LAUSR

Abstract Turkey is exposed to forest fires damaging thousands of hectares of forest every year. Earlier studies indicate that Turkey would be one of the most affected Mediterranean countries due to increased fire risk as a result of climate change. Therefore, early detection and suppression of forest fires are very important to minimize fire damage. The main goal of this study is to analyse and model the fire risk using remote sensing technology. For this purpose, multi-temporal remote sensing data obtained pre-fire was integrated with ancillary data in Geographic Information System (GIS) to evaluate the spatial and temporal patterns of forest fire risk in Menderes region, Izmir. First, the rapid fire risk of the study area was assessed using Land Surface Temperature (LST) changes and in-situ meteorological measurements. LST and humidity were found to be the most important indicators for fire risk. In order to generate a fire risk map in the study area, six fire risk variables, determined statistically in previous studies, were considered. A linear model that uses predetermined weights for Turkish forest ecosystems was applied. According to the model, 22% of the burned area was found to be high risk area, while 77.5% was identified as moderate high risk area. The model is validated by overlaying all 292 forest fires obtained from the NASA-FIRMS website between 2002 and 2018 on the fire risk map.