Friday, December 27, 2013

The Winter Storm that Never Was

Ah, winter. Cold weather, numerous holidays, and...weekly winter storms that never actually occur. Ok, so obviously winter storms do happen on some weeks. It's those storms-that-never-were that have been getting a lot of attention lately.

I would venture to say that come late November or early December and continuing through late-spring, a winter storm can be found at some point in any forecast cycle of the GFS. The problem is, it is often found beyond the 7-day / 180hr timeframe and it often doesn't play out like the GFS (or other models) claims it will. As Meteorologists, we know this. But, let's be honest, it is really fun to look at the GFS beyond 180 hrs to see the "eye candy" as I've heard many call it. I am not a doctor, but I hear that candy isn't always the best for you and yet it is still being passed out on a weekly basis.

My concern is that as Meteorologists we are doing a disservice to the ones it is our duty to serve. Can I go as far as saying that it is like a dentist passing out toothbrushes AND candy? Some great maps and charts ("toothbrushes") are spread around prior to an upcoming event, but then those end-of-the-world snow maps (ie. the candy) are mixed in. I am no authority in the weather community, but I just feel like we have to think of the potential consequences of passing that candy around.

In my opinion, one of our core responsibilities (regardless of sector) is to serve our clients, whether that be the public, a paying customer, or a business. When we post, share, or link to a model's "blizzard-of-the-year" output at 384 hours, what are we accomplishing?

I read a great blog post recently by Victor Gensini on this very issue. Check out his blog post HERE. He hit on a lot of the core issues with these posts, so I won't rehash that. But, a few thoughts to add...

1) If a model solution matches up closely with your own forecast, then fine. Maybe you don't have a fancy map-making tool and it is just easier to post a model's output with some sort of caption that clarifies the map shown is a model forecast, but that it matches closely with your thinking. But, how many of us actually make detailed forecasts for our clients 384 hours out? I know some do, but for the rest, maybe we should wait until the model's output is within our forecast period.

2) Post a map of your own, if possible. Many NWS offices, media outlets, and private sector company's do this. Right or wrong, it is their personalized forecast that incorporates analysis of many model solutions by the forecaster(s) on duty. I believe our clients are best-served when WE analyze each model's solution and come up with a best-guess forecast as opposed to posting all the solutions. And oftentimes, it seems like the worst-case-scenario maps are shown which just doesn't seem to be a good service (unless, of course, you believe the worst-case will happen with a particular event...and those DO happen).

3) I am not against looking at model output beyond 180 hrs. I think it does have some use. Personally, I like to look at longer-range output sometimes just to get an idea of what the weather pattern might be like beyond DY 7. Maybe the GFS and the ECMWF show a deep trough developing across the eastern US 10 days may not happen exactly that way, but if a certain pattern is advertised consistently from run to run, maybe it will help with confidence in your forecast once that pattern comes into play within your forecast period.

4) The idea of posting images and explaining them or comparing with other model solutions is an interesting one. I have mixed feelings about this. I like the teaching motivation and I think it is a good one. But, even good intentions can go awry. I'm not sure yet if I am against this idea, but it might be an interesting case study to see whether a majority of the people reading the post focus on the learning aspect or whether they focus on the fact that one model is showing a huge storm.

5) I realize that a lot of hyped posts are done by those who are not actual Meteorologists. To be clear, I don't have something against people who make such posts. The internet and social media are great, but it does make all kinds of weather info available / shareable (for better or worse). Personally, I can't hold a non-Met responsible for those posts (although, I do wish they would stop). But, I believe that we, as Meteorologists (just like dentists with candy) should be held to a higher standard when it comes to passing around worst-case scenario posts that we don't actually believe will happen.

My desire is to see the weather community at large continue to serve others as effectively as possible. If posting certain model output isn't effective, maybe those posts should be as rare as the occurrence of those blizzards shown 2 weeks out...

No comments:

Post a Comment