-
Implement a synthetic weather data generation tool using Google's NeuralGCM framework. This tool should model spatiotemporal relationships in weather data (e.g., temperature, humidity, wind speed) to …
-
From the NeuralGCM documentation, "all_forcings" uses the function "forcings_from_xarray" for all ERA5 data to create the forcings. It is stated that “ERA5’s sea surface temperature and sea ice concen…
-
Semi-Lagrangian advection is a key part of efficient spectral atmospheric models.
In ECMWF’s model, it allows for [6x large time-steps](https://www.ecmwf.int/sites/default/files/elibrary/2014/9054-…
-
Hybrid sigma-pressure coordinates are used by most modern dynamical cores. Switching from sigma to hybrid coordiates would make it possible to run the Dinosaur dycore on native ECMWF/UFS grids, and pr…
-
Hello! Thanks for another great contribution.
While reading the NeuralGCM documentation (https://neuralgcm.readthedocs.io/en/latest/), I wasn't able to find instructions on reproducing the decadal…
-
Hi,
Thanks for open-sourcing the NeuralGCM. I noticed sea level pressure is used to detect tropical cyclones, but it seems SLP is not a standard output. How should we output SLP?
Also, is preci…
-
Is the full code for the training of the NeuralGCM provided？ If so, how can we use it? Is it possible to re-train another NeuralGCM utilizing a subset of the variables mentioned? Thanks for your work.
-
I would love to be able to install NeuralGCM’s Zarr/tfds reader as its own python package.
https://github.com/google-research/neuralgcm/blob/d32e8b6365bd6096a76b531a14df1a27723d5e2c/neuralgcm/refe…
-
The NeuralGCM paper contains some analytics that depend on the surface information. For example, the precipitable water calculation involves a vertical integral that needs surface pressure. The tracki…
-
I am currently trying to train NeuralGCM on some custom data, but the computational resource requirements are too high for my setup. The model (NeuralGCM-2.8deg) currently requires 16 TPUs, which is b…