Why are weather apps still so bad?

Technologically, we live in a time of plenty. Today, I can ask a chatbot to provide it the Canterbury Tales As if it was written by Taylor Swift or to help me write a factually inaccurate biography. With three swipes, I can summon almost everyone listed on my phone and see their confused faces over an impromptu video chat. My life is a mixed bag of information, and I follow the “all you can eat” plan. But there’s one particular area where technology hasn’t kept pace: weather apps.

Weather forecasting is always a game of prediction and probabilities, but these apps seem to fail more often than they should. At best, they perform as well as meteorologists, but some of the most famous ones perform much worse. For example, my favorite Dark Sky app, which was shut down earlier this year and rolled into Apple’s Weather app, accurately predicted the high temperature in my zip code just 39 percent of the time, according to ForecastAdvisor, which evaluates… Online weather service providers. In comparison, the Weather Channel app comes in at 83 percent. Apple’s app, while not rated by ForecastAdvisor, has a reputation for unpredictable forecasts and has been consistently criticized for providing faulty radar displays, mixing up precipitation totals, or, as happened last week, crashing altogether. Dozens of times, Apple Weather has lulled me into a false sense of security, leaving me wet and disheveled after a run, bike ride, or golf game.

People love to complain about the weather forecast, dating back to a time when local news meteorologists were the primary source for those planning their morning commute. But the apps have produced a new level of frustration, at least judging by it favorable to eccentric Tweet Over a contract. Nearly two decades into the smartphone era — when anyone could theoretically harness the power of government weather data and analyze dozens of complex graphs and models in real time — we’re still stuck in the rain.

Not all weather apps are the same. There are tens of thousands of them, from the simply designed Apple Weather app to the expensive, complex, and data-rich Windy.App. But all of these forecasts rely on similar data, which is pulled from places like the National Oceanic and Atmospheric Administration (NOAA) and the European Center for Medium-Range Weather Forecasts. Traditional meteorologists interpret these models based on their training as well as their intuition and past regional weather patterns, and various weather apps and services tend to use their own secret mix of algorithms to make their predictions. On any typical day, you’ll likely see similar forecasts from one app to another and on your TV. But when it comes to how people… Feel In terms of weather apps, these emergencies – which usually occur during severe weather events – are what stay on a person’s mind. “Eighty percent of the year, the weather app will work fine,” Matt Lanza, a meteorologist who runs the Space City Weather app in Houston, told me. “But the 20 percent where people get burns is a problem.”

There is no person on the planet who has a more torturous and conflicted relationship with weather applications than those who interpret forecast models for a living. “My wife is married to a meteorologist, and she will ask me directly if her favorite weather app says something different than my forecast,” Lanza told me. “That’s how ingrained these services are in most people’s lives.” The fundamental problem with weather apps, he says, is that many of them remove a crucial element of good, reliable forecasts: a human interpreter who can relay warnings about models or provide a set of results rather than a final forecast.

Lanza explained the human touch of a meteorologist using the example of a so-called high-resolution forecast model that can only forecast 18 hours out. He told me that he’s generally pretty good at predicting rain and thunderstorms — “but often it’s too hot and over-indicates the odds of a bad storm.” This model, if left to its own devices, would project showers and thunderstorms covering the area for hours, when in reality, the storm might only dump rain for 30 minutes in an isolated area of ​​the area mapped. “The problem is when you take form data and push it directly into the application without any human interpretation,” he said. “Because you’re not going to get the nuances from these apps at all. That can mean the difference between the chance of rain all day and the chance it will rain all day.”

But even this interpretation has caveats. All weather apps are different, and their forecasts have varying levels of sophistication. Some pipe model data is entered directly, while others are organized using artificial intelligence. The company’s app includes “billions of weather data points,” Peter Neely, director of weather forecasting science and technology at The Weather Channel, said in an email, adding that “our expert team of meteorologists oversees the process and corrects it as needed.”

