Boaters heading east from Pascagoula to Mobile Bay can't tell exactly where Jackson County, Miss., ends and Mobile County begins. From the water, the marshes and bayous look pretty much the same.

Yet if federal flood guidelines are to be believed, there's a dramatic change in the landscape that's impossible to see with the naked eye. In one area of the Alabama coast, federal flood maps predict that a once-in-a-century hurricane would cause a 14-foot storm surge. Right across the Mississippi line, the same hurricane would produce 19 to 20 feet of flooding, according to interim flood maps.

Those 5 or 6 feet are the difference between hurricane predictions of the 1980s—the basis of Alabama's flood maps—and the predictions after Hurricane Katrina, on which Mississippi's maps are based. In reality, the topography is the same across that artificial line, and there's no reason to expect the surge of a 100-year hurricane to be any greater or less on one side or the other.

Katrina pushed water far above the flood plain in Mississippi and Louisiana, destroying houses outside the high-risk area. To guide rebuilding, the Federal Emergency Management Agency rushed to create advisory flood heights that better reflect storm surges. Those advisory elevations are significantly higher than before, and higher than Alabama's.

The agency is now revamping flood maps in Mississippi and Louisiana with modern science and information gleaned from recent storms.

Maps for Mobile and Baldwin counties, on the other hand, remain firmly rooted in the past. Even new maps for coastal Alabama, which are being finalized now, are based on old science and outdated storm information, a study by the Press-Register indicates.

Those maps offer coastal residents little guidance on how to respond to storms like Katrina, which caused record flooding in wide areas of Mobile County despite striking some 70 miles away.

The maps don't account for hurricanes Ivan and Frederic, either. As with Hurricane Katrina, flooding from both of those storms exceeded 100-year levels.

The 100-year flood plain is the area that has a 1-in-100 chance of being flooded in any year. Homes within that area are considered to have a high risk of flooding. They must be elevated and insured for flooding if they are mortgaged.

Less damage, less attention

FEMA officials told the Press-Register last fall that Alabama wasn't included in their post-Katrina revisions because damage wasn't as extensive as in Mississippi and Louisiana. They also said they didn't believe that Katrina's flooding exposed problems with Alabama maps.

A Press-Register study of storm data, however, indicates that the Alabama coastline saw three of FEMA's 100-year storms from 1969 to 2005.

This spring, FEMA officials acknowledged problems with Alabama flood maps and pledged to overhaul them using the same methods employed in the other two states. FEMA says the maps won't be ready before 2010, and local building officials expect it will be even longer.

Whenever FEMA's maps emerge, most of the Alabama homes destroyed by Katrina and Ivan will have been rebuilt to the older standards—standards that failed to protect those areas before.

Some scientists and flood plain managers said the delay in updating Alabama flood maps reveals FEMA's habit of reacting to disasters rather than planning for them. Critics say the agency focuses on the hardest-hit areas and doesn't address places nearby that dodged the most catastrophic effects of the storm.

Those lucky areas face the same risks as the towns that found themselves in the bulls-eye, said retired Louisiana State University professor Joe Suhayda.

"If you guys get hit by another Ivan right up Dauphin Island, you're going to see a heck of a lot of interest, and all of sudden the studies are going to be revised," Suhayda said. He is now consulting on several projects related to flood plain mapping and coastal preservation.

"The assumption is, any of these storms could have occurred anywhere along this coast," he said. "We want to know what the actual risk is, not whether I rolled a six last time or a two."

Across the line

Steve Crockett was planning on building a new vacation home on Point aux Pins in southwestern Mobile County before Katrina destroyed the existing house, which had survived Frederic in 1979.

Mobile County requires that Crockett's new house be at least 17 feet above sea level, which is 2 feet above the 100-year flood prediction of 15 feet. Yet if he were building a few miles away on Heron Bayou in Jackson County, Miss., his house would have to be 21 feet above sea level.

"This elevation difference at the state line is not physically possible," said University of South Alabama professor Scott Douglass, who specializes in coastal engineering. "The difference between Pascagoula and Bayou La Batre is zero. They're all the same."

