Damien Robinson
Damien Patrick Robinson
Qualifications
BSc. in Geography, Queens University Belfast 2008.
PhD Student
Email: drobinson13@qub.ac.uk
Address
Room 02 044, Elmwood Building
School of Geography, Archaeology and Palaeoecology (GAP)
Queen’s University
Telephone
+44 (0) 28 9097 3929
Current Teaching:
Demonstrating :
310GGY2037 - Acquisition and analysis of geographical information
310GGY3023 – Advanced Geographical Information Systems
_____________________________________________________________________
Current Research: Environmental Change
Adapting geostatistical approaches to mapping air pollution in the
There are various spatial interpolation methodologies that can be used to predict data values at un-sampled locations, based upon existing data. Geostatistics is an umbrella term referring to the many variant forms of kriging. There is much debate as to which techniques create the most accurate, true representation of un-sampled locations. In most studies it seems, different approaches can be used depending on the nature of the data, hence there is no overall “best” interpolation technique. By and large the more data that can be availed of, then potentially the more accurate a prediction can become. Multivariate geostatistics (e.g. Simple kriging with a locally varying mean and Collocated CoKriging) avail of additional data to inform prediction, and therefore, on this basis would be expected to outperform univariate geostatistical methods (e.g. Simple Kriging and Ordinary Kriging).
In terms of pollution, the focus remains on nitrogen dioxide (NO2) and oxides of nitrogen (NOx). Data is collected through automatic monitoring and diffusion tube analysis at local government level. This is collated through the UK Air Quality Archive. Emissions are derived from various sources (e.g. Road networks, Industrial Areas, Airports etc.), using various modelling techniques. This data is collated, analysed and made available through the National Atmospheric Emissions Inventory. Dispersion modelled emissions data is made available through UK Air Quality Archive, and its regional derivatives, for use in Local Air Quality Management (LAQM). Geostatistics can account for any unexplained variation in dispersion models, and in theory should lead to a further increase in prediction accuracy.
A pilot study has been undertaken in N.Ireland, using all available annual mean data from 2004. All sites that were used were either urban or rural background sites, coupled with a secondary dispersion modelled data. Global/local regression and geostatistical analyses were undertaken. The results of this study have been submitted for publication.
Future work to include, alternative local regression approaches, alternative geostatistical techniques,
Publications:
Posters/Presentations
Robinson, D.P., 2009. Regional Case Study: Using geostatistics to increase the accuracy of NO2 mapping for government reporting in
Robinson, D.P., 2010. Adapting geostatistical approaches to mapping air pollution in the
Additional Information:
Memberships:
Member of Irish Organsiation for Geographical Information (IRLOGI)
Association for Geographical Information (AGI) and AGI Northern
