Construction of urban low-carbon development model based on digital economy: local practice to resist global warming

As global warming intensifies, cities face the challenge of fragmented pathways for low-carbon transformation. This paper establishes a multi-source carbon data system to clarify governance roles. It integrates transportation, building, and energy data to construct a digital twin city model, dynamically simulating emission reduction policies. It also uses graph neural networks to reveal coupling relationships and combines blockchain with federated learning to achieve multi-sector collaborative governance. Experiments demonstrate a collaborative governance index of 8.40, a 100% compliance rate for key indicators, and a 5.7% prediction error, validating the advantages of systematic low-carbon governance.