Climate impacts and where to find them: insights from text mining

Climate extremes, such as droughts, floods, and heatwaves, often trigger compound and cascading impacts due to interdependencies between coupled natural and social systems. Yet, our knowledge of these interactions remains limited mainly due to the lack of comprehensive impact data. Research typically considers only one isolated impact, system, socioeconomic sector, and/or hazard at a time, often ignoring dependencies between impacts as well as how they interact with response and adaptation measures.

Against this backdrop, the unprecedented abundance of digital texts and cutting-edge machine-learning tools has opened new research avenues for impact assessment research. In this talk, I will demonstrate how we can leverage natural language processing (NLP) and large language models on different text types to infer how climate extremes impact society. I will discuss the potential of unconventional data sources, such as meeting minutes, newspaper articles, and reports, to monitor the consequences of extreme events in near real-time.