遙測估算建物形變及其影響因素評估──以臺北市為例

Remote-Sensing Based Estimation of Building Deformation and Evaluation of Affecting Factors–A Case Study of Taipei City

為了城市中人民生命及財產的安全，實施長期的建物及基礎設施監測是一項基本工作。本研究基於合成孔徑雷達(Synthetic Aperture Radar, SAR)影像，利用長時間序列之高解析度COSMO-SkyMed影像，配合永久散射體合成孔徑雷達差分干涉(PS-InSAR)技術，得到2011年5月至2013年7月臺北市內之建物及地表變形成果，再進一步從中萃取建物之變形。接著，為得知建物變形與週遭環境因素的關係，本研究使用邏輯斯迴歸(Logistic Regression)，評估人為及自然環境因子與建物形變之關係。迴歸成果指出，建物的下陷確實會與鄰近之人為因素(如：施工捷運路線、興建建物)及自然環境因素(如：土壤液化潛勢)有顯著關係。

For the safety of people’s lives and property in cities, the implementation of routine and long-term health monitoring of buildings and infrastructure is a basic task. Considering the characteristics of large range of space and time coverage of Synthetic Aperture Radar (SAR) images, together with the capability to estimate displacement based on the interferometric SAR (InSAR) processing, we used time-series high-resolution COSMO-SkyMed SAR images, combined with Persistent Scatterer InSAR (PS-InSAR) analysis, to obtain displacement occurred in Taipei City from May 2011 to July 2013. From which the deformation of the building was further extracted. Next, in order to understand the relationship between building deformation and surrounding environmental factors, we employed Logistic Regression to evaluate the relationship between building deformation and man-made and natural environmental factors. The results indicated that buildings were significantly affected by nearby man-made factors (such as construction of metro routes, elevated roads, building construction sites) and natural environmental factors (such as soil liquefaction), increasing the probability of building deformation.