[ad_1]
Firstly protein foldingWeather forecasting now: London-based artificial intelligence firm DeepMind continues its work by applying deep learning to difficult science problems. Working with the Met Office, the UK’s national weather service, DeepMind has developed a deep learning tool called DGMR that can accurately predict the probability of rain in the next 90 minutes, one of the toughest challenges in weather forecasting.
In a blind comparison with available tools, several dozen experts decided that DGMR’s forecasts were the best for a range of factors 89% of the time, including forecasts for the location, extent, movement and intensity of rain. Results published in a Nature newspaper today.
DeepMind’s new tool no AlphaFold solving an important problem in biology It’s what scientists have been dealing with for years. Yet even a small improvement in forecasts is significant.
Predicting rain, especially heavy rain, is critical to many industries, from outdoor activities to aviation and emergency services. But doing this well is difficult. Figuring out how much water is in the sky and when and where it will fall depends on a number of weather processes such as temperature changes, cloud formation and wind. All these factors are complex enough on their own, but taken together they are even more complex.
The best available forecasting techniques use massive computer simulations of atmospheric physics. These work well for the longer term forecast, but less good at predicting what will happen in the next hour, known as the present forecast. Previous deep learning techniques have been developed, but these often do well at something like estimating location, at the expense of something else like estimating density.
CONTEMPLATION
“Predicting rain now remains a major challenge for meteorologists,” says Greg Carbin, chief of forecasting operations at the NOAA Weather Prediction Center in the US, who was not involved in the study.
The DeepMind team trained their AI on radar data. Many countries post frequent snapshots throughout the day of radar measurements that monitor the formation and movement of clouds. In the UK, for example, a new reading is published every five minutes. Putting these snapshots together provides an up-to-date stop-motion video of how rain patterns move across a country, similar to the forecast images you see on TV.
[ad_2]
Source link