Accuweather says NHC is much less confident than they are

[Anonymous]

Originally posted by Ryan McGinnis
The irony, of course, is that a good portion of the area shaded red (90+%) did not get sustained hurricane winds, as the storm landfell east of their projection. I dunno about you, but I like to be REAL sure about something before I say 90+%. Maybe that's why the experimental forecast from the NHC went with such low probs that far out?

If Accuweather knew half as much as they thought they did, they'd be dangerous.

Don't forget, too, that there was a 12 hour difference in the time of issuance. These jokers were deliberately misleading their audience.

 

[Robert Dewey]

Originally posted by Anonymous
Actually, he wasn't just hyping himself and \"accu\"wx, he was going a step further and bashing the NHC...

Okay, I see... The link to the article that is provided at the beginning of this thread shows nothing about bashing NHC - rather just a link to Accuweather's various forecasts. I assumed that the person bashing NHC was a PR person who wrote an article, not Joe. Given that, I too think an appology is in order.

Even so, I'll still subsribe to Joe's winter weather outlooks. It used to be free, but then Accuweather started charging for it - even Joe had some "beef" with that at first (until they waived a bigger paycheck in his face, I assume). The guy has come up with some good theories that seem to work year after year, and his analysis of the ensembles and global patterns are pretty well thought out. To top it off, I have talked to the guy via e-mail regarding certain winter weather events, and he never tried to conceal any mis-forecasts, and generally seemed like a nice guy (probably when the boss isn't looking). That's why I just hated to see him get bashed or flamed...

But, to each their own I suppose...

 

[Gest Aaron]

i guess what makes me kind of mad about this whole accuweather thing is that, it seems to me, they base a lot of their stuff off of noaa and the nhc and then alter it a bit to make it look like their own(kinda like TWC). What i mean is, whenever the NHC changes a prediction, they make a forecast "update". Whever NOAA issues a tornado warning for a tornado in a spiral band, they take it as their own.And i mean, lets face it, without NOAA and the NHC, these guys wouldnt be getting all of the aircraft info, and stuff like that. I believe it was steve miller that said, why bite the hand that feeds you?

 

[David Schuttler]

Accu-weather is on/was on Larry King , watching replay now

 

[Jeff Snyder]

Yeah, the Accuweather guy said that there needs to be some reconsideration of NHC policies in regards to issues watches/warnings. He said watches and/or warnings are only issued 24/36 hours ahead of landfall, wondering if it gives officials and people enough time to evacuate. The NHC issued the hurricane watch (originally west of Cameron) at 7pm Wednesday (~55 hours before landfall), granted the track shifted eastward with time and areas that experienced the eastern part of the storm were originally in the tropical storm watch. I don't think we can issue watches/warnings too much farther ahead of time based on the fact that track/intensity forecasts just aren't (usually) good enough to provide watches or warnings with much confidence. Same reason why we don't see tornado warnings issued 2 hours before a tornadic supercell comes through, or why we don't see severe thunderstorm watches issued 6-7 hours prior to expected initiation of severe thunderstorms.

 

[Mike Hardiman]

Originally posted by Tim Vasquez
I'm puzzled why no one has done a study on them to measure their (Accuweather's) skill vs. that of the NWS.
... snip ...
Probably not a big enough sample size to measure hurricane accuracy though.

Tim, I have wondered the same thing. It would be fairly easy to glean the numbers for perhaps a dozen or so major cities (and even outlying areas) and just crunch the numbers. It would make a good paper for the NWA Digest, should they ever get around to publishing another journal (2 years behind now?).

Perhaps the number one reason nobody has taken on a project like this is they don't want to get sued. I can almost guarantee that my employer (NWS) certainly would not allow me to publish such a paper while citing any affiliation with them. Management, especially at HQ, doesn't seem willing to go to bat for the NWS.

Hurricane skill would be more difficult. Sample size wouldn't be an issue if you just wanted to prove/disprove one of Accweather's claims to having more skill with one particular storm than NHC. Problem is, their cheesy online graphics very often show a gigantic hurricane symbol covering about a 100 mile stretch of coastline at their landfall point. They don't post raw coordinates as far as I can tell.

Each source (AccuWx, NWS) have had their victories and falls in the past... however Accuweather is famous for tooting their horn saying they did a better job (even when it is simply FALSE) while completely ignoring their numerous failures. They also do not divulge any objective verification statistics to the meteorological community to back up their claims.

Of course, my opinions are my own, not my employer's, and I am typing this at home, at an ungodly hour, on my own computer, etc, etc.

-Mike

 

[Evan Stubbart]

Hawaii

I remember going through Iniki in Hawaii. I lived on Oahu and AccuWeather said that the eye was going to pass right over our island while NHC, NOAA all said that it was going to go further west and miss Oahu (which it did), but AccuWeather kept saying that they were 100%. What a joke...