考量時空變化之都市洪災韌性指標定量化研究

Abstract

氣候變遷與都市化加劇極端降雨事件發生頻率，使都市面臨巨大的災害風險，防洪策略若能納入韌性概念，將有效減輕災害衝擊。關於韌性措施實際提升洪災韌性在過去文獻較少被提及，原因在於韌性是一個主觀的概念，其評估因子難以定義和量化。此外災前的預警、災中的應變以及災後的復原都會對於防災工作扮演重要角色，但既有方法難以完整評估所有階段，顯示在制定災害防治策略時需額外考量韌性的時變性。傳統上利用淹水潛勢地圖與風險矩陣的定性手段進行減災整備規劃及收容場所區位適宜性分析，然而上述方法僅提供物理淹水總量體的參考值，無法反映各區在災害不同階段的需求，也忽略了社會經濟和基礎設施對於韌性產生之影響。目前大部分相關研究對於都市防洪韌性僅提出質性評估架構，對於政策決策者在災害管理上較難提供有效且科學化的依據。 本研究以ISO 14091氣候變化風險評估的定義作為韌性評估框架，危害度、暴露度和敏感度為指標考量依據，有別於單純考慮物理因素因子，整合物理、社會經濟及基礎設施因子來建立洪災韌性指標(Flood Resilience Index, FRI)，並定義災害韌性衝擊值(FRI impact)評估FRI整體表現。透過3Di洪水演算模式模擬真實事件獲取淹水物理量的危害度子指標，然後結合非物理性因子影響的暴露度和敏感度之子指標來組成FRI。研究區域為新北市中永和地區、臺中市北屯區及臺南市新化都市計畫區，除探討不同因子對於FRI的影響，測試韌性防災措施低衝擊開發（Low Impact Development, LID)對於防災韌性的提升效益，藉由子指標與整體FRI之間的交互作用探討不同因子對於都市洪災韌性的影響與防災策略。 研究結果指出FRI和危害度曲線的最低點存在一時間差，此將為災害救援和減災策略制定提供重要根據，顯示僅考量物理因子將會低估災害所產生的衝擊。整體的防洪韌性會受到降雨型態的影響，高強度、長延時及左偏雨型將對應越大的韌性衝擊值。此外洪水期間，都市基礎設施的損壞程度會影響韌性表現，滯洪池和下水道系統的密集程度將影響非都市區域的韌性表現，而路網和電網系統在災中的損壞率對於都市區域的韌性影響更大。在高密度開發區域，經濟發展及自主防災程度越高將提升都市防洪韌性，顯著影響洪災回復期，產生更快的恢復速度和更高的韌性。 應用FRI探討都市計畫區的LID對於防災之效益，研究結果指出LID能在短延時降雨事件中減少淹水面積和深度，但長延時降雨的效果不佳。然而從防洪韌性角度來分析，即使在長延時降雨情境下LID還是能有效縮短洪水災害的回復期，增加都市區域的韌性表現。將研究區域根據是否為都市地區及洪水嚴重程度進行分類，本研究利用交叉相關函數(CCF)和耦合協調度(CCD)來建立標準化的防災策略，由量化FRI與子指標間的相互作用來智慧化輔助決策支援系統。在高危害度地區中，適應途徑的減災措施必須以減少物理洪水為主，並且考慮社會經濟因子，以提升災害回復效益減少風險。在輕度和中度危害地區，救援工作的策略需依賴主導危害度的因子和維持社會經濟穩定的因子。主導FRI的子指標會隨著開發程度不同而有所差異，不論在都市或非都市區域暴露度皆是韌性的主要決定因素，但在非都市區域中敏感度主導FRI程度會隨著淹水嚴重程度越高而降低。 本研究所提出的時變性FRI，能夠捕捉到各種韌性因子隨時間的複雜相互作用，改善目前韌性工具無法針對整個災害階段進行評估之不足，時變特性允許更適宜的規劃和防災策略，也可以適應洪災事件過程中的各種風險因素。未來洪災韌性策略可以考慮更多的適應性，且FRI提供一個通用的架構供後續研究者持續納入更全面的評估因子，透過不同情境中的FRI證實其可以實際使用於防災策略的制定，並將動態韌性措施整合到都市規劃和政策中，更能增強都市應對洪水的能力。

