Spatiotemporal CO₂ emission prediction modeling methods: a comprehensive literature review

Abstract Spatiotemporal prediction of carbon dioxide ( $$\:{CO}_{2}$$ ) emissions plays a crucial role in understanding the dynamic distribution of carbon dioxide emission sources and supporting targeted mitigation and enhancing climate modelling accuracy. Despite its growing significance, research in this area remains fragmented. This study presents the first comprehensive systematic literature review on spatiotemporal $$\:{CO}_{2}$$ emission prediction, conducted using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework. A total of 45 peer-reviewed studies published between 2000 and 2025 were identified and analysed to evaluate data types, modelling approaches, influencing factors, and performance metrics. The review reveals rapid advances in data-driven, hybrid, and remote-sensing-integrated models, yet highlights substantial gaps in fine-scale prediction, temporal generalization, and cross-scale model integration. The findings underscore the urgent need for unified spatiotemporal frameworks that can enhance predictive accuracy, address data scarcity, and strengthen the scientific basis for carbon management and climate policy formulation.