Earthquake loss database

Example: A tabular database documenting economic losses from a historical earthquake event, with values in USD and breakdowns by loss type and geographic area.

Step-by-step guidance

1. Dataset-level metadata

Select the following values when describing your dataset:

  • Risk data type: loss

  • Title: “Economic losses from [earthquake event name/date]”

  • Description: Brief description of the earthquake event, affected regions, and data collection methodology

  • Publisher: Organization that compiled the loss data (e.g., government agency, World Bank, insurance association)

  • License: Appropriate license

2. Resources

Add resources for your loss database files:

  • Format: csv, xlsx, or geopackage

  • Spatial resolution: Administrative level, building-level, or grid-based

  • Coordinate reference system: EPSG:4326 (if spatial) or not applicable for tabular data

3. Loss metadata

Under the Loss section:

Loss category

  • Category: economic

Hazard event reference

Link to the specific hazard event:

  • Hazard type: earthquake

  • Event name: Official name or designation of the earthquake

  • Event date: Date of occurrence (ISO 8601 format: YYYY-MM-DD)

  • Event magnitude: Moment magnitude (Mw) or other scale

  • Event location: Epicenter coordinates or affected region

Loss metrics

Define what loss types are measured:

Metric 1 - Direct structural damage:

  • Loss type: direct

  • Dimension: structure

  • Unit: Currency code (e.g., USD)

  • Reference year: Year for currency valuation

Metric 2 - Direct contents damage (optional):

  • Loss type: direct

  • Dimension: content

  • Unit: Currency code

Metric 3 - Business interruption (optional):

  • Loss type: indirect

  • Dimension: business_interruption

  • Unit: Currency code

Metric 4 - Casualties (optional):

  • Loss type: human

  • Dimension: fatalities or injuries

  • Unit: people or count

Temporal information

  • Assessment date: When losses were assessed or reported

  • Time period: If losses accumulated over time (e.g., business interruption duration)

4. Spatial coverage

Define the geographic extent of losses:

  • Scale: sub-national or national

  • Countries: Select applicable ISO 3166-1 alpha-3 country codes

  • Administrative regions: Provinces/districts affected

  • Bounding box: Coordinates of affected area

Example data structure

Your loss database should include:

For spatially-aggregated losses:

  • Geographic unit ID (e.g., admin code, district ID)

  • Geographic unit name

  • Geometry or coordinates (if spatial)

  • Direct structural losses

  • Direct contents losses

  • Indirect/business interruption losses

  • Number of buildings damaged by damage state

  • Population affected

  • Fatalities and injuries

Example CSV structure:

District_Code,District_Name,Direct_Loss_Structure_USD,Direct_Loss_Content_USD,Buildings_Damaged,Population_Affected,Fatalities
YEM001,Sana'a,125000000,45000000,2500,150000,45
YEM002,Taiz,78000000,28000000,1800,95000,28
YEM003,Aden,52000000,18000000,1200,60000,15

For building-level losses:

  • Building ID

  • Location (coordinates)

  • Building type/taxonomy

  • Damage state

  • Structural loss value

  • Contents loss value

  • Occupants affected

Key considerations

  • Specify the currency and reference year for all monetary values

    • Include exchange rates if values were originally in local currency

    • Note if values are adjusted for inflation

  • Document the source of loss estimates

    • Government assessments

    • Insurance claims data

    • Field surveys (PDNA, DALA)

    • Remote sensing damage assessment

    • Engineering estimates

  • Distinguish between different loss types:

    • Direct losses: Physical damage to assets (structure, contents, inventory)

    • Indirect losses: Business interruption, production losses, economic disruption

    • Intangible losses: Loss of life, cultural heritage, environmental damage

  • Include information about:

    • Assessment methodology and assumptions

    • Coverage (insured vs. uninsured losses)

    • Disaggregation (by sector, building type, damage state)

    • Uncertainty and confidence levels

  • Reference the specific earthquake event clearly:

    • Official designation or name

    • Date and time

    • Magnitude and depth

    • Link to earthquake catalog or USGS event ID

  • Note any exclusions or limitations:

    • Partial coverage areas

    • Excluded sectors or loss types

    • Time cutoffs for loss accumulation

Linking to hazard and exposure data

When possible, provide linkages to:

  • Hazard: Reference the earthquake event in a hazard catalog

  • Exposure: Link to building inventory or exposure dataset for affected area

  • Vulnerability: Reference damage functions used if losses are modeled rather than observed