My Stuff

Source

Predicting weather more than holding finger in the air

Today's question:

How is it they can predict that tomorrow's temperature will be 63 or 82 or whatever? Why not 60 or 65 or 80? How can they guess an exact number, and how often do they hit it correctly?

I always feel kind of bad when people complain about the forecasts they hear on radio or TV or read on their computers or in the paper. I like real meteorologists. If I wasn't as dumb as I am, I might have taken up meteorology. Or maybe archaeology. I think a lot of "ology" sciences are pretty cool as long as they don't involve dealing with real people. Like psychology.

Anyway, weather forecasting these days is pretty good and pretty scientific, but it still is not a 100 percent exact science. One thing they use is something called persistence forecasting. That pretty much means that if it was 75 today, it's probably going to be around 75 tomorrow.

That obviously doesn't work for the weather we've had lately — temperatures dropping by 20 or so degrees overnight. (Isn't that great?)

So, forecasters also use stuff such a seasonal averages and computer models, and they look at cloud cover and wind speeds and barometric pressure and at approaching high-pressure or low-pressure systems and see what kind of weather they are bringing and what's happening as they move along. Then, forecasters make the best predictions they can.

I am told that the National Weather Service and the American Meteorological Service figure that if their predictions are within 4 degrees of the actual temperature, they are on target. And I am told that forecasters from those two groups are on average correct 80 percent of the time.

Unless, of course, I had taken up meteorology. That might have lowered the average.

Reach Thompson at clay.thompson@arizonarepublic.com or 602-444-8612.