This simple study provides the technique to detect fire burnt in oil palm plantations, Riau Province, Indonesia.
The image acquisition date was June 25, 2013. As known, on late June 2013, smoke blew from burning forest in Sumatera Island to Malaysia dan Singapore. Indonesia then seen as the main perpetrator of this smoke due to forest conversion into oil palm expansion.
However, if we overlay the GIS datasets consist of Multinational Companies (MNCs) of Oil Palm Plantations, we can conclude that the fire burnt sources are came from these MNCs, which is the head office of the MNCs located in Malaysia dan Singapore.
The algorithm employed in this study are as follow:
1. Atmospheric correction using Dark object subtraction (DOS) technique. DOS searches each band for the darkest pixel value. Assuming that dark objects reflect no light, any value greater than zero must result from atmospheric scattering. The scattering is removed by subtracting this value from every pixel in the band. This simple technique is effective for haze correction in multispectral data, but it should not be used for hyperspectral data. DOS employed to all bands of Landsat 8 image.
2. Then the Normalized Burn Ratio (NBR) is used to enhance the spectral response of fire affected vegetation. The Normalized Burn Ratio is calculated from L8 bands 5 and 7 as: (B5 – B7)/ (B5 + B7) where B5 represents the near-infrared spectral range (0.845 -0.885 μm) and B7 represents the shortwavelength infrared spectral range (2.10 – 2.30 μm)
3. Finally, Red-Green-Blue band combination used to highlighted the fire burnt. Red layer is thermal band 10 (10.30 – 11.30 μm), Blue layer is NBR, and Green layer is shortwavelength infrared band.
4. Visualization using linear 2%