Improving the accuracy of near real-time seismic loss estimation using post-earthquake remote sensing images

Xinzheng Lu,Xiang Zeng,Zhen Xu,Hong Guan
DOI: https://doi.org/10.1193/041417EQS072M
2018-01-01
Earthquake Spectra
Abstract:With the rapid development of remote sensing technology, satellite or aerial images from the disaster area become available within 24 hours after an earthquake. The collapsed buildings can be easily identified from these images. In this work, a framework for near real-time seismic loss estimation for regional buildings is proposed, which improves the accuracy of nonlinear time-history analysis (THA)-based loss estimations by taking advantage of the identified building collapse scene of the disaster area. Specifically, a series of THA are performed for the target regional buildings, thereby generating a number of simulation results. Those simulation results that bear strong similarities to the identified collapse scene are identified as the optimal solutions, which will be used to estimate the seismic loss. The simulation results of the case studies signify that the use of the identified building collapse scene leads to much closer estimations to actual economic losses.
What problem does this paper attempt to address?