Looking for a weather forecast? Pick someone you trust

Don Schwenneker Image
Wednesday, September 13, 2017

Irma. It seems like we've been talking about it forever. It's actually been two weeks, but it seems like forever.

One of the reasons we started talking about it so soon? Some people chose to post one model, 10 days in advance. This particular post garnered over 4 million views and was shared over 166,000 times!!!!

It claimed that we would see 50 foot waves and 100+mph winds in North Carolina. We did not. It also asked right at the top to 'Follow' that person. Hopefully, you did not. It's what we call 'Click Bait'.

If you clicked the 'About' section on that person's profile, you found they have no particular weather training. If you listened to the video, the person claimed 'local media will not tell you this'.

I want you to know something. My first priority at work is you, the viewer. You are the reason I have a job and I take that responsibility VERY seriously. If I think a storm will hurt you, your family, or your property, I WILL tell you and not hold anything back.

There is nothing wrong with amateur meteorologists looking at models and sharing their thoughts. This isn't about me protecting my turf. But when someone is claiming I'm hiding information, then it gets personal.

And let's talk about the picture of the model that was shown. This particular visualization is made by Ventusky, a company out of the Czech Republic. To be honest, I really don't know much about the company. The graphics are impressive, but what are the parameters they use to generate them? It appears to be based off the German Model, ICON, with a little High Resolution modeling thrown in. But it is just one source. If the above poster had shown ANY other models that day, he would've seen a bunch of totally different solutions, and NONE of them hit North Carolina.

The two models I tend to look at first are the GFS (American Model) and the European Model and their various iterations. I will look at others as well (Canadian, NAVGEM, RPM, etc.). But 10 days out, there just isn't ONE model that can predict the exact location of a Hurricane that far away.

So if someone says there is one model that is right, I believe they're lying, or they want something, like new followers. With as fast as computers are getting, and more and more data being collected every day, we may not be far away from a model nailing a forecast 10+ days out. But for now, it just doesn't happen.

BTW, as we head toward winter, one model CANNOT predict a major snowstorm 10+ days out either, no matter what an amateur meteorologist posts on his/her Facebook page.

Bottom Line? Find someone you trust and follow them and listen what they have to say. If it's not me, that's OK. Really! Just make sure the person you follow is educated on how to use various models and what a forecast of 50 foot waves and 100mph wind really means to people. You wouldn't want someone to handle your money just because they are good at math. And PLEASE make sure aren't just looking to make you another follower...

Related Topics