Prediction of Carbon Emissions Based on PCA and Neural Network Models under the "Dual Carbon" Background

With the continuous increase in global greenhouse gas emissions, climate deterioration and energy shortages have become pressing issues each year. As a major emitter of carbon, China's specific timeline for reaching its peak carbon emissions and the peak value itself have garnered significant attention from the international community. Due to China's vast territory and uneven economic and cultural development across regions, there are substantial differences in carbon emissions. To address these disparities, it is crucial to implement energy-saving and emission-reduction strategies that are tailored to the development conditions of each region. This study focuses on key representative provinces and cities in four regions: Central, Eastern, Western, and Northeastern China. The prediction of carbon emissions under the "dual carbon" context is of great practical significance for regional ecological civilization construction, the high-quality development of the national economy, and the improvement of people's quality of life. This study employs a combination of Principal Component Analysis (PCA) and neural network methods to predict carbon emissions. It establishes three scenarios to forecast the carbon emissions of four provinces and cities from 2020 to 2050, followed by analysis and comparison.