If you’ve been watching weathercasts over the past few years, you’re probably hearing a lot more of the following: “the models show”, “the computer models indicate”, “the models show Tropical Storm Karen will go this way”, and so forth.
You’ve also probably figured out that when we’re talking about models, we’re not discussing anyone gracing magazine covers or surrounding Drew Carey on “The Price is Right”.
We’re talking about computer-generated weather forecasts that model what the atmosphere may do in the future.
In short, there are a handful of these programs that will take current weather data (not just what’s happening at the ground, but throughout the atmosphere) and run that data through a lot of equations that meteorologists use to predict all atmospheric motion. The result is a forecast of where all the atmospheric features (fronts, high and low pressure systems, etc.) will be in the future.
Models have been around for years. Increases in both computer technology and in our understanding of the atmosphere have led to increased accuracy in these models…translating to better weather information for all of us.
HOWEVER: these models are not perfect. And it’s likely they never will be.
There are two main reasons for this. First, the equations themselves don’t always represent a “perfect” understanding of what the atmosphere does. This is simply due to some limitations of our understanding of certain atmospheric processes.
And secondly, the models will always be limited by less than perfect data input. For the models to do their best job, they need current atmospheric readings from everywhere. While we can tell you what current weather conditions are at GSP, do you know what the temperature, humidity, and wind speed are at 10,000 feet over the airport? How about 20,000 or 30,000 feet? Using several methods (aircraft, satellites, weather balloons) we can make a really good guess as to what is happening up there right now, but we’re not 100% certain in many cases.
And if we’re not 100% certain about this data, we can’t be 100% certain about the output. It’s likely we’ll never know with full certainty about everything that’s going on in the atmosphere at any given moment. So the models will never be 100% accurate, all the time.
This is where humans come in. Good meteorologists use their knowledge of weather in a particular area and their experience with these models to come up with a forecast. Sometimes that forecast will in fact match what the models show. Sometimes the best forecast is not what any of the models are indicating.
The models are also called “computer guidance”. Which is exactly the correct term; they should be guiding the meteorologist, not necessarily dictating the complete forecast. Don’t confuse good meteorology with someone’s constant talk about the models…they may be letting those models do all the work.
You do want to hear a meteorologist use their own thought process. You want a meteorologist who will say “here’s what I’m forecasting” or “here’s what I think is going to happen.”
You want to see a meteorologist explain the possibilities if there’s doubt in what the models are showing.
You never want to hear a meteorologist say that something will happen just because the models say so.
I’d like to think I’m practicing what I preach. I hope someone lets me know (nicely!) if I stray from that path.
250 International Dr.
Can't find something?
All content © Copyright 2000 - 2014 Media General Communications Holdings, LLC. A Media General Company.