Google DeepMind's GenCast model has been shown to be more accurate than traditional methods by outperforming the top-performing supercomputer in 99.8% of predictions.
Engineer Ilan Price emphasizes that ensemble forecasts generated by GenCast provide a fuller understanding of possible weather conditions, helping decision makers to better prepare for upcoming weather events.
GenCast uses AI to identify patterns in historical weather data, creating 50 possible outcomes for a high degree of confidence in predictions.
While both GenCast and ENS are probabilistic models, GenCast departs from traditional methods by avoiding direct simulation of atmospheric physics, which enhances speed and accuracy.
Collection
[
|
...
]