本研究利用WRF模式模擬柯羅莎颱風，探討西北向侵台颱風登陸前路徑打轉的機制。控制組實驗（CTR實驗）模擬結果和之前研究結果一致，颱風在靠近地形時，颱風西側風速有增強的情況，顯示出通道效應的影響。此外，在地形的東南側有一個正渦度場生成，伴隨渦度變率極大值，導致颱風中心向南偏移。由擾動氣壓可知，正渦度場可能是由於此處的低壓所造成。除了通道效應，台灣東南側的低壓也是造成颱風路徑南偏的機制。根據空氣質點軌跡顯示，部份氣流過山後下沉增溫，造成低壓之生成。根據渦度收支分析，渦度平流為此渦度變化的主要貢獻。本研究另外使用中央大學中尺度模式進行理想個案的模擬，探討西北向颱風渦旋侵襲理想地形的路徑變化。模擬結果顯示，理想個案和上述真實個案有相似的機制，出現了通道效應以及在地形東南側的低壓，因此打轉機制不只是為柯羅莎颱風所特有。但颱風登陸位置若南移一些，即無低壓於地形東南側生成，颱風於登陸時並無向西南方偏移之現象。而在初始颱風較小實驗中，雖然較小颱風仍可伴隨低壓之生成，位於颱風內核圈之西南側，但渦度變率最大值位於西側，主要貢獻為渦度平流項，造成小颱風和大颱風路徑偏移情形不太相同。

This study employed the Weather Research and Forecasting Model to simulate Supertyphoon Krosa (2007) and examined the physical processes responsible for the significant track deflection and the looping track prior to landfall. When the Typhoon Krosa approached Taiwan, the wind west of the typhoon became stronger than that east of typhoon as a result of the channeling effect as found in other studies. A positive vorticity field forms at the south-eastern side of Taiwan and results in a maximum of vorticity tendency. Because a low is co-located at the same place of the positive vorticity field, it's inferred that the relative low leads to the positive vorticity and might facilitate a southward movement of the center causing a looping track. The backward trajectories indicate that the processes leading the formation of the low are associated with leeside subsidence warming. According to analysis of the vorticity budgets, both divergence and advection terms contribute to the vorticity change. To verify the mechanism in real cases, this study also used the NCU mesoscale model to simulate north-westward typhoons impinging an idealized island similar to Taiwan. The result shows that the mechanism of the looping track in idealized cases is similar to real cases. Both the channeling effect west of the typhoon and the generation of positive vorticity southwest of Taiwan are present in idealized cases. Therefore, the looping mechanism may occur not only in the Krosa (2007) but also in other cases with similar conditions. However, no positive vorticity is found southeast of the island for the typhoons when approaching further south and no southward track deflection is found prior to landfall. The sensitivity study on the typhoon size shows that the low also forms in the experiment of smaller typhoons but isn't strong enough to significantly affect the movement of the typhoons. The maximum of vorticity tendency dominated by the vorticity advection term west of smaller typhoons is the main reason for track deflection different from bigger typhoons.