The risk of flooding didn't deter James and Sandy Jarrett from buying a home on Fowl River. But 100-year storms came twice in 11 months, along with a lesser storm that also ruined everything. Their new home on the river is well above the flood plain.

James and Sandy Jarrett knew that their Fowl River home sat below the federal government's predicted 100-year flood level, so they bought flood insurance and figured they'd get wet —once every 100 years.

"I had no idea we'd lose everything three times in 11 months," Sandy Jarrett said.

The federal flood map puts a 1-in-100 annual chance on a hurricane or rainstorm pushing Fowl River over its banks until it climbs 8 feet above sea level, enough to send 2.5 feet of water inside the Jarrett home.

Instead, within one year, the river reached or exceeded 8 feet during Hurricane Katrina and an April 2005 rainstorm, and almost hit 6 feet during Hurricane Ivan. Each time, the Jarretts got soaked.

They may have been unlucky, but they were clearly misinformed, according to a Press-Register analysis. The newspaper's research suggests that government flood maps underestimate coastal flooding throughout Mobile and Baldwin counties, in large measure because the federal government's definition of a 100-year storm appears to have little basis in reality.

Predicting what a 100-year storm would look like can be a complicated process. Scientists study the impacts of past storms and make statistical calculations that predict how common such storms have been.

But Press-Register research indicates that federal flood predictions don't take into account some of the Gulf Coast's most significant storms.

The net result is that so-called 100-year storms are occurring several times a century in Alabama, flooding some people over and over.

The newspaper compiled government high-water marks from hurricanes spanning 36 years —1969 to 2005 —and compared them to current federal maps of local flood zones. The comparison showed that along the Alabama coast, hurricanes have pushed water above and beyond the 100-year flood plain repeatedly.

Due to poor mapping, creaky science and outdated data, federal flood maps simply don't appear to reflect the reality of flooding in both counties.

The deficiencies in Alabama are similar to those exposed in coastal Mississippi, where Katrina destroyed thousands of homes built too low, including many that were uninsured for flooding.

The Federal Emergency Management Agency is in charge of the flood maps and is redrawing them for Mississippi and Louisiana. Alabama's maps have not been addressed in the same way. FEMA officials pledge to take the same steps here, although it's not clear when such maps will be ready.

Few homeowners ever look at a FEMA flood map, but those documents influence where they build their homes, how they construct them and how much they pay for insurance.

Mortgage companies require all homes within the 100-year flood plain to be covered by the National Flood Insurance Program, which FEMA runs. And homes near open water must be elevated above storm surges and constructed to withstand the pounding of waves.

Inaccurate maps can influence people to forego flood insurance, either because they don't believe they're vulnerable or because their lender doesn't require it. Although the federal government urges everyone at risk to buy flood coverage, numerous interviews suggest that people make their decisions based on the official 100-year flood zones, regardless of the actual history of flooding in the area.

Outdated flood maps have led entire communities to build homes to standards more than two decades old. Back then, Camille was the most destructive storm to hit the coast, but was widely assumed to be an extraordinarily rare event. Thus the 100-year flood predictions developed after Camille didn't require builders to plan for a similar surge.

Likewise, the federal government considered Hurricane Frederic, which struck Dauphin Island in 1979, worse than a 100-year storm, so flood predictions were not raised to match that storm's surge.

A history of flooding

Storm data, however, fail to support the idea that Frederic was a hurricane so intense that it would come around less than once in a century, as claimed in FEMA's current Mobile County Flood Insurance Study. That study shows the scientific backing for flood predictions reflected on federal flood maps.

Over a 36-year period, two storms—Frederic and Katrina—have pushed water above the predicted 100-year level on Dauphin Island and the southern half of Mobile Bay's western shore. (Katrina's flooding extended farther north and inland than any other storm in the Press-Register's research.) Moreover, several other storms—Camille, Ivan in 2004, Georges in 1998, Elena in 1985—also surpassed the 100-year prediction in a few areas.

Baldwin County has a similar track record: Three storms in 36 years, Frederic, Ivan and Katrina, flooded land on the Gulf beyond the 100-year level.

Such storms are etched into the collective memory of the Mobile area, serving as benchmarks when people assess the latest hurricane brewing offshore. Yet even storms that left little impression in Alabama have exceeded the government's expectation for a once-in-a-century storm.

For instance, although 1995's Hurricane Opal struck Santa Rosa Island, Fla., it pushed water more than 16 feet above sea level—about 5 feet above the 100-year level—and inundated Baldwin areas outside the official flood plain.

But FEMA's Baldwin County Flood Insurance Study, revised in 2002 and again this year, downplays Opal's impact in Alabama, simply contrasting it with the 14- to 21-foot surge near landfall in Florida. "Baldwin County experienced substantially lower surge levels," the report states.

Such is the disconnect between flood insurance studies and post-storm surveys of high-water marks.

The insurance studies are updated periodically, but continue to carry forward old assumptions and data. Meanwhile government employees and contractors fan out after major storms and record flood levels and damage, many of which seem to cast doubt on the 100-year predictions.

For example, federal employees and contractors have returned several times to note flooding at the intersection of Cabana Beach Road and Pompano Way on Fort Morgan Peninsula.

According to the flood map of the area, water will reach 8 feet above sea level there every 100 to 500 years, on average.

