Articles | Volume 5, issue 9
https://doi.org/10.5194/acp-5-2431-2005
https://doi.org/10.5194/acp-5-2431-2005
21 Sep 2005
21 Sep 2005

Inverse modelling of national and European CH4 emissions using the atmospheric zoom model TM5

P. Bergamaschi, M. Krol, F. Dentener, A. Vermeulen, F. Meinhardt, R. Graul, M. Ramonet, W. Peters, and E. J. Dlugokencky

Abstract. A synthesis inversion based on the atmospheric zoom model TM5 is used to derive top-down estimates of CH4 emissions from individual European countries for the year 2001. We employ a model zoom over Europe with 1° × 1° resolution that is two-way nested into the global model domain (with resolution of 6° × 4°. This approach ensures consistent boundary conditions for the zoom domain and thus European top-down estimates consistent with global CH4 observations. The TM5 model, driven by ECMWF analyses, simulates synoptic scale events at most European and global sites fairly well, and the use of high-frequency observations allows exploiting the information content of individual synoptic events. A detailed source attribution is presented for a comprehensive set of 56 monitoring sites, assigning the atmospheric signal to the emissions of individual European countries and larger global regions.

The available observational data put significant constraints on emissions from different regions. Within Europe, in particular several Western European countries are well constrained. The inversion results suggest up to 50-90% higher anthropogenic CH4 emissions in 2001 for Germany, France and UK compared to reported UNFCCC values (EEA, 2003). A recent revision of the German inventory, however, resulted in an increase of reported CH4 emissions by 68.5% (EEA, 2004), being now in very good agreement with our top-down estimate. The top-down estimate for Finland is distinctly smaller than the a priori estimate, suggesting much smaller CH4 emissions from Finnish wetlands than derived from the bottom-up inventory. The EU-15 totals are relatively close to UNFCCC values (within 4-30%) and appear very robust for different inversion scenarios.

Download
Altmetrics
Final-revised paper
Preprint