利害相關者對颱洪災害預警系統規劃要素之價值評估：多屬性效用理論之應用

Evaluating stakeholders’ preferences in typhoon-flood hazard warning system planning using multiple-attribute value theory

氣候變遷與極端天氣帶來的挑戰，使許多國家致力於建置災害預警系統，以降低傷亡與風險，及促進公共衛生與安全。然預警系統規劃過程，需更廣納利害相關者參與及不同價值觀，以利於溝通與降低衝突。本文主要目的，在於探討利害相關者對颱洪災害預警系統價值判斷，及對其構成要素之偏好選擇，以提供災害風險管理決策參考。文中應用多屬性效用理論(multi-attribute utility theory)，透過政府官員與專家問卷調查，引用價值函數與多變量分析，比較受訪者對預警系統不同構成因素重要性之判斷與偏好。分析結果發現政府官員與專家對預警系統設計考慮的要素，多數無顯著差異；多偏重在技術、容易標準化及「由上而下」之因素。認為主導預警系統運作的核心，仍以災害資訊提供及緊急應變為主，較忽略末端使與者反應與調適行為。因災害潛勢與衝擊充滿不確定性，公共安全與災害風險管理難完全依賴科技。而需更有效掌握居民脆弱度與調適力分布，及藉由雙向溝通與利害相關者參與，建立居民導向之預警與緊急救助系統。

Many countries are developing hazard warning systems to confront the impact of climate change and other extreme events in order to reduce the risk of disasters and casualties, as well to enhance public health and security. As part of policymaking, a warning system also needs to more effectively minimize conflicts while communicating the risk of disaster to policymaking by inviting the participation of various stakeholders and incorporating their values into the process of system planning. The aim of this study was to explore the value preferences of stakeholders with regard to the components of typhoon-flood hazard warning systems. This study used multi-attribute utility theory to model stakeholder values for typhoon-flood hazard warning system choices. We interviewed officers and experts to determine weight and value preferences for the components of a warning system, and these were compared using value function and multivariate approaches. The differences in most value preferences between officers and experts were insignificant. They assigned higher weights to the provision of information and emergency preparedness functions in building warning systems. This implied that the warning system plan tended to be conceived largely as a ‘top-down’ and standardized process, with relatively little engagement of the end-users in terms of their characteristics, responses and adaptive capabilities. The uncertainties of the impact of hazards require that warning systems function not only on a sound technical basis, but also effectively consider the vulnerability, adaptive capacity and participation of the people exposed to risk. A people-centered warning system should be built to empower communities to prepare for, respond to and adapt to disaster risks. These findings could improve the process of planning warning systems to maximize public security and risk management.