You look out your window to see immense water pouring down, the roads are flooded, and you can’t leave the house.
That’s the situation that Texans found themselves in during Hurricane Beryl, a category 5 Atlantic hurricane.
In early July, Hurricane Beryl hit the Caribbean. One of the top European weather agencies predicted it would hit Mexico next, based on observations and calculations from the agency’s supercomputers. At the same time, AI engineers running an AI program pinpointed a landing in Texas using data on previous patterns of the atmosphere. A mere four days later, the hurricane rushed into Texas, near the area the AI calculated, with immense force. 36 people have been confirmed dead in Houston so far. Roads are flooded and power is down in nearly 3 million homes. AI would have at least gave them a warning to prepare, but the weather agency gave an inaccurate location.
Weather forecasting is used in daily life, not just for hazardous disasters like hurricanes. For many, it helps them decide what to wear. Will today’s outfit be a large fluffy coat and nice warm boots? Or will it be a tank top and some sandals? Maybe somebody is planning an outdoor event and needs to find a sunny day, but one that is cool enough to stay outdoors. Whether it has an impact on the whole world, or just a single person, weather forecasts are crucial.
Current technology in forecasting relies on current readings, data from a lot of sources, and millions of calculations made by supercomputers to make predictions about the weather. However, people need quick updates for their daily life. In addition to aiding in choices of what to wear, a particular area may need updates quickly, as weather could change from stable to hazardous in a short amount of time. If the weather changes agencies must respond, and updates are crucial. Our current forecasting speed is failing to meet those demands, crunching in the numbers could take more than an hour because there are multiple factors that have to be considered for forecasting. This is also a reason why calculations are so complex for reliable weather forecasting. Current technology is therefore lacking in the speed it needs to keep up with user demands.
AI approaches weather forecasting differently. It does not need current readings or complicated calculations, AI operates on patterns already learned from the past. It can even detect the most subtle patterns, an important tool because even the smallest of patterns can eventually lead to drastic outcomes. AI programs can make predictions in seconds that supercomputers would need at least an hour for. According to the BBC, “Artificial Intelligence could save lives by warning where a hurricane will hit land much sooner than traditional forecasting systems, researchers say.” AI models can also run on desktop computers, a large upgrade from the room-sized supercomputers currently used, making equipment more accessible for small forecasting stations.
Even with AI’s speed and accuracy at landing points being extraordinary, many people are wary of AI taking over weather forecasting completely. While AI is better than current technology in many ways, DR Molina of the University of Maryland explains that “All models are wrong to some extent, and might get the hurricane track right, but what about rain, maximum winds and storm surge?” Weather experts predict that AI models will be used along with the supercomputer forecasting since each has its own strengths and weaknesses and together will allow for a forecasting that’s more reliable. People are also concerned with the loss of jobs resulting from a shift to AI. Over 4.5K people work as meteorologists and would have their employment threatened by a shift to AI Jamie Rhome, deputy director at the National Hurricane Center says that “Having a human at the table to apply situational awareness is one of the reasons we have such good accuracy,” affirming that humans still play a very significant role in forecasting despite the promise of AI.
Whether it’s AI, the supercomputer, or the forecasters, they are all making large contributions to providing reliable weather forecasting at remarkable speeds.
Image Credit by Pixabay
That’s the situation that Texans found themselves in during Hurricane Beryl, a category 5 Atlantic hurricane.
In early July, Hurricane Beryl hit the Caribbean. One of the top European weather agencies predicted it would hit Mexico next, based on observations and calculations from the agency’s supercomputers. At the same time, AI engineers running an AI program pinpointed a landing in Texas using data on previous patterns of the atmosphere. A mere four days later, the hurricane rushed into Texas, near the area the AI calculated, with immense force. 36 people have been confirmed dead in Houston so far. Roads are flooded and power is down in nearly 3 million homes. AI would have at least gave them a warning to prepare, but the weather agency gave an inaccurate location.
Weather forecasting is used in daily life, not just for hazardous disasters like hurricanes. For many, it helps them decide what to wear. Will today’s outfit be a large fluffy coat and nice warm boots? Or will it be a tank top and some sandals? Maybe somebody is planning an outdoor event and needs to find a sunny day, but one that is cool enough to stay outdoors. Whether it has an impact on the whole world, or just a single person, weather forecasts are crucial.
Current technology in forecasting relies on current readings, data from a lot of sources, and millions of calculations made by supercomputers to make predictions about the weather. However, people need quick updates for their daily life. In addition to aiding in choices of what to wear, a particular area may need updates quickly, as weather could change from stable to hazardous in a short amount of time. If the weather changes agencies must respond, and updates are crucial. Our current forecasting speed is failing to meet those demands, crunching in the numbers could take more than an hour because there are multiple factors that have to be considered for forecasting. This is also a reason why calculations are so complex for reliable weather forecasting. Current technology is therefore lacking in the speed it needs to keep up with user demands.
AI approaches weather forecasting differently. It does not need current readings or complicated calculations, AI operates on patterns already learned from the past. It can even detect the most subtle patterns, an important tool because even the smallest of patterns can eventually lead to drastic outcomes. AI programs can make predictions in seconds that supercomputers would need at least an hour for. According to the BBC, “Artificial Intelligence could save lives by warning where a hurricane will hit land much sooner than traditional forecasting systems, researchers say.” AI models can also run on desktop computers, a large upgrade from the room-sized supercomputers currently used, making equipment more accessible for small forecasting stations.
Even with AI’s speed and accuracy at landing points being extraordinary, many people are wary of AI taking over weather forecasting completely. While AI is better than current technology in many ways, DR Molina of the University of Maryland explains that “All models are wrong to some extent, and might get the hurricane track right, but what about rain, maximum winds and storm surge?” Weather experts predict that AI models will be used along with the supercomputer forecasting since each has its own strengths and weaknesses and together will allow for a forecasting that’s more reliable. People are also concerned with the loss of jobs resulting from a shift to AI. Over 4.5K people work as meteorologists and would have their employment threatened by a shift to AI Jamie Rhome, deputy director at the National Hurricane Center says that “Having a human at the table to apply situational awareness is one of the reasons we have such good accuracy,” affirming that humans still play a very significant role in forecasting despite the promise of AI.
Whether it’s AI, the supercomputer, or the forecasters, they are all making large contributions to providing reliable weather forecasting at remarkable speeds.
Image Credit by Pixabay