Measurement of Soil Carbon Storage in Natural Ecosystems Based on Random Forest Model:Take Jingzhou Miao and Dong Autonomous County as an Example
This article takes Jingzhou Miao and Dong Autonomous County as an example,com-prehensively using remote sensing data and field sampling data,and using the random forest machine learning algorithm to calculate the soil organic carbon storage in the area.Through fitting analysis with field test sample data,the accuracy of the model evaluation is verified,and the following research results are obtained:(1)The accuracy verification result shows that the determination coefficient R2 of the random forest model is 0.90;(2)The soil organic carbon storage calculated by the random forest model is 8.99Tg,with an average carbon density of 4.18kg/m2;The overall distribution pattern of soil organic carbon density in Jingzhou shows a"low in the middle and high in the surrounding areas";(3)The soil carbon storage of several ecosystems in descending order is:forest land(7.77 Tg),farmland(0.92 Tg),wetland(0.33 Tg),grassland(0.01 Tg).This article proposes a set of high-precision and practical methods for the background investigation of natural ecosystem carbon storage,which not only provides scien-tific basis for carbon storage investigation and carbon sink monitoring in Jingzhou,but also provides method reference for soil organic carbon storage calculation in other areas with simi-lar geographical and environmental characteristics.
万方数据
