Data sources
and methodology.
Climate Atlas preserves FEMA county ratings, expected annual loss, exposure and resilience fields in a traceable sample. This methodology explains the geographic unit, source release, missing statuses and extraction checks. Use it to interpret the workbook, retain evidence beside your findings and identify which questions need separate location research.
What does the sample contain?
The sample extracts five county records from the supplied FEMA National Risk Index (NRI) workbook. County codes (FIPS) use five-digit text; rating labels stay unchanged. Fields cover seven hazard ratings, fifteen modeled loss figures, three exposure measures and two resilience scores. Each metric retains its unit and observation window.
Source portal: FEMA National Risk Index ↗. Release: December 2025. Extraction: September 6, 2026. The extraction date is not each observation’s date.
What can county data show?
County ratings cannot establish an individual building’s risk. Location checks require verified coordinates, relevant map dates and confirmation of coverage.
Which layers are still under review?
Flood-zone checks, NOAA event history and further wildfire context remain planned layers. The sample includes no verified property findings from these layers.
NOAA storm event records ↗ include county and forecast-zone reporting. Compare records within the same reporting unit. Further wildfire context remains under review.
How are missing values handled?
Not Applicable, No Rating and Insufficient Data remain visible. Failed location matches and missing maps require review. We do not infer heat-day counts from risk labels or insurance premiums from county scores.
What does this research support?
Research screening and source review. It is not a certified flood determination, building inspection, insurance quote or future damage prediction.
Inspect the source sample ↗How are the distribution and completeness charts constructed?
Profiles, the comparison table and completeness tiles use the five sample rows and reviewed metrics. Distribution charts count all 3,232 workbook records, including missing and non-rating statuses. Bars show county counts, with equal weight per county. Profile positions show categories, not numerical risk scores.
Download the chart inputs and counts. These are independent charts, without government endorsement. The FEMA National Risk Index information provides the dataset license reference. The FEMA source data is public-domain US government data. Our compiled files are for internal use unless redistribution is agreed in writing.
Which checks support the published values?
- Retain the original workbook, sheet name and field names.
- Normalize FIPS to five-digit text and check that the five selected IDs are unique.
- Independently read the source and reconcile each CSV, JSON and XLSX sample value.
- Count labels across all 3,232 source rows. Check chart totals, including non-rating statuses.
These checks verify source values and their presentation, not FEMA’s model or a property’s exposure. The rating guide explains how the labels are used.
How are freshness and corrections handled?
This fixed development snapshot has no live feed or promised refresh schedule. Before replacing an input, check its release, geography, definitions, category counts and sample rows. Record changes and rebuild downloads together.
For a suspected error, keep the county FIPS, field name, source filename and downloaded value. Check the source row before changing anything. Keep unexplained differences in the review queue. Email suspected errors to [email protected].
Who is responsible for the research?
Climate Atlas extracts and compares the data and explains the workflow. FEMA publishes the underlying NRI material. Climate Atlas operates independently of government agencies. No independent expert review or certification is claimed.
The portfolio checklist shows our proposed way to retain a source, date, geographic unit and follow-up for every research result.
What is the difference between Not Applicable and No Rating?
They are distinct FEMA source statuses, not interchangeable risk categories. Not Applicable identifies a field marked as not applicable in the source. No Rating identifies a field without an assigned rating. Neither means Very Low or a measured absence of loss. Preserve the supplied label and consult FEMA field documentation before deciding why it appears. Keep Insufficient Data and blank fields separate too.
Which decisions need evidence beyond this sample?
This is county screening evidence. It is not scenario analysis, a climate projection or a property-level assessment. It does not establish compliance with an ESG reporting framework. Future pathways, site conditions and asset vulnerability require additional evidence and a defined assessment method.
What else should I know about the methodology?
Are the charts live or forecast data?
They show a fixed extraction of supplied county labels and counts. The charts contain no live event feed or future property scenario.
Why are county counts and sample completeness different?
The distribution counts all source counties for one hazard. Completeness counts 35 fields across 5 sample counties and 7 hazards. Neither measures the number of covered properties. Inspect the sample rows alongside the charts.
Are flood maps and NOAA events included in the export?
No. Those layers are still being validated. The current export contains the listed FEMA county rating fields. The flood guide points to the separate official map workflow.
When was this methodology last reviewed?
Last reviewed: October 23, 2026.
- October 23, 2026: clarified missing statuses, internal reuse, contact and screening limits.
- September 6, 2026: sample extraction recorded.
How to cite: Climate Atlas, Data sources and methodology, reviewed October 23, 2026, https://climateatlas.us/methodology/.
For proposed bulk fields and delivery scope, inspect the data-access field registry.