High resolution modelling of particulate matter air quality in the UK with a focus on carbonaceous aerosol
The Earth’s atmosphere consists of both gaseous and condensed-phase components, the condensed-phase material is called particulate matter (PM). The effects of atmospheric PM include adverse health impacts, as well as climate forcing. Both qualitative and quantitative knowledge about PM is necessary to assess these effects, and to devise best mitigation strategies. Understanding the distribution of atmospheric particulate matter is complex because much of it is of secondary origin rather than from primary emissions. Furthermore, there are multiple anthropogenic and natural sources of the contributing precursors, and all these processes are influenced by atmospheric conditions and transport. In this work, one of the major constituents of atmospheric PM - carbonaceous aerosol - is studied. A regional application of the EMEP MSC-W atmospheric chemical transport model - EMEP4UK - was used to model air pollution over the British Isles with a horizontal resolution of 5 km x 5 km. One-way nesting was used from the European computational domain of 50 km x 50 km to the finer spatial grid of EMEP4UK. Several model experiments were devised in order to investigate the well-known deficiency that models currently underestimate organic aerosol (OA) concentrations compared with observations. The model experiments were evaluated with comprehensive year-long novel measurements from the Clear Air for London (ClearfLo) campaign in 2012. Several sources of organic aerosol that are either missing, greatly underestimated, or may be spatially misplaced in official emissions inventories were re-evaluated. Firstly, missing diesel-related intermediate volatility organic compound (IVOC) emissions from diesel vehicles derived directly from field measurements at the urban background site during the 2012 ClearfLo campaign were added into the model. According to the model simulations, these diesel-IVOCs can explain on average ~30% of the annual secondary organic aerosol (SOA) in and around London. Furthermore, the 90- th percentile of modelled daily SOA concentrations for the whole year was 3.8 μgm-3, constituting a notable addition to total particulate matter. More measurements of these precursors (currently not included in official emissions inventories) is recommended. Secondly, spatially and temporally resolved emissions of cooking OA (COA; emissions from meat charbroiling, or frying and deep-frying) were developed. These emissions are currently neglected in European emissions inventories, yet measurements point to significant COA contribution to ambient PM concentrations (up to 2.0 μgm-3 on annual average for central London). The final COA emission source strength derived here (320 mg person-1 day-1) was spatially distributed to workday population density (as opposed to residential population density). The impact of COA on surface concentrations is spatially very limited, however, as the modelled concentrations dropped markedly outside of urban areas. For example, annual average modelled concentration for the Harwell location was just 0.1 μgm-3. Thirdly, redistributing 50% of non-industrial wood and coal burning emissions to residential population density (thus over-writing, in part, the assumption made by the national emissions inventory that only smokeless fuels are burned in smoke control areas) increased the modelled solid fuel OA (SFOA) concentration at the London North Kensington site to 0.8 μgm-3, from the Base run value (using the emissions’ spatial distribution and total as officially reported) of just 0.3 μgm-3. For comparison, the measured annual mean concentration of SFOA at this site was 1.0 μgm-3. Based on the model evaluation presented, redistribution of SFOA emissions into smoke control areas is justified, but further refinement of the amount, as well as the temporal emission profile of this component is necessary. The total effect of the three refinements undertaken in this work increased the model estimate of the annual mean OA concentration at the London North Kensington site from 1.8 μgm-3 to 3.8 μgm-3, which is much closer to the observed value of 4.2 μgm-3. Thus, this work has provided relevant insight into the nature and magnitude of missing, under-represented, and spatially inappropriately-distributed emissions of primary OA and OA precursors. Although the study area was focused on pollutant concentrations over the British Isles, all of the components examined here are of great relevance to the air quality in other countries as well — in Europe and globally. Therefore, the inclusion of these improvements into other air quality models and official emissions’ inventories is advised.