Miami Herald special series on hurricanes Part 2 (5 stories)

[HurricaneJoe22]

Money woes stifle hurricane science

BY DEBBIE CENZIPER

dcenziper@herald.com

The clues to help forecasters figure out why hurricanes defy them sit untouched inside a dusty storage room off Miami's Rickenbacker Causeway.

Here, at the government's Hurricane Research Division, the life spans of dozens of storms are documented on nine-track tapes stacked floor to ceiling, wall to wall, from Claudette to Fabian to Irene to Andrew.

There are trillions of bytes of data. But the division's 16 remaining researchers don't have time to study them.

While the National Oceanic and Atmospheric Administration has been warning the nation about the rise in dangerous hurricanes, the agency has denied requests from its hurricane researchers for more scientists, modern equipment and backing for breakthrough projects, a Herald investigation found.

Since 1995, NOAA's Hurricane Research Division lost 11 scientists and has replaced just four, leaving 31 people and a base budget that hasn't topped $3.5 million in more than two decades.

A former director and two senior researchers say they've pleaded for 10 years for an increase of at least 50 percent, but NOAA has granted only incremental bumps that barely kept pace with inflation -- or no increases at all.

''Our requests were dead on arrival,'' said former Hurricane Research Division Director Hugh Willoughby, who quit the post in 2002 after seven years of denials. ''Basically, it was a fool's errand.''

It's difficult to track how much money researchers have asked for and why it was denied. That's because requests for budget increases made by NOAA agencies including the Hurricane Research Division are considered a part of internal planning -- kept outside the view of Congress and the public. NOAA administrators decide which projects to support, then send an agency-wide request every year to the Department of Commerce and eventually to the president.

''It means Congress can think they're funding hurricanes properly when they're not,'' said research meteorologist Mike Black, a 20-year division veteran.

Through more than 20 e-mails, spreadsheets and other Research Division reports, The Herald documented the decadelong budget struggle -- which continues even as powerful hurricanes, most recently Katrina and Rita, batter the United States.

At the Research Division, it has come to this: Black studies the intricacies of storms that killed thousands of people on a rigged personal computer because the one NOAA gave him was 8 years old and obsolete. Key data from hurricane hunter flights is stored on a 10-year-old computer; there's no money to replace it.

The screen that researchers use to dissect satellite images is a hand-me-down from a lab in Colorado. Devices they rely on to test ocean temperature were Cold War leftovers donated by the Navy.

But nothing is as frustrating as the possibility of progress, just out of reach.

''We are the world's experts on hurricanes, but we're desperate. We need help,'' Black said.

The nine-track tapes, so far unexplored, are filled with data from radars and hurricane hunter planes, documenting how eye-wall changes made hurricanes stronger, how cool water made them weaker, how cloud patterns produced torrential rainfall.

Researchers believe unraveling the behavior of past hurricanes will help their colleagues at the National Hurricane Center predict storms of the future more precisely.

Forecasters say the tapes, among other benefits, could detail such dynamics as hurricane eye-wall replacements. That's when an outer ring of wind and rain forms outside the main eye wall, choking off the wall and weakening the storm, or in some cases, contracting and intensifying, producing a stronger hurricane.

By examining dozens of storms, scientists also hope to find patterns that can help them pierce two of the biggest blind spots in hurricane forecasting:

• The sudden strengthening of storms just before landfall, like Hurricane Charley in 2004, Hurricane Iris in 2001 and Hurricane Keith in 2000, all mild-mannered hurricanes that unexpectedly exploded.

• Rainfall prediction, a particularly unsettling mystery because freshwater flooding has become the No. 1 killer during hurricanes. Deadly flooding sideswiped Richmond, Va., during Hurricane Gaston in 2004, South Florida during Hurricane Irene in 1999, and Honduras and Nicaragua during Hurricane Mitch in 1998. During that hurricane alone, more than 9,000 people died.

Researchers have another goal, too: feeding more weather information, like wind data from radars, into NOAA's computer models, used by the Hurricane Center to predict the path and strength of storms.

That requires faster computers, but NOAA is falling behind. In November 2000, the computers NOAA used for weather prediction were the sixth most powerful in the world -- today they're 69th, according to a 2005 ranking of the world's fastest supercomputers compiled partly by the U.S. Department of Energy.

Upgrading the computer models also requires expert help; the Research Division wants to hire at least four more scientists to build programs that would allow more detailed weather-observation data to feed into the models. But NOAA has refused requests for more staff.

