Published on: July 6, 2008 at 10:25 PM

The National Hurricane Center has released its annual Forecast Verification Report for 2007. Atlantic track forecasts in 2007 established new records for accuracy in the 36-96 hour period. Official track forecasts were better than the consensus models, but not as good as the best dynamic models. Intensity forecasts were less accurate than pervious averages.

Track forecast error is defined as the distance between the forecast position and the best track position determined after the season. Skill is assessed by comparing the track forecast error against the error from the CLIPER5 model, which uses climatology and persistence, but no information about the atmosphere. Intensity error is the difference between forecast and actually observed intensity according to the best track. Skill is determined by comparing the difference against the DSHIFOR5 model.

Mean track errors in 2007 were 33 n mi at 12 hours and 258 n mi at 120 hours. These values are smaller than the previous 5 year average by 7%-24%. The 36-96 hours values were new record lows. Since 1990, the 24-73h track forecast errors have been reduced by more than 50%. Track forecast skill in 2007 was comparable to the previous 5 year levels. Increases in forecast skill have changed little since the significant improvements at the beginning of the decade.

The Federal Government sets goals for performance under the Government Performance and Results Act of 1993. The goal for the 48 hr track forecast in 2007 was 110 n mi. The actual result was 86 n mi.

Mean intensity forecast errors in 2007 ranged from 8 kt at 12 hours to 30 kt at 96 and 120 h. These were larger than the 5 year means, by 25% or more at the longer periods. Rapid strengthening scenarios led to the large errors, and there were many instances of rapid intensification in 2007.

Forecasts of tropical cyclone formation were done internally for the first time in 2007. The experiment was successful enough that the NHC team decided to begin releasing the forecasts to public as an experiment in 2008. They can now be found on the NHC website.

So, official NHC track forecasts are continue to get better, but intensity forecasting is not improving at the same rate. No matter what, we salute the fine heroes and heroines at the National Hurricane Center.