ProPublica

Journalism in the Public Interest

Cancel

Why So Many Flood Maps Are Still Out of Date

A Q&A with Professor David Maidment on what makes today’s maps 10 times more accurate than the ones much of the country is still stuck with

Flood waters engulf homes after Hurricane Ike passed through September 14, 2008 in Bridge City, Texas. (POOL/AFP/Getty Images)

The Federal Emergency Management Agency’s flood insurance maps are critically important for millions of Americans who live in flood-prone areas. The maps determine the annual premiums for flood insurance, which is required by law for homeowners with federally backed mortgages who live in high-risk areas. But many of the nation’s flood maps are woefully out of date.

ProPublica talked with David R. Maidment, a professor at the University of Texas at Austin who has advised FEMA on flood mapping, about why the data behind modern maps is 10 times as accurate as the older data and why half of Texas still doesn’t have up-to-date maps.

How did you become involved with flood mapping?

I was the drainage engineer for a small community out in Rollingwood, Texas. It’s a small suburb of Austin, and it was hit by a big flood in 1981, which is when I joined the University of Texas at Austin. They didn’t have much of any regulation at the time, and so people did more or less whatever they wanted, and they put buildings wherever they wanted. It was clear that we needed a better system.

Why is it important to have up-to-date flood maps?

Flooding is the natural disaster that impacts people more than any other.

Since federal disasters started being declared in 1952, two-thirds of them have involved flooding. The next one after that is fire. There have been about 300 federal disasters declared from fire and 1,600 from flooding. Fire is a very horrible thing, but flooding occurs much more frequently, and it’s also much more widespread. That’s the fundamental reason. Floods devastate human lives to a greater extent than any other national hazard.

What’s changed since you started studying flooding?

The big change, which happened about 10 years ago, was the conversion of maps from a paper system into a digital system. That was a very large investment of federal funds. It’s considered to be the largest civilian mapping program in the world.

How up to date are FEMA’s flood maps now?

It varies quite a bit. The data in cities is generally pretty good. It’s once you get outside of the cities that things start falling away. The original question that we were asked to investigate was whether it was possible to produce updated flood maps with out-of-date terrain and panametrics information. Flood maps have two-dimensional information that says, ‘Where are things on the ground?’ And they have three-dimensional information, which says, ‘How high is the land?’ We found that the two-dimensional data — where are things horizontally? — is actually pretty accurate, because it’s being done with image mapping now, and the emergence of Google Earth and other things like it have made image mapping really accurate. What we found was out of date was the nation’s three-dimensional elevation data, which in many cases was drawn from U.S. Geological Survey topographic maps that were developed 30 or 40 years ago, using mapping processes that aren’t as accurate as what is now possible.

What makes the new maps more accurate than the old ones?

The big improvement that’s happened is in the vertical elevation data. There’s a new technology that’s emerged called lidar, and what it involves is having an airplane flying over the land and firing out a bunch of laser pulses in kind of a beam. The beam comes out of the airplane and hits a mirror that goes left and right as the airplane goes forward. So there are billions of pulses that hit the ground and bounce back to the airplane again. About a third of the country has been covered with lidar now. The three-dimensional elevation data that it generates is about 10 times more accurate than the old data.

If you live somewhere prone to flooding, what do these improvements mean?

There’s a much more precise definition of what the real flood risk is. Also, they make it possible to define how deep the water will be. In the old flood plain maps, the only question they answered was, ‘Is your building anywhere within the floodplain boundary?’ And with the more accurate three-dimensional data, different questions can be asked: What is the likelihood that your building will flood? And if it does flood, how deep will the water be? It wasn’t possible to answer any of those questions before.

And FEMA couldn’t do that with the old maps?

No. The old maps were 2D — they were paper maps. There was no 3D aspect to them at all.

How much of the country has these newer, more accurate flood maps?

The original intention was to cover the whole United States, coast to coast, with digital flood maps. But it turned out they had a billion dollars — $200 million a year, over five years — to do that, and it wasn’t a billion-dollar job. It was probably a $10 billion job, or a $20 billion job. It turned out to be much more expensive to do that than what had been anticipated at the beginning. And so FEMA cut back on the original goals, and instead focused on the areas of highest flood risk, which generally are associated with high-population zones and zones near the coast. They quantified the flood risk in about 65 percent of the country that contained 92 percent of the population. That means that in the populated areas, the flood maps are actually in pretty good shape now. But in the unpopulated areas, in many cases there are no updated maps at all. Half of Texas, for example, has no flood maps. And I’m distressed about that, because we’re still not where we need to be.

This interview has been edited for clarity and length.

Why no mention of data from radar imaging satellites?  The US government has been creating very precise elevation maps of the entire globe for decades, and now this same data is available commercially from a variety of sources including the Canadian Radarsat-2 satellite.