''Significant improvements are possible right now,'' said longtime meteorologist and former Penn State University dean Charles Hosler, whose former students include top administrators at the Weather Service. ''Even a very small expenditure would mean a lot for this country -- and we're talking millions, not billions.''

NOAA chief Conrad Lautenbacher counters that many research arms of the agency compete for money -- and that's in limited supply. NOAA's 2005 budget request included $6 million to assess the stock of monkfish and blue marlin, $12 million to research endangered salmon, $5.5 million to start air quality ozone forecasts and $6.6 million to study aerosols, records show.

There was no additional money for the Hurricane Research Division. NOAA's 2006 budget request also included no new money.

''Sitting at the top of the pyramid of my empire, I'd certainly like to have more money, too,'' said Lautenbacher, appointed by President Bush in 2001. '' . . . I'm a scientist. I'd like to see as much as we can get them.''

Between 1998 and 2005, the Research Division's base budget grew by about 14 percent, from $2.6 million to $3 million, an increase that barely kept pace with inflation. The division also receives grant money from within NOAA and from other agencies, but that money is doled out on a year-by-year basis and is not permanent.

BUDGET WOES
RESEARCHERS LEAVE AMID FRUSTRATIONS

The lack of support for hurricane study has frustrated researcher Stan Goldenberg, who fumed in a 2004 e-mail to a colleague: 'What will it take, a Category Three/Four (hurricane) hitting New York City?''

The Research Division's budget troubles heated up when Willoughby became director in the mid-1990s.

Records show he regularly told his bosses about deficits the division faced simply to keep operating at basic levels, according to a review of his budget spreadsheets.

Still, no money.

Top scientists left, including Mark DeMaria, who had created the only computer model capable of helping to predict a storm's strength.

''You could see that this was an agency that was not growing,'' said DeMaria, who left research for the Hurricane Center and now works for another NOAA agency in Colorado.

Even after a landmark study in 2001 by the Research Division's scientists warning of far busier hurricane seasons, NOAA continued to turn down requests.

Former Hurricane Center directors were so concerned that they decided to take an unprecedented step: They appealed directly to Congress.

''This is not the time to allow funding for this vital organization to continue decreasing,'' five former directors wrote in 2001.

Willoughby left a year later. He took a job across town at the International Hurricane Research Center at Florida International University.

It wasn't always this bad, he said.

Forty years ago, leading hurricane experts were clustered on the University of Miami campus -- forecasters at the Hurricane Center, scientists at the Research Division and academics with the university's meteorology department. Hurricane hunter research planes were stationed nearby at Miami International Airport.

But NOAA changed that, moving the researchers to a building on Virginia Key, the forecasters across town to the FIU campus, the hurricane hunter planes to MacDill Air Force Base in Tampa.

The separation frustrated scientists, who were used to poring over satellite images and radar data together, or driving minutes away to the airport for regular hurricane hunter missions.

Despite the problems, researchers have managed with limited resources to advance the science of hurricane forecasting.

They helped design new equipment, including an acclaimed stepped frequency microwave radiometer that for the first time measures surface wind speeds from hurricane hunter planes. On a shoestring budget, the project took 20 years, ending last year when the device was brought into regular use. The equipment has been so useful to forecasters, Congress is paying $10 million to install the device on all hurricane hunter planes.

The Research Division was also the first to discover that sampling the winds in the environment around a storm would dramatically improve track forecasts. Former Hurricane Center Director Bob Sheets persuaded Congress after a decade of discussions to buy a $43 million jet devoted exclusively to those missions; the jet is now credited with improving track forecasts by as much as 25 percent.

Researchers want to do more, but say the competition for dollars has grown even tougher under Lautenbacher.

A retired Navy vice admiral, Lautenbacher was president of the nonprofit Consortium for Oceanographic Research and Education before taking the helm of NOAA.

Some suggest Lautenbacher focuses more heavily on ocean sciences, given his background.

''Hurricane researchers are competing with porpoises and tuna fish and codfish, meaning in a bureaucracy, who gets funded isn't necessarily based on merits but on politics,'' said Hosler, who has studied NOAA for Congress.

Tornado research has also drawn support.

NOAA's National Severe Storms Laboratory in Oklahoma, the nation's authority on tornado research, has a $19 million budget this year -- more than $15 million came from NOAA, the rest from grants. The lab has 44 NOAA employees plus 75 from the University of Oklahoma who receive grant money.

That's four times the staff at the Hurricane Research Division. The lab's mission is broader than the Research Division's, and researchers acknowledge that tornadoes impact more states.

But advocates for hurricane research argue the United States is experiencing a dangerous surge in tropical weather -- nine of the past 11 years have posted above-normal hurricane seasons -- and research dollars have not kept pace.

