ClimateCause: Complex and Implicit Causal Structures in Climate Reports

arXiv cs.CL / 4/17/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces ClimateCause, a new expert-annotated dataset built from science-for-policy climate reports to capture complex, higher-order causal structures beyond what existing datasets provide.
It normalizes and disentangles cause-effect expressions into individual causal relations, while adding metadata for correlation type, relation type, and spatiotemporal context to support graph construction.
The authors show that ClimateCause can be used to quantify the readability of climate statements by linking it to the semantic complexity of the causal graphs.
Benchmarking with large language models indicates that causal chain reasoning is a particularly challenging problem compared with correlation inference on this dataset.
Overall, the dataset and experiments aim to improve causal reasoning and evaluation for climate-change understanding and science-to-policy communication.

Abstract

Understanding climate change requires reasoning over complex causal networks. Yet, existing causal discovery datasets predominantly capture explicit, direct causal relations. We introduce ClimateCause, a manually expert-annotated dataset of higher-order causal structures from science-for-policy climate reports, including implicit and nested causality. Cause-effect expressions are normalized and disentangled into individual causal relations to facilitate graph construction, with unique annotations for cause-effect correlation, relation type, and spatiotemporal context. We further demonstrate ClimateCause's value for quantifying readability based on the semantic complexity of causal graphs underlying a statement. Finally, large language model benchmarking on correlation inference and causal chain reasoning highlights the latter as a key challenge.