Starting this month, the tens of millions of Americans browsing the real estate website Zillow will encounter a new type of information.

In addition to disclosing a home’s square footage, school district, and walkability rating, Zillow will begin informing users Climate risk – the probability that a major weather or climate event will occur in the next 30 years. The focus is on risk from five types of hazards: floods, wildfires, high winds, heat and air quality.

The data has the potential to change how Americans think about buying a home, especially because climate change is likely to worsen many of these dangers. About 70% of Americans look at Zillow at some point during the home buying process. according to the company.

“Climate risks are now a critical factor in home-buying decisions,” Skylar Olsen, Zillow’s chief economist, said in a statement. “Healthy markets are markets in which buyers and sellers have access to all relevant data for their decisions.”

That’s true – as long as the information is correct. But can home buyers actually trust Zillow’s climate risk data? When climate experts closely examined the underlying data Zillow uses to assess climate risk, they weren’t convinced.

Zillow’s climate risk data comes from First Street Technologya New York-based company that uses computer models to estimate the risk that weather and climate change pose to homes and buildings. It is by far the best-known company focused on modeling the physical risks of climate change. (Although First Street was originally founded as a charitable foundation reorganized as a for-profit company and accepted $46 million in investments earlier this year.)

But few experts believe that tools like First Street’s are capable of actually modeling the dangers of climate change at the level of individual objects. A report from a team of White House scientific advisers completed last year that these models were of “questionable quality” and a Bloomberg Investigation found that different climate risk models could provide completely different disaster estimates for the same object.

Courtesy of Zillow

Not all First Street data is considered equally suspect. His estimates of heat and air pollution risk generally drew less criticism from experts. But its estimates of flood and wildfire risk — the most catastrophic events for homeowners — are widely viewed as inadequate at best.

So while Zillow will soon tell you with seeming precision that the chance of a given home being exposed to a wildfire in the next 30 years is 1.1%, potential homebuyers should view that kind of estimate with “a lot of caution,” says Michael Wara. a senior research scientist at the Stanford Woods Institute for the Environment told me.

Here’s a quick guide to thinking through Zillow’s climate risk estimates.

Neither First Street nor Zillow immediately responded to requests for comment.

If Zillow tells you that a property recently flooded, take it seriously.

Zillow said once the data is available, it will tell users whether a particular home recently flooded or burned in a wildfire. (It also indicates whether a home is near a source of air pollution.)

Homebuyers should take this information seriously, Madison Condona Boston University School of Law professor who studies climate change and financial markets told me.

“If the home has flooded in the recent past, that should be a big red flag for you,” she said. Homes that were recently flooded would most likely be flooded again, she said. Only 10 states Requiring a home seller to notify a potential buyer of a flood.

Use Zillow’s flood risk assessment as a starting point — but contact your local government to learn more.

First Street claims its physics-based models can determine the risk of flooding on individual properties. But the ability to determine whether a particular home will flood depends on detailed knowledge of local infrastructure, including stormwater drains and what is happening on other properties, and that data does not appear to be present in any model at this time, Condon said.

When Bloombergcompared Based on results from three different flooding models, including the First Street model, they agreed on results for only 5% of the properties.

If you’re concerned about a home’s risk of flooding, contact your local government and ask if you can look at a flood map or even talk to a flood manager, Condon said. Many cities and towns maintain flood maps in their records or on their website that are more detailed than First Street can provide, she said.

“The local flood manager who has inspected the property will almost always understand flood risk better than the big, top-down national model,” she said.

Don’t hesitate to get flood insurance.

In some cases, Zillow recommends that a home buyer purchase federal flood insurance. That’s generally not a bad idea, Condon said, even if Zillow comes to that conclusion using national modeling data that has errors or errors.

“It’s simply true that more people should buy flood insurance than is generally believed,” she said. “Therefore, a general overcorrection would be good.”

Zillow likely underestimates a home’s wildfire risk, especially in the West.

“If you’re considering buying a home in an area at risk of wildfires, particularly in the western United States, you should generally assume that Zillow underestimates wildfire risk,” Wara, the Stanford researcher, told me.

That’s because computer models for estimating wildfire risk are in their relatively early stages of development and are improving rapidly. Even the best academic simulations lack the detailed structural-level data that would allow them to predict a property’s future wildfire risk.

That’s actually a bigger problem for homebuyers than it is for insurance companies, he said. A household contents insurance company decides every year whether to insure a property. If it considers new science and concludes that a particular city or structure is too risky, it may increase its premiums or even refuse to cover a property altogether. (State farm stopped (Selling homeowners insurance in California last year, in part because of the wildfire threat.)

But when homeowners purchase a home, their lives and assets are tied to that property for 30 years. “Maybe your kids go to the school district,” he said. It’s much harder to sell a home if it’s not insured. “You have an illiquid asset and it’s much harder to move it.”

That means First Street’s wildfire risk data should be viewed as an “absolute minimum estimate,” Wara said. In an area at risk for wildfires, “the actual risk is most likely much higher” than the models say.

Wildfire models are particularly poor at predicting the most devastating types of fires.

In recent years, out-of-control wildfires have killed dozens of people or destroyed tens of thousands of homes Lahaina, Hawaii; Paradise, California; And Marshall, Colo.

But in these cases, once the fire started burning houses, it was no longer a forest fire but a building-to-building fire. The fire began to jump from house to house like a book of matches, burning down entire neighborhoods within minutes.

Modern computer models are particularly poor at simulating this transition — the moment when a forest fire becomes a city fire, Wara said. Although it only occurs in perhaps 0.5% of the worst fires, these fires are responsible for the destruction of most homes.

But “how this happens and how to prevent it is not yet fully understood,” he said. “And if they are not well understood from a scientific perspective, that means they are not included in the (First Street) model.”

Even the university’s best wildfire models don’t have good data on the structural details of each property – such as the material of its walls or roof consist of – that would make it vulnerable to fire.

Your entire neighborhood is important when assessing flood or wildfire risk.

When assessing whether your home is at risk for wildfires, its structure is very important. But “you also need to know what your neighbors’ houses look like within about 250 meters. So that’s your entire neighborhood,” Wara said. “I don’t think anyone has that data.”

A similar principle applies to flood risk considerations, Condon said. Your home may not be flooded, she said, but it’s also important whether the roads to your home are still passable or if the power lines are down. “It doesn’t make much sense to have a house that’s resistant to flooding if the entire neighborhood is going to flood,” she said.

Experts agree that the most important measures to prevent forest fires – or floods – must take place at the local level. Although few communities currently implement mandatory burning or fuel reduction programs, some do, Wara said.

But because no one collects data on these programs, national risk models like First Street’s do not factor these programs into a region’s wildfire risk, he said. (In the rare case that there is a governmentFor example, when dumping fuel or conducting mandatory firefighting in a city, the risk of wildfires there may actually be lower than Zillow says, Wara added.)