渦旋動力初始化方案應用於全球高解析度模式MPAS之颱風模擬;Impacts of Dynamic Vortex Initialization scheme of a Global Variable-resolution Model MPAS on Simulations of Typhoons

NCUIR > College of Earth Sciences > Department of Atmospheric Sciences and Graduate Institute of Atmospheric Physics > Department of Earth Sciences > Item 987654321/82706

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/82706

Title:	渦旋動力初始化方案應用於全球高解析度模式MPAS之颱風模擬;Impacts of Dynamic Vortex Initialization scheme of a Global Variable-resolution Model MPAS on Simulations of Typhoons
Authors:	林家洋;Lin, Chia-Yang
Contributors:	大氣科學學系
Keywords:	跨尺度天氣預報模式;渦旋初始化;MPAS;Dynamic Vortex Initialization
Date:	2020-01-21
Issue Date:	2020-06-05 16:46:06 (UTC+8)
Publisher:	國立中央大學
Abstract:	全球跨尺度天氣預報模式(MPAS)是由美國國家大氣研究中心(NCAR)所發展的新一代天氣預報模式，本篇研究將使用MPAS的60-15及60-15-3公里可變解析度全球網格模型模擬近年來的七個颱風，包含蘇迪勒(2015)、梅姬(2016)、尼莎(2017)、瑪莉亞(2018)、山竹(2018)、米塔(2019)、利奇馬(2019)。本篇論文裡的動力初始化方案將會建構real-case的渦旋作為初始場，並且積分一小時，在這一小時裡，模式會重新產生一個新的渦旋，我們便可將這個新的渦旋重新移到一小時前的位置，由此反覆來回積分，讓渦旋強度達到接近觀測的強度，使強度或路徑預報能夠接近觀測，以利於修正預報。本篇研究的relocation方法在純量場是使用反距離權重法，在向量場是使用MPAS模式內建的投影法。在本篇研究中我們將使用不同物理參數化以及不同初始場強度和不同模式解析度的渦旋在使用渦旋初始化的差異，並會針對路徑預報、氣壓強度、風速強度等等進行探討，以及在渦旋靠近地形時要如何處理地形，使地形對於渦旋初始化的影響達到最小，以此測試MPAS模式在使用這個渦旋初始化方法可以能會遇到的問題。從結果中我們可以看到我們可以將初始場較弱渦旋透過渦旋初始化的方法改善渦旋的強度使其和觀測一致。透過渦旋初始化的方法，我們可以發現模式解析度60-15公里和積分強度上的限制，只能使渦旋增強到935hPa、48m/s，超越這個強度的渦旋須改用模式解析度60-15-3公里。從不同物理參數化中我們可以看到在路徑預報上使用渦旋初始化對於使用mesoscale_reference改善情況較convection permitting明顯，而在強度預報的部分使用convection permitting改善情況較mesoscale_reference明顯。從結果中我們也看到在路徑誤差的部分在某些個案中有明顯的改善、其餘個案則是些微改善。在強度預報中，不管是氣壓還是風速的預報，在所有個案都能被改善，除了發生強度最大的時刻會有稍微提早的現象。我們也發現初始場就和觀測差不多強度的渦旋使用渦旋初始化的結果期改善幅度就不明顯。 ;This study develops a dynamical vortex initialization for a global variable-resolution model (MPAS) in application to simulations of westbound typhoons approaching Taiwan. The MPAS employs 60-15-3 km variable-resolution with 3-km resolution in the vicinity of Taiwan. Seven westbound typhoons are investigated including Soudelor (2015), Megi (2016), Nesat (2017), Maria (2018), Mangkhut (2018), Mitag(2019), Lekima(2019) with different tracks and intensities. Dynamical vortex initialization schemes have been proposed and presented in literatures for regional models with uniform grids. In this study, the dynamical vortex initialization scheme conducts a real-case vortex (in a radius of 600 km of the typhoon center) as downscaled from the initial global operational analysis and then implants the generated vortex into the observed location in several tens of 1-h cycling model integration in 3-km resolution of hexagonal grids. In each 1-h cycle, the model re-generated vortex can be relocated and artificially amplified according to the best-track observations. In this study, the mothed of the relocation in scalars is use Distance Weighted Interpolation method and vectors is use projection method. We will use different physics suites, initial strength, and model resolution to discuss the tracks and intensities in different typhoons, and we also conduct the situation when the vortex near the terrain, and we have to add some limit in vortex to make the influences of terrain minimize. From the results, the vortex can be intensified to reach observation by the DVI scheme and can give a better structure and intensity than initial condition. The improvement of tracks error in mesoscalse reference physics suite is better than convection permitting physics suite, and v-match is slightly better than p-match. The improvement of intensities error in convection permitting physics suite is better than mesoscalse reference physics suite, and v-match is slightly better than p-match. For the most case, the intensities error can be improved, especially initial intensities error, but the tracks error is only improved in some case.
Appears in Collections:	[Department of Atmospheric Sciences and Graduate Institute of Atmospheric Physics ] Department of Earth Sciences

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	214	View/Open

社群 sharing

Loading...