ZHANG Guojun, MA Wentao, WEI Xiaoyan, ZHANG Huwei, WANG Xiao, TAO Qingrui, SU Liping, ZHANG Rong, WANG Liming, XIAO Peiqing, WANG Zhihui
In order to explore accurate accounting methods for carbon sequestration in soil and water conservation measures and to provide technical support for carbon trading in soil and water conservation projects, this study focused on Ningxia as the research area. By comprehensively utilizing literature, remote sensing technology, artificial intelligence algorithms, and statistical analysis methods, a retrieval method for ecosystem carbon sequestration rates and carbon storage of soil and water conservation measures based on multimodal data and artificial intelligence algorithms was built. The 2022 carbon sequestration status of soil and water conservation measures in Ningxia was quantitatively assessed. The results show that a) among the six artificial intelligence machine learning algorithms, such as Random Forest (RF), Artificial Neural Network (ANN), Support Vector Machine (SVM), Lightweight Gradient Boosting (LGB), Extreme Gradient Boosting (XGB) and Extreme Random Tree Regression (EXT), the Extreme Random Tree Regression algorithm has the highest accuracy in inverting vegetation carbon density and soil organic carbon density in Ningxia. b)The spatial distribution of ecosystem carbon sequestration rates in Ningxia exhibits a pattern of higher values in the south and lower values in the north, with a regional average of 24.53 g/(m2·a). The carbon sequestration rates for soil and water conservation forests, enclosure management, economic forests, artificial grasslands, and terraced fields are 26.65, 27.25, 27.28, 18.80, and 22.68 g/(m2·a), respectively. c) The total carbon sequestration of soil and water conservation measures in Ningxia in 2022 ranges from 2.269 5 to 2.332 6 million tons per year, including an increased carbon sequestration of 2.080 3 million tons per year, soil carbon retention of 0.630 4 million tons per year, and emission reductions of 0.189 2 to 0.252 3 million tons per year. Of the increased carbon sequestration, 80.8% is distributed in vegetative measures, 79.5% of soil carbon retention is distributed in terraced fields and 79.5% of emission reductions are also distributed in terraced fields.