When Zillow recently removed private climate-risk scores (powered by First Street Foundation) from its listings, the public debate quickly became polarized: Were the models inaccurate? Were they scaring buyers? And was Zillow protecting the real estate market?
Instead of debating it abstractly, I realized this was the perfect moment to do something simple but revealing: evaluate my own home using multiple flood hazard/risk data sources, including climate-forward projection data, to see for myself how the confusion is playing out. What I found was a microcosm of what’s happening nationwide: four sources, four different “truths,” and no single place where a homeowner can understand the full story.
Here’s what I learned when I tried to answer the basic question: “Does my house in Boulder have flood risk?”
FEMA told me, “you have no flood risk.”
When I put my address into FEMA’s National Flood Hazard Layer, the message was simple:
- Not in a Special Flood Hazard Area
- Zone X (unshaded)
- No base flood elevation
- No mapped hazard on my parcel
If I were an ordinary homebuyer, I’d interpret that as: “Safe. No flood risk. No further thinking required.” But FEMA’s maps only show river or creek overflow. They do not map rainfall flooding, storm drain overload, hillside runoff, street flooding, groundwater rise, or basement infiltration. And in 2013, my neighborhood did flood—not because Bear Canyon Creek overtopped, but because of rainfall and stormwater failures. FEMA maps cannot show that.
Understand, FEMA’s map wasn’t wrong. It was simply too narrow and incomplete to describe my real risk.
First Street told me: “1 out of 10 (minimal risk).”
Next, I checked First Street Foundation’s FloodFactor tool—the same data Zillow removed after public pressure. My score: “1 out of 10 (minimal risk).”
The model told me:
- “The structure is unlikely to experience flood damage.”
- “Flooding may occur in the neighborhood, but water is unlikely to reach the building.”
Again, reassuring. But here’s what readers must understand: First Street’s flood score estimates the probability and severity of flooding at the building level based on modeled surface-water inundation depth and expected structural damage. What it does not explicitly model are the many ways that homes actually flood, especially in urban and foothill communities like mine.
It does not fully consider finished basements (I have one, our master suite), below-grade seepage through concrete foundation walls or slabs, stormwater surges, groundwater rise, or water entering through window wells. So, the score wasn’t wrong, it was just framed through a specific definition of “damage.” A narrow lens, just like FEMA, but in a different way.
Argonne National Lab told me: “Expect 20–40% deeper rainfall‑driven flooding.”
Then I checked Argonne’s CLIMRR rainfall-based flood model, one of the most advanced federal climate tools available to the public. This model doesn’t evaluate structural damage; it evaluates future hazard intensity.
For my home’s watershed, Argonne projected a 20–40% increase in rainfall-driven flood depth by mid‑century. This was the first dataset that matched what we saw in 2013:
- Water flowed down my street.
- Storm drains were overwhelmed.
- Streets became channels.
- Basements, not main floors, took on water.
Argonne was showing the hazard behind the lived experience, a type of risk that sits outside FEMA’s regulatory flood maps and is not the primary focus of building-level damage models like First Street’s.
Neighborhood history told me: “In 2013, houses like yours flooded.”
Finally, I spoke with neighbors. The stories were consistent:
- “Our basement flooded.”
- “We had to tear out the drywall.”
- “Water came down the street—not from the creek.”
- “Storm drains couldn’t keep up.”
This was the reality that none of the formal flood resources had fully captured.
What these four lenses revealed
Flood risk in the US is a multi-layered puzzle because each dataset is measuring a different part of the hazard.
- FEMA shows regulatory river/coastal risks, which is narrow by law.
- First Street shows the probability of floodwater hitting the house using credible climate data but within a narrow structural definition.
- Argonne shows how rainfall-driven flooding will intensify regardless of whether it reaches a house.
- Local history shows how the watershed actually behaves, including hazards none of the maps capture.
Each dataset is valuable, but each one is incomplete. And none can tell the whole story alone. This is the real significance of Zillow’s reversal: millions of homeowners are navigating a fragmented and contradictory hazard-data environment, often without knowing it. And if this is how confusing it is for someone like me—someone who works in climate risk and professional liability—imagine the average American trying to decode their home’s real exposure.
For design professionals, the lesson is clear: no single dataset for weather hazards and risks is enough to understand a site’s true exposure. Checking multiple sources and communicating those nuances to clients strengthens both the project and the design professional’s legal defensibility in the event of a claim. Treating hazard and climate data as part of early-stage due diligence can be a core risk management strategy.