Drought Indices to Map Forest Fire Risks in Topographically Complex Mountain Landscapes

Abstract

Drought has the potential to lead to forest fires. Forest fires generally occur during the dry season when the mountain slope forest experiences a water deficit. Drought identification based on remote sensing is useful for mapping potential fires in Arjuno- Welirang Forest and TNBTS Forest (in Bromo Tengger Semeru National Park). This research used Landsat-8 images in 118/065 and 118/066 in August and November 2015-2018. Validation data were obtained using high resolution planet scope images and rainfall data. Three drought indices were tested to identify fires, namely TVDI, VHI and NDDI. The indices were tested visually using high resolution images and tested meteorologically using SPI. From the results of the accuracy test and correlation, TVDI had the highest accuracy in the Arjuno-Welirang forest (96% accurate), while the best index for TNBTS was the VHI index (96% accurate). Keywords: drought indices, TVDI, VHI, NDDI, forest fires, Indonesia