But after Frederic, surveyors recorded two high-water marks of 12 feet. Opal caused 11.85 feet of flooding in 1995. In 1998, Georges struck near Biloxi and shoved 12.4 feet of water on the Fort Morgan beach. Then Katrina flooded the area to 11 feet.

"If it's been exceeded four times in places in the last 50 years, we're nowhere near what the correct 100-year storm level is," said Scott Douglass, a coastal engineering professor at the University of South Alabama.

The Press-Register found similar clusters at Coden, Bellefontaine, Alabama Port, Fowl River, West Beach and the central beach of Gulf Shores, the north shore of Little Lagoon, Perdido Key, Cotton Bayou and Dauphin Island.

There could be other problem areas, but the Press-Register's research was limited by the availability of high water mark surveys. While the government reliably collects such information near landfalls of major hurricanes, the data can be spotty for lesser storms and areas farther from landfall, including creeks and rivers that can channel storm surge well inland.

Several scientists reviewed the Press-Register's findings, including Q. Jim Chen, a Louisiana State University professor who specializes in coastal flooding and storm surge modeling.

Chen said he was surprised at how poorly flood predictions correlated with actual high-water marks. For instance, though Dauphin Island was on Ivan's weaker, west side, two marks exceeded the 100-year level and one was recorded outside the flood plain.

He said he would be concerned if a third to half of the high-water marks in an area exceeded the 100-year level. According to Press-Register research, about 40 percent of high-water marks in Alabama for Frederic, Ivan and Katrina exceeded the 100-year level or were recorded outside that flood plain. (The newspaper research didn't include marks influenced by waves.)

Questions have been raised previously about FEMA's flood predictions for southwest Alabama. In a report for the agency summarizing Ivan's impact in Baldwin County, an engineering firm noted how many high-water marks exceeded the 100-year level.

"An evaluation is needed of the recurrence intervals of the surge conditions across the area," the report stated. "Preliminary evidence suggests that much of the area that experienced the most severe surge conditions was exposed to more than (FEMA's predicted) 100-year conditions."

Local government officials in both counties reviewed the Press-Register's research and said it confirms what they had long suspected: The flood maps are unreliable.

In fact, until FEMA's 2002 Baldwin County Flood Insurance Study, the federal government predicted that some areas of the beach would not be subject to the damaging effects of waves—which any viewer of The Weather Channel knows is a mistake.

"As far as FEMA was concerned, you could build on a shallow foundation on the coast," said Lannie Smith, the head building inspector for Orange Beach, referring to thin or weakly anchored foundations prone to failure during storms. "It was just wide open."

Some local governments have dealt with excessive hurricane flooding by raising their minimum building requirements above FEMA's 100-year height. That covers new construction and renovation within the 100-year flood plain, but has no effect beyond it.

So while people within the flood plain build higher, those just outside remain on the ground.

Most recent development in Gulf Shores, for example, has occurred outside the flood plain, where the city does not regulate elevations to deal with storm surge, according to Brandan Franklin, the city's chief building official.

The strike zone

Besides changing the federal government's flood maps, recent events have shifted how scientists predict hurricanes.

In their new research, they are focusing on storm data from the last 50 years and discounting older, more questionable data, said Joe Suhayda, a retired Louisiana State University professor who specializes in storm surge modeling. That, in effect, increases the predicted severity of 100-year storms, simply because the impact of powerful storms in recent years isn't diluted by decades of substandard data from the first half of the 20th century.

Scientists also are rethinking how much emphasis they put on historical flooding, Suhayda said. The standard practice has been to take the recorded storms in an area and use them in a computer model that simulates thousands of years.

But that focuses too narrowly on the place of a hurricane's landfall, ignoring neighboring areas that face the same dangers, Suhayda said. Wise coastal planning is based on risk, not merely the circumstance of one storm or another.

This is the problem with FEMA maps, according to some people involved in flood plain management. Hundred-year flooding can be caused by a moderate storm, such as 1995's Opal, hitting in just the right spot, but it also can be caused by the wide swath of a major storm, such as Katrina, that strikes 70 miles away.

If the damage spreads well beyond landfall, "why should (the 100-year storm) be based on the fact that that storm hit there?" asked John Eringman, a retired flood plain coordinator for the U.S. Army Corps of Engineers.

Current research indicates that the area between Grand Isle, La., and the Alabama-Florida line is at the highest risk on the Gulf Coast for Category 3, 4 and 5 hurricanes, Suhayda said. New Orleans has the highest probability.

Certainly the Jarretts know what it's like to be in the strike zone.

That year of flooding on Fowl River caught up with them. The county wouldn't let them put any more money into their property unless they elevated the remaining structure or demolished it and built a new, higher house.

"The slab and everything went to the dump," Sandy Jarrett said.

Because they had flood insurance—living in the flood plain in a mortgaged house, they had to—the National Flood Insurance Program paid $30,000 of the cost of building higher.

The new home had to be at least 9 feet above sea level. They built it to about 15 feet, as high as houses on the shore of Mobile Bay.

"We didn't want any water," Sandy Jarrett said. "If we get water now, the whole of Mobile will be wiped out."

(Press-RegisterStaff Reporter Dan Murtaugh contributed to this report.)

Originally published on June 9, 2007 in the Mobile Press-Register. Courtesy of the Press-Register 2007 © All rights reserved. Reprinted with permission.