The European Centre for Medium-Range Weather Forecasts (ECMWF) has opened the call for Code for Earth 2026. Developers, researchers and data specialists can submit project proposals until 9 April 2026. This round comprises 16 challenges across four thematic streams, including—for the first time—a regionally tailored Africa stream. The initiative is supported by Copernicus services, the EU Destination Earth (DestinE) initiative and the Africa Regional Centres of Excellence Programme (ArcX).
The programme aims to develop open prototypes for weather, climate and atmospheric applications. Transparency, reproducible workflows and collaborative development are central. Participating services provide data, infrastructure and technical mentorship so teams can address scientific questions in a practical, operationally relevant setting.
The four streams for 2026
– Data visualization and visual narratives: Preparing complex model and observational data for clear analysis and decision support. – Machine learning and AI for weather, climate and atmosphere: Methods for pattern recognition, improved forecasts and uncertainty quantification using large datasets. – Software development and open-source tools for Earth-system science: Modular, maintainable software for models, data pipelines and analysis tools. – Africa stream (new): Co-designed with regional partners; focuses on capacity building and locally relevant applications such as flood monitoring, land-use assessment and risk estimation.
Why openness matters
Open software and open data improve reproducibility and trust, encourage reuse and accelerate improvements through community peer review. Open solutions are also easier to adapt and transfer across regions.
AI with a measured approach
Machine-learning methods can fill gaps in observational records, detect patterns and refine predictions. Prerequisites include representative training data, measures to mitigate bias, and models whose behavior can be understood and audited. Computational cost and energy use must also be weighed. Open development practices increase transparency and facilitate sharing of best practices.
Project requirements
Successful projects combine domain expertise with solid software engineering: automated testing, clear documentation, containerization, interfaces to data services and implementation of the FAIR principles (findable, accessible, interoperable, reusable) for software and metadata. Code for Earth supports teams with infrastructure and mentoring.
Participation and selection
The call targets professionals in software development, data science, meteorology and climate research, as well as users from government agencies and civil-society organizations. Proposals will be evaluated on scientific relevance, technical feasibility and user benefit. Details on participation, funding and the individual challenges are available on the programme page.
Context and significance
Code for Earth bridges the gap between research and operational use by bringing together open data, joint development and targeted capacity building. The approach produces tools tailored to regional needs and helps translate scientific methods into operational practice—relevant for research, climate adaptation and risk management.
Copernicus Climate Change Service
The original source allows readers to verify the source material and subsequent updates.
Open source →
