Abstract
This study examines recent research trends in the application of artificial intelligence (AI) to climate change research. Using a refined keyword dataset extracted from the abstracts of 5,103 research articles identified through title- and author-keyword-based search criteria in the Web of Science Core Collection and published between January 2018 and June 2025, we applied Latent Dirichlet Allocation (LDA) topic modelling to identify the major thematic structures of the climate–AI literature. The analysis yielded 14 detailed topics, which were grouped into four major research areas: (1) AI-based climate system prediction and modelling, (2) AI-based climate change impact and risk analysis, (3) AI-utilizing climate change mitigation and policy analysis, and (4) AI methodology and data-driven techniques. The results indicate that solution-oriented themes, including mitigation, risk management, and methodological innovation, occupy a substantial portion of the recent AI–climate literature within the analyzed corpus. The findings provide a structured overview of how AI applications in climate change research are distributed across major thematic areas, highlighting the integration of heterogeneous data sources, methodological advancements, and cross-disciplinary applications. Overall, this study offers a corpus-defined perspective on the thematic structure of recent climate change research involving AI.