FEMA's Flood Maps Miss 200,000+ High-Risk Harris County Properties. Here's What the Texas A&M Study Found.
FEMA's Flood Maps Miss 200,000+ High-Risk Harris County Properties. Here's What the Texas A&M Study Found.
Ask three different sources how many Harris County properties face a serious flood risk and you get three different answers. FEMA's current maps say about 184,800. FEMA's proposed new maps say 345,400. And a Texas A&M research team says 389,500 - more than double what the official maps on the books today show.
That last number comes from the Houston Chronicle's July 30 investigation into a flood-risk model built by Texas A&M's Institute for a Disaster Resilient Texas. It's worth your attention, and it's worth understanding exactly what the gap between these numbers means before you trust any of them.
Three Numbers, Same County
Depending on which map you trust, between 184,800 and 389,500 Harris County properties face a 100-year-level flood risk. Here's where each number comes from:
| Source | Properties at 100-year flood risk | What it's based on |
|---|---|---|
| FEMA's current maps (in effect since 2007) | 184,800 | Engineering models built on 2001-era rainfall statistics |
| FEMA's proposed maps (released February 2026) | 345,400 | MAAPnext: Atlas 14 rainfall, modern terrain and ponding models |
| Texas A&M IDRT model (2026) | 389,500 | Machine learning trained on a decade of real flood insurance claims |
The gap between the first two rows is the county effectively admitting its maps were stale - the new maps nearly double the count on their own. The gap between the second and third is a genuine argument about what “risk” even means. Both matter if you own or are buying a home here.
Why the Official Maps Undercount
FEMA's flood maps were built to set insurance requirements and guide development rules - not to tell you whether your specific house will flood. Harris County's current maps were last comprehensively updated in 2007, on rainfall statistics from 2001. Harvey, the Memorial Day and Tax Day floods, and Beryl all happened on top of maps that treated storms like them as near-impossible (we covered the full history in our Harvey anniversary post).
Even FEMA doesn't claim the maps are complete. Asked about the A&M findings, an agency spokesperson told the Chronicle the maps do not capture every place that could flood:
"While maps help communities manage risk, they do not guarantee that flooding will not happen outside of these boundaries." - FEMA spokesperson, to the Houston Chronicle
The proposed MAAPnext maps fix a lot of this - they use NOAA's Atlas 14 rainfall benchmarks and, for the first time, model how water ponds in streets and neighborhoods instead of only how bayous rise. But local officials told the Chronicle they could still take years to finalize. Until then, the 2007 maps are the law.
How the Texas A&M Model Works
Instead of simulating how water should move, the A&M model learns from where flooding actually caused damage. That's the core difference.
The Institute for a Disaster Resilient Texas, led by Sam Brody - the lead technical expert on Gov. Abbott's post-Harvey commission - spent a decade building the tool, which the public can use through its Buyers Aware platform. It trains a machine-learning algorithm on federal flood insurance claims data that isn't available to the public, scoring every property on seven factors:
- Elevation - lower ground collects water and sits in the path of runoff
- Distance to coast - storm surge and high-tide exposure
- Distance to stream - overflow risk when channels leave their banks
- Height above nearest drainage - vertical separation from the nearest stream, lake, or coast
- Land roughness - vegetation slows water, though in low spots it can trap it
- Soil drainage - how fast saturated ground lets water through
- Imperviousness - how much surrounding concrete and pavement pushes runoff your way
Rice University engineering professor James Doss-Gollin, who was not involved in the work, explained to the Chronicle why this approach fits Houston: traditional floodplain logic assumes a river, a floodplain, then hills. Houston doesn't work that way.
"Houston is very flat, with a lot of concrete, and gets torrential rain, so it can flood anywhere." - James Doss-Gollin, Rice University
The Model Has Blind Spots Too
The A&M model's biggest strength is also its biggest weakness: it learns from insurance claims, and huge numbers of Houston homeowners don't carry flood insurance. Neighborhoods with low insurance uptake generate fewer claims, so the model has less to learn from there.
