Articles | Volume 17, issue 3
Atmos. Chem. Phys., 17, 1945–1983, 2017
Atmos. Chem. Phys., 17, 1945–1983, 2017

Research article 09 Feb 2017

Research article | 09 Feb 2017

The CAMS interim Reanalysis of Carbon Monoxide, Ozone and Aerosol for 2003–2015

Johannes Flemming1, Angela Benedetti1, Antje Inness1, Richard J. Engelen1, Luke Jones1, Vincent Huijnen2, Samuel Remy3, Mark Parrington1, Martin Suttie1, Alessio Bozzo1, Vincent-Henri Peuch1, Dimitris Akritidis4, and Eleni Katragkou4 Johannes Flemming et al.
  • 1European Centre for Medium-Range Weather Forecasts, Reading, UK
  • 2Royal Netherlands Meteorological Institute, De Bilt, the Netherlands
  • 3Laboratoire de météorologie dynamique, UPMC/CNRS, Paris, France
  • 4Department of Meteorology and Climatology, Aristotle University of Thessaloniki, School of Geology, Thessaloniki, Greece

Abstract. A new global reanalysis data set of atmospheric composition (AC) for the period 2003–2015 has been produced by the Copernicus Atmosphere Monitoring Service (CAMS). Satellite observations of total column (TC) carbon monoxide (CO) and aerosol optical depth (AOD), as well as several TC and profile observations of ozone, have been assimilated with the Integrated Forecasting System for Composition (C-IFS) of the European Centre for Medium-Range Weather Forecasting. Compared to the previous Monitoring Atmospheric Composition and Climate (MACC) reanalysis (MACCRA), the new CAMS interim reanalysis (CAMSiRA) is of a coarser horizontal resolution of about 110 km, compared to 80 km, but covers a longer period with the intent to be continued to present day. This paper compares CAMSiRA with MACCRA and a control run experiment (CR) without assimilation of AC retrievals. CAMSiRA has smaller biases than the CR with respect to independent observations of CO, AOD and stratospheric ozone. However, ozone at the surface could not be improved by the assimilation because of the strong impact of surface processes such as dry deposition and titration with nitrogen monoxide (NO), which were both unchanged by the assimilation. The assimilation of AOD led to a global reduction of sea salt and desert dust as well as an exaggerated increase in sulfate. Compared to MACCRA, CAMSiRA had smaller biases for AOD, surface CO and TC ozone as well as for upper stratospheric and tropospheric ozone. Finally, the temporal consistency of CAMSiRA was better than the one of MACCRA. This was achieved by using a revised emission data set as well as by applying careful selection and bias correction to the assimilated retrievals. CAMSiRA is therefore better suited than MACCRA for the study of interannual variability, as demonstrated for trends in surface CO.

Short summary
We combine satellite observations of carbon monoxide, ozone and aerosols with the results from a model using a technique called data assimilation. The generated global data set (CAMS interim reanalysis) covers the period 2003–2015 at a resolution of about 110 km. The CAMS interim reanalysis can be used to study global air pollution and climate forcing of aerosol and stratospheric ozone. It has been produced by the Copernicus Atmosphere Monitoring Service (http://atmosphere.
Final-revised paper