Assessing atmospheric composition impacts using a chemical climatology framework: Case studies at the UK monitoring supersites
Malley, Christopher Stuart
In the mid-1800s, monitoring networks were established to investigate atmospheric composition impacts, and the conditions giving rise to them. The development of these networks, in terms of coordination and standardisation between contributing sites, has resulted in large advances in knowledge of the nature of atmospheric composition. Currently thousands of sites collect high quality atmospheric composition measurements globally. This thesis contends that in order to maximise the information derived from these measurements, a further advancement in standardisation is required to encompass the interpretation of monitoring network data. Currently there are limited examples of a common interpretation of data applied across all sites in a monitoring network, especially in relation to specific atmospheric composition impacts. In this thesis, a ‘chemical climatology’ framework is outlined which provides a common basis for targeting analysis towards identifying the linkage between a specific atmospheric composition impact and its causal drivers. Case studies apply the chemical climatology framework to demonstrate its utility in deriving scientific and policy relevant conclusions using measurement data from the UK monitoring supersites located at Harwell and Auchencorth. Prior to this, the representativeness of each site is quantified through the application of cluster analysis to ozone data at 100 rural European sites to identify groupings of sites with similar ozone variation. Harwell was representative of rural locations within 120 km of London, while Auchencorth was representative of a larger, transboundary spatial domain including the remainder of the rural UK. The first case study links the impact of ozone on human health (quantified by SOMO10 and SOMO35 metrics) and vegetation (flux-based PODY) to meteorological and emissions drivers. Between 1990 and 2013 at Harwell, there was a significant decrease in the contribution of European ozone to determining the impacts. Improvement in the human health impact was heavily dependent on the choice of metric (SOMO35 decreased, no change in SOMO10), and the vegetation impacts had not improved as high ozone episodes frequently coincided with plant conditions which reduced ozone uptake. These chemical climates emphasise the need for ozone mitigation on larger (hemispheric) scales than currently implemented. Secondly, the impact of 27 measured VOCs on the extent of the regional ozone increment is assessed. The photochemical loss of VOCs is then linked to reported gridded VOC emissions using air mass back trajectory analysis. Ethene and m+p-xylene had the largest diurnal photochemical loss during maximum monthly regional ozone increment, but the key conclusion was the limitation introduced through the reporting of gridded VOC emissions in heavily aggregated source sectors. Finally, the conditions producing the long term health impact of particulate matter (quantified by annual average PM10 and PM2.5 concentrations) at each site are derived through integration of measurements of PM10 and PM2.5 with measurements of PM constituents. It is shown that the frequent, moderate PM10 and PM2.5 concentrations made a larger contribution to annual average values compared to the relatively infrequent high, episodic concentrations. The contribution of PM constituents and the contribution of local vs regional emissions to the range of PM concentrations is investigated. It was concluded that similar reductions in the contribution of secondary inorganic aerosol to the moderate PM10 and PM2.5 concentrations could be achieved from both the reduction of frequently traversed, smaller emissions sources, and less frequently traversed, larger emissions sources. The final chapter demonstrates the benefits from the extension of this framework to an entire monitoring network. It is envisioned that for each atmospheric composition impact, a standard set of statistics would be calculated which quantify the ‘impact’, ‘state’ and ‘drivers’ of that chemical climate. Calculation of ozone human health chemical climates across 100 European monitoring sites demonstrate this concept. This standardised interpretation of monitoring network data not only allows consistent comparison of an impact, but the common basis for determining how the impact is derived allows for the consideration of novel mitigation strategies and their spatial applicability.