Deforestation Detection in Indonesian Tropical Forest Using Multi-Temporal ALOS/PALSAR Backscatter Data

  • I GD Yudha Partama Postgraduate Program on Regional Planning and Environmental Management, Universitas Mahasaraswati Denpasar
  • Tasuku Tanaka Graduate School of Science and Engineering, Yamaguchi University, Yamaguchi, Japan
  • Kakuji Ogawara Graduate School of Science and Engineering, Yamaguchi University, Yamaguchi, Japan
  • Takahiro Osawa Center for Remote Sensing and Ocean Science (CReSOS), Udayana University
  • Putu Sujana Postgraduate Program on Regional Planning and Environmental Management, Universitas Mahasaraswati Denpasar

Abstract

This study tested the effectiveness of multi-temporal Phased Array L-band Synthetic Aperture Radar (PALSAR) data to detect the deforestation in tropical forest region, Riau Province, Indonesia. The image differencing method by using a threshold was utilized for automatic detection of deforestation mapping. In addition, the influence of SAR polarization and seasonal changing due to rainfall event on accuracy of deforestation detection was investigated. Overall, we obtained the following results: (1) the accuracy of deforestation detection mapping using horizontal transmit-vertical receives (HV) polarization is giving better results than horizontal transmit-horizontal receives (HH) polarization. (2) The seasonal changing due to rainfall event affected the radar backscattering signatures of deforested areas and deforestation detection accuracy. High-accumulated precipitation decreased the sigma nought (σ0) difference between deforested areas and natural forests, causing decreased accuracy of deforestation detection. The best selection
of PALSAR data before detection and fixing a threshold value can be one of alternative to reduce the error in automatic detection. The overall accuracy of deforestation detection mapping using a fixed threshold was 92.21%. The deforestation mapping method proposed in the study can be utilized for assessment of yearly changes in forested areas and is useful for tracking changes in forest cover on large scales.

Published
2020-11-30