Forest fire risk modeling and fire simulation in the southeast of Urla peninsula (Seferihisar-Menderes)

Forest fires, defined as natural or man-made disasters, cause great damage to the ecosystem, cause loss of life and property, and cause significant damage to fauna and flora. If the intervention is not carried out in a planned manner during the fire, the losses can reach much larger levels. In disaster risk studies, early detection of forest fires and determination of risky areas and modelling the movement of the fire will be vital. In fire simulation research, models implemented with real data obtained from fire areas can be viewed with the help of Geographic Information Systems (GIS) and evaluated within disaster risk decision support systems. In this study, the forest fire that occurred in the Urla region of Izmir in 2019 was evaluated within the framework of plant fuel properties, topographic features, landscape features, meteorological features. Accordingly, fire risk models were created, and fire simulations were constructed in the light of the findings obtained within the scope of the study. The large area of the red pine forests and dry bushes existing in the study area increases the risk. The dryness of the climate and weather conditions in summer leads to an increased risk of fire. Topographic features existing in the study area and the location of transportation roads in case of fire are another factor that increases the risk. Since the Urla peninsula is located in a suitable location for tourism, the human factor and associated risk manifests itself. According to the fire risk map data obtained following the evaluation of all parameters, it has been determined that the high risk areas cover an area of 70%, 28% of areas with medium risk value and 2% of areas with low risk. In agreement with produced model, it was obtained as a significant finding that fires occurred in the study area in 2008, 2009 and 2019 were located in high risk areas. Accordingly, the actual fire locations are compatible with the study model. In the study, besides the risk model, simulation studies were also carried out to determine the behaviour and spread property analyses of a fire that will occur in the area. By simulating the 2019 fire in the model, it was observed that there was a 90% compatibility between the real fire damage and the model damage. In addition, systematic tests were conducted using different literature models in simulations. In this way, the risk model was examined under various conditions. In the research, it has been determined that the most important factor in determining the parameters used in fire risk determination can be realized by testing different perfectives and variables under different conditions. Another important finding obtained from the study is that it is predicted that it will make a significant contribution to the determination of disaster risk or disaster mitigation studies evaluating in the decision support mechanism. Keywords: 2019 Urla Fire, Fire Risk Modelling, Flammap, Fire Simulation, Geographical Information Systems