Derek - addendum to the what is/isn't a seasonal forecast...

[Steve]

If you recall our disagreement over the NOAA's 'forecast' a few weeks ago, you will rememeber that I disputed what they were doing - range of numbers and probability scheme - as something that cannot fail and can be spun any way (60% chance of a landfall or 40% chance of no landfall in a region = you win either way). I referenced Accuweather's approach the last couple of years as a new type of model for seasonal forecasting. Well they've bettered it again. In previous years they'd forecast landfalls in 10 regions with numbers of landfalls and an intensity rating for those landfalls. The intensity rating was based on a range of intensity between .5 and 6 depending on whether an area was affected by a TD thru Cat-5.

This year, they have changed their ratings from 1 (TD) to 64 (Cat-5). They note some very interesting patterns that develop when using this type of system. Will they be close or right? Who knows? But they have a history over the last couple of years. If they attain a fairly-close-to-predicted score again this year, those interested in storms - especially those who forecast them - will have to stand up and take note that a true forecast system is available to the public.

Steve

[Derek Ortt]

If they show some long-term results (standard period being over 10 years), then they will deserve the credit. However, I have seen that they are not the only organization making these types of forecasts. http://www.hurricanealley.net also makes landfall forecasts.

[rainstorm]

looks like hurricane alley is going along with me saying the east coast is safe

[Steve]

Thanks for the link. I think their landfall scenarios look good overall. The only deficiency is that they are using range-based intensity at landfalls. Otherwise, I'd say this is a good analysis, and I like the way they arrive at their analogs.

Thanks again.

Steve

[cycloneye]

Good landfall anaylisis from them overall but now let's wait for the real thing to see if all of what they say pans out.

[Derecho]

Well they've bettered it again. In previous years they'd forecast landfalls in 10 regions with numbers of landfalls and an intensity rating for those landfalls.
Steve

You have stood up and taken notice, of course, that AccuSux/Bastardi, ironically, when it comes to forecasting track and intensity of individual storms, do so in the vaguest of vague hand-waving fashion, in a manner that is physically impossible to numerically quantify or compare to NHC?

[Steve]

That's not what I'm talking abuot. I'm talking about a seasonal forecast, not individual storms. As a company, their tropical stuff lacks, but you'd have to look at comments during those events. Let's put it this way, Joe B. did ultimately make the LA call for both Isidore and Lili when things were unknown. But that's a whole nother ball of wax.

Steve