區域淹水風險之分級分區

Abstract

台灣的地理位置處於太平洋與亞洲大陸的交界地帶，因海島地形、緯度與季風因素，容易遭受如颱風和暴雨而導致的水災。面對因洪災不斷造成的經濟損失，政府能提供應付的臨時緊急重大事項經費卻十分有限，因此本研究探討在面臨淹水風險災害時，如何利用保險的分散和移轉風險功能，引進民間保險業者和資本市場支援，提高民眾保險意願，建立有效的淹水保險制度，除了減輕政府的洪災風險，降低國家整體經濟損失之外，也可讓民眾有一定程度的洪災保障，並使災後的復建能夠迅速進行。 本研究以二維零慣性漫地流模式與雨水下水道模式模擬汐止地區年平均淹水深度，以此作為淹水風險潛勢評估標準，使用Natural-Break與自訂方式決定不同等級之淹水風險，再以路網與村里形成之區域為依據進行淹水風險分區，各分區住戶負擔各自區域淹水風險。同時設定各種淹水保險模擬情境，改變一樓住戶與二樓以上住戶負擔保費比例，導入以電話訪查獲得之淹水保險意願曲線，觀察在各種情境下能得到的意願保險住戶與可收得之保費，並將不分區、不分級的情況比較之。 模擬顯示，投保意願住戶以Natural-Break在分級分區下，一樓住戶負擔保費為二樓以上住戶之兩倍為最多，共有20977戶；同樣情境下之自訂分級之投保意願住戶有20920戶；不分區則有19202戶。顯示對不同淹水風險等級區域進行空間上的分區，以各分區淹水損失由各分區住戶自行負擔其淹水損失為原則，依樓高進行住戶分級，在一定程度上能夠提高人民保險意願，協助建立淹水保險。唯效果並不十分顯著，可建立其他指標判定個別區域適合之初始淹水風險指標分級，並建立更精確之淹水保險意願曲線，應能得到更完善之淹水保險分級分區制度。

Situated at the border area between the Asia Continent and the Pacific Ocean, Taiwan is conducive to floods brought by typhoons and rainstorms because of its insular topography, latitude and monsoon climate. While floods have constantly been damaging the economy, the government can only provide a quite limited expenditure on emergencies. Therefore, the present research probes into the ways in which we can take advantage of the functions of risk segregation and transfer of insurance while facing flood hazards, so as to introduce the support of civil insurance companies and capital markets, to enhance citizens’ willingness to carry insurance, and to establish an effective flood insurance system. In this way, the flood risks of the government and the overall economic loss of the country can be reduced. The present research simulates the average annual flooding depth of Shi-Jr District, so as to establish an assessment standard for the potential of flood hazard. Besides, Natural-Break and manual methods are utilised to determine the different levels of flood hazards, and areas formed by roads and villages are used as the basis for flood hazard zoning, in which the residents respectively burden different flood hazards in different zones. Meanwhile, after setting up various simulated conditions for flood insurance and changing the proportions of insurance fee that first-floor residents and residents above the second floor should pay, this research also adopts a curve of the willingness to participate in flood insurance obtained by phone surveys, so as to observe the number of residents willing to assume insurance and the amount of insurance fee thus collected under various circumstances. According to the simulation, in a condition where the zoning and risk classification are done with Natural-Break and where the first-floor residents burden twice the insurance fee than residents above the second floor, there are most residents willing to assume insurance; the number of such residents is 20,977. Under the same circumstance except that the risk classification becomes manual, there are 20,920 residents willing to carry insurance. In situations without zoning, there are 19,202 such residents. As the simulation shows, citizens’ willingness to participate in insurance and assist in establishing a flood insurance system can be enhanced to a certain degree as areas with different levels of flood hazards are zoned spatially. However, the effects would not be too remarkable merely in this way. A more completed system for the zoning and classification of flood insurance can be achieved by establishing other index to judge the primary classifying index of flood hazards suitable for individual areas.