The Comparison of Landscape Metrics Derived from Remote Sensing Data before and after Wenchuan Earthquake

Abstract

Satellite remote sensing images monitoring before and after the Wenchuan earthquake were becoming of great significance in the field of disaster control and mitigation. This paper compared the landscape metrics derived from remote sensing data before and after Wenchuan earthquake. Firstly, two ASTER imageries acquired before and after earthquake were geo-referenced. Then, the two imageries were segmented and converted into vector files. Finally, six landscape metrics were computed using multi-resolution segmentation vector derived from the two imageries and the computation results before and after earthquake were compared. Comparing results, it concluded that the landscape metrics before and after earthquake are significantly different. After earthquake, the fine surface structure was destroyed due to numerous landslides and rockslides, so Perimeter Area Fractal Dimension and Splitting Index which represents shape complexity decreased. Perimeter Area Fractal Dimension and Splitting Index are more effective indexes to detect the land cover change before and after earthquake.