「天然災害老人弱勢族群社經脆弱度評估指標」之建立與空間聚集性分析應用

Constructing Social and Economical Vulnerability Assessment Indicators for an Disadvantaged Elderly Population and Application of Spatial Clustering Analysis

對於自然災害（如風災、水災、地震等）的受災者而言，年老的長輩可能遭受影響的風險 相對較高，為了降低老人弱勢族群面臨災害時的威脅，在各類災害管理的對策上皆應有特殊之 考量與作為。然而，不同的背景屬性所可能產生的災害威脅皆有所差異，即使是性質相同的弱 勢族群，也會因不同的背景屬性而有不同的影響程度，亦即同一弱勢族群暴露（expose）在不 同災害風險下所產生的脆弱度（vulnerability）差異。目前國內外關於災害管理的相關研究， 多對於不同的族群有差異性的考量，但對於如何清楚區隔同質族群脆弱程度的評估條件上，則 仍存在相當大的討論與整合空間。本研究以老人族群為研究對象，探討同質之弱勢族群在天然 災害威脅下的脆弱程度差異，據以建立一套區隔脆弱程度的評估指標；研究中運用模糊德爾菲 法確立指標項目，並透過層級分析法決定指標權重；再將評估結果經由空間統計分析（spatial statistic analysis）瞭解其空間聚集性（spatial clustering），研究結果顯示台灣本島各區域的鄉 鎮市區確實存在顯著的空間聚集。此一成果可作為相關研究於進行老人弱勢族群在災害威脅下 的脆弱度分析之參考，以及強化政府相關部門對弱勢族群的防救災能力，減少弱勢族群致災機 會及受災時之傷亡損失程度。

When the elderly are victims of natural disasters, such as typhoons, floods, and earthquakes, they may be at high risk for injury and property losses. To minimize threats to the disadvantaged elderly, special strategies are needed for different disaster stages. Notably, different types and scales of disasters have varying threats. For the same group under different scales of disasters, disadvantaged populations have different degrees of vulnerability. Many disaster-related studies have different risk factors for different populations or groups. However, how to assess the degrees of vulnerability of a single group remains unclear. In this study, disadvantaged elderly are the research subjects. This study discusses various degrees of vulnerability for disadvantaged elderly under threats of natural disasters, and develops a set of assessment indicators for social and economic vulnerability. These indicators are examined and weighted using the Fuzzy Delphi process and the Analytic Hierarchy Process (AHP). After establishing this set of indicators, an application of empirical data validation includes spatial statistical analysis, utilizes these indicators to determine the degree of vulnerability of the elderly, as well tests the significance of spatial clustering. Analytical results demonstrate that statistically significant relationships exist for elderly vulnerability among townships in Taiwan. This study can be a reference for analyzing elderly vulnerability under threats of natural disasters, and for reviewing the suitability of governmental disaster management strategies to reduce possible casualties and property losses.