Ryan Maue graciously made the figure above for me (click on it to enlarge). It shows an number of things related to North Atlantic hurricane activity. First, the curved line of red dots shows the running climatological total of Accumulated Cyclone Energy (ACE, a measure of intensity) by date. The stair-stepped black line of dots (look closely!) shows this year's running total up until yesterday. The red and black dashed lines and the blue box show the UK MetOffice, Gray/Klotzbach and NOAA forecasts for the predicted 2010 seasonal ACE totals. For comparison, on this date in 2005 the season total ACE was already above 70!
It is early to say much about forecast skill. However, we will certainly be able to say something meaningful within a few weeks, such as what sort of events will be necessary for the forecast to show skill (as compared to climatology, which is an uncontroversial naive baseline here, though some might argue for a shorter climatological period). The hurricane season peaks (using ACE) on September 11. This example shows that it is possible to begin to evaluate forecast skill even before a forecast period is up -- conditionally and/or probabilistically of course.
If Ryan will humor my requests for updates (he extracts a mean price, Thanks Ryan!;-) I'll ask him to add new data to this graph as the next days and weeks progress. Obviously, this year's forecast of a very active season is being watched closely. At this moment there are some hints of activity in the Atlantic. Stay tuned.
0 comments:
Post a Comment