Forest definitions collaboration based on global remote sensing data products
Forest definitions and remote sensing datasets provide a conceptual basis for monitoring forest change.In this study,we present an overview of forest definitions from the views of land use and land cover,introduce the forest categorization from three aspects:growth mode,forest age and canopy density,and review the evolution of forest remote sensing dataset from single sensor to optical and microwave remote sensing.Additionally,the differences in forest definition between various remote sensing data sets were compared from three perspectives:the threshold for forest elements,the level of classification and the spatial resolution.The shortcomings in product accuracy verification were summarized using data consistency,validation samples,and regional accuracy differences.In the future,the forest definition should be further coordinated based on the forest definition framework of"perspective-factor-threshold",and the area estimation bias caused by the different factor thresholds in the forest definition should be minimized.Meanwhile,deep learning and multi-source remote sensing data should be applied to produce accurate forest remote sensing data sets,especially for identifying various forest species.Finally,platforms for forest remote sensing datasets sharing need to be built to clarify the forest definition,spatio-temporal resolution,and data accuracy of the datasets.
NETL
NSTL
万方数据
