極端降雨量其空間分布之群集特性分析

A Study on Clustering Characteristics of Spatial Distribution of Extreme Rainfall

本文旨在探討降雨型態空間分布之群集特性，以定義空間屬性均一區，做為區域雨量分析之依據。以臺灣北部地區38 個雨量站樣本作為研究對象，選擇主要影響降雨空間分布之因子；如測站之空間位置、年雨量變異性、短延時降雨型態等物理特性做為分析變量。在主成份分析結果中，顯示前三個主成份可說明原始29 個變量之92.9%訊息，其中第一主成份佔82.7%為降雨型態因子、第二主成份佔6.1%為測站空間位置因子、第三主成份佔4.1%為年降雨變化因子。最後，在主成份空間中，經比較兩階段群集分析與模糊聚類分析之結果，將研究區劃分為三個降雨均一區，其分區結果除能結合地理位置之相鄰性外，又與氣候分區結果相吻合，顯現分區之合理性，將有助於雨量區域分析工作之進行。

The aim of this study is to identify the spatial distribution of extremerainfall in Northern Taiwan and further to divide this area into three homogeneousregions as the basis of regional analysis. Some physical characteristicsof 38 recording rain gages, representing typical spacing, annual rainfallvariations, and short period rainfall patterns, were chosen to be analyticvariables. By principal component analysis, these 29 variables were reducedto three independent principal components as clustering variables.The similarities of clustering variables between rain gages were calculatedby using Euclidean distance in the principal component space. The resultsof two-stage cluster analysis are better than fuzzy cluster analysis, andthree homogeneous regions were determined. The first subregion lies inTaipei and Tao-Yuan. The second includes Hsin-Chu and Ta-Tung andNan-Ao in I-Lan. The third encompasses the northern coast, Kee-Lung and I-Lan. These three regions not only have well-defined geographic boundaries,but also correspond to climatic characteristics. The first three principalcomponents account for 92.96% of original variation, and depict the rainfallpatterns, positions, and annual variation respectively. The study revealsthat there was a reasonable result from grouping 38 rain gages into threehomogeneous regions and the results could be an aid in regional analysis ofNorthern Taiwan.