A climatology of particulate pollution in Christchurch
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
The research presented in this thesis provides a quantitative analysis of atmospheric influences on particulate matter pollution in Christchurch across a wide range of spatial and temporal scales. A complex interaction of low level flow characteristics that form in response to local and regional features of complex terrain, together with an urban setting that is characterised by low density housing, mostly comprised of single storey dwellings that are poorly insulated, regularly leads to nocturnal smog events during winter in Christchurch. Provided synoptic flow is weak, the above mentioned flow interaction promotes flow stagnation over the city, when nocturnal katabatic drainage flows and day-time north-easterly on-shore winds converge over the city. Additionally, undercutting of the density currents promotes highly stable atmospheric stratification close to the surface, so that, in combination, both horizontal and vertical air movement is suppressed. As particulate emission release from solid fuel burning for home heating coincides with this poor atmospheric dispersion potential, particle concentrations can increase substantially so that national air quality guidelines are regularly exceeded during winter in Christchurch.
At the core of this thesis is a classification based approach that examines the day-to-day probabilities of breaches of the national air quality guideline for PM over the last decade at a single location in Christchurch as a result of variations in meteorological conditions alone. It is shown that, based on variations in temperature and wind speed, up to 85% of exceedence occurrence can be explained. From this, concentration trends over time, when meteorological variability is kept to a minimum, are assessed and evidence is found that recent regulatory measures to enhance air quality are beginning to show positive effects. Atmospheric processes that control pollution dispersion on the mesoscale are investigated through means of atmospheric numerical modelling in a novel approach that assimilates observational climatic wind field averages to drive low level flow for two idealised case studies. It is shown that this approach is able to reproduce the observed diurnal concentration patterns very well and that much of these patterns can be attributed to mesoscale circulation characteristics and associated atmospheric dispersion potential, namely flow stagnation and recirculation of contaminants. When timing of stagnation and subsequent recirculation is such that it occurs within a few hours after peak emission release, concentration increase is enhanced and dilution is delayed, thus severely exacerbating the problem. Links between exceedence probabilities and synoptic situations that favour the degradation of air quality are established and various synoptic transition scenarios are examined with regard to local air quality. The progression of anticyclones across the country is identified to be the dominant synoptic control mechanism and it is shown that latitudinal variation in the progression path determines the extent of expected exceedence probability. On interdecadal hemispheric scales, it is found that a particular combination of local and synoptic atmospheric conditions that favours air quality degradation, shows a re-occurring pattern of frequency maxima (and minima) with a periodicity of approximately 14 - 16 years. For the synoptic part of this interdecadal variability, a close relationship to Southern Hemispheric pressure anomalies in high latitudes is revealed. Finally, for verification of the combined findings and to assess their prediction capability, a validation case study is given which shows that the applied methodology is able to capture day-to-day variations in pollution levels with acceptable (statistically significant) accuracy.