Weather forecasts have come a long way over the past couple of decades. Meteorologists can give you a deadly accurate five-day forecast today, when forecasting the weather beyond a day was a feat 30 years ago. Forecasts today are accurate to a fault: people expect too much, and get angry and disappointed when their friendly local weatherperson can’t deliver.

Everyone jokes about meteorologists getting the forecast wrong more often than not (“it’s the only job where you can be wrong all the time and still get paid!”), but it’s the other way around: forecasts are scary accurate these days. Big busts are rare, and even forecasts people perceive as having busted are still pretty right on—remember the historic blizzard New York City was supposed to see at the end of January that people screamed and moaned about because it didn’t pan out exactly as forecast? The track of the storm moved just a couple of miles farther east than anticipated, keeping the western extent of those huge, feet-deep snowfall totals just a few miles east of the city.

But hey, if it didn’t happen here, it didn’t happen at all.

We have a couple of weather models that excel at predicting convection (thunderstorms), and sometimes they’re eerily accurate in predicting the location, shape, and strength of organized complexes of thunderstorms. Here’s an example from this past March, showing a line of supercell thunderstorms powering through Missouri.

This was the early afternoon run of the High-Resolution Rapid Refresh (HRRR) model that day:

And this is what the radar looked like five hours after that model run, showing almost exactly what the model predicted:


One of the problems with such accurate forecasts is that people expect too much from them. They expect accuracy that we haven’t achieved just yet, and it’s both a frustrating and dangerous side effect of advances in meteorology.

Now that we’re entering the warm season, pop-up thunderstorms will dominate the landscape between derechos and hurricanes. These pop-up thunderstorms go by many names—popcorn, garden-variety, pulse—but it’s all the same principle: these storms bubble up in a seemingly-random spot, rage on for an hour or so, then die off and give way to what’s left of the day’s sunshine.

You can forecast the environment in which these storms will form, but unless you’re in Mobile, Alabama, watching the sea breeze push ashore, it’s very hard to predict exactly where these pop-up thunderstorms will form. The general public doesn’t understand these limitations, so they bombard meteorologists and weather nerds with the same line of impossible-to-answer questions: Will I see a thunderstorm today? Will it rain over Jaycee Park at 3:30 PM on Sunday? Will I be okay walking to the bus stop this evening, or should I take my car?

Sometimes we can answer these questions with ease. You can sit there and watch the radar to figure out roughly what time squall lines or other complexes of organized thunderstorms will arrive at your location. This is how various weather apps can tell you it’ll start raining in exactly 36 minutes—they have algorithms that analyze radar data and figure out how fast showers and thunderstorms are moving, using this data to figure out how long it’ll take for precipitation to reach your location. When it comes to these summertime pop-up storms, they form in random locations and barely move, so predicting when and where they’ll form is virtually impossible with today’s technology.

If you go on Facebook and ask your local broadcast meteorologist what time it’ll start raining, unless it’s an organized storm system heading your way, you probably won’t get a straight answer, because there really isn’t one. The underlying problem here is that saying “I don’t know” is stigmatized. We’ve convinced ourselves that “I don’t know” is the worst thing someone can admit.

When a meteorologist says that he or she doesn’t know, it’s even worse. The pinpoint, personalized forecasts the public demands are dangerous because they create absurdly high expectations that ultimately erode overall trust in forecasts because people see “I don’t know” as some sort of proof that all forecasts are guesses or just plain wrong. In the warped mindset of so many people, the entire science of meteorology is a fraud because Skippy McHail at WXXZ can’t tell them exactly what time a thunderstorm will rain on little Jack’s crucial soccer game, if at all.

It’s summer. Thunderstorms can bubble up with little or no notice. A gully-washer could flood the town next to yours while it’s bone dry in your front yard. You could lose a tree to a gust of wind while two streets away, it was a quick burst of rain and nothing more. Forecasts have advanced by leaps and bounds over the years, but we still have a way to go before meteorologists can personalize your pop-up thunderstorm experience.

We all want answers ahead of time, but summertime weather isn’t always conducive to the instant I-want-to-know-now luxury that we’ve gotten used to. Sometimes we have to take matters into our own hands when it comes to staying out of the rain. It sounds flip, but when pop-up thunderstorms are possible, the best thing to do is to keep looking at the clouds (if you see towering, dark clouds looming overhead, the odds of rain are good!) and checking the radar to see what’s going on nearby. There are dozens of reliable apps and programs that provide radar imagery, not to mention great websites like Wunderground and the National Weather Service.

[Images: AP, GREarth]

You can follow the author on Twitter or send him an email.