The Cambridge p-TOMCAT model is a 3D global Eulerian model, driven by meteorological fields (winds, temperatures and humidities) from the operational analyses of the European Centre for Medium-range Weather Forecasts. For the purposes of this work, it was run at a resolution of ~2.8° x ~2.8° on 31 hybrid sigma-pressure levels. The levels stretch from the surface to 10hPa with a spacing of approximately 100m in the boundary layer and 1-1.5km in the vicinity of the tropopause. The model’s chemistry includes 52 species and 174 reactions that together describe the gas-phase HOx/NOx chemistries of methane, ethane, propane and isoprene; isoprene oxidation follows the Mainz Isoprene Mechanism (Pöschl et al., 2000).

Five experiments are carried out. The first comprises a PI model run, employing appropriate trace-gas emissions (Valdes et al., 2005); the remaining four comprise highly idealised experiments derived from the PI model run. In the first two, the annual-mean concentration of methane in the Antarctic boundary layer is reduced from its PI value of ~710 ppbv to ~345 ppbv, roughly the concentration of methane present at the LGM, either by increasing the production of OH by a factor of 2.5 (Sink 1) or by uniformly scaling-down the emissions of methane by a factor of 0.55 (Source 1). In the remaining two experiments, the concentration of methane is reduced to ~555 ppbv by removing all non-methane emissions from vegetation (Sink 2) or by scaling-down the emissions of methane by a factor of 0.83 (Source 2).

In each experiment, the model is run to equilibrium with repeated meteorological fields from a single year (1997). The model is subsequently run for three consecutive years (1998-2000) to explore inter-annual variations. It is during these three years that the data contained in this dataset are gathered. The combination of five experiments and three years of meteorology gives rise to a total of fifteen files, each containing fully 3D monthly-mean chemical fields in addition to some meteorological fields (see below). They follow the nomenclature, desire_bas_p tomcat_EXPERIMENT_METEOROLOGY.nc, where EXPERIMENT=’pi’, ‘sink1’, ‘source1’,’sink2’ or ‘source2’; and METEOROLOGY=’1998’, ‘1999’ or ‘2000’).