Source apportionment and chemical characterisation of airborne fine particulate matter in Christchurch, New Zealand

Environment Canterbury has a proposed plan to address elevated wintertime particle concentrations in Christchurch and other Canterbury towns. The proposed plan sets out air quality objectives for Canterbury and specifically targets the main source of wintertime particles, residential heating. Environment Canterbury is reliant on the plan strategies to meet the targets specified, and to meet the recently established National Environmental Standard for PM₁₀. The contribution of residential heating, relative to other sources, has been quantified by emission inventory (coupled with a box model). The appropriateness of measures contained in the proposed plan is debated by some, who are sceptical of the inventory results. Progress in air quality management in Christchurch would be facilitated by improving the understanding of PM ₂.₅ and its sources by undertaking an alternative approach to source attribution. A source apportionment study was therefore conducted during 2001/2002 to address these issues. This aimed to determine the chemical composition of fine particles (PM ₂.₅) in Christchurch, investigate and trial a receptor model, identify the main sources contributing to PM ₂.₅ and estimate relative source contributions with a view to using the technique to monitor plan strategy effectiveness. This is the first time that the Positive Matrix Factorisation (PMF) receptor model has been applied to seasonal PM ₂.₅ data in New Zealand. PM ₂.₅ samples were collected at St Albans, Christchurch during 2001 and 2002. The filters were analysed for gravimetric mass, elemental species and inorganic ions (24-hour averages). Organic and elemental carbon were measured using an automatic carbon analyser. The final dataset comprised 159 observations and 15 chemical species. PM ₂.₅concentrations were seasonally distributed with maximum concentrations occurring in winter. Major elemental constituents were organic and elemental carbon, sulphate, nitrate, sodium, chlorine and sulphur. Combustion-related species were highest in winter and those derived from natural sources (e.g. marine aerosol) predominant in summer. Five factors were resolved using the PMF receptor model: these were identified as wood combustion, marine aerosol, motor vehicles, secondary particulate and aged aerosol. On average, 92% of PM ₂.₅mass was explained by the resolved factors. The most important source of PM ₂.₅, on average, during summer was aged aerosol and in winter was wood combustion. The results were evaluated against inventory-derived data; both methods clearly identified wood combustion as being the predominant source of peak wintertime PM ₂.₅ (92% and 89%, respectively). These results provide independent verification that residential heating is the greatest contributor to wintertime particles, adding to the "weight of evidence" required to justify strategies adopted in Environment Canterbury's proposed plan. Further use of receptor models was recommended as they provide source information unavailable from emission inventories/including contributions from natural sources and secondary particulate, and may be used to track effectiveness of policy implementation