Skip to content

OceanDataStore |

About

OceanDataStore is a Python library designed to streamline writing, updating and accessing ocean model and observational data stored in cloud object storage.

Writing Ocean Data to Cloud Object Storage

For writing and updating files to S3-compatibible cloud object stores (e.g., JASMIN Object Store), OceanDataStore includes a commmand line interface (CLI) with the following commands:

  • send_to_zarr: Send local file(s) to a new zarr store in cloud object storage.
  • update_zarr: Update an existing zarr store in cloud object storage with local file(s).
  • send_to_icechunk: Send local file(s) to a new Icechunk repository in cloud object storage.
  • update_icechunk: Update an existing Icechunk repostitory in cloud object storage with local file(s).
  • list: List the objects found in a cloud object store bucket.

Accessing Ocean Data in Cloud Object Storage

To access ocean model and observational data stored in cloud object storage, OceanDataStore includes the OceanDataCatalog API with the following features:

  • Interfaces with a Spatio-Temporal Access Catalog (STAC) to expose available collections of ocean model & observational data stored in the JASMIN Object Store.
  • Search catalogs by collection, variable names or platform (grid type).
  • Subset & open Analysis-Ready Cloud Optimised (ARCO) datasets as lazy xarray Datasets.

Quick Start 🚀

Installation

We recommend downloading and installing OceanDataStore into a new virtual environment via GitHub.

After activating a new virtual environment, pip install OceanDataStore from GitHub: 

pip install git+https://github.com/NOC-MSM/OceanDataStore.git

Helpful Tip...
  • We strongly recommend setting-up a virtual environment before installing OceanDataStore with pip.

The simplest way to create a new virtual environment is to use venv:

python3 -m venv "env_ods"

Alternatively, using an existing miniconda, mamba or miniforge installation:

conda env create -n env_ods python=3.13

Learning More...

To learn more about writing ocean data to cloud object storage using the OceanDataStore CLI, see the CLI section of the User Guide. Alternatively, explore some typical workflows by visiting the Examples page.

To get started exploring the available ocean model outputs generated by the National Oceanography Centre, see the OceanDataCatalog section of the [User Guide].