Hurricane Isaias / Operational test

From regional warnings to local priorities.

As Isaias neared the northern Gulf Coast, the Resilitix team ran EmergenCITY AI as an operational test, briefing a diverse set of users, including federal, state and local emergency managers, on where water, wind and power loss could meet people: eight briefs in four days.

Operational test

Oct 7–10, 2026

Storm surgeInland floodingWind and outagesFive states

Prepared by the Resilitix team.
Briefed to a diverse set of users, including federal, state and local emergency managers.

01 / Documented work

<3 h

from the first question to a completed brief

Oct 7: first question 7:08 AM CDT; first brief completed 10:00 AM CDT

8 briefs

in four days: seven before landfall, one after

Oct 7 (two), Oct 8 (three), Oct 9 (two), Oct 10 (one)

12 ZIP codes

to look at first on landfall day, from warnings in five states

in seven places from Pascagoula to Panama City Beach; the post-landfall brief listed 10 ZIP codes in its first tier

Preparation times are based on recorded run and brief-completion timestamps. They are not measures of partner response time.

The operational challenge

A broad threat.
Limited time to assess it.

The model interprets. Analytical tools compute. The system shows its work. Human teams validate and decide.

The warnings cover five states.

They ran from Louisiana to Georgia. No team can check every coast, river and town.

The hazards overlap.

Storm surge and inland flooding get counted twice, or one gets missed.

The track keeps moving.

Each advisory shifted the track. A manual assessment can be out of date before it is done.

After landfall, data disagree.

Outage feeds, utility reports, shelter counts and food access do not line up.

02 / IN ACTION · OCT 9, THE PRE-LANDFALL BRIEF

Landfall day: from warnings in five states to 12 ZIP codes

01
5 states under warnings
hurricane, storm surge and tropical storm warnings, Louisiana to Georgia
02
432,547 residents
in forecast surge cells inside the Storm Surge Warning in Florida, Alabama and Mississippi
03
≥148,491 residents
in the cone's forecast surge or inland flood cells, each cell counted once
04
12 ZIP codes
to look at first, in seven places from Pascagoula to Panama City Beach

These are successive screening views with different geographic scopes; their population counts are not a single reduction funnel.

Completed at 6:54 PM CDT, about an hour and a half before landfall. Isaias came ashore near Destin at about 8:30 PM, inside the cone the runs used.

Map from the Oct 9 3:59 PM run: NHC peak storm surge forecast along the coast from Pascagoula to Panama City Beach, the cone the runs used drawn dashed, and five numbered first-look areas — Oct 9, surge forecast and cone (dashed): 1 Pascagoula, Moss Point, Gautier · 2 Pensacola · 3 Milton · 4 Navarre · 5 Panama City Beach. Forecast exposure, not confirmed impacts.

Explore another view from the brief

Map from the Oct 10 post-landfall brief: Florida Panhandle and southern Alabama counties shaded by the share of customers without power, with six numbered act-first places and blue rings where flood risk also applies — Oct 10: customers out by county (utility reports); numbers mark where power loss meets thin grocery access; rings, flood risk. Food access is screened, not confirmed. This archival brief excerpt includes preliminary outage values. Holmes County is shown as 99% out in the brief; the later review recorded a 95.5% peak. Those values have not been reconciled.

03 / Briefing cadence

Priorities that moved with the cone

Each brief re-ran the screens. On Oct 9, as the cone moved east, Mobile and Baldwin ZIP codes left the list and Navarre joined Pensacola and Milton; after landfall, Crestview and DeFuniak Springs joined the act-first list.

8 briefs, Oct 7–10, 2026. B1: Oct 7, Morning; B2: Oct 7, Afternoon; B3: Oct 8, Morning; B4: Oct 8, Morning update; B5: Oct 8, Afternoon; B6: Oct 9, Morning; B7: Oct 9, Pre-landfall; B8: Oct 10, Post-landfall. — Brief-completion sequence. Enlarge the timeline to inspect timestamps; the archive below gives the date and scope of each brief.

