GeoMIP Analysis Using the CMIP6 Google cloud Store

GeoMIP Analysis Using the CMIP6 Google cloud Store#

Overview#

This section contains a Jupyter notebook demonstrating how to work with GeoMIP (G6sulfur) and CMIP (e.g. ScenarioMIP SSP2-4.5 and SSP5-8.5 simulations) data which is stored on the Google Cloud in a cloud-optimised, zarr, form. The CMIP6 archive was produced as part of the LEAP project. In August 2025, Reflective organised the addition of G6sulfur data from all 6 participating models, for a set of around 1000 commonly used outputs. This resource demonstrates usage of the store, which supports fast and user-friendly cloud-based workflows.

Acknowledgements#

We thank Julius Busecke for his work ingesting the GeoMIP data, and the LEAP project for providing computing resources.

Notebook Location#

The notebooks shown here is also stored under /shared/Code_examples on the Cloud Hub.

Notebooks#

1. Data Access and a plotting example#

G6sulfur_from_cloud_demo.ipynb

This notebook demonstrates how to read in data using the intake-esgf package, and as an example, recreates figure 1 from Visioni et al., 2021.

What you’ll learn:

  • Searching for GeoMIP data on the CMIP6 GCS store

  • Working with zarr stores via xarray and intake-esm

  • Iterating over datasets to quickly reproduce common analyses

Key Benefits#

Direct Data Access#

  • No pre-downloading required - Access data directly from the zarr store

  • Real-time data discovery - Search and find datasets as needed

  • Lazy loading - Download only the part of the data you need, only when you need it

Reproducible Research#

  • Version control - Track exactly which data versions you used

  • Portable code - Share notebooks that work anywhere with ESGF access

Additional Resources#