Lautenbacher counters: ''Many of my research divisions would like more money . . . I can give you a long list of national priorities, like defense and homeland security.''

THE FUTURE
NO ADDITIONAL MONEY AFTER 2007

NOAA officials also say the Research Division isn't the only arm of the agency that studies hurricanes, and that overall, hurricane research funding NOAA-wide grew from $5 million to $8 million since 2000.

''If you compare that to a lot of other federal agencies, they'd be envious,'' said NOAA budget director Steve Gallagher.

Yet, a NOAA official confirmed in late September that the Research Division's most recent request for more money to study why hurricanes rapidly intensify has been denied.

In NOAA's budget plan for 2007 through 2011, there's no additional money for the Research Division.

Said former director Willoughby: ''We see the outlines of what needs to be done, but before it comes to fruition, more people could die.''

------------------------------

The next big thing may be short-handed

A 'next generation hurricane prediction system' is expected to bring widespread improvements to forecasting, but more scientists are needed to get the project done.

BY DEBBIE CENZIPER

dcenziper@herald.com

In a bustling Maryland office, National Weather Service scientists are building a state-of-the-art computer system expected to bring unprecedented improvements to hurricane forecasts.

When it's complete, the ''next generation hurricane prediction system'' will simulate conditions in and around hurricanes in greater detail than any of the dozen different computer models the National Hurricane Center now uses to predict the path, strength and size of storms.

But the number of scientists working on hurricane modeling has dropped in recent years, according to government officials, and the amount of time needed to build the new hurricane system is daunting.

Naomi Surgi, who is overseeing the development of the system for the Weather Service's Environmental Modeling Center, said she hopes to have the project complete by 2007.

But she acknowledged that more scientists are needed in her group to work on the system.

''The number of scientists needs to be dramatically increased,'' Surgi said.

Hurricane Center Director Max Mayfield said he's concerned about the staffing problem because the project offers the most promising possibilities for better forecasts.

Current computer models are capable only of using crude representations of weather conditions to predict a hurricane's track and strength. The new system will incorporate many more details that for the first time will give forecasters a look at the entire structure of a simulated storm. Forecasters will be able to follow a hurricane step by step, change by change, as it approaches land.

''The hope is that we'll see a significant improvement in both track and intensity forecasting,'' Mayfield said. ``We've got to start somewhere, and quite frankly, we all have concerns about not having enough dedicated resources to really make this work.''

-----------------------------------

EQUIPMENT

Data to plot strength elusive

BY DEBBIE CENZIPER

dcenziper@herald.com

The nation's hurricane researchers are measuring the force of dangerous storms using hand-me-down devices from the Navy manufactured during the Cold War and coated in ash from a 1991 volcanic eruption in the Philippines.

And that's the upside.

The bigger problem: Researchers have no way of getting the data to the National Hurricane Center even though they believe it would help forecasters better predict the sudden strengthening of storms.

''There is no hope of accurately predicting the intensity change of a hurricane without accurately being able to predict and forecast the conditions in the upper ocean,'' said research meteorologist Mike Black, with the Miami-based Hurricane Research Division, which supports the Hurricane Center during storms.

The Navy donated the devices to the Research Division in the early 1990s. Some came from a base in the Philippines and were still covered with ash from the eruption of Mount Pinatubo.

Though the devices were manufactured 15 years ago, they were welcome gifts to the government's hurricane researchers, who have long lacked basic equipment.

Researchers decided to drop the devices, called the Airborne Expendable Bathythermograph, during hurricane missions on the government's specially equipped turboprop planes, the WP-3D Orion.

The three-foot expendable probes capture the temperature of the ocean down to 1,000 feet. Heat fuels hurricanes -- storms are generally stronger when the water is warmer in the upper layer of the ocean.

The more modern dropwindsondes, also released from planes, don't measure ocean temperature; satellite data is not nearly as precise and measures only the ocean's surface temperature.

Problem is, researchers can't get the data to forecasters and the computer models they use to plot the path and strength of hurricanes.

Software must be developed, at an estimated cost of $200,000. But researchers say the National Oceanic and Atmospheric Administration won't support the request.

Researchers also want the hurricane hunter planes flown by the U.S. Air Force Reserve to drop the devices. Modifying the planes would cost an estimated $5 million.

----------------------------------

Center signs off on its own storm forecasts

Forecasters at the National Hurricane Center often verify their own predictions. Some say it's a conflict of interest.

BY DEBBIE CENZIPER

dcenziper@herald.com

Some meteorologists are calling on the National Hurricane Center to launch a rigorous examination of its forecasts -- and they want the work done by scientists outside the National Weather Service.

