Mariner's 1-2-3 rule and the cone of uncertainty for Katia

Hurricane tracks for the 2005 season, from NOAA

Updated: 9/5/11--new graphic of Katia's projected path added.

I've been curious about the likely track that Hurricane Katia will take, and thought that it would be interesting to look at the tracks that other hurricanes have taken when the have been about where Katia is today (longitude about 55 W, latitude about 20 N.)  It didn't take long to figure out that hurricanes seem to be distinct individuals with a mind of their own! No wonder hurricane forcasting is so difficult! One NOAA scientist interviewed this week commented that they had had the track of Irene predicted fairly accurately, but that getting the intensity right is the difficult part.

Katia and its project path, Saturday Sept. 3 to Thurs Sept. 8
NOAA

It's difficult to control formatting in Blogger, but I've tried to make these two images similar in latitude scale. Notice how the cones of uncertainty for Katia and Tropical Storm Lee in the Gulf of Mexico are just overlapping.

Trajectory predicted 9/05/11
NOAA

The "Mariner's 1-2-3 Rule" says that the National Hurricane Center forecasts have uncertainties of 100-200-300 nautical miles at times ahead of 24-48-72 hours respectively. This rule determines the basic "cone of uncertainty" that you see in all the forecasts. In practice, it looks like NOAA scientists do something slightly more complicated. For a five-day forecast, they place a set of imaginary circles along the forecast track at the 12, 24, 36, 48, 72, 96, and 120 h positions. The size of each circle is set so that "it encloses 67% of the previous five years official forecast errors."  It is broadened beyond the strict 1-2-3 rule to reflect estimated boundaries of the maximum tropical storm force winds (34 knots) at each of these times. The entire track of a tropical cyclone remains within the cone roughly 60-70% of the time.

More information is available here.