[Prediction and Path Analysis of Carbon Peak in Heilongjiang Province]

Heilongjiang Province, as a Chinese heavy industry base and major grain producing area, has faced notable challenges in recent years due to climate change. To effectively predict the trend of carbon emissions, the Bayesian optimization algorithm, isomap algorithm, and support vector machine algorithm were integrated to construct a carbon emission prediction model. The results demonstrated that the integrated prediction model exhibited excellent generalization ability, with a correlation coefficient of 0.99 during the training phase, and maintained generalization ability with a correlation of 0.81 during carbon emission fluctuations for satisfactory prediction accuracy. Carbon emissions in Heilongjiang Province were forecast to peak at 381 million tons by 2035 through low-carbon development, five years ahead of baseline projections with a 6.39% cut. This peak would come ten years sooner than that under high-speed growth, reducing emissions by 17.06%. Analysis with the Tapio Decoupling Index Model （2012~2019） showed an unstable decoupling between economic growth and carbon emissions but a positive trend in energy intensity vs. emissions. Discussions are underway based on these forecasts and decoupling insights to outline pathways for Heilongjiang to reach its carbon peaking goals earlier.