Application of Polarimetric Decomposition and Interferometric SAR Using ALOS-2 PALSAR-2 Data to Detect Potential of Combustible Peatland Areas

Abstract

Forest fire in Indonesia occurred mostly in peatland area when the peatland areas were dried with groundwater table more than 40 cm. This peatland condition has become degradation areas with high potentials to fire. Some previous research utilized optics data remote sensing to detect the potential combustible peatland area while others concerned on backseat-tering information of synthetic aperture radar data compared with Forest Fire Danger Rating System (FDRS) data to identify a peat fire risk area. Peatland is prone to fire, usually associated with land that is open, close to the road, and dry conditions. In this research, polarimetric decomposition and interferometric SAR techniques have been used to determine the potential of combustible peatland area. Polarimetric Decomposition is easier to use to recognize open peatland areas. Besides that, it can also be used to identify roads and canals in peatland areas. Based on the research carried out this time, by implementing the Yamaguchi three-component model-based decomposition, we easily distinct areas that are exposed to peatlands. The dominant surface scattering marks open spaces. By using interferometric SAR technique, we also easily map areas experiencing subsidence. Regions that experience subsidence in peatland areas are usually dry areas with low groundwater conditions. Based on the results of interferometric synthetic aperture radar by using ALOS-2 PALSAR-2 data, subsidence conditions have been found in this area of peatland. SAR ALOS-2 PALSAR-2 image acquired on May 9, 2015, is used as master image and image received on March 25, 2017, is used as a slave image. Phase interferogram generated with multi-looking 5 × 5 pixels, Goldstein filtering 5 × 5 pixels, and normal baseline -43 m, and H-ambiguity 1049.4. Annual subsidence rate average was 2.8 cm/year, minimum 2.5 cm/year, and a maximum of 3.5 cm/year. The subsidence rate then converted to groundwater level information based on Woosten model and validated by using groundwater table measurement from the field. The simulation of groundwater table in average is 69.4 cm, with minimum value 63 cm and maximum value 87 cm. Based on validation and compared to the field data with correlation 0.85, and the area confirmed as high potential of combustible peatland area.