渦旋植入對不同降水物理參數法颱風路徑系集預報之影響研究

The Influence of the Bogused Vortex on the Typhoon Track Ensemble Forecasts

周昆炫(Kun-Hsuan Chou) ； 黃柏智(Bo-Chih Huang)

《大氣科學》 38卷4期 (2010/12) Pp. 301-329

摘要

本研究使用WRF模式，選用四種不同積雲參數法及微物理參數法組合成16種不同降水物理過程產生路徑系集預報，並與世界主要全球預報模式NCEP、JMA、UK和EC等比較優劣。此外，由於目前全球模式資料所解析之颱風強度仍不足，本研究中額外使用渦旋植入法來加強颱風的初始強度，以期得到較佳的強度模擬，並比較較佳初始強度下所模擬的路徑系集預報有何差異。另外，特別分析不同降水物理過程參數法對於颱風強度、結構與路徑模擬之影響。
在沒有使用渦旋植入實驗中的研究結果顯示，不同降水物理過程參數法組合所得到的路徑系集預報，可產生扇形的路徑分布，可視為颱風未來移動方向的參考依據。此外，當積雲參數法使用Kain-Fritsch時的四組參數法組合，可以產生最佳的路徑系集預報；而積雲參數法選擇Kain-Fritsch及微物理法選擇WSM3時則可以獲得最小平均路徑誤差，為未使用渦旋植入模擬中最佳降水物理參數法之組合。另外，在使用渦旋植入實驗中的研究結果顯示，選用Grell-Devenyi積雲參數法及WSM3微物理參數法時可以得到最小平均路徑誤差，可作為渦旋植入後最佳之降水物理參數法組合。根據有無渦旋植入的路徑誤差分析顯示，當模式初始場中使用渦旋植入後，可有效縮小系集預報路徑的分布，並對於預報後期之路徑預報可有效改善。
至於在比較不同降水物理參數法所模擬之颱風強度、結構與環境流場分析顯示，當積雲參數法使用KF時可有效且快速加強颱風強度與大小，並容易偏北移動且移動速度較快；而當微物理參數法使用Kessler時，可能導致預報路徑易偏東移動。此結果主要為降水物理參數法對於太平壓高壓環境流場及颱風本身結構具有影響，進而使颱風路徑出現變化。

關鍵字

WRF ； 颱風路徑系集預報 ； 積雲參數法 ； 微物理參數法

並列摘要

A series of numerical simulations are conducted by the Weather Research and Forecasting model (V3.0.1) to examine typhoon track ensemble forecast. Four different convective parameterizations (CPs) and four different microphysical parameterizations (MPs) are chosen as the combination of the precipitation parameterizations. The track error statistics for all members, ensemble mean and four main operational global model forecasts (NCEP, JMA, UK, and EC) are also investigated in this study. Besides, due to the intensity of tropical cyclone analyzed by the global analyses is generally too weak to present the actual strength of tropical cyclone, the implantation of bogused vortex are applied in this study to obtain the better intensity simulation and examine its impact on the performance of the ensemble forecasts. Furthermore, the impact of different precipitations on the intensity and structure of typhoon as well as the environmental flows are also investigated.
Based on the experiments without the implantation of bogused vortex, the results show that ensemble forecasts by using different precipitation parameterization schemes could obtain a track spread that regard as an area that typhoon might approach. Furthermore, the ensemble track from four tracks of fixed Kain-Fritsch CP the better typhoon forecast track. Moreover, the combination of Kain-Fritsch CP and WSM3 MP precipitation parameterization could lead to the smaller averaged track errors and this result indicates that this combination of precipitation parameterization will provide the best track forecast. Meanwhile, for the experiments with the implantation of the bogused vortex, the Grell-Devenyi CP and WSM3 MP precipitation parameterizations could be the best combination of precipitation parameterization which leads to the smaller averaged track errors. Furthermore, the results show that the implantation of the bogused vortex efficiently shrink the spread of the ensemble tracks, and significantly improve the track forecast in the later part of the forecast period.
For the impact of different precipitation parameterization schemes on the simulated intensity and structure of typhoon as well as the environmental flows. The model tends to simulate more intense and larger size typhoon as well as obvious northward movement track when the Kain-Fritsch CP are chosen, while appears to simulate more weaker and smaller size typhoon as well as obvious eastward movement track when the Kessler MP are used. The reason is raised from that the steering flows are mainly dominated by the strength of the Pacific high and the structure of typhoon where are significantly influenced by the precipitation parameterization schemes in the numerical model.

並列關鍵字

WRF ； Ensemble forecasts ； Convective and Microphysical parameterizations