America’s Flood Forecasts Reach Nearly Everyone. Can Cities Turn Warning Into Action?
Expanded NOAA flood forecasts and AI research exposing gaps in federal hazard maps offer cities new intelligence, but protecting residents and infrastructure depends on local coordination, reliable data and timely decisions.
By The Spatial Signal Editorial Board
· Week of November 2, 2026 · 8 min read
Listen · shortened audio edition

Image created by AI
CLIMATE & RESILIENCE | GEOSPATIAL INTELLIGENCE
# America's Flood Maps Can Now Look Five Days Ahead. Are Cities Ready?
DEK: A major expansion of NOAA's flood forecasting system means nearly every American community can now see where river flooding may spread before it happens. New AI research suggests millions of people are missing from existing federal flood-hazard maps. Together, the developments could transform how cities prepare for disasters and protect infrastructure.
For generations, communities facing rising floodwaters have relied on river gauges, weather warnings and local experience to decide when to close roads, protect infrastructure or evacuate neighborhoods. Increasingly, they have another tool: maps that show where the water itself may go, days before it arrives.
Yet the value of those advances ultimately depends on what happens after a forecast appears on a screen.
A major expansion of the National Oceanic and Atmospheric Administration's flood forecasting system, announced September 24, has brought experimental flood-inundation mapping to nearly 100% of the U.S. population.
The system now covers approximately 3.3 million miles of rivers and streams, transforming predictions of rising river levels into geographic estimates of which surrounding areas may flood.
For much of the country, the forecasts extend five days into the future. Instead of knowing only that a river is expected to crest at a certain height, emergency managers can examine the neighborhoods, roads and infrastructure that may lie in its path.
The expansion marks a dramatic change in the availability of flood intelligence. When NOAA introduced the service in 2023, it covered approximately 10% of Americans. Coverage grew to 60% in 2025 before reaching near-national availability this September.
But the technology's growing reach raises a larger question: Will cities actually use this new intelligence to make better decisions before disasters strike?
### A second discovery exposes millions of overlooked risks
The NOAA announcement comes amid growing evidence that America's existing flood maps may provide an incomplete picture of long-term exposure.
In June, researchers published findings in the journal Nature Communications suggesting that approximately 11 million additional Americans and 4.1 million buildings may be exposed to flood hazards not represented in the federal mapping baseline they examined.
The researchers, Abraham Noah Wu, Ye Zhang and Rudi Stouffs, used a generative artificial intelligence model to produce a continuous, high-resolution flood-hazard map of the contiguous United States.
Their system examined relationships between terrain elevation and existing FEMA flood-hazard boundaries, then estimated potentially flood-prone areas in places where federal mapping was incomplete or unavailable.
Compared with the 2023 federal baseline, the model suggested that the number of people potentially exposed to the modeled flood hazard could be 69% higher than previously represented. The estimated number of exposed buildings increased by 81%.
The findings do not mean that 11 million people have suddenly entered officially designated flood zones. They are estimates generated by an AI model, not new FEMA regulatory determinations.
Nevertheless, the implications are significant for communities making decisions about housing development, roads, bridges, public facilities and infrastructure investment.
The research points to a fundamental vulnerability: areas outside existing federal flood-hazard boundaries are not necessarily free from flooding, particularly where mapping coverage has been incomplete.
The study's findings also come with limitations. The model was less reliable when applied to unfamiliar terrain, and it did not comprehensively capture rainfall-driven surface flooding, such as water accumulating in urban streets when drainage systems are overwhelmed.
Still, it offers a potentially valuable way to identify places that deserve closer investigation, particularly for smaller communities without the budgets to commission extensive flood studies.
Read the original Nature Communications study
### The difference between knowing a river will rise and knowing where the water will go
NOAA's expansion addresses a different, more immediate problem.
Its Flood Inundation Mapping service combines hydrological forecasts with elevation data to estimate how rising rivers may spread across surrounding terrain.
The service includes frequently updated analyses of recent conditions and forecasts showing the potential maximum extent of flooding over the next five days across much of the continental United States.
The new coverage also extends to Hawaii, Puerto Rico, the U.S. Virgin Islands and parts of Alaska, although forecast horizons and geographic coverage differ.
These are experimental predictions rather than guarantees of what will happen at any individual property. They also do not replace FEMA's legally effective flood maps, which serve separate regulatory and insurance-related purposes.
Even with those limitations, the geographic information can be valuable.
A river forecast that once appeared primarily as a predicted water level can now be visualized as a potentially flooded landscape. For emergency managers, that could mean identifying a threatened roadway, vulnerable neighborhood or critical public facility before the water arrives.
And in at least two American communities, that capability has already been put to use.
### In Arkansas, flood maps helped provide days of warning
In April 2025, the small city of Hardy, Arkansas, faced the prospect of major flooding along the Spring River.
The National Weather Service's Little Rock office shared flood-inundation information with local officials as the threat developed.