FEMA contractors recorded two high-water marks of about 19 feet on Mississippi's Heron Bayou after Katrina. Now, the agency's recommended building heights at Heron Bayou are 19 or 20 feet above sea level, and Jackson County tacks on another foot.

FEMA's recommendations extend to the state line, but end there. The agency hasn't called for any changes in Alabama.

FEMA surveyors didn't measure the height of Katrina's surge near Crockett's house. The closest high-water mark was recorded about two miles northeast on Henderson Camp Road. There, the surge reached 11 feet, even after traveling nearly a mile inland, beyond the 100-year flood plain.

Mindful of what happened in Mississippi, and seeking a break on his flood insurance premium, Crockett said he probably will end up building to 20 or 21 feet above sea level.

"If that's what they did in Mississippi ... I can guarantee you that's what needs to be done here," he said. Alabama didn't get the brunt of Katrina, he said, "but the risk is the same again."

Flood predictions don't match across the Alabama-Florida line, either. The 100-year flood prediction increases abruptly by 1 to 3 feet as soon as you cross from the white sand of Orange Beach to the white sand of Escambia County, Fla.

The discrepancy there has nothing to do with Katrina or Ivan. For reasons unexplained, flood elevations haven't matched across the Alabama-Florida line for 20 years, according to a review of prior flood maps.

Taken together, the differences to the east and the west of Alabama would lead one to believe that the state is in some kind of meteorological safe zone—which, scientists say, isn't the case.

"I have not seen a trough in hurricane probability in Alabama," said Don Resio, a U.S. Army Corps of Engineers scientist. He is helping develop a single method of predicting storm surge that will be used by both FEMA and the corps.

In fact, Suhayda said new research indicates that severe hurricanes are more likely to strike between New Orleans and the Florida line than anywhere on the Gulf Coast.

New science vs. old studies

The shortcomings in Alabama's maps originate in little-read FEMA publications called "flood insurance studies." They outline the storm history for an area and describe how the 100-year flood levels were developed.

To predict the 100-year flood level in an area, scientists take data from sample hurricanes of varying intensity and enter it into a computer model. The model uses those storms to simulate thousands of years of storms.

FEMA's flood insurance studies for Mobile and Baldwin counties have been periodically updated, most recently this year. But each new report over the last two decades carried forward outdated flood predictions.

Since those flood insurance studies were first completed, mapping technology has improved to the point that ground elevations can be measured to within a foot, compared to older maps that were accurate, at best, to 5 feet. Old maps are notoriously bad at delineating the slight elevation changes that determine what gets wet and what stays dry in a coastal area.

Computer modeling can predict a hurricane's effects at 2 million places on the seawater's surface every second—precision unheard of 25 years ago. Equally important, researchers can plug the past 25 years of data into their models, which basically doubles the amount of reliable hurricane data available in the early 1980s.

These modern techniques are being employed for the first time in coastal Mississippi and Louisiana, and they are yielding higher 100-year flood predictions. Alabama's probably will rise, too, whenever its study is finished.

A look at FEMA flood insurance studies for Mobile and Baldwin counties shows how obsolete the current predictions have become. Those reports—including the most recent versions completed this year—rely on hydrologic studies completed between 1976 and 1984, aside from some minor revisions in 1998 that added the effect of waves on Baldwin's shores.

The sample storms used to create the Mobile and Baldwin predictions are significantly weaker than those used in post-Katrina research for Mississippi and Louisiana, Suhayda said.

For instance, the worst storm in the Mobile and Baldwin models, predicted to occur 1 percent of the time, had a central pressure of 942 millibars. (More intense hurricanes have lower central pressures.)

But such hurricanes aren't so uncommon on the Gulf Coast. Frederic had a central pressure of 946 millibars at landfall. Opal in 1995, which caused significant coastal flooding, had a central pressure of 942 millibars, according to the National Oceanic & Atmospheric Administration.

Looking back in time

It's not just the storm data that's out of step. Information about Mobile County's economy and infrastructure has been carried forward in FEMA flood insurance studies without updates for more than 20 years.