Climate change and urbanization have intensified the frequency of extreme rainfall events, exposing metropolitan areas to significant disaster risks. Incorporating resilience concepts into flood prevention strategies can effectively mitigate disaster impacts. Traditionally, qualitative methods utilizing flood hazard maps and risk matrices have been employed for disaster prevention decision-making and resource allocation. However, flood hazard maps only provide reference values for total volumes and cannot reflect the varying needs of different areas over time. Furthermore, conventional flood risk analyses typically only consider physical inundation conditions, neglecting socioeconomic and infrastructural disparities, leading to uneven resource distribution. This study establishes a quantitative assessment from a resilience perspective on both temporal and spatial scales, proposing time-varying indicators to enhance urban disaster prevention efficacy. Using the ISO 14091 definition of climate change risk as a resilience assessment framework, hazard, exposure, and sensitivity serve as the basis for indicator consideration. Diverging from indicators that solely consider immediate hydrological factors, this study integrates hydrological, socioeconomic, and infrastructural factors to construct a Flood Resilience Index (FRI). The 3Di flood simulation model is employed to simulate real events and obtain hazard sub-indicators of physical flood quantities. These are then combined with exposure and sensitivity sub-indicators influenced by social, economic, and infrastructural factors to compose the FRI. The study areas include the Zhonghe-Yonghe district of New Taipei City, the Beitun district of Taichung City, and the Xinhua urban planning area of Tainan City. In addition to exploring the impact of different factors on the FRI, the study identifies dominant factors through the interactions between sub-indicators and the overall FRI across different regions, while also investigating the influence of various factors on urban flood resilience and disaster prevention strategies. Under the interplay of hazard, exposure, and sensitivity, the research results indicate a temporal discrepancy between the FRI and the lowest point of the hazard curve, providing crucial basis for disaster relief and mitigation strategy formulation. Moreover, during flood events, the extent of urban infrastructure damage affects the overall disaster impact. Compared to retention ponds and drainage systems, road and power grid networks have a more significant influence on disasters, implying that infrastructure substantially impacts the FRI. In high-density development areas, economic development and autonomous disaster prevention have a certain influence on resilience. The overall disaster impact is affected by rainfall intensity, with rainfall skewness dominating in the initial stages of a disaster. The FRI curve shape is determined by the position of its lowest point, while population age distribution and economic conditions significantly influence the flood recovery period, with better socioeconomic conditions leading to faster recovery rates and higher resilience. Comparing urban and non-urban areas reveals that exposure is the primary determinant of resilience in both contexts, with sensitivity having a relatively greater influence in urban areas. Classifying the study areas based on urbanization and flood severity, this research employs cross-correlation functions and coupling coordination degree to establish standardized disaster prevention strategies, utilizing quantified FRI and sub-indicator interactions to intelligently assist decision support systems. In high-hazard areas, adaptive mitigation measures must primarily focus on reducing physical flooding while considering socioeconomic factors to enhance disaster recovery efficacy and reduce risks. In low and moderate hazard areas, rescue strategies should rely on factors dominating hazard levels and maintaining socioeconomic stability. Applying the FRI to examine the disaster prevention benefits of Low Impact Development (LID) in urban planning areas, the results indicate that LID can reduce flood area and depth in short-duration rainfall events but is less effective for long-duration rainfall. However, FRI analysis shows that even under long-duration conditions, LID can effectively shorten the recovery period of flood disasters, thereby increasing flood resilience. The time-varying FRI proposed in this study can capture the complex temporal interactions among various resilience factors. Its temporal characteristics allow for more appropriate planning and disaster prevention strategies, adapting to various risk factors throughout the flood event process. Future flood resilience strategies could consider greater adaptability, including comprehensive risk assessments, integrating dynamic resilience measures into urban planning and policies to further enhance urban flood response capabilities.