At 7pm on wednesday, 2/6/2013, the weather man is predicting a historical storm that is expected to hit the Northeastern U.S. during the weekend. People in Boston, New York, and other northeastern areas should get ready for it. In this video when meteorologist Chad Myers talked with host Erin Burnett, he used a football play to describe the weather system.
It is a great job that Chad explained the weather (and storm) forecasting in such a simple and understandable way. At the end, Erin and Chad has an even more interesting chat on our weather foresting models:
Chad: This is 36 hours before the storm even starts. That’s why I’m so wish-washing. I can’t say 1 to 12, because that was what the computer are. The computers are literarily for New York city could be 1 inch, mostly rain, and I’m saying. I will show you what is scary in one second. I will just hold on to the graphics here. Boston on this computer, 21 inches of snow. I think that’s a pretty good number, especially for the burghers. Look at this thing, this computer says, OMG, you know what, it is not going to warm up at all. The rain is not going to happen, it is going to be all snow, 23. So literarily we have 1 to 23 inches for snow possible for New York city. That’s why we have to wait for another computer run before we really understand what going to happen here. I hate to be wish-washing about it, but it is a big storm for somebody anyway!
Erin: But, but, … I guess I’m so confused that you seemed to me that we always knew what the weather would be, right? A few days in advance. Nowadays, every time there is a big storm, we have no idea. Could be big, could be little, what is going on? Why is this because … Climate change? I don’t know. What?
Chad (with smile): It is because over population of models and over thinking it. Now we have so many models. Look at this one, oh, look at this one. Before we had two, Elephant and GM that’s all we had. We look at one and we decide one or the other. Now we had like nine, so we don’t know which one to pick.
Erin: You know what, sometime the plenty of choices is not so great.
Chad: That’s right!
We know less and we are less certrain about our prediction because of “over population of models and over thinking”. NICE!
Watch out for the storm of computation power, good or bad.
Learning From Data 