Abstract:
Forest and land fires are a common phenomenon in several regions of Indonesia. It is assumed that most of the forest and land fires originate from human activities. This study aims to statistically test the spatial correlation between the number of hotspots or the frequency of forest and land fires, to the distance from various types of landuse in the Riam Kanan subwatershed. The data used in this study are landuse and hotspot data. The spatial correlation analysis in this study was conducted using Euclidean Distance and single regression. Euclidean Distance is used to measure the flat distance between the fire location and the location of human activities. Meanwhile, single regression is used to measure the correlation between the number of fire occurrence points and the flat distance from the location of human activities. The single regression models used are linear, power, exponential, logarithmic, and polynomial. The results showed that the frequency of forest and land fires had a very strong spatial correlation with human activities, especially in the sub-watershed area of Riam Kanan. So it can be stated that the frequency of forest and land fires does have a strong correlation with human activities. The lowest spatial correlation is the distance from the rice fields, and the highest spatial correlation is the distance from the river. However, the number of hotspots increases drastically the more distance it is from the road, and almost approaches zero the farther the road is. Thus, although the spatial correlation with roads is not as high as other land uses, this drastic increase in the number of hotspots indicates that road accessibility has a strong contribution to forest and land fires.