After each hurricane, the Hurricane Center verifies the accuracy of its forecasts. But in almost every case since 1992, the center has allowed forecasters to verify their own work, The Herald found.

Some meteorologists say it's a conflict of interest, and the Hurricane Center should bring in independent experts to produce detailed studies after storms.

That hasn't happened.

Former Hurricane Center Director Jerry Jarrell, who retired in 1999, said he pushed to keep better track of errors but was pressured to set the issue aside.

''I had stones thrown at me early on,'' he said. ``The Weather Service didn't particularly like to be advertising errors.''

The Hurricane Center's post-storm reports contain only a brief section about forecasting flaws -- and they're almost always written by the same meteorologist who crafted some of the forecasts.

After Hurricane Iris pummeled Belize in 2001, the meteorologist who verified the center's performance authored seven of the 20 discussions issued during the storm's life span.

The report did point out the track forecast was ''unusually large'' and that the storm's strength was underestimated. But the report, along with most of the others, did not answer the most important question: why forecasts go bad.

During Iris, for example, the Hurricane Center did not fly its acclaimed Gulfstream jet to measure the steering currents around the storm, which forecasters and researchers say could have improved the forecast.

Hurricane Center Director Max Mayfield argues that all the reports are passed around in-house for reviews.

He said he'd bring in outside help, but he doesn't have the money.

''There's no incentive there for us to fake it one way or another,'' he said. `` . . . We just want to get it right.''

------------------------

FLOODING

When surprises turn deadly

Researchers want to launch extensive studies on rainfall during hurricanes but say they've been turned down by the National Oceanic and Atmospheric Administration even though thousands of people have died from freshwater flooding.

BY DEBBIE CENZIPER

dcenziper@herald.com

HANOVER, Va. - The last thing Elizabeth Reavis saw before the water sneaked up behind her, sucked her under and sent her tumbling downstream -- battered and disoriented and freezing cold -- was her children.

They were curled around a tree a few feet from where their truck had skidded along the banks of a flooded creek, just three miles from home.

It was raining hard and had been for hours, but Reavis and her neighbors figured they were getting hit merely by a strong summer storm.

What they didn't know: Hurricane Gaston, which had rolled inland after its strike on South Carolina the day before, was drenching Virginia with torrential rainfall and flash flooding.

Forecasters have struggled for years to predict rainfall during hurricanes -- it's a particularly troubling blind spot because drowning from inland flooding has become the No. 1 killer during hurricanes.

''There's no doubt we need to do a better job forecasting precipitation, absolutely no doubt,'' said National Hurricane Center Director Max Mayfield.

In September 2004, 3,000 people died when Hurricane Jeanne stalled over Haiti, producing torrential flooding and mudslides.

Entire families drowned, bodies unearthed as floodwaters receded.

In one hospital, every patient died when walls of water and mud overwhelmed the building.

Mayfield considered hiring counselors for his staff.

''That one just broke my heart,'' he said.

For years, scientists at the Miami-based Hurricane Research Division, which supports the Hurricane Center, have pushed to launch more extensive studies to improve rainfall forecasts.

But they say they've received little support from the National Oceanic and Atmospheric Administration.

One key project: tapping into the power of a unique NASA satellite to better study rainfall globally, and in the end, help researchers understand where and when rain will fall during hurricanes.

''We're just scratching the surface,'' said Hurricane Research Division Director Frank Marks.

In some parts of Virginia in 2004, Gaston dumped 12 inches of rain within eight hours, catching local Weather Service meteorologists by surprise. By the time they started calling the downpour ''the remnants of . . . Gaston'' in weather advisories that grew more urgent as the night wore on, creeks and roadways had flooded. Eight people drowned.

Reavis and her two grown children had just left their house to pick up a family friend when their truck became trapped in rising waters along a creek bed. They climbed through the windows, but the water hurtled Reavis downstream.

She grabbed a branch and pulled herself up, calling into the wind, ``I'm all right! I'm all right!''

No one answered.

Alone, beneath a canopy of oak trees and a black sky, her shorts and water shoes torn away by the water, Reavis told herself that her children were strong. Her daughter, Janai, was 26; her son, Jamie, 19. Two hours later, a motorist heard her cries and pulled her from the creek.

''I found a girl,'' he told Reavis. ``She's here.''

''But my son,'' Reavis said. ``Where is my son?''

Jamie Reavis, a tattoo artist two days away from buying his first car, drowned as he struggled to buttress his sister against a tree.

His body was found at daybreak.