Spatio-temporal vulnerability of forest fire to climate change in Korea

Climate is known as crucial factor on forest fire occurrence. Although most of forest fires occurred by human activities, climate factor is the most important cause for forest fire due to moisture containment of fuel. The objective of this study is to estimate forest fire occurrence according to the past forest fire and to assess the vulnerability with forest cover map. Spatial data of forest fire was prepared from forest fire record of Korea Forest Service (KFS). The forest fire record consists of date, time and XY coordination of the forest fires between 1991 and 2006. Past climate data obtained at 75 weather stations was rearranged with Inverse Distance Weighted Interpolation. The climate data included temperature, relative humidity, cloudness and precipitation. We considered the laps rate temperature depending on elevation. The 10 days accumulation of rainfall data was used to the precipitation data. Forest fire point data and climate data were matched according to time and location. Prepared data was analyzed in statistical method, Poisson regression. And also, we estimated the probability of forest fire occurrence with future climate data which was simulated by A1B scenario of IPCC. This future climate data was prepared with ECHO-G model by Korea Meteorological Administration (KMA). Finally, the forest fire risk map was overwrapped on forest cover map. Because fire occurrence depends on the species of forest, the fire occurrence probability was weighted according to the forest cover information. In this way, the vulnerability was assessed. Our results showed that relative humidity has a significant effect on forest fire. Using these results, forest fire danger map of the future was derived.