[sureklix]

Very good question! These models are trained on ~40 years of ERA5 data –you can think of past forecasts from numerical models integrated with real observational data to have a continuous distribution of weather parameters (temperature, wind etc.) Therefore model resolution is 0.25 degrees (28km x 28km at the equator).

The way accuracy is measured is through picking targets (say temperature at 2 meters, at x,y lat & lon and forecasted 24h ago) and comparing them on RMSE and ACC (anomaly correlation co-efficient). For instance, in Google Graphcast paper they pick 1380 targets and the model out performs NWPs in 90% of them.

[m13uz]

To add to this, there are other ML models with higher resolution. For example, Google's MetNet-3 uses satellite radar images and ground measurements, and its resolution is 1km x 1km. And we are currently working on training a "nano" version of this!

Metnet-3: https://news.ycombinator.com/item?id=38122631