Abstract:
Forest fires once get started cause huge damage. In Thailand, most forest fires are caused by human activities, most likely found in the dry season. This study aims to apply remote sensing technology based on data from Landsat 8 OLI satellite to investigate areas burned by forest fires in Doi Suthep-Pui National Park, Chiang Mai province, Thailand. Differences of spectral indices in 4 patterns are used, i.e. Normalized Difference Vegetation Index (NDVI), Normalized Burned Ratio (NBR), and Burn Area Index (BAI) in April 2021. How the study was conducted included 1) collecting data from Landsat 8 OLI satellite, 2) analyzing the difference of spectral indices in 4 patterns, i.e. NDVI, BAI, and NBR, and 3) analyze data accuracy using statistical methods. The study results revealed that BAI gave the most accurate data for investigating areas burned by forest fires, Kappa Statistics shown was 0.87, followed by NDVI showing Kappa Statistics equal to 0.77, and NBR showing Kappa Statistics equal to 0.67.
Page(s):
1039-1045
DOI:
DOI not available
Published:
Journal: ARPN Journal of Engineering and Applied Sciences, Volume: 18, Issue: 9, Year: 2023
Keywords:
Remote Sensing
,
Digital Image Processing
,
Electromagnetic Wave
,
spectral indices