Enhancement of fire danger rating system for a better land/forest fire warning in South Sumatera Province, Indonesia

Abstract

Forest Fire Danger Rating System (FDRS) developed in Indonesia is based on the Canadian Forest Fires Danger Rating System. The Meteorology, Climatology, and Geophysics Agency operates and publishes a daily Fire Weather Index system on its website as part of the FDRS. The so-called SPARTAN system is based on weather elements of rainfall, air temperature, wind speed, and humidity and does not consider soil conditions. This research aims to improve the Fire Weather Index system by adding information on land conditions. In this study, the area of interest was South Sumatera Province of Indonesia and the period of analysis was 2019. The normalized difference polarization index (NDPI) derived from the Synthetic Aperture Radar (SAR) data of the Sentinel-1 satellite and land cover changes and fire incidents derived from the optical data of the Sentinel-2 satellite are used to represent land conditions. Since NDPI shows a good correlation with the degree of soil moisture, the NDPI is considered for the soil moisture conditions. Furthermore, integrating soil moisture conditions and land cover changes into the FWI system provides better early warning information for land/forest fires. Fire hotspot data and in-situ fire information are used to validate the results. This study concludes that adding information on land conditions will provide detailed and better land/forest fire warnings.