates the ensemble forecast generated from the perturbations in the initial/boundary conditions of four different global models: GFS, GEM, ECMWF, and GSM in the E-RTFDDA for several flooding events in China. Comparisons with satellite and rain gauge will be considered.