宜蘭縣環境災害脆弱度與恢復力之研究

A RESEARCH OF VULNERABILITY AND RESILIENCE TO ENVIRONMENTAL HAZARDS IN YILAN COUNTY

宋健豪(Sung Chien-Hao)

指導教授 ： 廖學誠

國立臺灣師範大學/文學院/地理學系/博士(2022年)

https://doi.org/10.6345/NTNU202200678

摘要

本研究旨在探討台灣東北區域宜蘭縣境內，環境災害脆弱度與恢復力之空間分布模式，受到地形與降雨型態之影響，洪患與土石流為本研究區中最為常見之自然災害，本研究應用社會脆弱度指數(Social Vulnerability Index，SoVI)、社區恢復力基線指數(Baseline Resilience Indicator of Community，BRIC)、回復力-脆弱度空間剖析模式(Spatially Explicit Resilience-Vulnerability model; SERV)模式量化研究區域中對於洪患與土石流之災害脆弱度與恢復力，本研究應用主成分分析整合具相關性、共線性之變數，已提供更為簡潔之成果，除此之外，本研究應用空間自相關探討脆弱度以及恢復力之空間分布模式以及空間差異，並透過空間落遲迴歸、空間誤差迴歸以及地理加權回歸驗證SoVI、BRIC以及SERV模式之效力；，空間自相關成果顯示研究區域中之山區為極度脆弱並缺乏恢復力之區域，反之，平原中之都市區域為低脆弱度及高恢復力之區域，根據空間落遲迴歸、空間誤差迴歸以及地理加權回歸成果顯示SoVI、BRIC以及SERV，就全縣尺度三項指標具尚可接受之判定係數均及解釋能力，並無顯著偏誤；就平原區域而言，三項指標均具備較佳之判定係數均及解釋能力，並無顯著偏誤；就山區而言，整體有待提升；本研究發現，宜蘭縣境內山區與平原之存在顯著之空間差異，地形條件為造成顯著空間差異的主要因素之一，研究區域內山區之海拔高度與陡峭坡度為主要影響其社、經發展障礙的主因，由於山區不利於社、經發展的因素，而使其成為極度脆弱並缺乏恢復力之區域，其他地區如平原城市地區，由於其有利於社會經濟發展的地形，而使其具備低脆弱度及高恢復力。

關鍵字

脆弱度 ； 恢復力 ； 社會脆弱度指數 ； 社區恢復力基線指數 ； 回復力-脆弱度空間剖析模式 ； 空間自相關 ； 空間落遲模式 ； 空間誤差模式 ； 地理加權迴歸 ； 空間差異

並列摘要

This research aims to explore the spatial pattern of vulnerability and resilience to environmental hazards in Yilan County, northeastern Taiwan. We apply the Social Vulnerability Index (SoVI), Baseline Resilience Indicator of Community (BRIC), and Spatially Explicit Resilience-Vulnerability model (SERV) to quantify the vulnerability and resilience to environmental hazards, including flood and debris flow events, which are the most common environmental hazards in our case study area due to the features of topography and precipitation. We apply the Principal Component Analysis (PCA) to aggregate the variables with correlation and collinearity. Moreover, we use spatial autocorrelation analysis to analyze the spatial pattern and spatial difference. To validate the effectiveness of SoVI, BRIC, and SERV, we also adopted the authentic debris flow and flood events as dependent variables and also applied the Spatial Lagged Model (SLM), Spatial Error Model (SEM), and Geographically Weighted Regression (GWR). The result of spatial autocorrelation analysis shows that the mountain areas are extremely vulnerable and lack enough resilience. In contrast, the urban regions in plain areas show low vulnerability and high resilience. According to the results of SLM, SEM, and GWR, on the county-wide scale, the three indexes adopted in this research have relatively acceptable R2, which means these three indexes have an acceptable ability to explain the occurrence of authentic events of debris flow and flood without significant bias. In plain areas, the three indexes adopted in this research have relatively higher R2 and the ability to explain the occurrence of authentic events of debris flow and flood without significant bias. In mountain areas, overall, the R2 and the ability to explain the occurrence of authentic events are poor. This research found that the spatial difference between the mountain and plain areas is significant. The topography is the most significant factor in the spatial difference. The high elevation and steep slopes in mountain areas are significant obstacles to socioeconomic development. This situation causes consequences of high vulnerability and low resilience. The other regions, the urban regions in the plain areas, have favorable topography for socioeconomic development.

並列關鍵字

Vulnerability ； Resilience ； Social Vulnerability Index (SoVI) ； Baseline Resilience Indicator of Community (BRIC) ； Spatially Explicit Resilience-Vulnerability model (SERV) ； Spatial Autocorrelation Analysis ； Spatial Lagged Model (SLM) ； Spatial Error Model (SEM) ； Geographically Weighted Regression (GWR) ； Spatial Difference

參考文獻

Adger W.N., Brooks N., Bentham G., Agnew M., and Eriksenet S. (2004). New indicators of vulnerability and adaptive capacity. Tyndall Centre for Climate Change Research, Norwich, United Kingdom.

Adger W. N. (2006). Vulnerability. Global Environmental Change, 16(3):268-281.

Anselin, L. (1988). Spatial Econometrics: Methods and Models. Kluwer Academic Publishers, Dordrecht, Netherlands.

Anselin, L. (1995). Local indicators of spatial association-lisa. Geographical Analysis, 27:93-115.

Anselin, L. (2005) Exploring Spatial Data with DeoDa: A Workbook. Center for Spatially Integrated Social Science, Santa Barbara, United States.