European Reanalysis 20th Century (ERA-20C)

European Reanalysis 20th Century (ERA-20C)
Name	European Reanalysis 20th Century
Acronym	ERA-20C
Producer	European Centre for Medium-Range Weather Forecasts
Period	1900–2010
Resolution	~125 km (T159)
Variables	3D atmospheric fields, surface pressure, temperature, winds, humidity

European Reanalysis 20th Century (ERA-20C) is a global atmospheric reanalysis covering 1900–2010 produced by the European Centre for Medium-Range Weather Forecasts using early-observation assimilation to produce consistent historical estimates of atmospheric states. It supports studies in climate change, historical climatology, paleoclimatology, and meteorology by providing long-term, gridded datasets of pressure, temperature, winds, and humidity. ERA-20C underpins research linking early 20th century variability to later observational networks, and serves as a reference for comparisons with other reanalyses and coupled climate models.

Overview

ERA-20C is an observationally constrained global atmospheric dataset spanning 1900–2010 produced by the European Centre for Medium-Range Weather Forecasts as part of the ERA series alongside ERA-Interim and ERA5. It uses a 20th-century observing system with surface and marine pressure observations to reconstruct atmospheric circulation at synoptic and climatic scales, providing fields such as geopotential height, temperature, and wind at multiple pressure levels. ERA-20C has been used in studies involving El Niño–Southern Oscillation, North Atlantic Oscillation, Arctic Oscillation, Pacific Decadal Oscillation, and investigations of extreme events like the 1922 Knickerbocker Storm, 1930s Dust Bowl, and Great Storm of 1987.

Development and data sources

The development of ERA-20C built on archival efforts by institutions including the International Maritime Organization, World Meteorological Organization, National Oceanic and Atmospheric Administration, and national meteorological services such as the Met Office, Deutscher Wetterdienst, and Météo-France. Primary observational inputs included digitized station records from the International Surface Pressure Databank, marine logbooks from the Hudson's Bay Company and Royal Navy, and early aerological soundings archived by the U.S. Weather Bureau and Deutscher Wetterdienst. Satellite-era datasets from NOAA and European Space Agency were excluded to preserve a consistent pre-satellite observing system. Collaborative projects such as the Atmospheric Circulation Reconstructions over the Earth (ACRE) initiative and data rescue efforts from the National Aeronautics and Space Administration archives contributed digitized pressure and temperature observations.

Methodology and assimilation system

ERA-20C employed the Integrated Forecasting System (IFS) model developed at the European Centre for Medium-Range Weather Forecasts coupled with a 4D-Var data assimilation framework adapted for century-scale reconstruction, leveraging versions of the IFS used in ECMWF operational forecasting in the early 2010s. The assimilation ingested surface pressure and marine pressure observations to constrain the model’s initial conditions while excluding satellite radiances to maintain homogeneity across the record. Background error covariances, model physics schemes, and observation operators were tuned using techniques refined in reanalysis efforts such as NCEP/NCAR Reanalysis and JRA-55. Ensembles of assimilations and sensitivity experiments addressing aerosol forcing, sea surface temperature forcing from HadISST, and assimilation window choices were used to assess robustness.

Validation and performance

Validation of ERA-20C involved comparisons with independent datasets including radiosonde records archived by the International Civil Aviation Organization, station temperature series such as Central England Temperature, and marine datasets like the ICOADS collection. Performance metrics showed ERA-20C reproduced large-scale indices like the North Atlantic Oscillation and Southern Annular Mode with reasonable fidelity compared to reanalysis peers, though skill varied regionally and temporally. Quantitative assessments referenced comparisons with ERA-Interim, ERA5, NCEP/NCAR Reanalysis, and MERRA, as well as event-based evaluations for historical storms studied in archives from the Met Office and the Royal Meteorological Society.

Applications and use cases

ERA-20C has been applied in diverse studies: attribution of long-term climate trends in analyses associated with the Intergovernmental Panel on Climate Change; reconstruction of historical cyclone tracks informing research by the International Best Track Archive for Climate Stewardship; driving historical land surface and hydrological experiments in models used by the European Commission and World Climate Research Programme; and providing boundary conditions for regional climate downscaling efforts by groups at ETH Zurich, University of East Anglia, and CNRS. It also supports interdisciplinary work linking meteorological reconstructions to societal impacts in World War I and World War II archival studies, as well as analyses of early aviation weather hazards.

Limitations and uncertainties

Limitations of ERA-20C arise from sparse early-20th-century observations, inhomogeneities in station coverage, and omission of satellite radiances, leading to larger uncertainties in the pre-1950 period over data-sparse regions such as parts of the Southern Ocean, Antarctica, and interior Africa. Biases may result from uncertainties in sea surface temperature datasets like HadISST and from changes in marine observing practices documented by the International Maritime Organization. Users should consider ensemble spread, known weaknesses compared to ERA5, and caution when interpreting small-scale features or trends that may be sensitive to assimilation choices influenced by archives from the National Meteorological Center era.

Availability and access

ERA-20C data are distributed by the European Centre for Medium-Range Weather Forecasts through the ECMWF data portal and have been made accessible to researchers and institutions such as Copernicus, World Data Center, and university consortia. Data access typically requires registration with ECMWF and complies with licensing agreements used for other ECMWF products; subsets and derived diagnostics are available through portals operated by C3S, EDA nodes, and partner repositories including the British Atmospheric Data Centre and the NOAA National Centers for Environmental Information.

Category:Climate datasets Category:Reanalysis datasets Category:European Centre for Medium-Range Weather Forecasts