Prediction and Analysis of Carbon Emissions in China Based on ARIMA-BP Model

Abstract

Based on the energy consumption data of 30 provinces in China from 2000 to 2021, this paper estimates and predicts the total carbon emissions of 30 provinces in China from 2000 to 2035 using ARIMA model and BP neural network model. ArcGIS and standard elliptic difference are used to visually analyze the spatio-temporal evolution characteristics, and LMDI model is further used to decompose the driving factors affecting carbon emissions. The results show that: (1) China's total carbon emissions increased year by year from 2000 to 2035, but the growth rate of carbon emissions decreased gradually; The carbon emission structure is "secondary industry > residents' livelihood > tertiary industry > primary industry". the growth rate of carbon in secondary industry and residents' livelihood is relatively fast, while the change trend of primary industry and tertiary industry is relatively small. (2) the spatial distribution of carbon emissions in China's provinces presents a typical "eastern > central > western" and "northern > southern" distribution pattern, with the carbon emission center moving to the northwest; (3) The regions with higher development level of digital economy, industrial structure and new quality productivity have relatively less carbon emissions, with significant group difference effect; (4) Energy consumption intensity effect is the main factor to drive the continuous growth of carbon emissions, per capital GDP and energy consumption structure effect are the main factors to curb carbon emissions, and the impact of industrial structure and population size effect is relatively small. Based on the research conclusions, policy suggestions are put forward from the aspects of energy structure, industrial structure, new quality productivity and digital economy.