Use this URL to cite or link to this record in EThOS:
Title: Adapting geostatistical approaches to mapping air pollution in the UK
Author: Robinson, Damien Patrick
ISNI:       0000 0004 2742 5041
Awarding Body: Queen's University Belfast
Current Institution: Queen's University Belfast
Date of Award: 2012
Availability of Full Text:
Full text unavailable from EThOS.
Please contact the current institution’s library for further details.
Nitrogen Dioxide (N02) is detrimental to human health. It is difficult to map accurately, as concentrations can vary greatly over small distances. Current mapping methods do not fully utilise monitored data, and as such the objective of this project was to apply new techniques based on both modelled and measured data. This was achieved through geostatistical estimation and simulation techniques. Measured N02 from automatic and diffusion tube monitoring was the primary dataset. The secondary dataset consisted of dispersion modelled NOx data at a lkm2 scale for the UK. NOx is useful for informing N02 prediction, given that N02 is a constituent pollutant of NOx• The key geostatistical estimation techniques investigated were simple kriging (SKL ordinary kriging (OK) and simple kriging with a locally varying mean (SKim). The geostatistical simulation analysis applied Sequential Gaussian Simulation (SGSIM). SK and OK can only utilise a single dataset. SKim differs from in that additional data can be inputted to inform prediction, and hence potentially reduce uncertainty. Supplementary secondary data can be highly beneficial, particularly when the primary dataset is not evenly spatially distributed. The secondary data (NOx dispersion modelled data) were used to define the locally varying mean in SKim, for both the geostatistical and simulation analyses, using two regression approaches: (i) global regression (GR) and (ii) geographically weighted regression (GWR). SKim with GWR defined locally varying means was further investigated through examination of various parameters. Throughout, the effects of other variables that may affect prediction were also discussed. The integration of the local model into the geostatistical algorithm produced an increase in estimation accuracy (measured through a reduction in uncertainty), in comparison to other techniques. SKim with GWR defined locally varying means, produced the most accurate predicted and simulated outputs of all approaches tested in this research.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available