Browse all 8 brief covers Dates, editions and coverage

B1, Oct 7, Morning brief cover: Tropical storm, NHC Advisory 3 — B1 · Oct 7
Morning
Tropical storm, NHC Advisory 3

B2, Oct 7, Afternoon brief cover: Hurricane watches, Advisory 5 — B2 · Oct 7
Afternoon
Hurricane watches, Advisory 5

B3, Oct 8, Morning brief cover: Surge and inland flooding on one grid — B3 · Oct 8
Morning
Surge and inland flooding on one grid

B4, Oct 8, Morning update brief cover: Advisory 8 — B4 · Oct 8
Morning update
Advisory 8

B5, Oct 8, Afternoon brief cover: Advisory 9 — B5 · Oct 8
Afternoon
Advisory 9

B6, Oct 9, Morning brief cover: Category 3, Advisory 11 — B6 · Oct 9
Morning
Category 3, Advisory 11

B7, Oct 9, Pre-landfall brief cover: Advisory 13 — B7 · Oct 9
Pre-landfall
Advisory 13

B8, Oct 10, Post-landfall brief cover: Outages, food access, shelters — B8 · Oct 10
Post-landfall
Outages, food access, shelters

These are cover images from the briefing archive. Full briefs and supporting methods can be requested from the team.

04 / Operational value

What the work helped make visible.

The examples below describe the assessments and briefing outputs prepared during this test.

Prioritize limited attention

Know where to look first.

In this event: on landfall day, warnings in five states narrowed to 12 ZIP codes in seven places.

→ Focus the next round of local checks.

Recognize compound risk

See where hazards and lifelines interact.

In this event: 4,535 residents in cells where surge met inland flooding, each cell counted once.

→ See overlapping exposure without double-counting.

Iterate rapidly

Keep priorities current as the forecast moves.

In this event: eight briefs in four days, tropical storm to post-landfall, each checked against the latest advisory.

→ Priorities that keep pace with the storm.

Expose uncertainty and gaps

Separate what is known from what is not.

In this event: the post-landfall brief explicitly marked Holmes County food access as not yet screened while reporting extensive outages.

→ Make the next verification question visible.

05 / POST-EVENT REVIEW · IMPACTS REPORTED THROUGH OCT 11

What a post-event review found

The Resilitix team compared seven pre-landfall briefs with 528 reported impacts and 20 hotspots identified in its post-event dashboard, compiled through Oct 11. Earlier briefs had named 16 of those 20 hotspots in their own ZIP codes, a median of 41.5 hours before landfall. The final First and Next tiers contained 19 ZIP codes and 8 of the 20 hotspots, including the dashboard’s top two, Pensacola and Milton. The list covered 2,335 km²—an area equivalent to about 3% of the cone’s land across Alabama, Florida, Georgia and Mississippi. Some listed ZIP codes extended beyond the cone. The review identified gaps inland, east of the track, and in Georgia.

16 of 20

hotspots named in their own ZIP codes before landfall

of the 20 hotspots in the team’s post-event dashboard; 19 counting neighboring ZIP codes

41.5 h

median from that first naming to landfall

across the 16 own-ZIP matches; range 10.5–58.5 h from brief completion to landfall

4 of 4

highest-ranked counties on the first brief’s priority list

ranked by the dashboard’s impact index; Oct 7, 58.5 h before landfall; 8 of its top 10 were flagged

47 of 62

major reported impacts in areas flagged beforehand

Florida 31 of 33; Alabama 16 of 21; Georgia 0 of 8

Retrospective review of one event. These figures describe reported impacts within areas flagged beforehand; they do not establish predictive accuracy or operational decision outcomes. Non-impacted flagged areas and false-alarm burden were not assessed. Method and impact list available on request.

Bars for the seven pre-landfall briefs showing how many of the 20 hotspots had been named in their own ZIP code, cumulative: 3, 8, 11, 12, 13, 16, 16, with hours before landfall from 58.5 to 1.6 — Cumulative matches to the dashboard’s 20 hotspots: own ZIP code (dark), plus neighboring ZIP codes (light). Times show hours from brief completion to landfall.

