An article in the New York Times yesterday discussed a new flood-risk website called FloodFactor.com. The article’s headline trumpeted “New Data Reveals Hidden Flood Risk Across America.” The new website was developed by First Street Foundation, which defines itself and its mission as, “a non-profit research and technology group defining America’s Flood Risk.” Past. Present. AND FUTURE. That’s your first clue as to what this new site offers that FEMA doesn’t.

According to the Times, FloodFactor estimates that 14.6 million properties are at risk from what experts call a 100-year flood, far more than the 8.7 million properties shown on federal government flood maps. 

FloodFactor takes into account things like sea-level rise, increased rainfall, and flooding along smaller creeks not mapped federally.

New York Times

Purpose of This Post

This article will examine the advantages and limitations of both of the FEMA and FloodFactor sites. Together they can give property owners a fuller understanding of their flood risk. Individually, each can mislead.

Pros and Cons of FEMA Maps

Until now, the gold standard for mapping flood risk has always been FEMA. But the process of adopting FEMA flood maps is highly politicized.

That’s both good and bad. Good that information is debated. Bad that sometimes pushback from landowners and local governments distorts true flood risk. No one wants to be reclassified INTO a flood zone that would require paying more for flood insurance. And no one wants his or her property to become undevelopable or unsaleable after a map revision.

Another flaw in the FEMA system: many areas remain unmapped. That allows developers to say they are not in a flood zone when, in fact, they are. We saw that loophole exploited by LJA Engineering with Perry Home’s Woodridge Village. They showed the Taylor Gully floodplain magically stopping at Harris county line. The area on the other side of that county line had not been mapped, therefore they claimed, it wasn’t in a floodplain. That enabled the developer to follow more lenient development guidelines.

FEMA maps also do not estimate future flood risk due to factors such as climate change.

Sometimes it even feels as though FEMA’s trying to catch up with the past.

FEMA maps in Montgomery County are currently based on flood data developed during the 1980s. However, the county has been one of the fastest growing in the country. New unmapped upstream development has significantly altered downstream flood risk, as we saw in Elm Grove twice last year.

Finally, FEMA maps, at present, don’t define risk for individual properties (although FEMA hopes to introduce that capability next year). “This leaves millions of households and property owners unaware of their true risk. There has long been an urgent need for accurate, property-level, publicly available flood risk information in the United States,” says FloodFactor.com.

Pros and Cons of FloodFactor.com

FloodFactor.com addresses many of FEMA’s shortcomings. That’s a huge benefit.

During Harvey, HCFCD says that of the 154,170 homes flooded, 70,370 were outside of the 1% (100-yr) and .2% (500-yr) floodplains. 105,340 or 68% were outside the 1% (100-yr) floodplain. 64% of the homes flooded did not have a flood insurance policy in effect.

Perhaps with a better understanding of true flood risk, many more home and business owners would have purchased flood insurance and avoided financial pain.

Among the many things that FloodFactor looks at, their maps factor in:

• Claims include insurance and disaster relief claims on National Flood Insurance Program policies
• FEMA Individual Assistance flood claims for those who do not have NFIP policies
• Disaster assistance provided by the Small Business Administration. 

However, during Harvey, of the 154,170 homes across Harris County that flooded, 68% did not have flood insurance. Many of those sought government assistance. But many also did not. Most government assistance is biased against upper income households.

Regardless, FloodFactor.com claims that it:

• Shows flood risk to individual properties
• Calculates street flooding risk due to heavy rainfall
• Incorporates areas unmapped by FEMA
• Compensates for adaptations such as levees and berms that protect people from flooding
• Projects future changes such as precipitation increases, stronger storms, and sea level rise
• Predicts risk 30 years into the future, theoretically at least, enabling users to gage the probability of flooding during a 30-year mortgage
• Intends to update its models annually
• Shows risk to individual properties on a sliding scale of 1-10, instead of a yes/no, “I’m in/out of the 100-year floodplain” as FEMA does.
• Shows users an estimate of risk in increments (500-year, 100-year, 20-year and 5-year floods or 0.2%, 1%, 5% and 20% annual risk).
• Explains what parts of a building are affected by floods of varying depths
• Presents flood mitigation solutions individuals can implement
• Provides area-wide “risk overviews” and predicts how they will change in the future

FloodFactor.com Looks at Texas and Houston

FloodFactor posted a national outlook with state-by-state discussions. The discussion on Texas says, “The First Street Foundation Flood Model calculates the number of properties facing any risk of flooding in Texas as 2,116,800 over the next 30 years. Of these properties, 218,700 were categorized as facing almost certain risk, with a 99% chance of flooding at least once over the next 30 years.”

