模式初始條件的設定對於颱風的預報有很大的影響。為了增加颱風的初始強度，前人研究加入虛擬渦旋以增強颱風的強度。位渦為一包含熱力及動力結構之物理量，經過反演可以得到三維的動力及熱力結構。本篇研究基於Kieu and Zhang(2010)的研究，利用虛擬位渦擾動反演出三維風場、氣壓場及溫度場，並使用三維變分同化(3DVAR)將虛擬位渦反演渦旋同化至2010年9月15日12 UTC颱風凡那比(Fanapi)的初始場中，並與植入虛擬渦旋之模擬結果做比較。位渦反演研究顯示，位渦擾動可以增強颱風眼牆外的風速，降低颱風中心氣壓，且促進暖核的發展。加入虛擬位渦擾動可以增強颱風結構及強度，較強的位渦擾動，反演後的風場較大，颱風中心氣壓較低；較弱的位渦擾動，反演後的風場較小，颱風中心氣壓較大，位渦擾動強度與颱風最大風速及颱風中心最低氣壓呈線性相關。眼牆傾斜的位渦擾動不會改變反演結果的強度，但會改變颱風垂直結構的分布。 本研究進一步將上述的虛擬位渦反演渦旋同化至2010年9月15日12 UTC颱風凡那比(Fanapi)的初始場中。實驗結果顯示，同化虛擬位渦反演渦旋及植入虛擬渦旋可以改善颱風初始場的結構與強度，但同化包含溫度資料的虛擬位渦反演渦旋在初始場颱風外圍產生強烈的反氣旋。96小時模擬顯示，同化虛擬位渦反演渦旋的實驗都可以改善颱風凡那比的路徑及強度模擬，其中以同化風場及氣壓場的結果最好。植入虛擬渦旋的模擬結果在本次實驗中並不理想，未來可以做進一步的討論。位渦擾動分布敏感度實驗顯示，同化較強的虛擬位渦反演渦旋，模擬的颱風初始場較強，同化較弱的虛擬位渦反演渦旋，模擬的颱風初始場較弱，而同化眼牆傾斜的虛擬位渦反演渦旋對颱風初始場強度的影響並不明顯，但可以改變颱風垂直結構的分布。眼牆斜率愈大的虛擬位渦反演渦旋，同化後的颱風初始場眼牆斜率愈大。本實驗顯示，對於2010年9月15日12 UTC颱風凡那比而言，虛擬位渦反演渦旋強度較弱、眼牆傾斜率較大的模擬結果最好。整體來說同化位渦資料可以有效改善颱風路徑預報及初始場強度，若要發展成一套作業系統仍需要更多個案的測試。The specification of the initial typhoon vortex structure plays an important role in the prediction of typhoon. Conventionally, vortex bogussing with a Rankine vortex profile is usually applied to specify the initial conditions for typhoons, thus losing a vertically consolidated vortex structure. In this study, the 3-D balanced vortex structure is obtained using the potential vorticity (PV) inversion developed by Kieu and Zhang (2010). This inversion method may provide useful variables such as 3-D wind, pressure and temperature most of which can be assimilated by the WRF 3DVAR to improve the typhoon vortex for Fanapi (2010). In the PV inversion, PV anomalies (PVAs) enhance cyclonic flows in the eyewall and reduce the minimum pressure in the eye. Application of the bogus PVAs can improve typhoon vertical structure and intensity. The results show that typhoon intensity is more sensitive to change in the amplitude of PVAs than in eyewall titling rate, but the latter may improve typhoon vertical structure. The PV-inverted bogus vortex was assimilated to improve the initial typhoon vortex for Fanapi. Assimilation with the PV-inverted vortex shows a more organized vertical structure and leads to a remarkable improvement in Fanapi’s track prediction, as compared to assimilation with a conventional Rankine vortex. Assimilation with a stronger (weaker) PV-inverted vortex gives a stronger (weaker) initial typhoon vortex. Assimilation with an eyewall-titling PV-inverted vortex also shows an initial typhoon vortex with an eyewall more titling with height. This study shows that Fanapi’s prediction will be improved as the assimilated PV-inverted vortex is more similar to the observed initial typhoon vortex. The new vortex initialization developed in this study facilitates implantation of a slantwise eyewall as well as an asymmetric vortex, thus potentially leading to improvement on typhoon prediction.
