Ensemble Simulation of Rainfall in Typhoon Sinlaku (2008) Based on EnKF Data Assimilation

Typhoon Sinlaku (2008) produced heavy rainfall over the mountain- area of Taiwan. The simulation and forecast of typhoon induced rainfall near Taiwan is rather challenging due to the interaction between typhoon circulation and the complicated topography of Taiwan. Wu and Kuo (1999) pointed out three challenging issues in the understanding and prediction of typhoons near Taiwan. First, the inadequate observations do not provide an accurate description of the structure and intensity of typhoons. Second, the prediction models have insufficient horizontal resolution to provide detailed descriptions of mesoscale wind and rainfall distribution. Finally, the effects of the Central Mountain Range on the movement and structure of a typhoon are very complicated.

During T-PARC (THORPEX-Pacific Asian Regional Campaign; Elsberry and Harr 2008) in 2008, abundant aircraft observation data was taken on Sinlaku. With a newly-proposed tropical cyclone vortex initialization method based on the ensemble Kalman filter (EnKF), a high-resolution and dynamically balanced vortex structure have been constructed in the Weather Research and Forecasting (WRF) model. Taking advantage of the dataset, this study focuses on the ensemble simulation of Sinlaku during its landfall period. The simulated ensemble mean track is consistent with the observation while there is a wide spread of the tracks among ensemble members. By analyzing all 28 members, the predictability of the rainfall associated with Sinlaku is examined. It is found that members with the track close to the best track are able to predict the realistic rainfall distributions though the total rainfall amounts and their cumulative frequencies are underestimated. In contrast, members with southward-biased tracks display higher rainfall amounts and cumulative frequencies which are more similar to the observation than those with more accurate tracks. Although the rainfall amounts and cumulative frequencies of the southward-biased members are consistent with the observation, the rainfall patterns are not consistent with those observed. In the simulation with finer resolution, the cumulative frequency remains nearly unchanged while the maximum rainfall amount increases.

In conclusion, for typhoons making landfall on Taiwan, the track is a critical factor in rainfall simulation, which is consistent with previous studies. It is important to show that through the ensemble simulation, the uncertainties which cause different rainfall patterns or amounts can be addressed from the variation of the ensemble members, thus providing more insights into the rainfall predictability associated with typhoons near Taiwan.