AbstractsAstronomy & Space Science

A multi-scale assessment of the air pollution climatology of Alexandra, New Zealand using a combination of field observations and statistical and physical based modelling

by Nicholas Joseph Price

Institution: University of Otago
Year: 0
Keywords: Air Pollution; Climatology; Particulate Matter; Alexandra
Record ID: 1305292
Full text PDF: http://hdl.handle.net/10523/5592


This thesis examines the relationship between atmospheric phenomena and particulate matter in Alexandra, a township located within an inland basin in Central Otago. The primary aim of this research is to define the air pollution climatology of Alexandra by characterising those atmospheric processes which control particulate matter concentrations at a variety of spatio-temporal scales. During the winter months, Alexandra suffers from some of the poorest air quality in New Zealand and currently exceeds the Air Quality National Environmental Standards (AQNES) for daily PM10 nearly fifty times a year. This provides a major challenge for the Otago Regional Council (ORC) as they look to meet the AQNES, which is three exceedances a year, by 2016. However, the relationship between emissions and PM10 is complex due to the climate of Alexandra being strongly influenced by its location within an inland basin. This makes the task of meeting the AQNES inherently difficult. Therefore, an understanding of the atmospheric processes that control PM10 is of importance to the ORC as it attempts to improve Alexandra’s air quality. This research utilised both long term meteorological data as well as observations from a field campaign, to represent atmospheric processes occurring at the synoptic, local and micro-scales. Long-term data from the ORC’s monitoring station (2006 – 2013) and a NIWA automatic weather station (2008 – 2013) were analysed to describe the relationship between PM10 and the synoptic and local-scale conditions over the winter months (May-August). A temporary automatic weather station was installed between 14 June 2013 and 20 August 2013 which included eddy covariance measurements in order to make inferences on the micro-scale processes present within the atmospheric boundary layer. Finally, a case study was also undertaken on 23 July 2013 which utilised vertical wind profiles observed by Sodar. Synoptic scale analysis indicated that poor air quality is not only linked to anticyclonic conditions, but was a feature of nearly all synoptic conditions, which highlights the effects of topographic sheltering on Alexandra’s climate. Even during times of strong westerly flow over the South Island, high PM10 concentrations were observed, indicating that the air mass within the Alexandra Basin is frequently decoupled from the gradient flow. The continuation of synoptic conditions was found to be an important control on air quality, with those circulations linked to high PM10 persisting longer and being preceded by synoptic conditions that are also linked to poor air quality. To define the local meteorological conditions that were most important to air quality in Alexandra, a Classification and Regression Tree (CART) analysis was used. A range of meteorological predictor variables recorded over seven winters were recursively partitioned in order to predict whether or not Alexandra’s daily average PM10 concentration exceeded the AQNES standard. The conditions defined by the CART analysis indicated that air temperature and wind…