You know the drill. It’s a steamy July morning, and you’re headed out for the day. You open up your go-to weather app to find out if you need to lug your umbrella around. The forecast says, “Sunny with a 30 percent chance of thunderstorms.” What does that mean, exactly? If you choose not to bring your umbrella, you have a one in three chance you’ll get drenched darting from your car back to the office after lunch? You may not like those odds. Better bring the umbrella, just in case.

It turns out the business of predicting the weather is not quite as precise as you might like. WRAL-TV Chief Meteorologist Greg Fishel is a self-described “weather geek” who says he could stare at weather models all day if he didn't have to go on camera at 6 o’clock and tell us what the weekend forecast is shaping up to be while we’re cooking dinner. But even with weather models and prediction tools that improve all the time (think hurricane tracking 20 years ago versus the hour-by-hour directional models of today), there are limitations.

While meteorologists are using calculus combined with historical data to make predictions on the weather, Fishel said the analysis will never be perfect. He pointed to problems such as:

• An insufficient number of weather stations around the world.
• Computers that can’t do calculus.
• Atmospheric processes that are poorly understood (for example, meteorologists can forecast the directional path of hurricanes but have a harder time understanding and predicting hurricane intensity).

In an industry where TV consultants are advising local stations to cut back on the weather segment and keep it simple, Fishel said he and his peers need to be honest with people about what they know and what they don’t know and tell them why. And that’s complicated.

So, until they figure it out, you better bring your umbrella.