--- title: "EM-DAT Disaster Classification" description: "Explore 43 EM-DAT disaster categories and evidence linking extreme-heat reporting, normalized climate-sensitive attention and hazard-specific narratives." url: "https://nosible.com/ontologies/em-dat" --- [Home](https://nosible.com/) [Ontologies](https://nosible.com/ontologies) EM-DAT Ontology field guide By NOSIBLE Research Updated 2026-07-18 # EM-DAT Disaster Classification EM-DAT separates disasters by cause, impact channel and financial exposure. Disaster counts are not interchangeable. Heat, floods, storms, earthquakes and industrial accidents create different loss channels, exposed assets and response windows. EM-DAT preserves those distinctions in four levels. World uses a three-level reduction so researchers can construct hazard-specific, point-in-time panels instead of relying on a generic disaster flag. Each cohort retains its distinct transmission mechanism. [[ 1 ]](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) The distinction matters for asset pricing and risk. Physical hazards affect property, production, insurance, credit and public balance sheets through different mechanisms. A classified event stream can show when each mechanism enters reporting and which entities or regions are exposed. It cannot replace physical observations, loss estimates or causal attribution from independent evidence. [[ 4 ]](https://www.ngfs.net/en/press-release/ngfs-publishes-latest-long-term-climate-macro-financial-scenarios-climate-risks-assessment) [[ 5 ]](https://doi.org/10.1111/jofi.13219) Categories 43 categories Structure 3 levels World events labelled 9.1% Stable codes Since v1 Release data CC0 On this page [01 Foundations](https://nosible.com/ontologies/em-dat#foundations) [02 Categories](https://nosible.com/ontologies/em-dat#vocabulary) [03 Example usages](https://nosible.com/ontologies/em-dat#trends) [04 Downloads and references](https://nosible.com/ontologies/em-dat#downloads) Foundations ## EM-DAT separates disaster mechanisms but does not measure physical hazard intensity EM-DAT divides disasters into natural and technological groups, then into subgroups, types and subtypes. World stops at type. The hierarchy separates meteorological, hydrological, climatological, geophysical and biological events. This matters because heat, floods, wildfire and earthquakes have different spatial footprints, lead times and financial transmission channels. Those differences determine the appropriate exposure map. [[ 1 ]](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) [[ 2 ]](https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11/) World classifies the event described in reporting. It does not determine whether an event meets the loss or mortality thresholds used by the EM-DAT disaster database. It also does not estimate physical intensity, insured loss or climate attribution. Validate every result against independent hazard, exposure and vulnerability data. [[ 1 ]](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) Categories ## EM-DAT uses three levels to distinguish natural, technological and complex disasters World implements two roots, nine intermediate groups and 32 disaster types across three levels. Each assigned event receives one path from group to type. The explorer exposes definitions, stable codes, full paths and World V1.2 counts. Search and bounded tree expansion keep the interface usable as larger ontologies are added. Search code, label, definition or path 2 shown · 43 total Labelled events **1,393,908** World coverage **9.1%** All World events **15,311,040** 1,393,908 of 15,311,040 World events carry labels from EM-DAT Disaster Classification. Select a category to see its count and both relevant shares. Hierarchy disaster_group / disaster_subgroup / disaster_type + Natural Natural hazards grouped by physical mechanism · 6 children + Technological Human, infrastructure and industrial failure events · 3 children Natural ### Natural Copy link #### Definition Disasters caused by geophysical, meteorological, hydrological, climatological, biological or extraterrestrial hazards. Potential research use Natural hazards grouped by physical mechanism Code Natural Events with label 661,379 Share of labelled 47.4% Share of World 4.3% **Events with label** is the selected count. **Label share** divides it by 1,393,908 assigned events; **World share** divides it by all 15,311,040 events. #### Representative World V1.2 events One strong classified example per available year, with up to ten years shown. 10 examples 1. 2015-10-04French Riviera Floods Kill At Least 16, Natural Disaster DeclaredCoverage 95 2. 2016-11-14World Bank: Natural Disasters Cost $520bn and Push 26m into PovertyCoverage 40 3. 2017-01-042016 Natural Disasters Cause Record $175 Billion in Global DamagesCoverage 20 4. 