臺北市山坡地人工邊坡災害風險空間統計與分析

Abstract

臺北市面積約為27,180公頃，而山坡地有15,004公頃，占總面積55％。因臺北市為盆地地形，高度都市化之下，平地不敷使用，進而往周邊山坡地進行開發。開發坡地時則會透過建造擋土設施，創造使用空間，及維護環境穩定安全。本研究「人工邊坡」一詞，為經過人為挖填、整地所形成，且具擋土或護坡設施之邊坡。人工邊坡有其使用年限，如未善加維護管理，對於坡地安全的影響甚鉅；並且臺北市山坡地地質環境複雜，加上近來年在氣候變遷之下，極端降雨更加頻繁，所引起邊坡崩塌之事件逐漸增加，每次事件的發生，都突顯山坡地人工邊坡管理，以及災害預防之重要性。因此臺北市政府於民國99年開始至民國103年，分階段進行全市山坡地範圍內人工邊坡調查建檔作業共計有三萬四千餘筆，其風險等級由高至低分為五個等級，針對不同風險等級與現況之人工邊坡，訂定後續因應對策與機制。

本研究利用臺北市普查之三萬多筆資料，首先透過敘述統計，了解人工邊坡基本特性及風險徵兆因子數量樣態，結果顯示護坡型人工邊坡其總數與徵兆數量，占整體相當大的比例；爾後針對其風險評分較高之高風險邊坡進行空間分析，首先核密度分析得知位於文山區及北投區有較明顯群聚，進一步透過熱區分析發現，位於北投區大屯山一帶及文山區為明顯之熱區，並以空間自相關分析發現高風險評分之人工邊坡具有高度空間相關性，且較明顯群聚區域亦位於北投區、士林區與文山區。後續以敘述統計之成果，以及相關文獻篩選環境因子與徵兆因子自變數，並將風險評分高低與否為依變數，進行羅吉斯迴歸統計。經研究發現，邊坡坡面破損之徵兆對於風險評分高低有顯著影響。人工邊坡調查迄今已近10年，邊坡現況是否與當時相符，抑或風險提升，需重新普查確認，但時間與人力為現階段面臨的課題。若將研究成果應用於管理實務，先以熱區分析之相關特定地區及有特殊徵兆之邊坡進行彙整，分年度執行抽查工作，檢視其風險等級之變化，後續進行滾動修正調整標的，或許能有減輕人力與加快普查時程之效益。The area of Taipei City is approximately 27,180 hectares, with slopes covering 15,004 hectares, accounting for 55% of the total area. Due to Taipei's basin terrain and high urbanization, flat land is insufficient, leading to development extending into surrounding slopes. When developing these slopes, retaining structures are constructed to create usable space and ensure environmental stability and safety. In this study, "artificial slopes" refer to slopes formed through excavation, filling, and landscaping, equipped with retaining or slope protection facilities.

Man-made slopes have a limited lifespan, and inadequate maintenance can significantly impact slope safety. Taipei City's mountainous terrain has complex geological conditions, exacerbated in recent years by more frequent extreme rainfall due to climate change, resulting in an increasing number of slope collapse events. Each event highlights the importance of managing Man-made slopes on mountainous terrain and disaster prevention. Therefore, from 2010 to 2014, Taipei City government conducted phased surveys and documentation of Man-made slopes within the city's mountainous areas, totaling over 34,000 entries. These entries were categorized into five risk levels, from high to low, with corresponding strategies and mechanisms developed based on different risk levels and current conditions.This study utilized data from over 30,000 records collected through Taipei City's census. Initially, descriptive statistics were used to understand the basic characteristics of man-made slopes and the quantity patterns of risk indicators. Results indicated that slope protection type man-made slopes accounted for a significant proportion in total numbers and risk indicators. Subsequently, spatial analysis focused on high-risk slopes, starting with kernel density analysis revealing prominent clusters in Wenshan and Beitou districts. Further hot spot analysis identified notable hotspots around Datun Mountain in Beitou and Wenshan districts, confirming through spatial autocorrelation that high-risk man-made slopes exhibit strong spatial correlation, with prominent clusters also found in Beitou, Shilin, and Wenshan districts. Following this, logistic regression analysis was conducted using descriptive statistical findings and selected environmental and indicator variables from relevant literature, using risk assessment scores as dependent variables. The study found significant impacts of indicators of slope damage on high and low risk assessments.The man-made slope survey has been ongoing for nearly a decade, necessitating reassessment to determine if current slope conditions match those from previous assessments or if risks have increased. However, time and manpower remain critical challenges at this stage. Applying research outcomes to management practices, focusing initially on areas identified through hot spot analysis and slopes with specific indicators, annual sampling and inspection can evaluate changes in risk levels, allowingfor continuous adjustments and potentially reducing manpower and expediting survey timelines.