To limit the impacts of climate change as much as possible, rapid climate action is essential. Although still insufficient to meet international climate goals, the breadth and speed of the climate-driven transformation make it challenging for climate scholars to follow developments across emerging research and innovations. Meanwhile, Natural Language Processing (NLP) technologies advance at an unprecedented scale. Particularly with the advent of Large Language Models, NLP is seen as a promising tool for tackling pressing societal challenges.
However, NLP and climate science are progressing in parallel: NLP is primarily disseminated through major conferences, while climate research is typically presented within journals, and more policy-oriented venues. The result is a fractured landscape, where key innovations in one domain may not easily translate into actionable insights or tools in the other.
This topical collection addresses this gap by showcasing frontier NLP methods for climate action, sparking ideas and building a community around responsible best practices for adopting NLP methods and LLMs. The collection will benefit both the climate community seeking to capitalize on advances in NLP and the applied NLP community facing growing pressure to focus on applications for public good and sustainability.
We particularly welcome submissions that use NLP or more broadly AI methods to:
• Develop and evaluate real-world tools to support climate action
• Provide new and actionable perspectives on climate change science
• Enable large-scale evidence synthesis and review
• Support adaptation, mitigation or other relevant tracking
• Improve transparency and accountability
• Detect or analyze narratives, framing, greenwashing and misinformation around climate change
In addition, we encourage submissions addressing other topics at the intersection of NLP (and more broadly AI) and climate action, including novel methods, datasets, applications, or perspectives that can help accelerate progress toward climate goals. We strongly encourage submissions that include publicly available code, data, or models to promote reproducibility. Papers should clearly articulate the climate science implications of their work.