Skip to main content

GeraSure / County Hazard Risk / Methodology

Gera County Hazard Score (GCHS) — Methodology

Full reproducible formula for the GCHS/100 index. Every number traces to FEMA National Risk Index public-domain data (November 2023).

What is the GCHS?

The Gera County Hazard Score (GCHS) is a 0–100 index that summarises the natural hazard risk of a US county into a single number. Higher scores mean higher risk. It is computed by Gera from three publicly available FEMA National Risk Index (NRI) percentile-rank scores:

  • EAL_SCORE (Expected Annual Loss) — percentile rank of a county's expected annual monetary loss from 18 natural hazard types. Higher rank = more loss.
  • SOVI_SCORE (Social Vulnerability Index) — percentile rank of social and demographic factors that affect a community's capacity to respond to and recover from disasters. Higher rank = more vulnerable.
  • RESL_SCORE (Community Resilience) — percentile rank of community infrastructure and economic capacity. Higher rank = more resilient (beneficial, so Gera inverts it).

The formula

GCHS = 0.50 × EAL_SCORE
      + 0.30 × SOVI_SCORE
      + 0.20 × (100 − RESL_SCORE)

Why these weights? Expected Annual Loss (EAL) carries the greatest weight (50%) because it is the direct measure of physical risk and economic exposure. Social Vulnerability (30%) amplifies the human impact of hazard events — high-vulnerability communities face disproportionate losses. The inverse of Resilience (20%) captures whether a community has the institutional and economic capacity to recover. These weights follow established resilience-science literature weighting EAL as primary and vulnerability/resilience as modifiers.

Range guarantee: Because all three FEMA inputs are percentile ranks in the range 0–100, and the resilience inversion (100 − RESL) maps RESL to the same 0–100 range, the weighted sum is also in 0–100 by construction. No additional normalisation step is required.

Worked example

Los Angeles County, CA (FIPS 06037):

InputValueWeightContribution
EAL_SCORE96.550%48.3
SOVI_SCORE76.630%23.0
100 − RESL_SCORE100 − 49.9 = 50.120%10.0
GCHS—100%91.3 ≈ 91.3

Actual published GCHS for LA County may differ by up to ±0.1 due to rounding in the intermediate FEMA scores. All FEMA scores are sourced from the NRI county CSV.

Population cap and state hubs

The FEMA NRI includes 3,143 counties with complete EAL, SOVI and RESL data. Gera publishes 480 individual county pages (the 480 most-populous counties, covering the majority of the US population). The remaining approx 2,663 counties are summarised as 49 state-level hub pages showing average, minimum and maximum GCHS and total population covered per state. This cap is documented in the generated data module comment block.

18 natural hazard types in EAL

FEMA's Expected Annual Loss score aggregates expected losses across all 18 natural hazard types included in the NRI:

  • Avalanche
  • Coastal flooding
  • Cold wave
  • Drought
  • Earthquake
  • Hail
  • Heat wave
  • Hurricane
  • Ice storm
  • Landslide
  • Lightning
  • Riverine flooding
  • Strong wind
  • Tornado
  • Tsunami
  • Volcanic activity
  • Wildfire
  • Winter weather

Data source and licence

Source: FEMA National Risk Index (NRI) — Harvard Dataverse

Publisher: Federal Emergency Management Agency (FEMA), US Government. Open data in the public domain — no copyright applies to US Government works.

Reference period: November 2023 (NRI release 2024).

Gera attribution: The GCHS formula, index name and this methodology page are original Gera work. The underlying NRI scores are FEMA public-domain data. Any citation of the GCHS should credit both Gera (for the formula and index) and FEMA (for the source scores).

Contains public sector information published by Federal Emergency Management Agency (FEMA) and licensed under the US Government open data, public domain. Source: FEMA National Risk Index (NRI) — Harvard Dataverse (November 2023, published 2024).

Reproducing the GCHS

  1. Download NRI_Table_Counties.tab from doi:10.7910/DVN/JSQ8KZ.
  2. For each county row, read EAL_SCORE, SOVI_SCORE and RESL_SCORE. Skip rows with missing values or zero population.
  3. Compute: GCHS = round(0.50 × EAL_SCORE + 0.30 × SOVI_SCORE + 0.20 × (100 − RESL_SCORE), 1).
  4. Sort by POPULATION descending. Top 480 → leaf pages. Remainder → state aggregate hubs.
  5. Verify a sample by picking any county in the published pages and confirming the GCHS matches the formula applied to the downloaded NRI CSV.
Browse county hazard risk →Explore insurance options