摘要
Changes in land productivity have been endorsed by the Inter Agency Expert Group on Sustainable Development Goals(IAEGSDGs)as key indicators for monitoring SDG 15.3.1.Multiple vegeta-tion parameters from optical remote sensing techniques have been widely utilized across different land productivity decline processes and scales.However,there is no consensus on indicator selection and their effectiveness at representing land productivity declining at different scales.This study proposes a fusion framework that incorporates the trends and consistencies within the four com-monly used remote sensing-based vegetation indicators.We ana-lyzed the differences among the four vegetation parameters in different land cover and climate zones,finally producing a new global land productivity dynamics(LPD)product with confidence level degrees.The LPD classes indicated by the four vegetation indicators(VIs)showed that all three levels(low,medium,and high confidence)of increasing area account for 23.99%of the global vegetated area and declining area account for 7.00%.The Increase high-confidence(HC)area accounted for 2.77%of the total area,and the Decline-HC accounted for 0.35%of the total area.This study demonstrates the accuracy of the high-confidence(HC)area for the evaluation of land productivity decline and increase.The“forest”landcover type and“humid”climate zone had the largest increasing and declining area but had the lowest high-confidence proportion.The data product provides an important and optional reference for the assessment of SDG 15.3.1 at global and regional scales according to the specific application target.
基金
was funded by the National Key Research and Development Program of China(No.2016YFC0500806)
the Strategic Priority Research Program of the Chinese Academy of Sciences(Grant Number XDA19090124).