What am I not understanding about this that makes it a multi-billion dollar project?  I’m not asking this in exasperation of a wasteful government, but rather assume I’m missing something.  We have what must be trillions of GPS data points and who knows how many pictures.  Add in the satellite pictures Brian mentions (which seem to do great work on elevations on Mars), it seems like the expense should be in physical confirmation where data conflicts or there’s limited data.

Our property abuts some swampy woods in Chesterfield County, VA.  We have lived here for 35 years.  We have asked the county and state for help to drain the swamp for most of those years because it is a mosquito-breeding health hazard, but we have received no help with it.  We have never experienced flooding on or anywhere within 10 miles of our property.  Then the flood map was redrawn.  The swamp became a “creek” in spite of the fact that it is stagnant, and our yard was declared “riparian.”  Next came the notice in the mail that we need flood insurance.  Yep, I’m real impressed with the new flood maps.

Interesting. We’re also starting to build weather radars that are closer to the ground, which can detect weather, including floods and tornadoes, more accurately. It’s called Collaborative Adaptive Sensing of the Atmosphere, or CASA. American Public Media’s Marketplace talked about what CASA can do after the Moore, Oklahoma Tornado: http://www.marketplace.org/topics/tech/weather-economy/what-will-it-take-make-tornado-prediction-better

However, CASA requires high speed internet (with accuracy and speed comes the need for a better network). US Ignite hopes those next generation networks continue to be built, from Chattanooga, to Kansas City, Austin, Provo, Burlington, Blacksburg, etc.! So we can continue to foster the creation of next generation apps, like CASA.

- US Ignite, @US_Ignite on Twitter, https://www.facebook.com/USIgnite on Facebook.

Brian and John ask some very cogent questions. I suspect the answer is in the scale of the data. For example, I don’t know the vertical and horizontal resolution of the radar data Brian mentions but lidar operates at much higher frequencies (light vs radio wave frequencies) and therefore is capable of better vertical and horizontal resolution—but at the cost of thousands or millions of times the number of data points.

The cost of storing and processing those data points is correspondingly several orders of magnitude larger than the cost of my home computer. (A million times the cost would be in the billions of dollars.) In addition, the cost of obtaining the data by airplane, as described in the story, would involved thousands of hours of flying time (cost of airplanes and pilots would be ???) to cover 1+ million square miles (1/3 the area of the US).

Satellite-based radar terrain maps with 1-meter resolution have been commercially available for years, and the military has much better (down to millimeter resolution).  And the data is refreshed quite often - a single satellite can cover the globe in few days and there are several satellites doing this mission.

The cost to store digital elevation data (DEM) is much less than the cost of storing optical imagery.  NOAA has an elevation map of the whole world at about 30 meter accuracy that can be downloaded from the Internet in a few hundred MBs:
http://www.ngdc.noaa.gov/mgg/topo/globe.html

I don’t see any need to do this with aircraft unless th flood maps need better resolution than what satellites offer.

Amorette Allison

July 9, 2013, 12:11 p.m.

The community where I live got clobbered with a new FEMA flood map.  It was done by mathematical modeling, which means water stops dead in the middle of some streets and there are places where you can be IN the ‘flood plain’ and look DOWN on parts OUT of the ‘flood plain.’  Our average depth of flood is 18 inches and the error for the map is 24 inches.  I just paid off my mortgage and am going to skip the stupid insurance next year but lots of people are stuck paying hundreds of dollars a year for mistakes.

Maidment is just plain incorrect when he states that there is no 3-D info on the FEMA maps. The engineering studies for our area (1968-1982), which is what FEMA law says is what determines flood zones, have elevation contours every two foot interval. So do the Rate Maps.
For us, the biggest problem with flood mapping is that banks make arbitrary flood zone determinations and they expand the zone, ignoring the FEMA maps, and there’s not a thing you can do about it. The banks can actually foreclose on your house if you don’t pay for “flood insurance” even though the FEMA map shows your house outside the zone. Wells Fargo tried that with us, and we’re still working on it eight years later, even though we have a letter from FEMA that states that our house was never in a flood zone. We’ve been screwed out of over fifteen thousand dollars. Pure, unadulterated theft.

For our country, too, the biggest problem with FEMA is the cost of the insurance. It is outrageously expensive, and every area FEMA decides to define as a flood area is condemned to become an “economic dead zone”. Every person in our country should watch this short video of Congressman Higgins(D) from New York. http://www.youtube.com/watch?v=H7_n16cstHY

This article is part of an ongoing investigation:
After the Flood

After the Flood: The Challenge of Rebuilding as the Climate Changes

In the aftermath of Hurricane Sandy, ProPublica is investigating the response to disasters as the climate changes.

Get Updates

Stay on top of what we’re working on by subscribing to our email digest.

optional

Our Hottest Stories

  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •