The Impact of Microphysics Schemes on Typhoon Simulation

本研究透過加入氣膠(凝結核)活化過程的參數法以及使用double-moment bulk scheme，以了解不同雲微物理參數法對模擬颱風路徑、強度及結構的影響。針對2013年天兔(USAGI)及潭美(TRAMI)颱風，選用WDM6、Morrison以及CLR參數法，並在各參數法中設定不同氣膠(凝結核)數量濃度進行真實颱風個案模擬。 結果顯示，氣膠(凝結核)數量濃度初始設定與雲微物理參數法的差異皆會影響天兔與潭美颱風路徑、強度及微物理結構。不同雲微物理參數法模擬綜觀天氣系統造成的差異雖然不大，但由於各參數法在水物轉換過程設計的不同，造成較小尺度模擬結果的差異，進而導致模擬的颱風路徑、移速及強度不同。此外，比較兩個強度不同的颱風模擬結果，不同雲微物理參數法之間的差異，在強度較輕的潭美颱風之模擬上較為明顯。在颱風路徑方面也會透過不同雲微物理過程導致小尺度模擬的差異，造成各組路徑的分歧，其中潭美路徑的差異較大。颱風移動方向與速度之差異大致上可透過颱風潛熱釋放水平分布之不均勻性來解釋。颱風中心南側的潛熱釋放大於北側時，會造成颱風向南移動的分量增加。 在參數法之間差異的方面，WDM6參數法不管是在模擬天兔或是潭美的強度結果都是最強的，此結果與該參數法中雲水自動轉換為雨水過程較為快速以及雪轉換成冰雹後透過融化過程產生較多的雨水有關，造成颱風中心附近潛熱釋放較大。此外，CLR參數法設定三組不同背景氣膠數量的模擬結果差異最大。當氣膠濃度較高時，颱風外圍雨帶會形成較多但形狀較小的雲滴，透過垂直上升運動到高層外流區形成較多小的冰晶，而在颱風外圍雨帶的冰晶成長潛熱釋放使得颱風中心加熱效率變差，導致模擬的颱風強度較弱。 本研究最後進一步討論雲─輻射交互作用過程如何影響颱風路徑及強度。在模擬天兔個案中，各組大氣中的短波輻射淨值在南北側差異不大，而大氣中的長波輻射淨值南北側差異最大的是CLR_marine，導致該組颱風中心北側較冷、南側較暖，有利於颱風移動方向偏南；在潭美個案中短波輻射淨值有較明顯的南北差異，其中CLR參數法模擬北側大氣中的短波輻射淨值較其他兩種參數法模擬的結果大，該結果使颱風北側大氣底層短波熱通量增加，有利於路徑偏北，但長波向上輻射通量也增加，應不利颱風向北移動。儘管如此，綜合水物相態轉換潛熱釋放與大氣中長/短波輻射淨值的結果，最後仍使得CLR參數法模擬潭美的路徑結果較WDM6與Morrison參數法模擬的結果偏北。由於本研究實驗並未將輻射參數法的使用與否作為研究的重點，因此尚未能證實影響颱風強度結果的差異是否主要由雲微物理參數法設計各水物形成的差異所貢獻，亦或是雲─輻射交互作用過程才是影響颱風強度的主因。Through the use of double-moment bulk scheme with aerosol activation process, this study was expected to realize the impact of microphysics schemes on typhoon simulation. For simulating typhoon USAGI and TRAMI (2013), this study selected WDM6, Morrison and CLR schemes and set different background aerosols number concentrations. The results showed that background aerosols number concentrations and microphysics schemes (MPs) themselves would influence the tracks, intensities and microphysical structures of typhoon USAGI and TRAMI. The differences of the simulated synoptic-scale systems with those MPs were small. However, those MPs used in this study applied different methods to the formation and phase transform processes of hydrometeors, which might influence the latent heat distribution of typhoon. Therefore, that result would affect typhoon’s track, moving speed and intensity. Besides, comparing the simulation results of two typhoons with different intensities, the simulation results of typhoon TRAMI with weaker intensity exist larger differences from those MPs. In terms of typhoon’s track, the results of TRAMI exist larger differences from different MPs causing small scale differences. The asymmetric distribution of latent heat release may further influence the moving directions and speeds of the simulated typhoon. For instance, a stronger latent heat release at the south side leads to a southward movement of the simulated typhoon. In terms of different MPs, the intensities results of using WDM6 scheme to simulate typhoon USAGI and TRAMI exist the strongest intensity. This result was related to the rapid cloud drops autoconversion process and more rain from more graupel melting process which might cause more latent heat concentrating at the center of typhoon. Besides, from simulations using CLR scheme we demonstrated significant aerosol impact on typhoon. For instance, urban type aerosol tends to produce more but smaller cloud drops, which lead to stronger mixed-phase processes and thus a stronger latent heat release. This effect is more obvious at TC’s outer rainbands, such that the enhancement of the outer rainband disrupted the moisture transport into the TC center. In consequence, urban type aerosol makes the typhoon weaker than that with marine type aerosol. Finally, this study further discussed how cloud-radiative interaction processes may possibly influence the tracks and intensities of the simulated typhoon. In the USAGI case, shortwave radiation in the atmosphere did not show significant differences between the northern and southern parts of the typhoon center. However, CLR_marine showed larger differences in atmospheric longwave radiation between the northern and southern parts of the typhoon center. This resulted in cooling to the north of typhoon center but warming to the south, and this favored typhoon’s moving southward. In the TRAMI case, shortwave radiation in the atmosphere showed significant differences between northern and southern parts of the typhoon center. Among different microphysical schemes, the CLR scheme simulated more downward shortwave radiation to the north of the typhoon center, which favored northward movement of typhoon. However, that scheme also simulated more upward longwave radiation to the north of the typhoon center, which was unfavorable for northward movement of typhoon. The combined effect of latent heat and radiation effects using the CLR scheme simulated a net northward track in TRAMI case. However, this study did not include a full coupling of the radiation scheme with the ocean, so no definite conclusion can be made regarding the cloud-radiative effects on the track and intensity of typhoon