TY - JOUR
T1 - Mapping dynamics of deforestation and forest degradation in tropical forests using radar satellite data
AU - Joshi, Neha
AU - Mitchard, Edward TA
AU - Woo, Natalia
AU - Torres, Jorge
AU - Moll-Rocek, Julian
AU - Ehammer, Andrea
AU - Collins, Murray
AU - Jepsen, Martin Rudbeck
AU - Fensholt, Rasmus
PY - 2015/3/1
Y1 - 2015/3/1
N2 - Mapping anthropogenic forest disturbances has largely been focused on distinct delineations of events of deforestation using optical satellite images. In the tropics, frequent cloud cover and the challenge of quantifying forest degradation remain problematic. In this study, we detect processes of deforestation, forest degradation and successional dynamics, using long-wavelength radar (L-band from ALOS PALSAR) backscatter. We present a detection algorithm that allows for repeated disturbances on the same land, and identifies areas with slow- and fast-recovering changes in backscatter in close spatial and temporal proximity. In the study area in Madre de Dios, Peru, 2.3% of land was found to be disturbed over three years, with a false positive rate of 0.3% of area. A low, but significant, detection rate of degradation from sparse and small-scale selective logging was achieved. Disturbances were most common along the tri-national Interoceanic Highway, as well as in mining areas and areas under no land use allocation. A continuous spatial gradient of disturbance was observed, highlighting artefacts arising from imposing discrete boundaries on deforestation events. The magnitude of initial radar backscatter, and backscatter decrease, suggested that large-scale deforestation was likely in areas with initially low biomass, either naturally or since already under anthropogenic use. Further, backscatter increases following disturbance suggested that radar can be used to characterize successional disturbance dynamics, such as biomass accumulation in lands post-abandonment. The presented radar-based detection algorithm is spatially and temporally scalable, and can support monitoring degradation and deforestation in tropical rainforests with the use of products from ALOS-2 and the future SAOCOM and BIOMASS missions.
AB - Mapping anthropogenic forest disturbances has largely been focused on distinct delineations of events of deforestation using optical satellite images. In the tropics, frequent cloud cover and the challenge of quantifying forest degradation remain problematic. In this study, we detect processes of deforestation, forest degradation and successional dynamics, using long-wavelength radar (L-band from ALOS PALSAR) backscatter. We present a detection algorithm that allows for repeated disturbances on the same land, and identifies areas with slow- and fast-recovering changes in backscatter in close spatial and temporal proximity. In the study area in Madre de Dios, Peru, 2.3% of land was found to be disturbed over three years, with a false positive rate of 0.3% of area. A low, but significant, detection rate of degradation from sparse and small-scale selective logging was achieved. Disturbances were most common along the tri-national Interoceanic Highway, as well as in mining areas and areas under no land use allocation. A continuous spatial gradient of disturbance was observed, highlighting artefacts arising from imposing discrete boundaries on deforestation events. The magnitude of initial radar backscatter, and backscatter decrease, suggested that large-scale deforestation was likely in areas with initially low biomass, either naturally or since already under anthropogenic use. Further, backscatter increases following disturbance suggested that radar can be used to characterize successional disturbance dynamics, such as biomass accumulation in lands post-abandonment. The presented radar-based detection algorithm is spatially and temporally scalable, and can support monitoring degradation and deforestation in tropical rainforests with the use of products from ALOS-2 and the future SAOCOM and BIOMASS missions.
U2 - 10.1088/1748-9326/10/3/034014
DO - 10.1088/1748-9326/10/3/034014
M3 - Journal article
SN - 1748-9326
VL - 10
JO - Environmental Research Letters
JF - Environmental Research Letters
IS - 3
M1 - 034014
ER -
