Artificial intelligence and machine learning can provide more accurate analysis and forecasts in the energy sector.
Artificial intelligence and machine learning can provide more accurate analysis and forecasts in the energy sector. One such application is weather forecasting. I find this very interesting and highly recommend this recent discovery published in Science.
The GraphCast artificial intelligence model, created by developers at Google DeepMind, predicts weather around the world up to 10 days in advance, determining the turning point, among other things. for renewable energy technologies.
GraphCast generates an accurate 10-day forecast in less than a minute on a single Google Cloud TPU v4 device, supporting applications such as predicting tropical cyclone routes, atmospheric changes and extreme temperatures. It predicts hundreds of weather variables over 10 days with 0.25° resolution around the world in less than a minute. Conventional methods take several hours and are energy-intensive. GraphCast is much cheaper in terms of power consumption.
The developers point out that traditional numerical weather forecasting uses increased computational resources to improve forecast accuracy, but does not directly use historical weather data to improve the underlying model.
“GraphCast significantly outperforms the most accurate operational deterministic systems in 90% of a group of 1,380 verification targets, and its forecasts support better prediction of severe events, including tracking tropical cyclones, atmospheric changes and extreme temperatures.”
Detailed information about the method can be found on Science: https://www.science.org/doi/10.1126/science.adi2336