Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.595200
Title: Spatial-temporal analysis of traffic-related ground level ozone
Author: Munir, Said
Awarding Body: University of Leeds
Current Institution: University of Leeds
Date of Award: 2013
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Prolonged exposure to elevated levels of ozone (O3) has been proven to adversely affect human health, agricultural crops and building materials. Therefore, ground level O3 is currently regarded as one of the most harmful air pollutants, and identified as a priority pollutant in almost all national and international air quality legislations. Historically, O3 has been regarded as a rural issue because relatively high levels of O3-depleting species like NO from traffic and other combustion sources have helped to reduce its impact in urban area. However, more recently, and most likely as the result of increasingly more aggressive and effective management of NOx, urban O3 levels have begun to rise much more rapidly than ‘background O3’ in rural areas. As a result O3 pollution is becoming an increasing important urban issue. Therefore, the main aim of this PhD project is to characterise the temporal - spatial variability of ground level ozone. As in other studies, rural time-series of O3 and meteorological data and spatial parameters are considered as part of the analysis. However, here the analysis is extended using recently collected urban data, most notably co-located O3 and traffic related pollutant time-series and road traffic characteristics (from double loop counters), to provide more insight into the mechanisms driving traffic-related O3 trends. Traditionally, many researchers have used linear, mean-centred (and parametric) statistical methods to ozone. However, here, in light of the non-linear association between ozone and its potential covariates, the focus has been on non-parametric methods. In particular two methods have been extensively used, namely Quantile Regression and Generalised Additive Modelling. Such approaches provide a much more comprehensive description of O3 trends because they are better able to represent the distinctly different behaviour of O3 at concentration extremes. Statistical analysis shows that on average ozone concentrations are about 26 % higher in rural areas than in urban areas. The urban decrement varies at different quantiles of ozone concentrations and ranges from 10.5 μg/m3 (25%) to 21.56 μg/m3 (30%) at quantile 0.1 and 0.99, respectively. The results of quantile regression model show that up to 90 % of ozone variations between rural and urban sites can be explained with the help of road traffic characteristics. It is shown that the strength and nature (positive or negative) of the association between ozone and its covariates change at different levels of ozone. The model result shows that ozone levels increase towards north and decrease towards east in the UK. Ozone concentration appears to be negatively correlated with the distance from the coast within a range of 0 to 60 km, a trend associated with relatively less dry deposition of ozone molecules on water surfaces. Furthermore, there is a positive association between the altitude of monitoring site and ozone level, most likely due to local topographical effects. Ozone temporal trends show significant variability at different statistical metrics (e.g., mean, median, maximum and selected quantiles). Urban and rural trends have different rates and indicate that urban decrement has been decreasing over the period. Ozone trends during 1993 to 2011 and 2004 to 2011 showed different patterns, i.e., average ozone trends at both urban and rural sites are positive for the former and negative for the latter case. NOx trends have been stabilised during the last 8 years in urban areas and could have caused the ozone trends to change from positive to negative.
Supervisor: Chen, Haibo ; Ropkins, Karl Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.595200  DOI: Not available
Share: