Abstract:

Within this study we introducea framework for assessing spatial fire risk and exposure to three important habitat types in Cephalonia island, Greece.Existing maps were used for plot allocation in orderto measure several fuel parameters in representative natural fuel complexes for site-specific fuel models development, as well as for collecting training and validation points for satellite data classification. The spatial extent of the fuel types and the canopy cover were delineated using a Landsat 8 OLI image acquired on 23-7-2015and the Support Vector Machines-(SVMs) machine learning algorithm. Subsequently, The Minimum Travel Time (MTT) algorithm, as it is embedded in FlamMap spatial fire simulation software, was applied in order to assess critical fire behavior parameters and exposure of Cephalonia's habitats under three different meteorological and fuel moisture scenarios. The outputs of this study may be used as an application of quantitative and probabilistic risk assessment for habitats conservation planning, prioritization and management of high value natural and cultural resources.