Ten years of CO₂, CH₄, CO and N₂O fluxes over Western Europe inferred from atmospheric measurements at Mace Head, Ireland

Abstract. We estimated CO₂, CH₄, CO and N₂O emission fluxes over the British Isles and Western Europe using atmospheric radon observations and concentrations recorded at the Mace Head Atmospheric Research Station between 1996 and 2005. We classified hourly concentration data into either long-range European or regional sources from Ireland and UK, by using local wind speed data in conjunction with ²²²Rn and ²¹²Pb threshold criteria. This leads to the selection of about 7% of the total data for both sectors. We then used continuous ²²²Rn measurements and assumptions on the surface emissions of ²²²Rn to deduce the unknown fluxes of CO₂, CH₄, CO and N₂O. Our results have been compared to the UNFCCC, EMEP and EDGAR statistical inventories and to inversion results for CH₄. For Western Europe, we found yearly mean fluxes of 4.1±1.5 10⁶ kg CO₂ km⁻² yr⁻¹ , 11.9±2.0 10³ kg CH₄ km⁻² yr⁻¹, 12.8±4.2 10³ kg CO km⁻² yr⁻¹ and 520.2±129.2 kg N₂O km⁻² yr⁻¹, respectively, for CO₂, CH₄, CO and N₂O over the period 1996–2005. The method based upon ²²²Rn to infer emissions has many sources of systematic errors, in particular its poorly known and variable footprint, uncertainties in ²²²Rn soil fluxes and in atmospheric mixing of air masses with background air. However, these biases are likely to remain constant in the long-term, which makes the method quite efficient to detect trends in fluxes. Over the last ten years period, the decrease of the anthropogenic CH₄, CO and N₂O emissions in Europe estimated by inventories (respectively −30%, −35% and −23%) is confirmed by the Mace Head data within 2%. Therefore, the ²²²Rn method provides an independent way of verification of changes in national emissions derived from inventories. Using European-wide estimates of the CO/CO₂ emission ratio, we also found that it is possible to separate the fossil fuel CO₂ emissions contribution from the one of total CO₂ fluxes. The fossil fuel CO₂ emissions and their trends derived in that manner agree very well with inventories.