2019-01-22Poll: Natural Disasters Shape American Climate Change ViewsCoverage 29 5. 2020-10-12UN: Climate Change Doubled Natural Disasters Since 2000Coverage 29 6. 2021-09-25Haitian Migrants Flee Natural Disasters Amid Global Focus on Del RioCoverage 27 7. 2022-09-12Kentucky Natural Disasters Surge in Frequency and CostCoverage 190 8. 2024-02-28Natural Disasters and Inflation Drive Rising Home Auto Insurance CostsCoverage 141 9. 2025-01-24US Natural Disaster Losses Hit $218 Billion in 2024Coverage 72 10. 2026-01-24US Winter Storm Cuts Natural Gas Output and Spikes Power PricesCoverage 55 Browse categories to compare definitions, World V1.2 statistics and representative classified events. Example usage 1 of 3 ### Example Usage: Extreme-heat reporting nearly doubles as NASA temperatures rise while cold coverage barely moves World's EM-DAT Extreme temperature category combines heat and cold events at one level. We split 23,139 English assignments into explicit heat, cold or unresolved reporting cohorts, then divide monthly counts by all English World events. Heat rises from 81.0 to 160.9 reports per 100,000 between 2015-2019 and 2023-2025. [[ 1 ]](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) NASA's global anomaly rises from 0.93°C to 1.21°C across the same periods. Heat coverage increases 98.7%; cold and severe-winter coverage increases 25.5%. The smoothed month-to-month relationship is modest (r = 0.22), so the evidence supports structural alignment, not a claim that global temperature mechanically determines monthly reporting throughout the geographically heterogeneous World reporting corpus. [[ 6 ]](https://data.giss.nasa.gov/gistemp/index.html) [[ 7 ]](https://www.nasa.gov/news-release/temperatures-rising-nasa-confirms-2024-warmest-year-on-record/) Heat reporting +98.7% Cold reporting +25.5% NASA anomaly +0.28°C Monthly relationship r = 0.22 **Heat coverage separates from cold as the measured climate warms** monthly rate per 100,000 English World events · two-month half-life Heat reporting Cold reporting NASA anomaly **Heat reporting rises 99%; cold reporting rises 26%** pooled rate per 100,000 English World events **Heat** 2015-2019 81.0 2023-2025 160.9 **Cold and severe winter** 2015-2019 39.5 2023-2025 49.5 Heat and cold are mutually exclusive analytical cohorts inside World's EM-DAT Extreme temperature assignments, not additional EM-DAT categories. Rates use all English World events as the denominator. NASA GISTEMP is a global temperature measure, so the lines test co-movement in reporting rather than local weather attribution or direct measures of local hazard exposure across entities, assets and locations worldwide. Example usage 2 of 3 ### Example Usage: Climate-sensitive disaster coverage rises 44% after normalization for the growing World corpus World assigns 21,706 events to Drought, Extreme temperature or Flood in 2015-2019 and 89,211 in 2023-2025. Raw growth is not comparable because the World corpus also expanded. After dividing by all events, the basket rises from 901.4 to 1,296.7 assignments per 100,000. That is a 43.9% increase beyond corpus growth. [[ 1 ]](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) Floods damage property and infrastructure. Extreme temperatures affect labour, power demand and physical assets. Droughts constrain agriculture, water and hydroelectric generation. The chart shows when coverage of these financial exposures rises. Researchers can then join each hazard to locations, issuers, insurers and sovereign balance sheets. [[ 2 ]](https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11/) [[ 3 ]](https://doi.org/10.3386/w30445) 2015-2019 901.4 per 100k 2023-2025 1296.7 per 100k Normalized increase +43.9% Classified events 151,772 **Normalized climate-sensitive disaster coverage rises 44%** rate per 100,000 World events · two-month half-life **Extreme temperature contributes the largest normalized increase** pooled rate per 100,000 World events **Extreme temperature** 2015-2019 395.8 2023-2025 548.4 **Flood** 2015-2019 296.5 2023-2025 428.2 **Drought** 2015-2019 209.1 2023-2025 320.1 The basket contains World events assigned to EM-DAT Drought, Extreme temperature or Flood. Monthly counts are divided by all World events and smoothed with a two-month half-life. The chart measures classified reporting, not physical disaster incidence, climate attribution or investment performance. Example usage 3 of 3 ### Example Usage: Earthquakes emphasize economic impact while wildfires concentrate responsibility and conflict across reporting EM-DAT categories separate hazards that a generic disaster flag combines. In a fixed first-quarter 2024 English cohort, Economic Consequences frames 78.5% of earthquake events but only 21.0% of wildfire events. Responsibility leads wildfire