Forest fire is a frequent disaster which happened in Palangka Raya, Central Kalimantan, Indonesia each year. On 2019, various fire incidents were happened in the area, affected human's health due to the fire smoke. Monitoring forest fire burn areas become important in order to observe affected areas and for mitigation purposes. Active remote sensing is suitable for forest fire identification, since it can penetrate through fog, clouds and smoke, hence the identification is better than passive remote sensing. Nowadays, with the availability of Sentinel-1 SAR free data with medium spatial resolution and high temporal resolution, monitoring large forest areas can be done easily. Burn areas can be extracted from Sentinel-1 with one of burn area index, namely Radar Burn Difference (RBD). In this study, burn areas identification was performed using RBD and threshold method. VV and VH polarization was used in identifying burn areas. Towards the best identification, some threshold values were tested. The results showed that threshold of μ-1σ for RBD VH had the highest overall accuracy about 88% in identifying burn areas, compared with reference data. It means the result is accurate enough in presenting burn areas. The results show a 15,935.197-hectare burned area from RBD VV and 15,679.835-hectare from RBD VH.
|IOP Conference Series: Earth and Environmental Science
|Published - 2023
|8th Geomatics International Conference, GeoICON 2023 - Surabaya, Indonesia
Duration: 27 Jul 2023 → …