Monitoring atmospheric composition & climate

MACC-II GRG will provide continuation of the global atmospheric composition service for the stratosphere and troposphere, further develop the dissemination of products, and continue the development of the C-IFS model including some first elements for source inversions.  The main service lines of GRG are:

  • analyses and forecasts of stratospheric ozone by the three indepent assimilation systems IFS-MOZ, BASCOE and SACADA (7);
  • continued reanalysis of total stratospheric ozone columns with the multi-sensor TM3DAM assimilation systems;
  • analyses and forecasts of global tropospheric composition using IFS-MOZ and IFS-TM5 with a focus (8) on ozone, NOx and CO, but including several other reactive gases compounds;
  • provision of global boundary conditions for regional air quality modellers world-wide
  • support of the analysis of extreme events through dedicated sensitivity runs and data analysis

Another major component of the work plan is the continued development of the integrated C-IFS system (Chemistry modules coded into the IFS model instead of running coupled in parallel), which is planned to become operational towards the end of MACC-II.  In particular, data assimilation will be enabled in C-IFS, mass conservation issues will be addressed and various physical and chemical parameterizations will be improved to make them more consistent with the IFS's physical parameterizations.  Furthrmore, ECMWF will improve portability of the C-IFS code with a view to it becoming a community model in the future.

The GRG work plan also includes linmited targeted research in order to investigate problems in the satellite data products, improve model parameterizations or find and fix model errors, adapt the system to new data sets or to include new features that were identified as important in research activities outside of MACC.  This includes testing inverse methods to constrain emission sources of CO (in collaboration with EMI, GHG and GDA).

All Stratospheric and tropospheric products products will be accessible through dedicated web portals (the stratospheric portal will be hosted by BIRA-IASB and the tropospheric portal will be hosted by ECMWF), as well as through the fully Open Geospatial Consortium (OGC)-compliant Web coverage service (WCS) at Forschungszentrum Julich.  Validation results from ECMWF's online verification as well as those from VAL project will be closely linked to the data products to facilitate the assessment of product quality for users.

GRG in MACC-II differs from G-RG in MACC through the seperation of production and development activities (in this sub-project) from validation activities (in the new VAL sub-project).  It is recognised that close interactions between these sub-projects will be necessary in order to ensure efficient communication about service quality and scientific model problems.  GRG continues to evaluate the model systems used in the production chain and those under development, however, the clear focus of these activities is on process characterisation and model improvemnet, while VAL places its emphasis on the independent assessment of product quality.

Satellite data in MACC-II GRG

The Copernicus service architecture  envisages that the Level-2 satellite data products that are used (i.e. assimilated) in the atmospheric core service will be provided by space agencies such as ESA, EUMETSAT, NOAA, NASA and JAXA.  However, there is a line of "research satellite data products" that were developed with ESA funding in the past, and that are generated at various research institutions in near real time.  As a result of the integration of PROMOTE and GEMS activities into MACC and for lack of alternative funding sources for these products, MACC includes the NRT production of these data sets.  In MACC-II, EUMETSAT (GOME-2) data will be provided through the Ozone and Atmospheric Chemistry Monitoring SAF (see OBS sub-project specification), but the GRG sub-project has to include continued provision of some research satellite data products from ESA instruments to ensure continuity.  The various liaison and coordination mechanisms available to the MACC-II consortium will be used dueing the project to clarify the arrangements for supply of Level-2 satellite products from future missions, especially from Sentinel series.

The satellite data to be assimilated in NRT by the GRG systems during MACC-II are set out in the tables below:


Data products to be assimilated into IFS-MOZ and IFS-TM5


Data products to be assimilated into SACADA


Data products to be assimilated into BASCOE



The following should be noted:

  • The above tables include some data that currently in MACC are assimilated passively, or whose assimilation is planned.  Active assimilation of these data in MACC-II will be dependednt on successful results of passive assimilation.
  • In order to enhance the quality of the assimilation (error characterisation, coverage) and to ensure robustness of the forecasts (availability of data for the analysis) it is important to have parallel data streams running for different instruments
  • The operation of Envisat is expected to terminate in 2013.  The GRG planning assumes that data delivery for Envisat will stop at the end of 2013
  • Experience in GEMS and MACC has shown that satellite SO2 columns are not accurate enough to constrain atmospheric SO2 concentration fields in general.  They can however be used to detect and monitor large pollution plumes, especially from volcanic eruptions.  Test experiemnets in MACC demonstrated successful assimilation of SO 2 columns dueing such events.  For this reason the NRT data stream of SO2 columns from SCIAMACHY by BIRA will be maintained.  The remark on the expected lifetime of Envisat applies here also.
  • Assimilation of HCHO (formaldehyde) column retrievals from GOME and SCIAMACHY was tested in GEMS and MACC.  In general the quality of satellite HCHO data is not sufficient to constrain the forecast model, but the data might be useful validation purposes.  BIRA-IASB has agreed to provide SCIAMACHY HCHO data in delayed mode to ECMWF free of charge.
  • Work package WP53 will investigate assimillation of TES satellite products for ozone and CO.
  • If the Sentinel-5 precursor is launched during MACC-II and the data becomes available early enough, trail assimilation runs with ozone and CO data from this instrument will be carried out.

Model systems in MACC-II GRG

The main operational model systems in the MACC-II GRG sub-project will be coupled IFS-MOZ and IFS-TM5 systems, which will be used for routine analyses and forecasts and for sensitivity experiments during extreme events.  The uncoupled model versions of TM5 and MOZART will be employed for specific model developments and for testing of model improvements.  The MACC-II stratospheric service consists of the three independent NRT systems IFS-MOZ, SACADA and BASCOE.  A fourth model system (TM3DAM) is used specifically for the reanalysis of total ozone columns,  Descriptions of all the models mentioned above and their set-up will be made available through the project web site.

C-IFS, the new chemistry model that has been developed during the first phase of MACC, has already demonstrated reasonable success in simulating tropospheric ozone concentrations.  However, further developments and careful evaluation are needed before the new model can replace the coupled model systems.  GRG planning is that C-IFS will begin to run in parallel with the operational models in the second half of MACC-II.  If the evaluation of C-IFS results show at least the same quality as the current systems, then the coupled models will be taken out of operation during the final months of MACC-II.  The RTD activities in MACC-II GRG largely concentrate on the following C-IFS developments:

  • Introducing a flexible interface for different chemistry schemes:
  • enabling data assimilation with C-IFS;
  • analysing seasonal trace gas budgets and mass conservation issues;
  • improving the mass fixer to balance mass losses/gains due to the transport scheme;
  • upgrading resolution from T159/L60 to T255/L91 (including elevated model top of 0.01hpa) and keeping track of changes in the operational IFS system;
  • interfacing with the IFS-physics to improve wet deposition, photolysis rates and radiation feedback;
  • updating emission data and deposition velocities to output of SUMO;
  • improving portability of the C-IFS code so that it may serve as a community model.