Ensemble Forecasting Of TC Track And Intensity Based On ETKF Initial Perturbation Scheme

Ensemble prediction mainly includes three technical issues:generating initial perturbation, using numerical models, extracting useful information from ensemble forecasting. The initial field directly affects the quality of numerical prediction, especially for such short-term small scale system like typhoon. The Ensemble Transform Kalman Filter (ETKF) scheme is a new method to generate initial perturbation, which may reflect the observational network information, and the error variance can be kept on all relevant directions.At present, the study of applying the ETKF method to Typhoon ensemble prediction is also less. In this paper, the WRF ensemble prediction system is established based on ETKF initial perturbations scheme, to study the quality and structure of the initial perturbation field, and to conduct typhoon "Sura" track and intensity ensemble forecasting experiment. The results showed that, the ensemble prediction system using ETKF method can maintain a more reasonable spread, and the significantly large value is in the vortex region; the ensemble mean forecast are superior to the control forecast on the effect of elements fields; the improvements of typhoon intensity forecast are limited, the forecast results need to be improved; for the prediction of typhoon track, there maintains considerable divergence between the ensemble forecast members of the typhoon track, Ensemble mean forecast for moving track and landing site were significantly improved, is a better forecasting.Then an ensemble prediction System based on Breeding of Growing Mode (BGM) method was established to conduct typhoon "Sura" track and intensity ensemble forecasting experiment in comparison with ETKF methods. The results showed that, divergence velocity of temperature and wind fields are smaller by using the BGM method. For the forecasting results of wind and temperature fields, ETKF method is better than the BGM method. For the prediction of typhoon intensity, the improved error rate of ETKF method is higher than that of BGM method, however, the forecast results of two methods are not satisfactory. For the prediction of typhoon track, the divergence between each member is limited, ETKF method is considerably better than the BGM method.