Model Verification so far with Dennis (CMC, NAM garbage)

[Derecho]

With Stormtrakker I can calculate model errors; some problems as some of the globals I'm inputting with modeltrakker by hand, also the storm really isn't that old of a storm, some models haven't been run the whole time, but a general idea can be given. I'd cut and paste the rankings if I could...

But looking at 72 hour performance so far....

Of individual models that have been run a lot, surpsingly the UKMET is tops, with 85 mile error at 72 hours.

The CONU consensus model has 69 mile error at 72 hours.

The NHC forecasts have been GREAT, they have a tiny 64 mile 72 hour error, better than any individual model.


The NAM has a disastrous 171 mile error at 72 hours.

And literally the WORST model has been the Canadian, with an embarassing 247 mile error at 72 hours (due to a number of runs that had it going up eastern Florida.)

Literally the only model worse than the Canadian has been "XTRP" which isn't even a model, just extrapolated motion

Just showing I'm not kidding when I tell people the NAM and Canadian should be ignored

[jkt21787]

I said ignore the CMC and NAM when it showed SW FL hits, but I practically got knocked off the board by everyone.