The 2007 preliminary flood insurance study for Mobile County does not mention any hurricanes since Frederic. One section notes how much freight went through the Alabama State Docks—in 1976.

The report also states that the completion of the "new" Dauphin Island bridge probably would accelerate redevelopment after Frederic. People have been driving on that bridge for almost 25 years.

Baldwin's new report describes Ivan, but surge modeling data wasn't adjusted to take it into account. Katrina doesn't get a mention.

One sentence in the preliminary 2007 Mobile County report (and its 1998 predecessor) is particularly dated: "Hurricane Frederic is estimated to have a recurrence interval of more than 100 years along the coastal areas."

If that were true, it would explain why Frederic's flooding exceeded 100-year levels in Mobile County. Yet Katrina came along 26 years later and matched or exceeded that flooding, casting doubt on how extraordinary those storms are.

Douglass said "it's very tempting" to write off damaging storms like Frederic.

Though Mobile and Baldwin's coastal studies have grown dusty, coastal research was updated for the 2006 revision of the flood insurance study for Escambia County, Fla.

The Escambia report states that earlier studies underestimated hurricane flooding because they used flood predictions made in 1982, about the same time as Mobile's and Baldwin's predictions were done. Even after elevations were raised to account for Frederic's flooding, the Escambia report states, Hurricane Opal suggested that those were suspect as well.

The 2006 Escambia report concludes that the 100-year elevation on the Gulf in that county should be 10.5 feet (not including the added impact of waves). Neighboring Baldwin County's report, revised in 2002 and again this year, calls for a 100-year elevation between 7.1 and 7.8 feet.

Likewise, a 2003 study for the Florida Department of Transportation recommended using that state's own predictions of 100-year levels rather than FEMA's. Florida's estimates ranged from about 4 feet higher in Destin to about a foot higher in Escambia County.

A difference of 3 or 4 feet is "huge" in flat coastal areas, Douglass said. He believes flood levels have been too low in Mobile and Baldwin counties for years.

He wrote a report for the Alabama Department of Economic and Community Affairs in 1999 in which he stated that 100-year elevations had been incorrectly estimated at about 8 feet on the Baldwin coast. "This water level has been exceeded, on the barrier island where the buildings are, at least four times in the last 30 years," he wrote.

Douglass estimated the correct level was at least 13 feet.

Brad Loar, acting mitigation director for FEMA's southeast region, said "unfortunately it's very common" for flood maps to be tweaked over the years without a re-evaluation of storm predictions. "We just haven't had the funds committed, funds appropriated, to do that complete restudy," he said.

Waiting for guidance

Since years will pass before Alabama's revamped flood maps are ready, people putting their lives together after Katrina and Ivan probably won't be able to wait to see the new 100-year predictions.

Doug Martin tried to wait. Katrina flooded his law office in Bayou La Batre and swept away his Coden home.

Bayou La Batre required him to raise his office by 1.2 feet, he said, "which wouldn't take me out of the water with Katrina."

Yet Martin saw that Mississippians had to build 3 to 8 feet higher than pre-Katrina heights. So he looked for advice on how high he should go, in the meantime working in an office stripped to the studs and going home every night to his FEMA trailer.

Bayou La Batre officials planned to simply follow whatever changes FEMA recommended, but FEMA didn't recommend any.

Finally, he decided he had to get on with his life. He raised his office 3 feet higher than Katrina's surge. Meanwhile, in the absence of new direction from FEMA, the city at last decided to raise its minimum building requirement by a foot.

On his property overlooking the Mississippi Sound, Martin is building a new house 21 feet above sea level—among the most elevated homes on the Mobile County coast.

"I don't want to do it again, and I can't afford to do it wrong," he said.

"Who knows? They're predicting at least 30 years of bad storms, and I think we need some guidance. But we're not getting it, so we just guess."

Originally published on June 10, 2007 in the Mobile Press-Register. Courtesy of the Press-Register 2007 © All rights reserved. Reprinted with permission.