應用圓柱座標MANDOP分析法反演颱風風場
Date Issued
2004
Date
2004
Author(s)
劉浙仁
DOI
zh-TW
Abstract
Many Doppler wind constrains can be obtained in multiple radars observation area, but the three orthogonal wind components can not be acquired in full domain. In order to compensate the lost information of observation, continuity equation or other constrains combined with many Doppler wind constrains are used. Besides, under the assumption of certain characters of the wind field, regression equations can be applied for representing the wind components to reduce the unknown numbers. Scialom and Lemaitre (1990) addressed the concept of a three dimensional analytical wind field and brought it into the cost function, and thus converted the variational problem into the regression problem which can be solved by the least square error method. The present research is devoted to the application of MANDOP analysis in retrieving the typhoon wind field and introducing the cylinder coordinates to improve the accuracy of the result.
The present paper adopts the cylinder coordinates in which the center axis is typhoon center and assumes that the analytical wind components, with respect to each coordinate, may be written as a product of three functions of each coordinate. The base functions are power series in radial, Fourier series in tangential and Lengendre series in vertical direction. In the respect of constrains, except for using Doppler wind and continuity constrains, the boundary condition at ground, top of troposphere and typhoon center are included.
Besides using uniform flow and vortex for testing, the simulated typhoon Herb data, produced from MM5 model, are used as the real typhoon wind field. Doppler wind observations are simulated from two to three radars which are at different places. The wind retrieval method above is used to evaluate the results and the affects by comparing the retrieval wind with the true wind under different conditions.
Subjects
反演颱風風場
圓柱座標
MANDOP
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-93-R91521306-1.pdf
Size
23.31 KB
Format
Adobe PDF
Checksum
(MD5):165ebd1d220c5e12606aba880383ced3