The cone of uncertainty continues to shrink
March 1, 2011
In a note on “Product changes for the 2011 hurricane season” , the NHC revealed the sizes of the cones of uncertainty that will be used in 2011 forecasts. They are a fair bit smaller than last season’s at the extended forecast times.
Recall that the forecast cone is defined as representing the probable track of a storm and is formed by circles sized such that they contain 2/3 of the forecast error for the time periods averaged over the past five years. The NHC first started using this method in 2007; at the time the size of the circle for 120 hours had a radius of ~325 nautical miles; this year it will be about 85 miles smaller. Below are the sizes for 2011 (Atlantic basin), with the change from last year in parentheses:
12 h: 36 nm (-)
24 h: 62 nm(-)
36h: 99 nm (-6)
48h: 98 nm (-10)
72h: 144 nm (-17)
96h: 190 nm (-30)
120h: 239 nm (-46)
When the 2010 forecast verification (the document reviewing the NHC’s forecast performance for the season) is released, it will be interesting to whether/how much of the shrinkage is from 2010 as oppposed to the 2005 year season “rolling off” the average. As the error trend chart shows, 2005 was the worst of the past five with regards to average forecast track accuracy.