凡那比颱風之海氣交互作用模擬分析 －海洋冷暖渦影響探討

Abstract

本實驗利用一維海氣耦合模式，進行颱風凡那比颱風(2010)生成發展中加入冷渦及暖渦的實驗；在模擬之前利用系集卡爾曼濾波器(EnKF)進行資料同化，在同化期間並加入了ITOP實驗的C-130飛機觀測資料，以能得到更接近觀測的颱風強度及結構資料來進行模擬，以便進一步探討在加入海洋渦漩並將海洋渦漩的混合層厚度、水平半徑大小以及與颱風路徑的相對位置等變因，對於颱風強度以及上層海洋回饋機制的影響進行探討，以及定量分析這些敏感性實驗中渦漩對於颱風強度的影響程度，主要的研究發現簡述如下： 透過定量計算分析海洋渦漩對颱風強度影響的程度，會和渦漩本身的水平大小成正比；而在混合層厚度的變化上，暖渦厚度越厚，冷渦厚度越薄，對於颱風強度的影響也越明顯，其中暖渦為正貢獻，冷渦為負貢獻。比較渦漩對颱風強度造成的直接影響與颱風通過渦漩期間的總強度變化的比例，發現越厚越大的暖渦以及越薄越大的冷渦對於風強度直接影響的能力就越強；而當颱風通過越小越薄的暖渦及越小越厚的冷渦時，颱風強度的影響便不再受這些渦漩明顯作用；而對海洋上層反應過程的變化，暖渦的厚度越厚水平尺度越大，抑制海洋負回饋機制的能力越佳，而冷渦的厚度越薄，越能加強海洋負回饋機制；而跟暖渦不同的是，冷渦實驗中將冷渦半徑改變對海洋負回饋機制的影響則不明顯。而路徑兩側的海洋渦漩對於颱風強度的貢獻作用在我們的定量計算分析上的結果是相當不明顯的，而比較左側與右側的冷渦與暖渦，唯有位於右側海溫下降敏感區域的海洋渦漩會對於海溫下降的機制產生影響，右側暖渦抑制海溫下降的能力隨著半徑增加厚度加大而增強，冷渦則是隨著厚度變薄而加大海溫下降的區域面積，即有助於增強負回饋機制，但因颱風僅經過這些渦漩的邊緣，渦漩影響時間亦較短，因此這些右側渦漩對海洋負回饋的機制的影響，並未反應在颱風強度上。 針對穿越暖渦的颱風而言，在離開暖渦後因更強的強度可引發較無暖渦狀態的海洋場更顯著的海洋冷卻效應，會進一步反過來削弱自身原先受暖渦正貢獻增加的強度，即可說是一種由暖渦間接導致的海洋負回饋機制；而在暖渦的厚度與抑制海溫冷卻的能力之間的關係，在颱風通過暖渦所在海域後增強，將會引發更強的洋流流入暖渦內，對於較厚的暖渦，此更強的洋流所引發的流切逸入作用導致的海溫下降作用仍可被抑制住，對於較薄的暖渦，則會無法抑制這種較原先颱風經過暖渦時，尚未增強時所驅動的洋流場更強的洋流所引發的更強流切逸入作用，導致海溫下降，且下降的程度比起無暖渦的實驗更大。 而目前我們所用的一維海洋模式無法反應出海洋內的平流過程，因此在颱風引發的海洋冷卻機制上有低估的可能，加上若是進行海洋渦漩敏感性實驗，也可能因無法考慮海洋渦漩的流場性質，而使得模擬出的海洋負回饋機制結果可能和真實情況有所差異，故後續研究希望能利用具有完整物理過程的三維海洋模式進行模擬，以期能更了解接近真實情況下，颱風與海洋交互作用的物理機制與過程。

This study uses a 1-D atmosphere-ocean coupled model to simulate Typhoon Fanapi (2010) with ocean warm and cold eddies. In order to have a reasonable initial TC structure and intensity, a new method of TC initialization based on ensemble Kalman filter (EnKF) is applied before conducting the coupled model simulation. And here we add the aircraft observation data during ITOP in the TC initialization period. With the improved initial TC structure and intensity, we change conditions of ocean eddies such as eddies’ mixed layer depth, horizontal radius and location relative to the TC track. In our research, the eddy feedback factor is then used to assess how the TC intensity is influenced by ocean eddies, including upper ocean response process.

By calculating the TC intensity influenced by ocean eddies, we find that for warm eddies, if the mixed layer depth is deeper and the horizontal radius is larger, the potential to increase the TC intensity would be stronger, and the ability to restrain SST cooling effect would be stronger, and vice versa for cold eddies. If the eddies are in the right hand side of the TC track, the SST cooling effect would be changed, but if the eddies are in the left hand side of the TC track, the SST cooling effect remains unchanged. Besides, eddies in both the right hand side and left hand side do not significantly affect TC intensity. When the TC passes by the center of warm eddies, the SST cooling effect outside warm eddies would be increased because the TC is more powerful in driveingocean current faster, mixing more cold water, and thus reducingSST. This effect would also decrease TC intensity when TC passes by warm eddies. If the faster current flows into the inner core of warm eddies, the thicker eddies can restrain the mixing effect so that the SST cooling effect would be reduced. While the eddy is thinner, it can’t restrain more powerful shear-driven entrainment. Therefore, the SST cooling effect in the location of thinner warm eddies will be more obvious. The 1-D ocean-coupled model we use in this study neglects the SST cooling effect by advection. In our future works, more complicated (3D) ocean models would be used to simulate the feedback effect from ocean. We can compare the results of 3D and 1D models to quantitatively evaluate the impact of advection (non-local effect) on upper ocean feedback and the TC intensity change.