本研究以The Fifth-Generation NCAR / Penn State Mesoscale Model (MM5)並利用四維變分資料同化方法(4DVAR)，同化衛星與虛擬渦旋資料，模擬2008年卡玫基颱風。資料同化所使用的三種衛星觀測資料，包括：Special Sensor Microwave Imager (SSM/I)垂直水氣積分(Integrated Water Vapor)與海面風速資料(Ocean Wind Speed)及Constellation Observing System for Meteorology Ionosphere & Climate (COSMIC)折射率(Refractivity)資料，期望同化觀測點分布較廣的衛星資料可對模式大氣有較大範圍之修正。經過同化虛擬渦旋之後，可在初始颱風環流中高層得到暖心結構、颱風中心附近亦有水氣輻合，同時能夠有較準確的颱風中心定位，進而改善颱風路徑與強度之模擬。 模擬實驗分為三組，第一組實驗中分別同化不同衛星資料，結果顯示，單獨同化衛星資料無法有效改善颱風的模擬路徑與強度。第二組實驗則只同化虛擬渦旋，當加入虛擬渦旋後可明顯改善卡玫基颱風在第二天與第三天之模擬路徑誤差。第三組實驗，在4個小時的同化窗區內，同時同化衛星與虛擬渦旋資料。實驗結果顯示，同化虛擬渦旋與COSMIC折射率資料，對於駛流場有較大之修正而更降低路徑誤差。當虛擬渦旋再加入同化SSM/I海面風速資料，則能夠維持颱風底層風場進而增加海氣交互作用產生較多潛熱，提供颱風發展較佳的環境條件。同化虛擬渦旋與垂直水氣積分之模擬，因水氣場受到修正，造成颱風的不對稱結構而影響颱風後期發展。 This study uses MM5 4DVAR to assimilate bogus vortex and satellite data to simulate Typhoon Kalmaegi (2008). There are three types of satellite data assimilated in the experiments. They are Special Sensor Microwave Imager (SSM/I) integrated water vapor (IWV), ocean wind speed (OWS) and Constellation Observing System for Meteorology Ionosphere & Climate (COSMIC) refractivity (REF) data. After applying bogus vortex data assimilation (BDA) based on 4DVAR, a cyclonic circulation with a warm core in the upper troposphere is produced. Moreover, moisture convergence near the typhoon center is induced and the location of the typhoon center is more accurate. Consequently, the simulated track and intensity of Kalmaegi (2008) were improved in these experiments with BDA. Three group experiments are designed to study the impact of different observations. In the first group experiment, SSM/I IWV, OWS and COSMIC REF are assimilated into the Kalmaegi case individually. However, there is no significant impact on Kalmaegi simulation when applying satellite data only. In the second group experiment, only the bogus vortex data are assimilated. When assimilating the bogus vortex data, a great improvement on the track simulation is found, especially after 24 simulation hours. The third group experiment assimilates not only bogus vortex data but also different satellite data. The results indicate that the bogus vortex with REF data leads to an adjustment of the steering flow and successfully improves the simulated track. On the other hand, the assimilation of the bogus vortex with SSM/I OWS data can enhance the low-level wind, resulting in more latent heating. Therefore, it generates an environment more favorable for typhoon development. However, the simulation with assimilation of SSM/I IWV causes the correction of water vapor. In simulation of Typhoon Kalmaegi, such correction results in a more asymmetric structure of wind field and latent heat flux which appears to weaken the typhoon intensity.
