New technique helps monitor fast-changing hurricanes near shore
When hurricane season kicks off on June 1, Wen-Chau Lee (EOL), Paul Harasti (UOP/VSP), and Michael Bell (EOL) will be watching with even more interest than usual. Effective this season, a technique they and colleagues developed becomes an official part of the forecasting toolkit at NOAA’s National Hurricane Center (NHC).
The new technique, called VORTRAC (Vortex Objective Radar Tracking and Circulation), provides details as often as every six minutes on hurricanes as they approach the Gulf and Atlantic coasts. Along with calculating the peak winds rotating around a hurricane and their distance from the storm center, VORTRAC also infers the barometric pressure in the eye of the hurricane, a very reliable index of its strength.
This downtown structure in Punta Gorda, Florida, suffered major damage during Hurricane Charley in August 2004. Parts of the Florida coast were caught by surprise when Charley’s strength increased rapidly as it neared land. Scientists applied VORTRAC retroactively and found that the technique would have accurately captured the storm’s quick burst in intensity. (Photo by Mark Wolfe, FEMA.)
Drawing on NOAA’s network of coastal Doppler radars, the system graphically displays real-time updates that will help forecasters monitor the hurricane’s central pressure and assess whether the storm is gathering or losing strength as it nears land.
“VORTRAC will enable forecasters, for the first time, to continuously monitor the trend in central pressure as a dangerous storm nears land,” says Wen-Chau. With the help of this tool, he adds, “Vulnerable communities can be better informed of sudden changes in hurricane intensity.”
Fast changes, big problems
To monitor the strength of a landfalling hurricane, forecasters now rely on aircraft to measure winds and drop instrument packages that gather data on winds and pressure. But due to refueling, personnel schedules, and other flight logistics, there can be gaps of several hours between aircraft missions. This means that coastal communities may not be swiftly alerted to changes in approaching hurricanes.
VORTRAC uses the Doppler radar network established by NOAA in the 1990s. About 20 of these radars are scattered along the Gulf and Atlantic coastlines from Texas to Maine. Each radar can measure winds blowing toward or away from it, but no single radar could provide an estimate of a hurricane’s rotational winds and central pressure until now.
Forecasters using VORTRAC can monitor the trend of hurricane intensity each time a NOAA Doppler radar scans the storm, which can be as often as every six minutes or so. Without a technique like VORTRAC, forecasters would need at least two coastal radars in close proximity to each other in order to obtain the same information. But most of the network’s radars are too far apart to qualify.
Last year, Hurricane Humberto struck near Port Arthur, Texas, after unexpectedly strengthening from a tropical depression to a hurricane in less than 19 hours. In 2004, parts of Florida’s southwest coast were caught by surprise when Hurricane Charley’s top winds increased from 110 to 145 miles (177 to 233 kilometers) per hour in just six hours as the storm neared land. For both Humberto and Charley, the rapid intensification was greater than predicted.
Wen-Chau and Michael applied VORTRAC retroactively to the two hurricanes and found that the technique would have accurately captured their quick bursts in intensity. “Our research shows that this technique can capture sudden intensity changes in potentially dangerous hurricanes,” says Michael says.
The roots of VORTRAC go back to the early 1990s, when Wen-Chau and Ben Jong-Dao Jou (National Taiwan University) jointly developed a software routine—the ground-based velocity track display (GBVTD)—that can extract hurricane circulation using data from a single Doppler radar. Paul began working on similar problems at the University of Toronto in 1994 and continued while he was a visiting scientist at NHC, NCAR, and the Naval Research Laboratory in the early 2000s.
Paul developed a separate single-radar technique—Hurricane Volume Velocity Processing—to estimate the primary hurricane circulation and the vertical variations of the environmental winds. The component of those winds that blow perpendicular to the radar beam cannot be resolved by GBVTD, Paul notes. If not included, he adds, those winds can produce a significant bias in estimating the central pressure of the hurricane.
Eventually, Paul and Wen-Chau joined forces. “We thought we should get together and have a combined method,” says Paul, who now works as a VSP project scientist at NRL. Paul, Wen-Chau, and Michael worked with Lisa Mauger, a former student assistant in EOL, to merge the techniques and develop a graphical user interface that clearly displays real-time hurricane updates.
After years of further refinement and testing, supported mainly by NSF and NOAA, VORTRAC is now ready for prime time, which makes Paul happy. “It’s a dream come true, really. By merging several techniques, we can now provide a missing link in short-term hurricane prediction.”
Science from SOARS
VORTRAC was evaluated at NHC through the multiagency Joint Hurricane Testbed (JHT), which carefully examines many techniques that could lead to improved hurricane forecasts.
Shirley Murillo, one of the first SOARS protégés, is now the JHT’s administrator. Shirley has a particular interest in VORTRAC, because during her years in SOARS she worked on improving an algorithm for locating a hurricane’s center in radar data. Wen-Chau served as her science advisor, and Michael brought Shirley’s work forward after she completed SOARS, producing the robust and reliable center-finding algorithm that is a key part of VORTRAC.
“It’s very rewarding to have worked on this research and seen it evolve into a tool that is now accepted into the operational forecasting environment,” says Shirley.
Forecasters can monitor hurricanes with VORTRAC for typically several hours before landfall, and sometimes as much as a day or longer. The specific monitoring time for each hurricane depends on its speed, trajectory, and size. Each NOAA Doppler radar can sample conditions out to about 120 miles (193 kilometers), so slower-moving hurricanes will spend more time closer to shore, where VORTRAC can analyze them.
In time, VORTRAC may also help improve long-range hurricane forecasts by using data from airborne Doppler radars or spaceborne radars to produce detailed information about a hurricane that is far out to sea. Forecasters could input the data to computer models to improve three- and five-day forecasts.
This display from a test involving Hurricane Humberto (2007) shows the usefulness of the VORTRAC system in monitoring Humberto as it quickly intensified before striking near the Texas/Louisiana coast early on September 13. The VORTRAC estimates agreed closely with Humberto’s actual central pressure, as measured by Air Force reconnaissance planes just before landfall. VORTRAC also captured the storm’s rapid intensification and indicates that the pressure began falling quickly near the end of a four-hour window when no aircraft data could be obtained. (Illustration courtesy Michael Bell.)