Wikimpacts 1.0: a new global climate impact database based on automated information extraction from Wikipedia
9 affiliations
- 1Department of Hydro Sciences, TUD Dresden University of Technology, Dresden, Germany
- 2Department of Water and Climate, Vrije Universiteit Brussel, Brussels, Belgium
- 3RISE Research Institutes of Sweden, Sweden
- 4Swedish centre for impacts of climate extremes (climes), Uppsala University, Uppsala, Sweden
- 5Department of Earth Sciences, Uppsala University, Uppsala, Sweden
- 6Department of Linguistics and Philology, Uppsala University, Uppsala, Sweden
- 7Center for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI), Dresden/Leipzig, Germany
- 8Department of Compound Environmental Risks, Helmholtz Centre for Environmental Research — UFZ, Leipzig, Germany
- 9Department of Meteorology and Bolin Centre for Climate Research, Stockholm University, Stockholm, Sweden
Abstract
Climate extremes like storms, heatwaves, wildfires, droughts and floods significantly threaten society and ecosystems. However, comprehensive data on the socio-economic impacts of climate extremes remains limited. Here we present Wikimpacts 1.0, a global climate impact database built by extracting information from Wikipedia using natural language processing. Our method identifies relevant articles, extracts the information using GPT4o, post-processes the information and consolidates the database. Impact data is stored at the event, national, and sub-national levels, covering 2,928 events from 1034 to 2024, with 20,186 national and 36,394 sub-national entries. The database shows low error scores (range from 0 to 1) for event-level information like timing (0.05), deaths (0.03), and economic damage (0.12), and slightly higher error scores for injuries (0.21), homelessness (0.25), displacement (0.29), and damaged buildings (0.28) compared to manually annotated data from 156 events. Wikimpacts 1.0 provides broader impact coverage on storms than EM-DAT at the sub-national level. In comparing impact values, 38 out of 234 matched events have identical data for deaths, and 7 of 94 for injuries. However, there are notable discrepancies in information on homelessness and damage. Our public database highlights the potential of natural language processing to complement existing impact datasets and to provide robust information on climate impacts.
Related
- Advances in intelligent extraction and identification of metallogenic information for mineral prospectivity mapping
- Chemical characterization and source tracking for aerosols: Development of an innovative selective sequential extraction protocol
- The Puzzle of Lost PAHs: A Step-by-Step Investigation of Pre-, During-, and Post-Solid-Phase Extraction from Water Samples
- Accurate determination of fluoxetine and its metabolites in Mediterranean mussel organs based on solid phase extraction clean-up and HPLC-HRMS
- Historical air quality database in Navarre, Spain (2011–2025): Harmonized hourly records and relational Star Schema of atmospheric pollutants
BibTeX
@article{li2026wikimpacts,
title = {{Wikimpacts 1.0: a new global climate impact database based on automated information extraction from Wikipedia}},
author = {Ni Li and Wim Thiery and Shorouq Zahra and Mariana Madruga de Brito and Koffi Worou and Murathan Kurfalı and Seppe Lampe and Paúl Muñoz and Clare M. Flynn and Camila Trigoso and Joakim Nivre and Jakob Zscheischler A and Gabriele Messori},
journal = {Natural hazards and earth system sciences},
year = {2026},
month = jun,
doi = {10.5194/nhess-26-2609-2026},
url = {https://doi.org/10.5194/nhess-26-2609-2026},
note = {CC BY},
}
RIS
TY - JOUR
TI - Wikimpacts 1.0: a new global climate impact database based on automated information extraction from Wikipedia
AU - Ni Li
AU - Wim Thiery
AU - Shorouq Zahra
AU - Mariana Madruga de Brito
AU - Koffi Worou
AU - Murathan Kurfalı
AU - Seppe Lampe
AU - Paúl Muñoz
AU - Clare M. Flynn
AU - Camila Trigoso
AU - Joakim Nivre
AU - Jakob Zscheischler A
AU - Gabriele Messori
PY - 2026
DA - 2026/06/04
JO - Natural hazards and earth system sciences
DO - 10.5194/nhess-26-2609-2026
UR - https://doi.org/10.5194/nhess-26-2609-2026
AB - Climate extremes like storms, heatwaves, wildfires, droughts and floods significantly threaten society and ecosystems. However, comprehensive data on the socio-economic impacts of climate extremes remains limited. Here we present Wikimpacts 1.0, a global climate impact database built by extracting information from Wikipedia using natural language processing. Our method identifies relevant articles, extracts the information using GPT4o, post-processes the information and consolidates the database. Impact data is stored at the event, national, and sub-national levels, covering 2,928 events from 1034 to 2024, with 20,186 national and 36,394 sub-national entries. The database shows low error scores (range from 0 to 1) for event-level information like timing (0.05), deaths (0.03), and economic damage (0.12), and slightly higher error scores for injuries (0.21), homelessness (0.25), displacement (0.29), and damaged buildings (0.28) compared to manually annotated data from 156 events. Wikimpacts 1.0 provides broader impact coverage on storms than EM-DAT at the sub-national level. In comparing impact values, 38 out of 234 matched events have identical data for deaths, and 7 of 94 for injuries. However, there are notable discrepancies in information on homelessness and damage. Our public database highlights the potential of natural language processing to complement existing impact datasets and to provide robust information on climate impacts.
SN - 1561-8633
SN - 1684-9981
ER -