Friday, December 24, 2010

If Only They Could Take a Retest -- Poor Grades for All

Ouch. It was a tough outing for the local weather mavens for the snowstorm just concluding. Not that we really needed a reminder of the fallibility of forecasters and their much-acclaimed weather models, but we certainly got one. A storm that had been anticipated for almost a week was essentially pronounced dead just 24 hours prior to its ultimate beginning. To borrow from Mark Twain, reports of the storm's demise were greatly exaggerated. Here are the grades for the Xmas Eve Eve storm of 2010. Remember to follow us on Twitter.

Piles of snow surround this golden retriever/cocker spaniel mix (courtesy Fuzzy Mutt Photos)
WCCO: F+ Declared "little or no snow from metro north and east" on Tuesday with "an extended period of dry weather ahead."
KSTP: F Not only missed the boat like everyone else, but also had the worst prediction on Thursday morning.
Fox: D+ Fox deserves a bit of credit for holding on to a forecast of "maybe up to an inch of snow here in the metro" on Tuesday night when others had written off the storm.
KARE: F On Tuesday night, KARE said, "If we see any snow, it would be in the southwest metro and be less than an inch.
Star Tribune: D First suggested the re-emergence of the storm on Wednesday night and came closest to nailing the forecast with its predictions on Thursday morning.
MPR: F Had the snow going to the southwest of the metro. And when the snow was back in the picture, their predictions for the metro were not clear.
NWS: F While they never took snow out of the forecast (maintaining a 30-40% probability), there was no mention of accumulations through Tuesday night. The introduction of a weather advisory was too late to serve its "heads up" purpose.

7 comments:

  1. This one was really tough to forecast well ahead of time. The models were a failure. Next storm up is a possible monster storm beginning on News Year's Eve. ALL the models agree someplace in the Upper Midwest is going to get clobbered, but good.

    ReplyDelete
  2. Perhaps you should grade the models and not the forecasters because they all report only what the models say.

    ReplyDelete
  3. Excellent point! It's the models, not the forecasters.

    ReplyDelete
  4. Your grading scale seems rather abitrary. You failed the forecasters for their forecast on Tuesday of an event that occurred on Thursday. The reasoning you give is quite vague, perhaps you could let us in on your methods in a bit more detail?

    --Kevin

    ReplyDelete
  5. Kevin,
    I don't use any scientific methods and it is somewhat arbitrary. I give weight (in my admittedly subjective way) to the forecast in the days approaching the storm (hence Tuesday's forecast for Thursday becomes relevant). I'm trying to provide some judgment for weather outlets that make claims early on that are not substantiated as the weather event occurs. The grading criteria link at the top of the site provides a little more background.

    Thanks for your input!

    ReplyDelete
  6. Thanks for the reply, Bill. I can appreciate "one person's opinion" just as much as the next. I merely don't want "one person's opinion" to become "another person's fact". I enjoy the site, keep it coming!

    --Kevin.

    ReplyDelete
  7. Thanks, Kevin! The site's for amusement purposes only, though I'm trying to be objective in my subjectivity. :-)

    ReplyDelete