Flood damage assessment using satellite observations within the google earth engine cloud platform

authored by

Mehdi Sharifipour, Meisam Amani, Armin Moghimi

Abstract

Floods cause significant damages to different assets every year and, thus, it is important to monitor floods and assess their damage using advanced technologies. In this regard, remote sensing systems, which provide frequent and consistent observations over large areas with minimum cost, are great resources. In this study, we developed a method to assess the damages to different Land Use/Land Cover (LULC) types caused by floods in the three countries of Iran, Ireland, and Sweden. The amount of flood damage to different LULCs was investigated using the flooded areas reported in the Emergency Management Service (EMS) and the generated LULC maps using the Support Vector Machine (SVM) algorithm and Sentinel satellite data within the Google Earth Engine (GEE) cloud computing platform. Overall Accuracies (OAs) for the LULC maps of Iran, Ireland, and Sweden were 84%, 88%, and 70%, respectively. The experimental results showed that cropland and barrens with 25,099 and 17,164 flooded areas were the most damaged LULC classes, respectively. The amount of damage for the tree class was 3,949 hectares.

External Organisation(s)

Tadbir Kesht Golestan Company
Wood Environment & Infrastructure Solutions
K.N. Toosi University of Technology

Type

Article

Journal

Journal of Ocean Technology

Volume

17

Pages

64-75

No. of pages

12

ISSN

1718-3200

Publication date

01.03.2022

Publication status

Published

Peer reviewed

Yes

ASJC Scopus subject areas

Ocean Engineering

Sustainable Development Goals

SDG 15 - Life on Land