Citation

Moharrami M, Javanbakht M, Attarchi S. Environ. Monit. Assess. 2021; 193(5): e248.

Copyright

(Copyright © 2021, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s10661-021-09037-7

PMID

unavailable

Abstract

Flood is considered to be one of the most destructive natural disasters. It is important to detect the flood-affected area in a reasonable time. In March 2019, a severe flood occurred in the north of Iran and lasted for 2 months. In the present paper, this flood event has been monitored by Sentinel-1 images. The Otsu thresholding algorithm has been applied to separate flooded areas from remaining land covers. The threshold value of -14.9 dB was derived and applied to each scene to delineate flooded areas. There was high variability of the inundated area; however, the presented threshold correctly represented the variation of the flood. The resultant maps were further verified by independent datasets. The overall accuracies were higher than 90%, confirming the applicability of the Otsu automatic thresholding method in flood mapping. The automatic approach is efficient in rapid fold mapping across complex landscapes.

Language: en

Keywords

Google Earth Engine; Sentinel-1; Aqqala; Flood; Otsu