NeuralGCM AMIP Climate Simulations: 2-Year, 20-Year and 40-Year Atmosphere-Only Runs

Provider
Google Research (Kochkov et al., NeuralGCM)
License
CC-BY-SA-4.0 (source)
Format
Zarr (in sampled files)
Stored at
Google Cloud Storage, gs://neuralgcm/amip_runs (us)
Access
Free and public. No account with the data provider is needed.
DOI
10.1038/s41586-024-07744-y
Category
Climate & Weather
Source
https://doi.org/10.1038/s41586-024-07744-y

About this dataset

Multi-year atmosphere-only (AMIP-style, prescribed sea surface temperature and sea ice) simulations from the NeuralGCM hybrid model of Kochkov et al. (Nature, 2024), for studying climate drift, seasonal cycles, tropical cyclone counts and precipitation climatology in a learned general circulation model. Three Zarr stores, measured 2026-11-16: v1_deterministic_2_8_deg/1980-to-2020 has 37 initial dates through 1980, each rolled out about 40 years at 6-hourly output (58,449 steps) on a 128x64 Gaussian grid with 37 pressure levels, about 18 TB; v1_precip_stochastic_2_8_deg/2001-to-2021 has 37 stochastic 20-year runs at 3-hourly output with cumulative precipitation and evaporation, about 1.6 TB; v1_deterministic_1_4_deg/2019-to-2021 has 37 two-year runs at 12-hourly output on a 256x128 grid, about 1.8 TB. Variables include geopotential, temperature, specific humidity, cloud ice and liquid water and u and v wind. Coordinates are time (initialisation) and prediction_timedelta. Open with xarray.open_zarr(path, storage_options={'token':'anon'}). Google releases these under a Creative Commons licence (docs name CC-BY 4.0 and link CC BY-SA 4.0; treat as CC BY-SA 4.0). Cite Kochkov et al. 2024.

How to access this dataset in Python

import gcsfs
fs = gcsfs.GCSFileSystem(token='anon')
fs.ls('neuralgcm/amip_runs/', detail=False)[:10]

Lists the files with anonymous access. Runs on any machine, with no cloud account.

Data structure

File types in sampled files: Zarr

Files

Example file: amip_runs/v1_deterministic_1_4_deg/2019-to-2021_256x128_gauss_37-level_stride12h.zarr/.zattrs

Analyze this dataset without downloading it

On Scigantic, NeuralGCM AMIP Climate Simulations: 2-Year, 20-Year and 40-Year Atmosphere-Only Runs opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Climate & Weather datasets