The Copernicus Climate Data Store (CDS) is the public access point for the data products of the Copernicus Climate Change Service (C3S) — the European Union’s operational climate-monitoring service, implemented by the European Centre for Medium-Range Weather Forecasts (ECMWF) on behalf of the EU’s Copernicus Earth observation programme. For researchers doing climate, environmental, agricultural, hydrological, or energy-systems work, the CDS is one of the largest single sources of free, open reanalysis and climate-projection data available — most notably ERA5, the reanalysis dataset now used as a baseline input across a huge share of published climate science.
This guide covers what C3S and the CDS actually are, what data holdings they provide, and the concrete steps a researcher or research-support office needs to get data out of the platform — plus the licensing and citation points a research data management (RDM) office should know before pointing a research team at it.
What Is the Copernicus Climate Data Store?
The CDS (cds.climate.copernicus.eu) is the technical platform through which C3S distributes its data: a searchable catalogue of datasets, a web-based download interface, and a programmatic API. It is not itself a separate organization — it is the delivery mechanism for C3S, one of six operational services that make up the wider Copernicus Programme (the others cover atmosphere, marine environment, land, security, and emergency management).
C3S is implemented by ECMWF under a delegation agreement with the European Commission, funded as part of the EU’s Earth observation programme. This distinction matters for attribution and for research administrators evaluating data provenance: the science and infrastructure are ECMWF’s, the mandate and funding are the EU’s, and “C3S” is the name of the service that sits between them.
What Data the CDS Provides
The CDS catalogue spans several categories relevant to research use. The most heavily used, by a wide margin, is reanalysis data.
ERA5 — the flagship global reanalysis
ERA5 is the fifth generation of ECMWF’s global atmospheric reanalysis, produced by C3S. A reanalysis combines historical observations with a modern numerical weather-prediction model, via data assimilation, to produce a physically consistent, gridded best-estimate of past atmospheric, land-surface, and ocean-wave conditions — effectively reconstructing the historical climate record at a resolution and consistency no observation network alone can provide. ERA5 supplies:
- Hourly data from January 1940 to the present (an earlier back-extension, ERA5 1940-1978, was completed to extend the record beyond the original 1979 start)
- Global coverage at roughly 0.25° (~31 km) horizontal resolution, with 37 vertical pressure levels
- Atmospheric, land-surface, and sea-state parameters, together with uncertainty estimates for each
An interactive exploration tool, ERA Explorer, lets users browse the ERA5 record visually before committing to a bulk download.
ERA5-Land and regional reanalyses
Alongside global ERA5, the CDS hosts ERA5-Land, a higher-resolution (~9 km) land-surface reanalysis better suited to catchment-scale hydrology, agriculture, and land-surface studies than the coarser global product. Two regional reanalyses extend this further for specific geographies: CERRA, a high-resolution reanalysis for Europe (from 1984), and CARRA, covering the Arctic (from the mid-1980s). Regional products trade global coverage for finer spatial detail in areas where global ERA5’s resolution is too coarse for local-scale analysis.
Climate projections
Beyond reanalysis of the observed past, the CDS also distributes forward-looking climate-model output: a substantial portion of the global projections produced under the Coupled Model Intercomparison Project (both CMIP5 and CMIP6 — the CMIP6 archive underpins the IPCC’s Sixth Assessment Report), and dynamically downscaled regional projections from the CORDEX (Coordinated Regional Climate Downscaling Experiment) initiative, now covering all 14 CORDEX domains worldwide. These are the datasets typically used for future-scenario impact studies, as distinct from ERA5’s historical reconstruction.
How Researchers Access CDS Data
Access to the CDS is free, but every dataset requires two things before data can be downloaded: a registered account, and explicit acceptance of that dataset’s licence terms.
- Register for a CDS account at cds.climate.copernicus.eu — no institutional affiliation is required.
- Locate the dataset via the catalogue search (e.g. searching “ERA5” surfaces the single-levels, pressure-levels, and land variants separately) and open its dataset page.
- Accept the dataset’s Terms of Use — this is done per dataset, at the bottom of that dataset’s download form, and must be repeated for each new dataset a project uses, not just once per account.
- Download either through the web form (select variable, time range, spatial extent, and file format interactively) or programmatically via the CDS API.
For anything beyond a small, one-off extract, most research groups use the CDS API rather than the web form, since it supports scripted, repeatable, and much larger bulk requests. The workflow is:
- Install the Python client:
pip install "cdsapi>=0.7.7" - Retrieve a personal access token from the CDS account page and store it in a local
$HOME/.cdsapircconfiguration file, alongside the API URL - Submit a data request from Python specifying the dataset name, variables, time range, and area — requests are queued and processed server-side, then retrieved once ready
Data requests, especially for multi-decade or global-resolution ERA5 extracts, can be large and can queue for processing time depending on CDS server load — research computing/RDM offices supporting a project should budget storage and wall-clock time accordingly, not assume near-instant delivery the way a small API call would behave.
Licensing, Attribution, and What RDM Offices Should Flag
CDS datasets are distributed under a Copernicus licence framework that permits free reuse, redistribution, and commercial use, conditional on attribution to the Copernicus Climate Change Service and, where specified, ECMWF. Because acceptance is tracked per dataset and per account, a data management plan citing CDS-sourced data should record: the exact dataset name and version used (e.g. a specific ERA5 product, not “ERA5” generically), the access date, and the required attribution statement — the same discipline CASRAI’s guidance recommends for any third-party dataset incorporated into a research output. See CASRAI’s guide on how to cite data for the general formats and principles this maps onto, and the FAIR Data Principles entry for the broader framework C3S’s open, well-documented, API-accessible catalogue is a practical example of.
Because CDS data is a re-published, quality-controlled derivative of many upstream observation sources rather than data collected by the researcher, projects incorporating it should also document it explicitly as a secondary/external dataset in the DMP’s data-sources section — distinct from primary data the project itself generates — consistent with the general secondary-data documentation practice covered in CASRAI’s types of research data guide.
Frequently Asked Questions
Is Copernicus Climate Data Store data free to use?
Yes. CDS access and the datasets themselves are free, including for commercial use, subject to accepting each dataset’s licence terms and providing the required attribution — there is no paywall or institutional-subscription requirement.
What is the difference between ERA5 and ERA5-Land?
ERA5 is the global atmospheric reanalysis at ~31 km resolution, covering atmosphere, land, and ocean-wave variables together. ERA5-Land is a higher-resolution (~9 km) reanalysis focused specifically on land-surface variables, better suited to local hydrological, agricultural, or land-use studies where ERA5’s coarser grid loses too much spatial detail.
Do I need to use the CDS API, or is the web interface enough?
The web interface is sufficient for small, one-off extracts. For anything requiring scripted, repeatable, or large-volume downloads — the norm for most research use — the CDS API (the cdsapi Python client) is the practical choice, and is what most published research using ERA5 data relies on for reproducible data acquisition.
Does CDS data need to be cited differently from data a researcher collects directly?
It should be documented as external/secondary data with the specific dataset name, version, and access date, plus the attribution the Copernicus licence requires — not just referenced generically as “Copernicus data” or “ERA5,” which under-documents exactly what was used for reproducibility purposes.
Related CASRAI Resources
- NOAA NCEI: Accessing Climate and Ocean Data — the equivalent US federal climate-data archive, for comparison
- How to Cite Data: Formats and the Data Citation Principles
- Types of Research Data: A Taxonomy for Data Management Planning
- FAIR Data Principles
- Open Data
