Deforestation monitoring in the Amazon River estuary by multi-temporal Envisat ScanSAR data

9th Symposium of the International Society for Digital Earth

地球科学;计算机科学

Chen, F.^1,2 ; Ishwaran, N.^2 ; Pezzuti, J. C. Brito^3

Key Laboratory of Digital Earth Science, Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, No.9 Dengzhuang South Road Haidian District, Beijing

100094, China^1

Intl. Centre on Space Technologies for Natural and Cultural Heritage under the Auspices of UNESCO, No. 9 Dengzhuang South Road Haidian District, Beijing

100094, China^2

Universidade Federal Do Pará, Guamá - Belem

PA

66075-110, Brazil^3

关键词: Amazon river;    Anthropogenic activity;    Change detection;    Multi-temporal;    Multi-temporal analysis;    Operational capabilities;    Optical satellite imagery;    Spaceborne synthetic aperture radars;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/34/1/012003/pdf
DOI  :  10.1088/1755-1315/34/1/012003

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In this study, we have capitalized on the all-weather, all-day operational capability of spaceborne synthetic aperture radar (SAR) systems and used multi-temporal (from 2002 to 2006), multi-track (track 174, 360 and 447) Envisat ScanSAR amplitude images for deforestation mapping and change detection in the Amazon River estuary. A synergistic approach to deforestation mapping was adopted using SAR backscattering anomalies, the neighbouring forest constraint and DEM-derived slopes based on the three following characteristics: (1) backscattering is reduced in regions suspected to have undergone deforestation; (2) open regions without neighbouring forests were identified for removal; and (3) false-alarms linked to water bodies are mitigated using the shape threshold of flat-slope objects. Our results show that deforestation in the Amazon River estuary continues to be a serious problem, particularly along the rivers, streams or roads, which are more susceptible to anthropogenic activities than other areas. Up to 2006, the deforested portion accounts for 4.6 per cent (3,096,000 pixels) of the entire study site of approximately 458,000 square kilometers (67,320,000 pixels). However, this figure, validated by Landsat ETM images, may have overestimated deforestation to some extent. Nevertheless, multi-temporal analysis using SAR systems, as done in this study, have a clear potential for surveillance of deforestation in the Amazon, particularly in light of the frequent cloud cover typical of the area and the limitations of deforestation monitoring by means of optical satellite imagery.