Single-event retrospective finding. False-alarm burden and decision outcomes were not assessed.

Bars showing the land area of each brief's priority ZIP list, from 4,200 and 5,421 square kilometers on Oct 7 down to 2,335 square kilometers on Oct 9, labeled with the number of ZIP codes and the hotspots inside the list — Land area of each First and Next priority tier, with the dashboard hotspots inside. For scale, the Oct 9 cone covered 78,888 km² of land in Alabama, Florida, Georgia and Mississippi. Listed ZIP codes were not all contained within the cone.

Single-event retrospective finding. False-alarm burden and decision outcomes were not assessed.

Map of the Florida Panhandle and southern Alabama: the final pre-landfall ZIP list and other named ZIP codes shaded, the cone dashed, reported impacts as dots colored by how their area was flagged beforehand, and the hotspots in the Panhandle and southern Alabama as numbered hexagons (hotspots 7, 10, 16 and 19 lie beyond the map) — Shaded: the final priority list (Oct 9, 6:54 PM CDT) and other ZIP codes named before landfall. Dashed: the cone used in the Oct 9 runs. Dots: reported impacts by prior flag. Hexagons: dashboard hotspots by rank; hotspots 7, 10, 16 and 19 lie beyond this map. Reported impacts, not screening output.

How to interpret the comparison

The map overlays reported impacts with the areas named in the earlier screening. Spatial overlap describes what was covered in this event; it does not establish predictive accuracy.

The recorded briefing outputs and the reported-impact list are different evidence sources.

Review method and definitions

Compared: seven briefs completed from Oct 7, 10:00 AM to Oct 9, 6:54 PM CDT, plus an Oct 7 lifeline edition, against 1,306 records cleaned to 528 impacts. Priority: county P1–P2, ZIP First and Next tiers, the Oct 7 priority ZIP lists and lifeline L1–L2. Hotspots are the dashboard’s 20 places ranked by severity-weighted reports. An own-ZIP match requires an earlier brief to name a hotspot’s ZIP code; neighboring ZIP matches are reported separately. The 41.5 h lead is from recorded brief completion to landfall near Destin at about 8:30 PM CDT on Oct 9, not a measure of time saved or partner action.

The retrospective measures use different denominators and timing references across these events. They should not be combined into a single accuracy score.

Request the method and impact list ↗

Carried forward

What to strengthen in the next event

Look for wind damage beyond the cone, in every state the storm will affect, using NHC wind-speed probabilities.
Follow the remnants after landfall, carrying the screens along their forecast path with forecast precipitation and severe-weather outlooks.
Screen for flash floods on that path, from forecast cumulative rainfall and flash-flood risk.
Give wind and power their own tier beside the water tiers, so places with high outage potential and little water can reach the list.

Bring your operational question

Apply the workflow
to your next question.

Explore an operational test for your jurisdiction or network with the Resilitix team.

Discuss an operational test ↗

Opens your email app · info@resilitix.ai

NEXT CASE STUDY

Southwest flash floods

Where to look first.

Read the case study →

EmergenCITY AI is designed for screening-grade resilience intelligence, planning, preparedness, monitoring, and decision support. It does not replace official forecasts, emergency command decisions, engineering studies, regulatory determinations, or local expert judgment.

The system runs the first pass. Your team makes the call. Resilience intelligence on demand.

emergencity.ai · info@resilitix.ai · EmergenCITY AI · Geospatial Resilience Intelligence

The operational test at a glance

A broad threat.Limited time to assess it.

The warnings cover five states.

The hazards overlap.

The track keeps moving.

After landfall, data disagree.

Landfall day: from warnings in five states to 12 ZIP codes

5 states under warnings

432,547 residents

≥148,491 residents

12 ZIP codes

Priorities that moved with the cone

B1 · Oct 7

B2 · Oct 7

B3 · Oct 8

B4 · Oct 8

B5 · Oct 8

B6 · Oct 9

B7 · Oct 9

B8 · Oct 10