They continued, “The city of Houston has the greatest number of properties at risk of flooding in the state with 186,500 currently at risk, or 32% of its total number of properties.” 

First Street Foundation

Interestingly, when FloodFactor.com modeled Texas flood risk, they used four historical storms that did NOT include Harvey. They did not explain why.

Shortcomings of Both FEMA and FloodFactor

As far as I can tell, neither FEMA, nor FloodFactor, account for upstream development in their flood mapping. However, FloodFactor does allude at one point to factoring in impervious cover. FloodFactor’s annual updates could help address that issue (if everything is really updated annually).

In December, the New York Times published a story about a company called Descartes Labs, which had trained computers to scan satellite images to detect changes in impervious cover. Descartes found that Texas had 9 of the top 20 counties in the US when ranked by the growth of impervious cover. Maybe that will become part of FloodFactor 2.0.

My background does not qualify me to critique FloodFactor’s methodology or documentation. But I have lived in the north Houston area for 35 years. So I do feel qualified to talk about their predictions for the Lake Houston area.

“Ground-Truthing” Accuracy Claims

I evaluated some areas that I know flooded during recent large storms to get a feeling for how well FloodFactor performed.

After all, if FloodFactor can’t predict the past, why should I believe it can predict the future any better?

I looked at four areas in Kingwood and SE Montgomery County. I was extremely familiar with homes and businesses that flooded or didn’t flood in each during major storms.

Elm Grove:

FloodFactor seemed to nail it. FEMA did not. FEMA shows no flooding north of the county line. FloodFactor even showed areas flooding on Perry Homes’ undeveloped property, plus around it.

However, the background map appears to be about 10 years old, judging by the extent of development in Woodridge Forest (lower left). Much of that area has since been built up by developers.

So has Woodridge Village (center forested area). It has been cleared for about two years now. Five large detention ponds have been constructed. And the dirt excavated from those has built up the remainder of the property. So in that sense, FloodFactor is outdated from the git-go.

It’s unclear whether Floodfactor’s predictions, say for the Taylor Gully area through Elm Grove (bottom center), take those new upstream detention ponds into account.

River Club/River Ridge:

Between FM 1314 and the West Fork just south of Northpark Drive, lie two small subdivisions called River Club and River Ridge. FEMA nailed it. FloodFactor missed. Badly.

FloodFactor (below) correctly showed street flooding in the higher elevations near FM1314, but missed West Fork flooding by a wide margin. They should have started with FEMA’s baseline.

North and South Woodland Hills in Kingwood:

Mixed results. FEMA nailed areas that flooded during Harvey, but missed some street flooding during Imelda. FloodFactor makes me scratch my head. For instance, FloodFactor showed most of North and South Woodland Hills flooding, presumably from street flooding, after climate change will make rainfalls even more intense. However, they don’t specify exactly where most of the risk comes from. These are among the two highest subdivisions in Kingwood.

Two Better than One

Given the fact that neither FEMA, nor FloodFactor, is perfect, it’s best to use them in conjunction with each other. With a knowledge of the limitations of each, one can gain a better understanding of flood risk. There’s just no substitute for talking to people in the neighborhood.

One more example in that regard.

Northpark Commercial Area Near Woodland Hills Drive

FEMA correctly shows historical flood risk. Note the location of red dot for reference below.

The red dot above shows the location of a building that was raised 3 feet before construction, but not listed with FEMA because it was far outside the 500-year floodplain.

One big benefit of the FEMA maps is that property owners who elevate their homes can have them reclassified out of the floodplain. It’s not clear how that might work with FloodFactor.

However, with input from both websites, if I were buying a property, I would know enough to start asking questions:

• What’s happening around this area to increase flood risk? (Upstream development in the Ben’s Branch Watershed? Environmental change?)
• Should I check the elevation of my property against FloodFactor’s estimate? (Yes!)
• Could floodproofing or elevation strategies help my property? (Yes!)
• Should I look again at getting flood insurance? (Yes!)

This is like getting a second repair estimate for your car. You are always safer with at least two estimates.

Make Your Own Comparisons

Find a neighborhood or property of interest on FEMA’s National Flood Hazard Layer Viewer.

Compare the same area on FloodFactor.

You can learn more about FloodFactor’s Methodology here.  

Here’s the Documentation for FloodFactor.

And here’s more about the model that FloodFactor uses.

Posted by Bob Rehak on 6/30/2020

1036 Days after Hurricane Harvey