Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Turquie → Forest Fire Analysis Using Satellite Imagery

İstanbul Teknik Üniversitesi (2011)

Forest Fire Analysis Using Satellite Imagery

ESEMEN Kerem

Titre : Forest Fire Analysis Using Satellite Imagery

Uydu Görüntüleri İle Orman Yangını Analizi

Auteur : ESEMEN Kerem

Université de soutenance : İstanbul Teknik Üniversitesi

Grade : Master of Science (MS) 2011

Résumé
Along the Mediterranean and southern Aegean regions of Turkey especially during July and August, high temperature combined with low humidity poses a grave danger for forest terrain in terms of wildfires. As the statistics verify that vast majority of the forest fires are human induced due to increasing demand for land, whether the cause of fire is natural or intentional the damage is rapid and should be monitored in stages of before, during and after. Remote sensing having a significant place in hazard monitoring and evaluation ; is able to provide multispectral satellites images of the fire zones for analysis and information extraction. Remote sensing and geographical information systems (GIS) together define a superior approach to fire monitoring and assessment. Especially for the stages of containment, damage analysis and monitoring, images acquired from remote sensing satellites provide synoptic data with rich spatial spectral information. In case of a high risk fire zone an information system is able to integrate key features such as metrological, transportation, physical, ecological and logistic data to carry out containment plans beforehand as well as damage analysis after a possible fire. Through out this study, an application of remote sensing and GIS in a wildfire damage analysis and assessment scenario was performed. Damage assessment for the wildfire that took place between July 31st and August 5th, 2008 effective at Antalya province, Manavgat and Serik municipalities was accomplished by processing a SPOT 4 satellite image and integrating with local forestry data within ArcGIS environment. Spectral analysis and NDVI transformation was applied on post fire SPOT 4 satellite image acquired on August 7th, 2008, having 20m spatial resolution. Multispectral classification methods were next applied and classification accuracy assessments were performed. After, geometric corrections were carried out for accurate area calculations ; classification results are integrated with local forest management data within ArcGIS to perform further analysis.

Présentation

Version intégrale (4,75 Mb)

Page publiée le 15 décembre 2020