Launch of Earthkit 1.0: Open-Source Tools for the Next Generation of Earth Science

Earthkit, ECMWF's open-source Python toolkit for weather and climate science, has reached a major milestone with the release of version 1.0. Designed to simplify everything from data access to analysis and visualisation, it offers researchers a stable, operationally proven foundation for working with large, complex earth science datasets.

What is Earthkit  

Developed by ECMWF as part of its long-term Software Strategy, Earthkit is an open-source Python ecosystem that simplifies weather and climate data workflows. Rather than a single application, it is a collection of interoperable Python packages that helps users retrieve, process, analyse and visualise Earth science data more efficiently. 

With the release of Earthkit 1.0, the core interfaces are now stable, giving researchers and operational users confidence that workflows built today will remain compatible in future releases. For scientists working with Copernicus data through WEkEO, that matters. 

Earthkit is central to ECMWF’s vision of a more open, interoperable and community-driven software ecosystem. Developed openly on GitHub, it has already attracted contributions from ECMWF Member States and international partners. Over time, it will replace long-standing ECMWF software packages, including Metview and Magics, bringing decades of operational expertise into a unified Python environment.

One ecosystem, many building blocks

Built on established open-source libraries including NumPy, pandas and Matplotlib, Earthkit integrates proven ECMWF technologies such as ecCodes, FDB and MIR. The result is a modern software stack that combines operational reliability with the flexibility of open-source software. 

Version 1.0 brings together six core packages covering key scientific applications:

• earthkit-data provides a format-agnostic Python interface for geospatial data, including accessing, inspecting and filtering data from a wide range of sources (including files, URLs, archives and remote data services) and formats (including GRIB, netCDF, Zarr, BUFR and many more).
• earthkit-plots creates publication-quality weather and climate visualisations with just a few lines of Python, supporting automatic data extraction and formatting based on metadata.
• earthkit-geo provides geospatial computations such as efficiently transforming data between different spatial grids using proven ECMWF interpolation technology.
• earthkit-meteo provides meteorological calculations and scientific post-processing functions. 
• earthkit-hydro extends the ecosystem with hydrological analysis tools for working with river networks, accumulations and more.
• earthkit-transforms offers aggregations, analysis and transformations across arbitrary data dimensions.

Together, these packages create a consistent workflow from data access to analysis and visualisation within a single ecosystem.

The Earthkit 1.0 ecosystem

Why it matters for WEkEO users

WEkEO provides access to vast volumes of Copernicus Earth observation data and Earthkit offers a streamlined way to turn those datasets into scientific products.

A single interface retrieves data from multiple sources while abstracting away file formats and storage. Once loaded, data can be inspected, filtered, converted to NumPy, pandas or Xarray, processed and visualised using the same platform.

Earthkit is also designed for modern computing. It supports efficient in-memory workflows, minimises unnecessary disk access, caches remote data where appropriate, and has been developed with GPU-enabled computing and machine learning in mind.

The software is equally suited to exploratory research and operational services, combining components already proven in ECMWF production systems with modern open-source development. For WEkEO users, it reduces the time spent on data wrangling and format compatibility, two of the more persistent frustrations of working with large earth science datasets. 

Getting started

Earthkit is open source, freely available and can be installed using pip, making it easy to integrate into existing Python environments alongside WEkEO data. 

Whether analysing Copernicus datasets, building operational services or developing machine learning applications, Earthkit provides a modern, scalable foundation that removes much of the complexity from weather and climate workflows. 

Explore the documentation, installation guide and the full Earthkit ecosystem on the Earthkit website, and start building faster, more reproducible workflows with Earthkit 1.0.