Why is Apple's weather app inaccurate?

Why is Apple's weather app inaccurate?

For the past seven weekends in New York City, it has rained. Never during the week. Never for just one hour. Never uncomfortable. At this point, it's as if the stormy weather has become fully aware, and knows the exact time on Friday to start ruining New Yorkers' plans.

During this time, this constant rain over the weekend only confirmed that Apple's weather app is largely useless. Personally, I've learned that the app can't differentiate between “light rain” and “rain,” the percentages it posts seem bogus, and I never trust it when it tells you when the rain will stop. I'm not alone. My friends and coworkers also have different stories about how the app let them down, or how sometimes it just doesn't work. Some even talk about Dark Sky, a weather forecasting app acquired by Apple in 2020, with a wistful, wistful sadness, like a lost love. Apple says Dark Sky's most beloved features have been integrated into its app, but Dark Sky fans aren't convinced. They say things were different then. Things were better.

My growing frustration led me to find out why Apple's weather app smells so bad. As I spoke with experts, I took comfort in the fact that there was an actual reason—algorithms specifically—to bother me. It's okay to be angry about something. But in my research I also discovered a new appreciation for local meteorologists and more about weather and weather forecasting than I had initially planned.

What we mean when we say Apple's weather app stinks

My serious complaint about Apple's Weather app is that it won't give me a straight answer when it comes to rain. Rain means wet socks, puddles, and moisture in my clothes hanging around all day. It also means dealing with people who say, “Oh, we needed that” with a polite smile.

Rain sucks.

My needs are simple: I want to know if it will rain, how much it will rain, when it will start, and when it will stop. Ideally, I wouldn't want to go outside to check if it was raining, otherwise why would I have a powerful computer in my hand if it couldn't tell me things that were happening around me?

Is this guy's weather app wrong? It depends on where you live!
Jaap Arens/NoorPhoto via Getty Images

“Apple's weather app isn't great at detail,” says John Hominuk, the meteorologist behind the app. New York Metro Weather. Homenuk has gained a loyal following in New York City for his accurate and cheerful daily weather forecasts. “Unfortunately, details are what we need if we're planning our lives. 'Do I need a jacket tonight?' Is it going to rain when I go sit on the roof later? She struggles with that kind of thing.”

Apple's Weather app, and weather apps in general, work using algorithms to interpret data — weather models, location, and current observations — drawn from different sources, Hominuk explained to me. Other experts I spoke to said the apps don't reveal what data they use nor how frequently the data comes from, which can lead to inaccurate readings.

These algorithms also have limitations. In weather forecasting, these limits arise because these equations rely on models that meteorologists understand to be imperfect.

“There's one big model that's used not only in applications, but also in weather data across the United States. It's called the GFS, the Global Forecast System,” Hominuk said, adding that the GFS tends to be wrong about speed, and sometimes predicts storms will go out to Sea and out of the area faster than expected Meteorologists who understand the GFS know its flaws, and use those errors and what the GFS predicts to provide more accurate forecasts.

“If there's a snowstorm… the app can say that four days from now it's going to be sunny and 45 degrees because the app uses GFS. But we know as humans that this model always does that,” Hominuk said, offering a hypothetical example. “The sea is always too far away from the storms, and we will be more careful.”

The GFS is just one of many models, each with its own tendencies and errors that humans can correct. Algorithms don't have that kind of distinction yet, which makes predictions from apps like rainfall and storms somewhat inaccurate. Algorithms also cannot compete with the human experience of living somewhere and knowing how the weather behaves in that particular area.

“Terrain can have a big impact on how these models perform,” Jeff Givens, a meteorologist based in Durango, Colorado, told me via email. Givens' accurate forecasts (especially when it comes to snow and storms) have earned him a following on his popular website Durango Weather Guy, because the San Juan Mountains tend to be less than general weather forecasts in his area. “The applications and models work best on flat terrain.”

Given this information, weather apps seem to work best in places with predictable rainfall patterns, as well as places without mountains or any kind of topographical feature to skew things. People in Southern California probably don't complain about Apple's app as much as someone in Durango or even New York City.

Apps are great when you put them into perspective

When I asked Alexander Stein, a professor in the Department of Earth and Climate Sciences at San Francisco State University, why Apple's weather app failed, he laughed at me.

“Don't you know if it's going to rain in an hour? I would say that's just a concern about where the peas are on your plate,” Stein said. “It's an incredible technological feat to know that it's going to rain all this week. I grew up in a world where weather forecasting was not accurate. We didn't have enough data. But over my lifetime, the skill of weather forecasting has increased amazingly. Talking with Stine gave me a new perspective on my issues with weather apps. When you consider how much these predictions have improved over time, these apps seem like a feat of technology, rather than a nuisance.

A person bikes through ankle-deep water along a flooded road near Prospect Park amid a coastal storm on September 29, 2023, in the Flatbush neighborhood of Brooklyn in New York City.

The National Weather Service does important work, such as modeling weather patterns and issuing warnings. But it's less of a worry if it rains tonight.
Michael M. Santiago/Getty Images

Everything we think about forecasting comes from the National Weather Service, a branch of the National Oceanic and Atmospheric Administration, Stein explained. Every six hours they run a simulation that then provides them with information for the next few weeks. The regional offices analyze this information practically, paying attention to historical data. Weather companies (such as Accuweather, Weather Underground, etc.) batch and modify that information to create forecasts.

“Ultimately, whatever someone puts on the app, they don't have access to different information than anyone else,” Stein said. “There's different information available to different weather forecasters. They all use the National Weather Service.

Stein says the models have improved dramatically over time, and are getting better every day. This is mainly due to more and more detailed data being fed into the equations over the years, to the point that there is more uncertainty in current satellite observations than in the prediction models themselves.

The basic idea: Everyone gets their weather data from the same place, and there shouldn't be radical differences between what weather companies and apps say. Also: stop complaining.

But Stein had a small concession. He clarified that my complaint was not related to the broad scope of weather forecasting which, as he noted, could have significant economic and governmental implications. He said my complaining had more to do with the trend he calls “acting now” — that's a whole different animal.

“The traditional weather prediction problem is the problem of understanding the fluid dynamics of the entire planet,” Stein said. “Whereas the issue is if it rains in five minutes, that's a very local (concern). This isn't something that, as far as I know, the National Weather Service is very concerned about. It's kind of neat, and maybe you can wear a coat, or maybe You can go out and put the bike in the garage.

Complaining about Apple's mistake about rain falling in Manhattan in seven minutes, when there have been huge advances in weather forecasting throughout Stein's lifetime (he's 49), feels like complaining about the way the peas were arranged on my plate. How Stein thinks about weather forecasting and how I before Stein thought about weather forecasting varies in size and scope.

But these disparate viewpoints find common ground when it comes to the importance of meteorologists.

As accurate as these models and forecasts are, meteorologists are key to understanding the weather around us, how it behaves, and the places we live. Apps will never, barring some kind of massive technological advance in the future, be as good at weather forecasting as meteorologists who understand how a certain combination of physics, mathematics, and geography works.

“It's part of understanding the value of meteorologists. This isn't me trying to defend my job,” said Homenuk, of NY Metro Weather. “Human input is needed to understand the complexities of weather.”

As of the time we spoke, Hominuk told me that he wasn't expecting any rain in the New York City forecast for Halloween. I'll check out the app, but I'll trust this one.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *