This investigation casts a wide net to identify climate information uses across climate action planning documents at state, local, and tribal levels. Our research team, which for this investigation includes collaborators at the Hebrew University, Tom Hope and Maayan Sharon, is employing document analysis and natural language processing to try to more quickly determine the types and sources of information planners use to inform their climate action planning documents.
Our document database currently consists of 405 climate adaptation, resilience, vulnerability, and action plans from hundreds of jurisdictions and dozens of tribes from around the U.S., with nearly 200,000 sentences containing “fingerprints” of potential examples of climate information use. Research team members are manually “tagging” a sample of these sentences to “train” a natural language processing model that may be then capable of analyzing climate information sources and uses across the larger set of documents.
While we hope this investigation will result in new insights and a new methodology, it has not been previously used for this type of evaluation. As such, it is a higher risk activity, but if successful, this method would enable more rapid and wide-angled exploration of how documented uses of climate information changes over time, both within and across sectors.
For more information on this investigation, contact firstname.lastname@example.org.