Forest Carbon Stock Estimation Based On GEE Cloud Platform

Forest nature reserves are the hotspots for terrestrial biodiversity conservation and the areas with the largest carbon storage and sink capacity among all types of terrestrial ecosystems,playing an important role in maintaining regional carbon balance.This study uses the Google Earth Engine remote sensing cloud platform to fuse Sentinel-1 radar data,Sentinel-2 optical images,elevation data and forest management inventory to classify the forest types and estimate the forest carbon sequestration capacity of the reserve,taking Anyi Qiaoling Provincial Nature Reserve as an example.The main results are as follows:(1)The forest types of Anyi Qiaoling Nature Reserve were classified based on the resampling of forest management inventory,elevation data and Sentinel multi-seasonal remote sensing images.The result shows that the random forest model has the best classification effect with an overall accuracy of 85.97%and a Kappa coefficient of 0.79 compared with the support vector machine and the stochastic gradient boosting algorithm;accordingly,the forest types of Anyi Qiaoling Nature Reserve can be classified as coniferous forest,broad-leaved forest,mixed coniferous,broad-leaved forest and shrub forest.(2)The study extracted the distribution and changes in the forest biomass from2017 to 2022,and the results showed that,the back-propagation(BP)neural network regression model is better than random forest regression model and multiple linear regression model.The regression models for coniferous,broad-leaved,and mixed coniferous forests all performed better than the model based on mixed forest picture elements,with R~2of 0.73,0.78,and 0.71,and RMSE of 25.13 t/hm~2,16.08 t/hm~2,and 18.56 t/hm~2,respectively.From 2017 to 2022,total forest biomass increased by71,127.16 t at a pace of 19.54%,with coniferous forests increasing by 19.76%,broad-leaved forests increasing by 20.18%,and mixed coniferous forests increasing by 16.71%.(3)The study calculated the forest carbon stock,carbon sink,and its economic value from 2017 to 2022,the results showed that the forest carbon stock of Anyi Qiaoling Provincial Nature Reserve was 186,117.73 t in 2017,with an average annual growth rate of 3.91%.The cumulative carbon sink of the forest is predicted to reach 61,103.46 t and 290,241.44 t by 2030 and 2060,respectively,if the annual carbon sink is maintained at 1.73 t/hm~2.The conclusions of this study provide basic data and recommendations for the conservation of forest resources in Anyi Qiaoling Provincial Nature Reserve,which could assist with achieving the expected carbon peaking and carbon neutrality targets of 2030 and 2060,and the methodology of this study is suitable for extrapolation to the watershed scale and large scale,providing a reference for assessing the carbon storage and carbon sink function of the mid subtropical forest ecosystems.