Striving for more accurate long-range forecasts

Hurricane Charley, the second major hurricane of 2004, vexed forecasters because it intensified from a Category 2 to a 4 in less than 12 hours.

“What Charley taught us is that good track forecasts are not enough. Charley taught us that peak wind is not enough to determine impact and that rapid intensity change is the most serious forecast issue,” said Frank Marks, director of the National Oceanic and Atmospheric Administration (NOAA). He was speaking at a hurricane research symposium in May at the University of Miami Rosenstiel School of Marine & Atmospheric Science. Charley’s rapid uptick was nothing. A Category 3 hurricane erupted in a large body of water at the school minutes after Marks’ presentation and it took barely a minute or two to whip itself to fearsome speed.

But this hurricane was controlled with the press of a button in a room cluttered with scientific equipment as Brian Haus turned still water into stormy water inside an Air Sea Interaction Saltwater Tank (ASIST). The long, rectangular tank can simulate a hurricane blowing winds equivalent to 120 miles-per-hour over the briny water in the tank.

“We can study the processes that occur at the air-sea interface,” said Haus, an associate professor in the division of applied physics at UM’s Rosenstiel School. “What’s unique about this facility is that we can generate a Category 3 hurricane and can make detailed process studies where air meets water and we can learn about how heat is transferred and how drag of the surface on the wind are important parameters that can go into hurricane forecast modeling.”

These forecast models have gotten better, Marks said. “Over the last 2 1/2 years we’ve seen acceleration.”

Today, the National Hurricane Center can forecast out to 48 hours, compared with a 24-hour reach in the mid 1970s. “We’re forecasting better at 48 hours now than we were at 24 hours then,” said Bill Read, director of the National Hurricane Center.

Super computers and hurricane-hunting flights that drop probes into storms to measure atmospheric pressure, wind speed and other indicators have helped improve models. These technological enhancements have helped forecasters give the public more advanced warning of pending storms and their tracks.

“We want to extend our forecast reliability out to seven days and want the same skill we have now at five days,” Marks said. That’s still 10 years in the future, he added. “The focus now is taking the data and making better use of it to improve the model patterns.”

Researchers from NOAA, the National Hurricane Center and UM’s Rosenstiel School recently gathered at the Science Before (and After) the Storm symposium to discuss advancements in hurricane model forecasting and new tools to study hurricanes, to, in effect, “give us a profile of the wind.”

The technology is helping scientists communicate better with the public. For instance, beginning with the initial weather disturbance of the 2011 season, which begins June 1 and runs through November, the National Hurricane Center will issue instant advisories on Facebook and Twitter.

This ability to personalize communications about threats could be a more important tool than it sounds, said Read. When the advisory hits the center’s webpage, a bot will automatically send out a feed to the social network sites.

One of the institutions’ main goals is to further study why storms intensify the way they do. Cone forecasts, the most popular measuring graphic used by the media, help predict where a storm might land when looking at the center line. These cones have gotten smaller and more precise by 16 percent at the five-day mark compared with 2003 during Hurricane Isabel, said John Cangialosi, a specialist with the National Hurricane Center.

But the study lags.

By 2020 researchers hope to have a day six- or seven-day forecast, up from today’s four- or five- day look and to indicate intensity changes. “Impact models get people to move. If we can come up with usable information—this is what is going to happen to you — we’ll get a better reaction from people,” Read said.

Most recently, these advancements in research tools have helped dispel a popular belief about the effect of global warming on the number and strength of storms, said David Nolan, a professor at the UM’s Rosenstiel.

“All other things equal, higher sea temperature means more hurricanes but in global warming all things are not equal. The atmosphere will warm with the surface. An unexpected result has been attained. Increasing sea temperature has had an opposite trend,” Nolan said, explaining that increased wind shears in the warmer climate pull hurricanes apart.

Studies found that it is “harder for hurricanes to form in a warmer world. This wind shear is more effective at suppressing storms in the warmer world,” Nolan said. “In the future world there will be fewer hurricanes but we believe the strongest storms will be stronger. It’s hard to say whether society wins or loses from that equation.”

Haus hopes that more information on such intensity and formation can come from continued research using the ASIST tank. Studies at UM on sea spray within the tank — “one of the real unknowns” — are ongoing as well as a study of interaction between wind and waves to coastal structures.