coverage at 37.4%, while Conflict reaches 38.0% for drought, across four preselected hazard cohorts. [[ 1 ]](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) [[ 8 ]](https://doi.org/10.1111/j.1460-2466.2000.tb02843.x) Those differences change the research question. Earthquakes create abrupt asset-damage and reconstruction windows. Drought creates persistent water, food and power constraints. Wildfire reporting emphasizes liability as well as physical loss. Hazard-specific classification therefore determines which entities, exposures and event horizons belong in a risk panel before returns, losses, spreads or portfolio outcomes are tested. [[ 4 ]](https://www.ngfs.net/en/press-release/ngfs-publishes-latest-long-term-climate-macro-financial-scenarios-climate-risks-assessment) Earthquake economic frame 78.5% Wildfire responsibility 37.4% Drought conflict 38.0% Wildfire non-economic 79.0% **Each hazard produces a different narrative-risk mix** share of classified English events · Q1 2024 Responsibility Economic Consequences Conflict Human Interest Morality **Wildfire** 377 **Flood** 237 **Earthquake** 247 **Drought** 805 0% 100% · event count at right **Only 21% of wildfire coverage is economically framed, versus 79% for earthquakes** share outside Economic Consequences **Wildfire** 79.0 % **Flood** 45.6 % **Earthquake** 21.5 % **Drought** 51.9 % The fixed cohort contains English World events from January through March 2024 with both an EM-DAT disaster-type assignment and a Media Frames assignment. Bars show within-hazard frame shares. They describe how reporting packages risk; they do not estimate losses, public sentiment or causal market effects. Data and sources ## Download EM-DAT and References ### Downloads Download categories and counts as CSV, or the complete machine-readable release as JSON. [**CSV** Definitions and counts Download ↓](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/nodes.csv) [**JSON** Full machine-readable release Download ↓](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/statistics.json) Release details and citation files [Manifest](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/manifest.json) [Release README](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/README.md) [CC0 license and scope](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/LICENSE.md) [Citation file](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/CITATION.cff) [Changelog](https://nosible.com/data/world-v1.2/ontologies/em-dat/v1/CHANGELOG.md) Need the complete World event schema? [Open the World data dictionary.](https://nosible.com/data-dictionaries#world) ### References 1. [ 1 ][Centre for Research on the Epidemiology of Disasters. EM-DAT disaster classification system, 2023 release.](https://doc.emdat.be/docs/data-structure-and-content/disaster-classification-system/) 2. [ 2 ][Seneviratne, S. I. et al. (2021). Weather and climate extreme events in a changing climate. In IPCC AR6 Working Group I, Chapter 11.](https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11/) 3. [ 3 ][Acharya, V. V. et al. (2022). Is physical climate risk priced? Evidence from regional variation in exposure to heat stress. NBER Working Paper 30445.](https://doi.org/10.3386/w30445) 4. [ 4 ][Network for Greening the Financial System. (2024). NGFS long-term climate macro-financial scenarios, Phase V.](https://www.ngfs.net/en/press-release/ngfs-publishes-latest-long-term-climate-macro-financial-scenarios-climate-risks-assessment) 5. [ 5 ][Sautner, Z., van Lent, L., Vilkov, G., & Zhang, R. (2023). Firm-level climate change exposure. Journal of Finance, 78(3), 1449-1498.](https://doi.org/10.1111/jofi.13219) 6. [ 6 ][NASA Goddard Institute for Space Studies. GISS Surface Temperature Analysis version 4 (GISTEMP v4).](https://data.giss.nasa.gov/gistemp/index.html) 7. [ 7 ][NASA. (2025). Temperatures Rising: NASA Confirms 2024 Warmest Year on Record.](https://www.nasa.gov/news-release/temperatures-rising-nasa-confirms-2024-warmest-year-on-record/) 8. [ 8 ][Semetko, H. A., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news. Journal of Communication, 50(2), 93-109.](https://doi.org/10.1111/j.1460-2466.2000.tb02843.x) Continue exploring ## Complementary ontologies [231 categories World Geography Locate disaster exposure by continent, region and country before joining it to assets or entities. Explore ontology →](https://nosible.com/ontologies/geography) [5 categories Media Frames Measure whether each hazard is presented through cost, responsibility, conflict or human impact. Explore ontology →](https://nosible.com/ontologies/media-frames) > Explore 43 EM-DAT disaster categories and evidence linking extreme-heat reporting, normalized climate-sensitive attention and hazard-specific narratives. **URL:** https://nosible.com/ontologies/em-dat