| 摘要: | 本研究利用IBM_VDRAS (Variational Doppler Radar Analysis System based on immersed boundary method)分析2017年7月7日的午後對流個案。分析之前我們新增了以下幾個功能，分別是科氏力、雷達波束遮擋和晴空回波同化以及新的極小化方法L-BFGS-B(Limited-memory BFGS for bound-constrained)。此外，為解決IBM_VDRAS低層風場高估的問題，在本研究中將地形下邊界條件從滑動邊界改為非滑動邊界，並和觀測比較孰優劣。個案分析會比較移動午後對流和一般午後對流的結構差別，並推測造成此差別的原因。
研究結果顯示，非滑動邊界能有效降低低層風速高估的問題，且風向的表現亦比較好。從IBM_VDRAS分析結果能看到，移動午後對流結構不對稱，類似中尺度颮線且對流後方存在強風區，冷池移速和強風區大小相當，因此對流移動應是由對流後方強風區平流所導致，而對流通過測站會有風速劇增的現象則是因為強風下沉至地表所導致；一般午後對流則是偏向垂直發展且沒有這個強風區，對流近似滯留。造成兩者的差異推測是環境風切的不同所導致，從ERA5再分析資料顯示，平原風場在有2-3公里有極值，0-3公里風切約8ms^(-1)，山區風速則偏弱，0-3公里風切只有3.5ms^(-1)，平原低層風切確實較大。
;This study utilized IBM_VDRAS (Immersed Boundary Method_Variational Doppler Radar Analysis System) to analyze the afternoon thunderstorm on 7 July 2017. Before we analyzed this case, a few new features were implemented to IBM_VDRAS. They included Coriolis force, radar beam blockage, assimilation of clear air echo, and a new minimizer called LBFGS-B (Limited-memory BFGS for bound-constrained). Furthermore, in order to resolve the problems associated with the overestimation of wind speed at the lowest level of IBM_VDRAS, we changed the lower boundary condition from free-slip to no-slip type, and compared the results against the observations. The results showed that no-slip type boundary condition is able to reduce the problem of wind speed overestimation, and generate more accurate wind directions than those from the free-slip type boundary. In the case study of this research we compared the differences between a fast moving convection over the plain and an ordinary convection developed in the mountainous area, and attempted to find out the reason causing such differences.
From the analyses of IBM_VDRAS, it can be seen that the structure of the moving convection was asymmetric, and was more like a squall line, with strong wind descending behind the convection to the ground. This is also confirmed by the surface station observations as the wind speed increased when the convection passed the station. In addition, the cold pool’s moving speed was approximately equal to the strong wind. Therefore, it is speculated that the movement of this moving convective system was driven by the advection of the strong wind. On the other hand, the ordinary convection developed vertically in the mountainous area without strong wind field, and was almost stationary. The difference might be attributed to the strength of the environmental wind shear. From the ERA5 reanalysis data, it was shown that the maximum wind speed in the plain area occurred at 2.0 ~ 3.0 km, and the low level wind shear below 3.0 km was about 8ms^(-1). By contrast, the wind speed above mountain was weaker, with a low level wind shear below 3.0 km of only 3.5ms^(-1). |
