With the increasingly severe global climate change problem, green supply chain finance, as an important financial tool to promote sustainable development, is facing unprecedented challenges. In this context, traditional credit risk assessment methods are unable to effectively address the complex environmental, social, and governance factors in green supply chain finance, as well as the multidimensional impact of climate risk. This study combines financial data, ESG scores, climate risk data, and supply chain network information of enterprises, and innovatively integrates graph neural networks and time series analysis models to construct a new credit risk assessment framework. The research results indicate that the ESG LimateGNN model exhibits superiority in multiple evaluation metrics, particularly in terms of discriminative ability, classification accuracy, and ranking ability, which significantly surpass traditional models and cutting-edge models in the field.