Hardy Mayor Ethan Barnes later told Arkansas PBS that the community had traditionally experienced only 12 to 24 hours of useful warning before severe flooding. This time, officials had approximately four to five days to prepare.
Local authorities used that time to distribute notices, warn residents and coordinate evacuations.
NOAA subsequently reported that Hardy achieved a nearly 100% evacuation success rate. Despite extensive flooding, no deaths were reported in the city.
The outcome reflects the combined efforts of forecasters, emergency responders and local officials rather than the maps alone. Nevertheless, it illustrates what can happen when advance geographic intelligence becomes part of an organized emergency response.
Watch the Arkansas PBS interview with Hardy's mayor
### In Texas, the forecasts supported road closures and evacuations
A second example emerged during the catastrophic Central Texas floods of July 2025.
As severe flooding developed around Georgetown, the National Weather Service's Austin/San Antonio office provided flood-inundation information to emergency management officials.
According to NOAA, those forecasts supported decisions involving road closures and evacuations.
Williamson County's contemporary emergency records confirm that officials activated emergency operations, evacuated vulnerable communities along the South Fork of the San Gabriel River and closed roads as conditions deteriorated.
The wider Central Texas disaster caused devastating loss of life and extensive property damage, demonstrating that better forecasting cannot eliminate the dangers of extreme weather.
Public records do not establish precisely how many closure or evacuation decisions resulted directly from NOAA's flood maps. But the case demonstrates how predictive geographic information can contribute to the coordination of emergency response.
Read Williamson County's official emergency declaration
### The next breakthrough could happen inside city GIS departments
The potential applications extend well beyond evacuations.
On September 25, one day after NOAA's expansion announcement, Esri revealed that the agency's flood-inundation forecasts were available through ArcGIS Living Atlas.
That brings the forecast information into a widely used geospatial environment where local governments already manage roads, bridges, buildings, water infrastructure, drainage networks and public facilities.
The implications could be substantial.
A city preparing for a major flood could overlay predicted inundation areas with its road network to identify potentially affected routes. Public works teams could review vulnerable bridges and drainage infrastructure. Water utilities could examine whether pumping stations or treatment facilities might require protective measures.
Emergency managers could assess which facilities and neighborhoods deserve priority attention, while planners could use separate long-term hazard estimates to identify locations where additional flood studies or infrastructure investment may be justified.
Esri has demonstrated how NOAA's forecast layers can be combined with building and bridge inventories to identify potentially exposed infrastructure.
But an important distinction remains: a bridge intersecting a predicted flood area is not necessarily impassable, and a building within a forecast polygon is not necessarily going to flood.
Local elevations, drainage conditions, field observations and engineering judgment remain essential.
See Esri's September 2026 flood-mapping announcement
### The technology is advancing faster than the systems built to use it
For many cities, the challenge may no longer be obtaining flood information. It is connecting that information to the people and systems responsible for acting on it.
Municipal infrastructure data is frequently distributed across departments. GIS teams maintain geographic records, public works manages roads and drainage, utilities operate their own networks, and emergency management coordinates disaster response.
Turning a predicted flood boundary into a timely operational decision requires those departments to work from reliable information and established procedures.
Forecast uncertainty adds another complication. NOAA cautions that its experimental inundation products should not serve as the sole basis for critical decisions without consultation with the appropriate National Weather Service forecasting office.
And national models cannot capture every local condition, from blocked culverts and changing drainage systems to urban flash flooding and coastal storm surge.
There is also no comprehensive public measure of how many American municipalities have integrated the expanded NOAA forecasts into routine infrastructure management, or what it costs to do so effectively.
That uncertainty may be the most revealing part of the story.
The federal government is making increasingly sophisticated flood intelligence available. Academic researchers are using AI to expose potential gaps in existing hazard maps. Commercial GIS platforms are making the information easier to access alongside infrastructure data.
Yet the value of those advances ultimately depends on what happens after a forecast appears on a screen.
For a city, the difference could be an inspection ordered earlier, a road closed before it becomes dangerous, a pumping station protected in time or an evacuation completed before residents become trapped.
America is rapidly gaining a more complete picture of where floodwater may go—both in the days ahead and over the longer term.
The flood map revolution will be judged not by how accurately it colors the landscape blue, but by whether communities can act on what those colors tell them before the water arrives.
### Sources
NOAA — September 24, 2026 NOAA Flood Mapping Tool Now Covers Nearly 100% of the U.S.
NOAA — National Water Prediction Service Live hydrological forecasts and flood-inundation mapping
Nature Communications — June 15, 2026 Deep Learning Completes US Flood Hazard Maps Revealing Millions Exposed to Previously Unrecognized Risk
Esri — September 25, 2026 NOAA Flood Inundation Prediction Maps Now in ArcGIS Living Atlas
Arkansas PBS — April 23, 2025 Flood Recovery in Hardy
Williamson County, Texas — July 5, 2025 Official Flood Emergency Declaration
FEMA — Official Flood Map Service Center Access effective federal flood maps