Consistency of reactive trace gases in the integrated reanalysis and forecast run within the GEMS system

An overview of the GEMS-system consistency from the point of view of the reactive trace gases is given, with a focus on the consistency between the different GEMS-GRG models, as well as between the GRG and the RAQ and GHG subprojects.
It is shown that the average magnitude and the seasonal variation in ozone columns in the reanalysis system are well captured. Also CO is on average well in line with the stand-alone model versions, with some exceptions over regions where different emission inventories are used in the stand-alone runs.
At higher latitudes in the northern hemisphere formaldehyde in MOZART results in higher columns compared to TM5, especially during autumn and winter months. With respect to tropospheric NO2, during summer months both TM5 and the global forecast run
show a low bias over the continents, compared to OMI. The very high NO2 levels in the eastern part of China are not captured, which points at the use of an outdated emission inventory. SO2 in the stand-alone versions of MOZART and TM5 are generally in line with one another. The monthly averaged OH zonal average concentrations between in the MOZART stand alone run shows a confined concentration at the surface, compared to TM5 and the Spivakovsky climatology, which is used in the GHG system to evaluate the methane-oxidation.
The monthly average burdens in TM5 and MOZART agree with the climatology concerning its seasonal cycle, but the total amount is lower by about 20 %.
Methane fields, which are prescribed at the surface in MOZART, are globally in line with a 4dVar analysis as performed in the GHG-subproject.