基于遥感技术的辽宁省半干旱地区森林地上碳储量估算OA

To understand the forest carbon storage and its spatial distribution in the semi-arid region of Liaon-ing and provide a basis for forest resource management and operation in the region,this study took the aboveground forest carbon storage in Chaoyang City as the research object.Using plot survey data,DEM da-ta,and simultaneous Landsat-8 remote sensing satellite images,four models were constructed:BP neural net-work,random forest regression,K-NN model,and support vector model.The results showed that the corre-lation coefficients between the predicted values and the true values of each model ranged from 0.55 to 0.84,among which the random forest(RF)model had the highest accuracy(coefficient of determination R2=0.84).The estimation results based on the RF model indicated that the average aboveground forest carbon density in Chaoyang City was 23.73 t·hm-2,with significant differences among counties and districts:Jian-ping County had the highest(28.14 t·hm-2),while Beipiao City had the lowest(17.91 t·hm-2).Combi-ning with the data forest resource management,it was found that the above-ground forest carbon storage in Chaoyang City is 10.505 million tons.