BURNED AREA ASSESSMENT USING BAMS: A CASE STUDY OF UPPER NORTHERN REGION, THAILAND.

Currently, the burned area statistics of Thailand use only fire suppression data from forest fire control station reports. However, forest fire phenomena are varied and some places are difficult to assess with ground surveys. The specific objectives of the study are: (1) to identify burned areas using burned area mapping software (BAMS) and assess the accuracy and (2) to quantify and analyze the characteristics of the burned areas of the Upper Northern region of Thailand. BAMS is supervised burned area mapping software based on Landsat data and two-phase supervised classification by using multi-temporal differences (pre- and post-fire scenes) of spectral indices. The research methodology consisted of 4 components comprising (1) data collection and preparation, (2) burned area extraction using BAMS, (3) accuracy assessment, and (4) overlay analysis. The results showed that in 2014, March had the most burned areas (1550.83 km²), followed by April (509.14 km²) and February (296.06 km²). Overall accuracy for 2 Landsat scenes, Path 130 Row 47 (Date 13 March 2014) and Path 131 Row 47 (Date 20 March 2014), with Department of National Parks, Wildlife and Plant Conservation (DNP) ground fire records in March 2014 were 93.23% and 90.72%, respectively. Based on overlay analysis, Chiang Mai had the highest total of burned areas (511.96 km²) followed by Lampang (466.92 km²) and Mae Hong Son (400.28 km²). In addition, there were 31 of 32 national parks affected by forest fires in 2014, the exception being Si Satchanalai National Park. The top 3 national parks which were most affected by forest fires were Op Luang, Mae Ping, and Mae Yom. Likewise, there were 21 of 22 wildlife sanctuaries affected by forest fires in 2014, the exception being Lamnam Nan Fang Khwa Wildlife Sanctuary. The top 3 wildlife sanctuaries which were most affected by forest fires were Lumnam Pai, Salawin, and Tham Chao Ram. According to the forest land use data in 2000 of the Royal Forest Department (RFD), the burned areas in 2014 inside the forest areas totaled 1569.52 km² or 66.62% and outside the forest areas totaled 786.51 km² or 33.38%. The types of forest most affected by fire were mixed deciduous forest and dry dipterocarp forest. The results from the study demonstrate that BAMS is suitable for burned area mapping in Thailand. Therefore, the DNP and the RFD under the Ministry of Natural Resources and Environment should apply BAMS as an operational tool to obtain annual burned area maps from Landsat data after the end of the fire season. [ABSTRACT FROM AUTHOR]