Climate and Earth observation data are increasingly important for health planning, early warning, forecasting, and research. But using these datasets often means working across different providers, formats, APIs, spatial grids, and update cycles.
The Open Climate Service (OCS) is an open-source project that makes this easier. It provides a consistent way to ingest, manage, process, and publish climate, Earth observation, and related datasets for a specific country or region.
In practice: a country team sets up one service that keeps a national rainfall, temperature, or population dataset updated, and serves it to DHIS2 Maps, the DHIS2 Climate app, and CHAP — so each of those applications doesn’t have to build its own integration with every upstream data provider.
An OCS instance can currently:
- Access datasets such as CHIRPS precipitation, ERA5-Land climate data, and WorldPop population data.
- Limit ingestion and storage to a configured country or regional extent.
- Harmonise and store datasets as standards-based GeoZarr (chunked Zarr v3).
- Publish data through open interfaces including STAC, Zarr over HTTP, and openEO.
- Run reusable processing workflows, including spatial aggregation to administrative or health-service areas.
- Automatically keep datasets up to date on a schedule, and trigger workflows when a dataset is updated.
- Produce DHIS2-ready data value set JSON and CHAP-ready CSV outputs.
- Support additional national or project-specific datasets, processes, and workflows through plugins.
OCS is designed to complement DHIS2 Maps and the Climate app while remaining independent of DHIS2. It can be deployed locally, in the cloud, or on national infrastructure, so countries and organisations keep control over their data and deployment choices — and can reduce dependence on external hosted Earth observation backends.
The broader goal is a reusable climate-data infrastructure layer that can serve DHIS2 applications, CHAP, GIS tools, analytical notebooks, and other systems.
Who might find it useful?
We would particularly like to hear from:
- HISP groups and DHIS2 implementers working with climate and health data.
- Ministries and national teams exploring climate-informed health services.
- CHAP users and modelling teams.
- GIS and Earth observation specialists.
- Developers interested in adding datasets, analytical processes, or reusable workflows.
- Projects that already have concrete climate-data integration challenges or national data sources.
Project status
OCS is under active development. APIs and data models may still change, and we are currently refining deployment, scheduling, workflow automation, and integration patterns. Community feedback at this stage is especially valuable.
Learn more and try it
- GitHub repository: https://github.com/dhis2/open-climate-service
- Documentation: https://dhis2.github.io/open-climate-service/
- Quick start: https://dhis2.github.io/open-climate-service/setup_guide/
- Instance and deployment guide: https://dhis2.github.io/open-climate-service/instance_guide/
- Roadmap: https://dhis2.github.io/open-climate-service/roadmap/
- DAC2026 presentations
We would welcome feedback on
- Which climate, environmental, population, or forecast datasets are most important in your context?
- Which DHIS2 or CHAP workflows should we prioritise?
- Are you already maintaining national or regional datasets that could be exposed through an OCS plugin?
- Would you be interested in testing an OCS instance or contributing to its development?
Getting involved
Please share your questions, use cases, and ideas in this thread — we would be very interested in learning what would make the Open Climate Service useful in real country implementations. For specific feature requests or technical issues, feel free to follow the usual DHIS2 way-of-work (a Jira ticket or a direct email); for wider discussion, this CoP thread is the best place.
