ERA-CLIM2 outcomes boost NWP and climate work

Roberto Buizza


ERA-CLIM2, a three-year European Union 7th Framework project that ended in December 2017, has made a major contribution to climate research and the development of climate monitoring services through its data rescue activities, post-processing of observations and work on coupled data assimilation methods. In particular, it has produced two coupled reanalyses, i.e. physically consistent datasets describing the evolution of the global atmosphere, ocean, land surface and cryosphere and the carbon cycle.

ECMWF played a key role in this collaborative project from the beginning: Dick Dee submitted the original proposal and acted as project coordinator until September 2015, when the author of this article took over. Magdalena Alonso Balmaseda, Manuel Fuentes and Patrick Laloyaux were work package leaders, and several other members of staff worked for the project on aspects of the science, technology, management and accounting.

Global 0–300 m ocean heat content anomalies with respect to 1958–2010. The chart shows results for the ORA-20C ocean reanalysis ensemble (10 members), the CERA-20C ensemble (10 members) and the ORAS4 ocean reanalysis ensemble (5 members), with faint lines showing the ensemble members and bold lines the ensemble mean. An OHC increase of 1x108 J/m2 corresponds to a temperature increase of 0.08 K averaged over the top 300 metres.

The project has been at the heart of a concerted effort in Europe to build the information infrastructure needed to support climate monitoring, climate research and climate services, based on the best available science and observations. More specifically, ERA-CLIM2 was one of the designated precursor projects of the EU-funded Copernicus Climate Change Service (C3S). It was intended to help build the infrastructure and production capabilities of C3S, which is being implemented by ECMWF. ERA-CLIM2 has achieved this goal by making substantial contributions to reanalysis research in four main areas:

  • Observation data rescue and post-processing: ERA-CLIM2 funded a major effort on data rescue for historical in situ weather observations around the world, and substantial work on the reprocessing of satellite climate data records to enable their use for reanalysis.
  • Data assimilation methods: ERA-CLIM2 led to the development and testing of coupled data assimilation methods, capable of including observations from different Earth system components (land surface, ocean, sea ice, atmosphere, chemical components, etc.) in a consistent manner, to produce a more consistent estimate of the evolution of the Earth system, especially at the surface. This will help to develop coupled data assimilation in numerical weather prediction (NWP).
  • Reanalysis production: ERA-CLIM2 generated innovative reanalysis datasets, such as the first European coupled ocean–land–atmosphere reanalysis of the 20th century (CERA-20C), which also includes the carbon cycle; a nine-year higher-resolution coupled reanalysis of the satellite era (CERA-SAT, see the separate article by Dinand Schepers in this Newsletter); and an ocean reanalysis of the 20th century, ORA-20C.
  • Evaluation and uncertainty estimation: ERA-CLIM2 has advanced our understanding of the quality of uncoupled and coupled reanalyses, and it has led to the development of methods for estimating uncertainty in reanalyses.

Understanding climate change is highly dependent on the availability of global satellite and conventional observational data for the atmosphere, the land, the ocean and sea ice, and on the development of coupled ocean–land–atmosphere models and assimilation systems that can ingest these data. A continuous cycle of research and development activities in data assimilation, data rescue, observation reprocessing, production, and diagnosis and evaluation is required to improve future reanalyses, so that they can provide a continuously improving depiction of the evolution of the Earth system.

Ocean heat content

CERA-20C and CERA-SAT are ensembles of coupled reanalyses, the first of their kind, a fact that opens new opportunities but also poses new challenges for evaluation. The distribution of the ensemble members should be a realistic representation of the uncertainty associated with the reanalysis. To check this, one can for example compare the reanalysis against other reanalysis products or against independent observations. A comparison of the ocean heat content (OHC) of the upper 300 m in the CERA-20C ensemble and in dedicated ocean reanalyses shows good agreement overall. The large spread among the 10 members of the CERA-20C reanalysis before 1950 is consistent with the sparse observation coverage during that time. It provides a measure of the uncertainty in the reanalysed state of the ocean during those decades. Short-term variations at other times are caused by internal variability and/or volcanic eruptions, of e.g. Agung (1963), El Chichon (1982) and Pinatubo (1991). These eruptions all led to a temporary cooling of the upper ocean. The increase in OHC since 1970 is consistent with observations and is the manifestation of global warming in the ocean state.

CERA-20C also performs well in comparisons of different ensembles against independent observations. For example, a detailed evaluation of precipitation in CERA-20C and ERA-20C (the uncoupled reanalysis produced in the ERA-CLIM project that preceded ERA-CLIM2) with rain gauge-based precipitation estimates from GPCC (Global Precipitation Climatological Center) indicates overall better quality for CERA-20C, particularly in Africa and the Indian Monsoon region.


The Copernicus Climate Change Service has been able to transfer the ERA-CLIM2-funded research and development into operational systems and services. It has also succeeded in integrating the outcomes of ERA-CLIM2 and the other four precursor projects – CLIPC (Climate information portal), EUCLEIA (European climate and weather events: interpretation and attribution), QA4ECV (Quality assurance for essential climate variables) and UERRA (Uncertainties in ensembles of regional reanalyses) – into user-oriented applications.

Examples of ERA-CLIM2 activities with a high potential for use in C3S are:

  • Development of a coupled assimilation system (three-dimensional ocean, sea ice, land and atmosphere) that can be used to generate climate reanalyses for the 20th century and consistently derived reanalyses of the global carbon cycle; the version of the CERA (Coupled ECMWF ReAnalysis) system used to generate CERA-SAT has been handed over to C3S so that it can be used as a prototype coupled assimilation system to generate the next C3S reanalysis, ERA6.
  • Development of a global data rescue registry to keep track of in situ climate data rescue efforts worldwide, based on activities started in ERA-CLIM and pursued in collaboration with the ACRE initiative and many other existing data rescue efforts and projects.
  • Development of a global in situ snow data collection based on work started in the CORE-CLIMAX coordination activity.
  • Further development of the ERA-CLIM Observation Feedback Archive and associated tools.

Numerous quality-controlled historical weather observations with potential high impact for climate reanalysis rescued and/or post-processed within ERA-CLIM2 have already been added to international public data collections such as ICOADS (International Comprehensive Ocean–Atmosphere Data Set) and ISPD (International Surface Pressure Databank), for use in reanalysis and other climate applications.

Website and data access

The ERA-CLIM2 project website ( includes information about the project and all reports associated with the project’s deliverables, links to data portals and links to relevant reanalysis websites. It also includes links to web pages on the four project General Assemblies, where all presentations can be accessed and downloaded.

ERA-CLIM2 reanalyses can be accessed from dedicated data portals, such as the MARS archive catalogue portal: Links to other sites where data can be accessed can be found on the ERA-CLIM2 website:

ERA-CLIM2 was funded under the European Union Grant Agreement No. 607029 under the title ‘European Reanalysis of the Global Climate System’. Its successful conclusion has been made possible by the outstanding commitment and work put in by the partner organisations and individuals involved in the project.

The ERA-CLIM2 Consortium:ECMWF (co-ordinating organisation) • UK Met Office • EUMETSAT • Universität Bern, Switzerland • Universität Wien, Austria • FCiências.ID, Portugal • All-Russian Research Institute of Hydrometeorological Information – World Data Centre • Mercator Océan, France • Météo-France • DWD, Germany • CERFACS, France • CMCC, Italy • Finnish Meteorological Institute (FMI) • The University of Reading, UK • Inria, France • Université de Versailles Saint-Quentin-en-Yvelines, France