Interactive chemistry in the Laboratoire de Meteorologie Dynamique general circulation model: Description and background tropospheric chemistry evaluation

[1] We provide a description and evaluation of LMDz-INCA, which couples the Laboratoire de Meteorologie Dynamique general circulation model (LMDz) and the Interaction with Chemistry and Aerosols (INCA) model. In this first version of the model a CH4 - NOx - CO - O-3 chemical scheme representative of the background chemistry of the troposphere is considered. We derive rapid interhemispheric exchange times of 1.13 - 1.38 years and 0.70 - 0.82 years, based on surface and pressure-weighted mixing ratios of inert tracers, respectively. The general patterns of the nitrogen deposition are correctly reproduced by the model. However, scavenging processes remain a major source of uncertainty in current models, with convective precipitation playing a key role in the global distribution of soluble species. The global and annual mean methane (7.9 years) and methylchloroform (4.6 years) chemical lifetimes suggest that OH is too high by about 19 - 25% in the model. This disagreement with previous estimates is attributed to the missing nonmethane hydrocarbons in this version of the model. The model simulates quite satisfactorily the distribution and seasonal cycle of CO at most stations. At several tropical sites and in the Northern Hemisphere during summer, the OH overestimate leads, however, to a too intense CO chemical destruction. LMDz-INCA reproduces fairly well the distribution of ozone throughout most of the troposphere. A main disagreement appears in the Northern Hemisphere upper troposphere during summer, due to a too high tropopause in the GCM. When the GCM winds are relaxed toward assimilated meteorology, a much higher variability is obtained for ozone in the upper troposphere, reflecting more frequent stratospheric intrusions. The stratospheric influx of ozone increases from 523 Tg/yr in the base case simulation to 783 Tg/yr in the nudged version.