Skip to content
DSML.KZ
Publications

Talk · IEEE International Geoscience and Remote Sensing Symposium · July 1, 2018

Use of Sentinel-1 Data for Earthquake Damage Assessment in Cases of Amatrice and Sarpol-E Zahab

Asset Akhmadiya, Qiming Zeng

Asset Akhmadiya
Open via DOI

Abstract

Earthquake damage assessment using multi-temporal medium-resolution SAR data with dual polarization is possible. In this study Sentinel-1 SAR data have applicated for earthquake damage assessment using existing change detection method. Sentinel-1 data are applicable due to universal availability of pre- and post-earthquake SAR imageries for maximum six days. The normalized difference of intensity correlation coefficient (NDIC), normalized difference of interferometric coherence coefficient (NDCC), co-seismic intensity correlation (IC) and difference in the backscattering coefficient (DBC) images were used in order to create damage maps. In this investigation, histogram of damaged and intact building in the values of NDIC, NDCC and statistical distribution of IC-DBC were analyzed with different window size of correlation coefficients r and local mean values m. It was found that damaged buildings are better detectable with NDIC and NDCC variables than IC-DBC. Artifical neural network (ANN) supervised classification method has better results of producer's accuracy for damaged building classes. Producer's accuracy of damaged building class is achieved 95.6%, however, there are misclassification between intact building and none built-up area, it is decreased by using GIS data having mask of built-up area. These results lend support to the idea of realization low-cost method of damage assessment after natural disasters based on Sentinel-1 and GIS data, which are required for insurance companies and governments. Amatrice and Sarpol-e Zahab cities were studied in this investigation, which have been more affected by the earthquakes in 2016-2017.

Synthetic Aperture Radar (SAR) Applications and TechniquesRemote Sensing and Land UseRemote-Sensing Image Classificationearthquake damage assessmentSentinel-1 SARmulti-temporal SAR

More by Asset Akhmadiya