Weather apps may be less reliable for another reason as well. When it comes to forecasting severe weather like snow, small changes in atmospheric moisture — the kind of change an experienced forecaster might notice — can cause large discrepancies in precipitation results. An application that does not require human processing may choose to average the range of model results, producing forecasts that do not reflect the dynamic situation on the ground. Or look at cities with microclimates: “Today, in Chicago, the lakefront will be in the lower 40s, and the suburbs will be 50+ degrees,” ABC 7 Chicago meteorologist Greg Dutra told me. “Often, the difference is more noticeable — a 20-degree swing just miles away.” These subtle temperature differences can sometimes mean a completely different forecast for people living in the same area, something one-size-fits-all weather apps don’t always capture.

Of course, meteorologists believe they’re doing better than predicting using an algorithm alone, but even weather app creators tell me the challenges are real. “It’s impossible for a weather data provider to be accurate everywhere in the world,” Brian Mueller, founder of Carrot Weather, told me. His solution to the problem of app-based inaccuracy is to give users greater ability to choose what they see when they open the Carrot app, allowing them to customize the specific weather information the app displays as well as the data sources the app will draw from. Mueller said he learned from Dark Sky’s success how important beautiful, detailed radar maps are — as a source of weather data and for entertainment purposes. In fact, it seems that meteorology is only part of the appeal when it comes to building your favorite weather app. Carrot has a great interface design, with bright colors and Easter eggs scattered throughout, such as its weather map-based geo-challenges. He’s also connected Carrot to ChatGPT to allow people to chat with the app’s fictional character.

But what if these detailed models and stunning maps, in the hands of weather experts like me, are the real problem? “The general public has access to more weather information than ever before, and I think that’s a bad thing,” Chris Messinis, a weather forecasting consultant in North Carolina who goes by the nickname “Weather Moose,” told me. “You can go to PivotalWeather.com now and view any simulation you want.” He says that this data is fine to look at if you know how to interpret it, but for people not trained to analyze it, it is at best worthless and at worst dangerous.

In fact, the outlook is better than ever, says Andrew Bloom, a journalist and author of the book The Weather Machine: A Journey Inside the Forecast, Tell me. He added: “But we are arguably less prepared to understand and act on this improvement, and forecasts are only as good as our ability to make decisions based on them.” In fact, even academic research on weather apps suggests that apps fail worse when they give users a false sense of certainty about forecasts. A 2016 paper for the Royal Meteorological Society argued that “the current way of communicating forecasts in the most common applications is guilty of ‘vulgarity’ (‘failure to acknowledge that forecasts may sometimes fail’) and ‘impoverishment’ (‘failure to address the wider context’ in which it is Make predictions’).

The conflicting relationship between people and weather apps may be just a manifestation of the information overload that dominates all aspects of modern life. These products give anyone with a phone access to a vast amount of information that can be very complex. Greg Dutra shared with me one of these generic, high-resolution models from the National Oceanic and Atmospheric Administration (NOAA), which was filled with indecipherable links to jargon terms like “maximum 0-2 km vertical vorticity.” Weather apps often seem to respond to this massive amount of data in two ways: by reducing it to a “partly sunny” icon, or by bombarding the user with information they may not need or understand. At its worst, the modern weather app seems to curry favor with people, entrusting them with doing their own research even if they’re not equipped. I’m not too proud to admit that some of the fun of playing with Dark Sky’s beautiful radar or Windy.App’s endless collection of models is the feeling of role-playing as a meteorologist. But the truth is, I don’t really know what I’m looking at.

What people seem to be looking for in a weather app is something they can justify their blind trust and let into their lives — after all, it’s often the first thing you check when you roll into bed in the morning. According to 56,400 reviews for Carrot in the Apple App Store, die-hard fans find the app entertaining and likable. “I love this psychotic and surprisingly accurate weather app,” says one five-star review. Although many people need reliable forecasts, real loyalty comes from a weather app that makes people feel good when they open it.

The duality of our weather apps is a strange relationship between feeling grateful for instant access to information and navigation at the same time as feeling guilty and confused about how the experience is also, somehow, unsatisfying — a bit like staring into an endless Netflix library and feeling like… If there was nothing to watch. Weather apps don’t get any worse. In fact, they are becoming more advanced, inputting more and more data and presenting it to us to consume. This of course may be the reason why they feel bad.