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

Accueil du site → Master → Grèce → Δημιουργία χάρτη εκτίμησης κινδύνου εκδήλωσης πυρκαγιάς για την ευρύτερη περιοχη του ΝΣ

National Technical University of Athens (NTUA) 2013

Δημιουργία χάρτη εκτίμησης κινδύνου εκδήλωσης πυρκαγιάς για την ευρύτερη περιοχη του ΝΣ

Tsiplakidis, Ioannis E. ; Τσιπλακίδης, Ιωάννης Ε.

Titre : Δημιουργία χάρτη εκτίμησης κινδύνου εκδήλωσης πυρκαγιάς για την ευρύτερη περιοχη του ΝΣ

Auteur : Tsiplakidis, Ioannis E. ; Τσιπλακίδης, Ιωάννης Ε.

Etablissement de soutenance : National Technical University of Athens (NTUA)

Grade : Master Thesis 2013

Résumé
In this thesis there has been an effort to formulate and apply an integrated model that can assess the danger of fire for a specific small area (Naval dock Salamis). That can be accomplished with the collaboration of remote sensing and cartographic data, in order to create-through an object-oriented image analysis approach-a thematic map of fire danger, using the company’s DEFiNiENS Imaging GmbH object-oriented image analysis software eCognition Initially a knowledge base is created with all the factors that influence in either a positive or a negative way a fire and a number or fire prediction indices were presented that quantifies in a different way the risk that reflect each factor. The need to take into consideration all the factors in the application model was covered by the use of two indices. The first one is called short index related to meteorological factors while the second one is called long index that has to do with all the other factors (topographic, ignition, vegetation, human activities). After choosing the two most appropriate indices, among those presented, the final map of danger was computed from the cumulative effects of risk that represent the two above indices. The composition and application of the model is followed by the evaluation. The evaluation was realized with known fire incidents which happened in the past. The results of the evaluation were satisfactory. The application model showed dynamic response to the changes of the factors. The only disadvantage was the limited number of events that participated in the evaluation procedure. In conclusion, it can be said that this study gives an application model to predict fires for a specific small area with the need of further observation and evaluation over time for better and safer conclusions.

Présentation

Version intégrale (14,5 Mb)

Page publiée le 7 mars 2019