Carbon emission prediction model of prefecture-level administrative region: A land-use-based case study of Xi'an city, China

Abstract

Climate change has become a global concern, and the prediction of carbon emissions is key to achieving carbon-reduction goals. The existing framework cannot accurately reflect the spatial distribution of carbon emissions, making it difficult to guide urban planning and management. Therefore, in this study, a carbon emission spatial simulation and prediction model was established. The model includes the GIS-Kernel Density sub-model for subdividing built-up area, the Land Use-Carbon Emission sub-model for establishing the correlation between land use and carbon emissions, the Multi Objective Programming-Principal Component Analysis-BP neural network sub-model for presetting development scenarios, and the Patch-generating Land-use Simulation sub-model for predicting. Xi'an was chosen as the study site, and two extreme scenarios were determined. A total of 373,318 development paths were segmented from the interval, and the optimal path was selected. All scenarios were simulated, and the carbon emissions and their spatial distribution were calculated. The results showed that the overall accuracy of the simulation exceeded 90%. Under the optimal path, Xi'an's carbon emissions reach 60.6 million tons at peak time, which will be reduced to 47.38 million tons by 2060. In addition, the model analyzed the temporal and spatial changes of carbon sources and sinks and drew up the path of carbon reduction by technology and urban planning. This model can provide a reference for regional carbon-reduction planning and carbon reduction technology implantation. It can propose strategies from the perspective of planning and management.