Centre de documentation
scientifique

Flood monitoring by integration of remote sensing technique and multi-criteria decision making method / Hadi Farhadi in Computers & geosciences, vol 160 (March 2022)  

Public [article]  inComputers & geosciences > vol 160 (March 2022) . - n° 105045 Titre : Flood monitoring by integration of remote sensing technique and multi-criteria decision making method Type de document : Article/Communication Auteurs : Hadi Farhadi, Auteur ; Ali Esmaeily, Auteur ; Mohammad Najafzadeh, Auteur Année de publication : 2022 Article en page(s) : n° 105045 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] analyse multicritère [Termes IGN] classification par séparateurs à vaste marge [Termes IGN] Electre [Termes IGN] image Sentinel-MSI [Termes IGN] inondation [Termes IGN] Iran [Termes IGN] Matlab [Termes IGN] rapport signal sur bruit [Termes IGN] surveillance hydrologique Résumé : (auteur) Traditional methodologies of flood monitoring are generally time-consuming and demanding tasks. In most cases, there is no possibility of flood monitoring in large areas. Due to the major drawbacks of conventional methods in flood detection of large districts, Remote Sensing (RS) has been efficiently employed as the best solution owing to its being synoptic view and cost-effective methodologies. One of the most challenging issues in RS technologies is choosing the optimal spectral bands to detect changes in the natural environment. In this research, Elimination and Choice Expressing Reality (ELECTRE), as one of the most widely used Multi-Criteria Decision Making (MCDM) techniques, was applied to select the optimal bands of Sentinel-2 satellite images for detection of flood-affected areas. For this purpose, the decision-making method was implemented during ten options and six criteria. The properties of the Sentinel-2 satellite images consisted of ten bands (with 10 and 20m spatial resolutions) and the criteria are the signal to noise ratio (SNR) related to sensor, standard deviation, variance, the SNR related to the bands, spatial resolution, and wavelength. Afterward, the ELECTRE technique was used to select six optimal bands among ten bands. The ELECTRE algorithm was programmed in MATLAB programming language that could make decisions with multiple options and multiple criteria. Furthermore, the Support Vector Machine (SVM) classification method, as one of the most powerful Machine Learning (ML) models, has been applied to classify the water bodies related to before and after the flood. According to the results of optimal bands classification, Overall Accuracy (OA) and Kappa Coefficient (KC) for the pre-flood classification were 93.65 percent and 0.923, respectively, and for the post-flood classification, the OA and KC values were 94.52 percent and 0.935 respectively. In the case of before and after flooding, the results of classification model for optimal bands had more accuracy levels in comparison with those obtained by original bands. Generally, it was found that the ELECTRE technique for selecting the best bands of Sentinel-2 satellite images and detection of flood-affected areas, in a short period of time with high accuracy, offers remarkable and consistent results. Numéro de notice : A2022-175 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1016/j.cageo.2022.105045 Date de publication en ligne : 29/01/2022 En ligne : https://doi.org/10.1016/j.cageo.2022.105045 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99814 [article]