The National Oceanic and Atmospheric Administration’s probable track cone for tropical cyclones will shrink slightly this year, thanks to more accurate forecasting.
The National Hurricane Center’s James Franklin, the branch chief, Hurricane Specialist Unit, told the Caymanian Compass that official track errors have been improving at a relatively steady pace over the past two decades.
‘[The track errors] are now half of what they were 15 years ago,’ he said.
The reduction in tracking errors is mostly due to advances in computer modelling, Mr. Franklin said. Today’s computer modelling programmes are more sophisticated, involve the use of satellite observations, and the computers themselves are faster, allowing the models to depict more detail.
As a result, the errors in tracking are getting smaller at every forecast time period. The NHC’s official track forecasts for the Atlantic basin in 2008 set records for accuracy at all times from 12 to 120 hours out.
The NHC’s mean track errors for tropical cyclone forecasts last year ranged from 28 nautical miles at 12 hours to 192 nautical miles at 120 hours, according the 2008 National Hurricane Center Forecast Verification Report, which Mr. Franklin authored.
The size of the probable track cone reflects official forecast error data over the past five years.
‘Since the cone is based on past NHC errors, it continues to shrink as the official errors get smaller,’ Mr. Franklin said.
For 2009, the probable track circle in the Atlantic basin will have a 36 nm radius at 12 hours; a 62 nm radius at 24 hours; and up to 302 nm radius at 120 hours.
Still some misses
Despite the improvements on tracking errors, some storms still confound the forecasters and the computer models.
In 2008, Tropical Storm Josephine and Hurricane Omar both had larger than average forecast errors.
Tropical cyclone intensity forecasts also remain troublesome. Over the past 20 years, official intensity errors have remained essentially unchanged, Mr. Franklin’s report stated. In 2008, intensity forecasts for Hurricanes Gustav, Omar and Paloma proved particularly problematic.
In addition, Mr. Franklin said forecasting the tracks of certain types of storms – like the Cape Verde storms that come off Africa – are easier than others.
‘It is true that the long-track Cape Verde systems are easier to forecast than late-season storms,’ he said. ‘Steering currents over the tropical Atlantic during the heart of hurricane season are less complicated than those surrounding late season storms, which often move northward out of the Caribbean and interact with strong weather systems coming out of the middle latitudes.’
During the National Hurricane Conference in Austin, Texas last month, National Hurricane Center Director Bill Read talked about how the average forecast errors have reduced significantly for the five-day forecasts.
When the NHC extended their forecast cones from three to five days in 2003, the average error exceeded 370 nautical miles. Last year, it was almost half of that at 192 nautical miles. Because of the higher degree of confidence in forecasting where tropical cyclones are heading, even five days out, Mr. Read said he wanted the NHC to extend the forecast to seven days.
Mr. Franklin said this could happen within five years.
‘The seven-day forecast is something that we are already starting to work toward,’ he said. ‘Before we issued five-day forecasts, we had a two-year internal test period, and my guess is that we are at least a couple years away from beginning such an internal test period for the seven-day forecast.’