The development of a satellite application to rapidly assess earthquake induced building damage

快速城市化对城市抵御地震的能力提出了巨大挑战。事件后紧急情况的有效管理对整体复原力作出了重要贡献。在最近的地震中（2010年海地地震，2015年尼泊尔地震），卫星图像的使用显示出巨大的潜力；然而，基于卫星的数据应用仅限于对完全毁坏的地区或非常严重受损的结构进行定性探测。该项目旨在开发一种基于卫星雷达数据的快速定量评估地震引起的建筑物损坏的工具。这将通过以下途径实现：A)基于卫星的位移测量分析；B)异构数据集（雷达、激光雷达和光学）的融合；C)设计和建造计算网络，以识别和监测建筑物损坏；D)与伙伴机构和研究小组合作，以确保强有力和全面的解决方案。这种工具在地震发生之前、期间和之后都有好处。首先，当新的SAR数据可用时，可以更新脆弱性情景，包括生命线、人口和建筑物分布以及逃生路线。在地震中，该工具可以将震后建筑损坏的评估时间从几天缩短到几小时。这对时间紧迫的救援行动是有益的。最后，该应用程序可以通过为未来的地震工程研究提供数据来通知重建计划。关键词：卫星遥感，机器学习，脆弱性评估，风险分析，地震工程

Rapid urbanisation poses an enormous challenge for the resilience of cities against earthquakes. The efficient management of the post event emergency gives an essential contribution overall resiliency. In recent earthquakes (Haiti 2010, Nepal 2015), the use of satellite imagery has shown tremendous potential; however, satellite based data applications have been limited to the qualitative detection of completely destroyed areas or very severely damaged structures. This project aims to develop a rapid tool for the quantitative assessment of earthquake induced building damage, based on satellite radar data. This will be achieved through; A) the analysis of satellite based displacement measurements; B) the fusion of heterogeneous datasets (Radar, Lidar & Optical); C) the design and build of a computational network to identify and monitor building damage and; D) working with partner institutions and research groups to ensure a robust and holistic solution. Such a tool has benefits before, during and after an earthquake event. Firstly, as new SAR data becomes available, vulnerability scenarios could be updated including life-lines, population and building distribution and escape routes. During the event, the tool could accelerate the assessment time of post-earthquake building damage from days to hours. This is beneficial for rescue operations whose efforts are time critical. Lastly, the application can inform reconstruction initiatives by feeding data into future earthquake engineering research. Key Words: Satellite Remote Sensing, Machine Learning, Vulnerability Assessment, Risk Analysis, Earthquake Engineering