The federal reference-data layer behind FEMA Public Assistance and CDBG-DR applications. A partnership briefing for HaloRRS.
Oahe Data · briefing prepared 20 April 2026
158
Killed in Joplin, 22 May 2011
~20 min
NWS warning lead time
$2.8B
Insured loss (costliest US tornado then)
7,500
Homes destroyed or damaged
75 mi
Furthest recovered medical record
HaloRRS was founded by people who lived this.
The evidence package that underwrote Joplin's FEMA Public Assistance reimbursement and CDBG-DR
rebuild was assembled, by hand, under a 180-day clock, by a school district that had just lost
its high school on graduation weekend.
Sources: NIST NCSTAR 3 (2014); NWS Central Region Service Assessment (2011); FEMA DR-1980-MO.
Act 1
Joplin in its climatological context, 1950–2025
The tornado that seeded HaloRRS did not arrive in a calm climate. Missouri's West Central climate
division (the division Joplin sits in) shows May 2011 as one of the wettest, most unstable
late-spring months in the 75-year record. The single most destructive weather event in modern
U.S. history did not come from a quiet atmosphere. It came from one that had been running hot
and wet for weeks.
Monthly precipitation (in)Monthly mean temperature (°F)22 May 2011 · Joplin EF-5
NOAA nClimDiv · Missouri West Central Division 2304 · monthly values 1950–2025
Why this is in the brief. Every FEMA Public Assistance Project Worksheet asks
the applicant to establish the declared event against a weather baseline. Every CDBG-DR application
asks for longitudinal community context. This is the reference layer Oahe can hand you ready-built
for any U.S. climate division: same primary NOAA source, same method, next district you walk into.
Act 2
The reference layer behind every Project Worksheet you write
Four counties. Four federally declared disasters HaloRRS has worked. The indicators below are
the ones a FEMA PA demographic-impact narrative and a CDBG-DR unmet-needs assessment are built
from. All four sit in Oahe's data layer today, for every county in the United States, already
processed through the Oahe refinery to the FP-STAN schema we use across every Oahe product.
Read Cross County, Arkansas. Wynne sits inside a county with the highest disability
rate, the highest food insecurity, and the lowest median household income of the four
engagement geographies shown. On paper, this is the county least equipped to self-assemble a
60-to-180-day recovery narrative. It is also the county where the reference-data layer matters
most, and where (if pre-built) it saves the most time in the window that counts.
Act 3a
Two maps, one partnership
Oahe's work on federal disaster payments renders as two 3D choropleths.
Above: the LIA2025 NFIP Named Storms map. 57 named storms 1979-2024,
county-level NFIP payments to private homeowners. Below: the companion
FEMA Public Assistance map. 86 Major Disaster declarations 1979-2024,
county-level federal share obligated to public entities (school districts,
counties, hospitals). Together they cover the private and public sides of
the same federal disaster-payment surface.
VERITAS MAPS
Truth Matters
NFIP Named Storms
County-level NFIP insurance payments per named storm event.
Source: FEMA NFIP · LIA2025 Coverage: 1979–2024
NFIP PAYMENTS
Decile 10 (Highest)
Decile 9
Decile 8
Decile 7
Decile 6
Decile 5
Decile 4
Decile 3
Decile 2
Decile 1 (Lowest)
Per-storm quantiles scaled to each event
Loading…
NFIP Insurance Payments by Named Storm
Storm
Extrusion10×
Veritas Maps · Truth Matters
Two palettes on purpose. The NFIP map uses a ColorBrewer
RdYlBu diverging palette (dark blue to deep red) over 57 named storms. The
FEMA PA map below uses an Oahe teal-to-gold-to-red diverging palette over 86
Major Disaster declarations. The different treatments reflect two distinct
Oahe products built on a shared FP-STAN-shaped data layer, not one forced
design. Consistency is in the data, not the skin.
Act 3b
FEMA Public Assistance, obligated
Oahe's LIA2025 NFIP Named Storms map shows what the National Flood Insurance Program
paid private homeowners after each storm. The choropleth below is the complementary layer:
what FEMA Public Assistance obligated to public entities (school districts, counties,
hospitals, tribal governments) after each federally declared disaster, 1979–2024. Four of the
disasters are highlighted because HaloRRS worked them. The gold ticks on the slider are those four.
Oahe DataEvidence Layer
FEMA Public Assistance
County-level federal share obligated, per declared disaster. Source: OpenFEMA v2 · PA Funded Projects Range: 1979–2024
PA Obligation
Decile 10 (Highest)
Decile 9
Decile 8
Decile 7
Decile 6
Decile 5
Decile 4
Decile 3
Decile 2
Decile 1 (Lowest)
Per-disaster quantiles scaled to each event
Loading…
FEMA Public Assistance · Federal Share Obligated
Disaster
Height10×
What these two maps are saying. Together, Oahe's two disaster-payment maps are the
full federal picture: NFIP private claims on one, FEMA PA public obligations on
the other. For any district HaloRRS takes on in 2026, both layers are already one query away, on
the same FP-STAN-shaped reference data, the same deck.gl animated 3D grammar, the same Oahe
refinery behind them. Two maps, one data layer, one partner.
The frame
Oahe already carries both halves of federal disaster payments. NFIP on the private side. FEMA
Public Assistance on the public side. For the schools, counties, and hospitals at the heart of
every HaloRRS engagement. Together that's the full federal payout surface for any hazard type,
already processed through the Oahe refinery and ready at the start of the 60-to-180-day
recovery clock, not at the end of it.
Methodology & data provenance
Act 1 climate strip: NOAA nClimDiv v1.0 division-level monthly series, Missouri West Central Division 2304, elements avg_temp / max_temp / min_temp / precip, 1950–2025. Loaded from data/noaa-nclimdiv/noaa_nclimdiv_division.parquet.
Act 2 vulnerability matrix: CDC PLACES 2025 (county-level health outcomes, 229K rows); AHRQ CLH county-level social determinants (40.7M rows). Both bundles processed through the Oahe refinery to FP-STAN v1.0 schema. Measure selection emphasizes indicators used in FEMA PA demographic-impact narratives and CDBG-DR unmet-needs assessments. Population from Census CoreGeo.
Act 3a NFIP choropleth: FEMA NFIP policy-claim data, county-level cumulative insurance payments per named storm event, 1979-2024. 57 events curated to named hurricanes, tropical storms, and major flood declarations with material NFIP payout. Per-storm quantile deciles. Rendered with deck.gl 8.9.36 GeoJsonLayer + MapLibre 3.6.2 on CARTO dark-matter basemap.
Act 3b PA choropleth: OpenFEMA API v2 endpoints DisasterDeclarationsSummaries and PublicAssistanceFundedProjectsDetails, pulled 20 April 2026. Curated to the largest Major Disaster declarations (DR-prefix) by declared-county count 1979–2024, plus all four HaloRRS-engaged disasters (DR-1980, DR-4332, DR-4407, DR-4698). County-level rollup sums federalShareObligated. Per-disaster deciles computed over positive-obligation counties only. Rendered with deck.gl 8.9.36 GeoJsonLayer extruded against TIGER county polygons. Same grammar as Oahe's LIA2025 NFIP map, so the two maps pair cleanly as halves of the same disaster-payment picture.