"Lots of people don't have flood insurance, so that's a big limitation." - James Doss-Gollin, Rice University
Brody's counter is that Texas floods often enough to compensate: "In Texas, it is so saturated. We have more data than we know what to do with." His team tested whether claimed-on properties systematically differ from the rest on the model's own factors. But the honest answer is that every one of these tools - FEMA's, A&M's, the commercial ones - is an estimate with gaps. UC Irvine professor Brett Sanders, whose 2024 paper found “concerning levels of uncertainty” in First Street's national flood data, called the A&M approach promising precisely because Houston's flood history gives it so much real data to learn from.
One thing the researchers themselves stress: when the models disagree, FEMA's designation still controls whether you're required to carry flood insurance. The A&M score is information, not regulation.
The Part That Should Actually Bother You
The risk data exists. The problem is that the people selling you a house don't have to show it to you - and in Houston, they actively chose not to.
Brody's institute spent about two years working with the Houston Association of Realtors to put its flood-risk ratings directly on HAR.com listings, one of the most-used real estate sites in Texas. Last year, shortly before launch, the association killed the project after agents complained about high-risk ratings on some properties and worried they could hurt sales.
"The consumer adamantly wants this information. They're going to find the information somewhere." - Bill Baldwin, Houston Association of Realtors spokesperson
Texas seller-disclosure law has its own holes. An owner who held a property briefly doesn't have to disclose flooding that happened before they owned it, and there's no central database that carries disclosures forward as homes change hands. The Chronicle's own searchable version of the A&M map sits behind its paywall. So the state of play in 2026: the science keeps getting better, and the information available to an actual buyer at an actual address barely moves.
What You Can Actually Do
You don't have to wait for the maps, the realtors, or the legislature. The address-level picture is available right now:
- Look up your address. A FloodMaps report shows your current FEMA zone and your proposed MAAPnext zone side by side, your exact ground elevation, nearby gauge history and 311 flooding reports, and whether Harvey damage was recorded at your address - the claims history the A&M model trains on, at the one address you care about.
- Read the proposed map, not just the current one. The February 2026 MAAPnext maps aren't law yet, but they're the best official picture of where risk is heading. Our homeowner's guide to the 2026 changes walks through how to read them.
- Buy flood insurance if you're anywhere near a boundary. Outside the mapped high-risk zones, policies are cheap, and locking one in before the new maps finalize gets you newly-mapped pricing with capped increases if you're reclassified. About a third of Harris County's Harvey damage happened outside the mapped floodplain.
- If you're buying, pull the history yourself. Ask for the seller's disclosure, then verify it - disclosure gaps mean the form alone isn't the whole story. Check the address before you make the offer, not after.
Iowa researcher Adam Pollack, whose 2026 paper found only 4% of climate-risk studies fully share their data and code, put the bottom line well: "Flood risk is really uncertain. But we can make good decisions in the face of uncertainty."
Three models, three answers, one county. Don't pick a favorite number - check your address.
Quick Answers
How many Harris County properties are at risk of flooding? FEMA's current maps place about 184,800 properties in the regulatory 100-year floodplain. FEMA's proposed February 2026 maps raise that to roughly 345,400, and a Texas A&M model trained on actual flood insurance claims estimates about 389,500 properties face a 1% annual chance of flood damage.
What is the Texas A&M Buyers Aware flood model? A flood-risk tool built by Texas A&M's Institute for a Disaster Resilient Texas. It trains a machine-learning algorithm on a decade of federal flood insurance claims and scores properties on seven factors, including elevation, distance to coast and streams, soil drainage, and surrounding impervious cover.
Why don't FEMA flood maps show all flood risk? FEMA maps were designed to set insurance requirements and development rules, not to predict damage at individual homes. Harris County's current maps date to 2007 and 2001-era rainfall statistics; FEMA itself acknowledges flooding happens outside mapped high-risk areas.
Do Texas home sellers have to disclose flood risk? Partially. Sellers must disclose floodplain status, past flooding they know about, and flood insurance claims - but short-term owners don't have to disclose flooding before their ownership, and no central database preserves disclosures as homes change hands.