客觀分析與正壓模式一理論與實作

劉廣英　鄧施人

Objective Analysis and Barotropic model一Theory and Practice

Koung-ying Liu　Shih-jen Teng

ABSTRACT

　　This is the first report of the two years project “A Study of Numerical Weather Prediction by the use of Electronic Computer”. This research includes two parts, the objective analysis and model testing. Due to the lack of expenmental study of objective analysis and initilization in Taiwan, more emphasis, in the first year, is on the first part. In this report we present two types, each involves four different kinds, of objective analysises. The experimental result indicates that the best method is

　　　　∆U=(W_mfΣW_i(U_f-U_i)/(ΣW_i)
Where
　　　　∆U = Correction of first-approximation at grid points.
　　　　W_mf = Maximum weighting factor.
　　　　W_i = Weighting factors.
　　　　U_f = Observed values.;　　U_i= Estimated values of stations.
and
　　　　W = exp(-AD²/R²); or W = (R² – D²)/(R² + D²)
where
　　　　R = Sean radius.
　　　　D = Distance between the grid point and the related stations.
　　　　A = An experimental constant.
　　In the second part of this report the procedues and results of 500mb prognostic contour charts which prepared by Limited Area Fine Mesh (LFM) model was presented. In order to give a complete discussion a simple theoretical derivation and some numerical experiments of relaxation method are also presen.
　　本文爲「利用電子計算機做數値天氣預備」硏究之初步結果，期重點有二，卽客觀分析，以求得網格點初値，及模式實驗與數個預報圖之製作。前者又爲執行中較偏重者，蓋因國內對數值預報之硏究工作，獨缺此項。經一年之努力，已完成實驗兩類八種求網格點初值之方法，而在數值預報圖製作方面亦已完成各種基本實驗及初步之預報圖，正式結果及與現用方法之比較則仍在進行中。
　　除上述重點工作外，對模式所必須之漸近(relaxation)法亦加以硏究，並做數値實驗。