Estimation of the Forest Fire Risk in Indonesia based on Satellite Remote Sensing

To minimize forest fires in tropical area is extremely important, because the fire has a large impact on global warming, biodiversity, and human society. In the previous study, Shimada and Ishibashi monitored the ground-water lever from the value of Normalized Difference Vegetation Index (NDVI) obtained in Kalimantan Island to predict where the forest fires will happen. We have developed a method to map the forest fire risk by calculating the value of Modified Soil Adjusted Vegetation Index 2 (MSAVI2). Moreover, we investigated the relation between the distance from a road as an artificial factor and the occurrence of the fire.First, calculating the MSAVI2 from Landsat 7 and 8 images of August, 2015 around Martapura in South Sumatra, Indonesia, we mapped the area where the plants were stressed. Next, we checked the degrees of matching between the area of low MSAVI2 and the forest fire points.As a result, half of the fires happened in the area having the MSAVI2 values of 0.20 to 0.35. When we focused on only the area which is over 5 kilometers far from a road, the degrees of matching became higher; it rose up to 62 percent.Those results indicate that the fire risks relate to the dry area calculated as low MSAVI2 in the case with less human activities. We need to consider an effect of artificial factors to estimate the whole risk of forest fire.In conclusion, the map of forest fire risk by calculating the value of MSAVI2 is applicable to an area with less artificial factor, while we have to take the effect of artificial fire factor into the consideration.