# Professor Duncan Lee

**Professor of Statistics**(Statistics)

**telephone**:
01413304047

**email**:
Duncan.Lee@glasgow.ac.uk

Room 342, Mathematics and Statistics Building, Glasgow G12 8SQ

## Biography

I received a first class BSc in Mathematical Sciences from the University of Bath in 2003, and a PhD in statistics with applications in environmental epidemiology from the same university in 2007. Following a 6-month temporary lectureship and a 1-year research assistant position, both at the University of Bath, I became a lecturer in statistics at the University of Glasgow in September 2007. I have been at the University of Glasgow ever since, and became a professor in August 2018.

## Research interests

My research interests are in developing spatio-temporal statistical methods for use in epidemiology in a predominantly Bayesian setting, focusing on **disease mapping**, **predicting air pollution concentrations**, and investigating the **effects of air pollution on human health**.

My first research area is in **disease mapping** whose aim is to estimate the spatial pattern in disease risk over a city or country, as well as determining whether there has been any increase or reduction in risk over time. These spatio-temporal risk estimates allow public health officials to address key policy questions, such as: (i) where is disease risk highest and which areas are exhibiting an increasing risk of disease over time; (i) how big are the health inequalities in disease risk between rich and poor communities, and how are they changing over time; and (iii) what are the impacts of covariate factors on disease risk.

My second research area concerns **air pollution**, both how to predict it in space and time, and what impact it has on public health. The first of these requires a high-resolution spatio-temporal prediction model to be developed, which allows air pollution concentrations to be predicted at unmeasured locations. The second of these requires regression models to be built that allow for residual spatio-temporal autocorrelation, which allows both the short-term and long-term health effects to be estimated.

### Research groups

## Grants

Grants as PI.

- ESRC grant ‘Allowing for cliffs and slopes in the risk surface when modelling small-area data’, Oct 2010, £52,556.
- EPSRC grant ‘A rigorous statistical framework for estimating the long-term health effects of air pollution’, Jan 2013, £269,755.
- MRC grant entitled ‘A flexible class of Bayesian spatio-temporal models for cluster detection, trend estimation and forecasting of disease risk’, Feb 2015, £304, 000.
- EPSRC grant 'Linking air pollution and health in Scotland to inform on locations for a Low Emission Zone', July 2017, £14,962.
- TB Alliance grant entitled 'Mathematical modelling of MDR, pre XDR, and XDR burden of the TB population', August 2017, £24,068.
- Scottish Government grant entitled `Estimating the effects of air pollution on the Covid-19 epidemic in Scotland', August 2021, £41,335.

Grants as CI.

- Environment Agency grant `'Spatiotemporal modelling of Nitrate and Phosphorus in River Catchments for England and Wales', March 2009, £61,710.
- Carnegie scholarship for PhD study for Craig Anderson, October 2011, £63,696.
- NERC knowledge transfer grant 'Statistics, environmental management, policy and regulation: developing the evidence base', August 2009, £295,868.85.
- ESRC grant 'Applied Quantitative Methods Network: Phase II', January 2013, £2,601,794
- NERC Postgraduate and Professional Skills Development Award, September 2013, £73,000.
- ESRC grant entitled 'Urban Big Data Research Centre', January 2014, £6,003,547.
- NERC grant entitled 'Advanced statistics training', September 2014, £73,000.
- Carnegie scholarship for PhD study for Eilidh Jack, October 2015, £70,000.
- NERC grant entitled 'Advanced statistics training', September 2015, £70,000.
- Innovate UK grant entitled 'A new integrated decision-making tool for urban health and policy evaluation: ‘QCumber-envHealth’', October 2015, £456,398.
- EPSRC grant (as part of the SECURE network) entitled 'Feasibility testing of low-cost sensors to represent spatio-temporal variability of ambient ground-level NO2 and O3 concentrations', October to December 2015, £25,000.
- CI for a NERC funded CASE PhD studentship, October 2016, £60,000.
- EPSRC grant (as part of the SECURE network) entitled 'Improved approaches for mapping and modelling blanket peatland extent and depth', May 2016, £24,962.
- Scottish Executive Health Department grant entitled 'Multimorbidity in UK Biobank', December 2016, £22,029.
- EPSRC grant entitled `Multilayer Algorithmics to Leverage Graph Structure (MultilayerALGS)', 2020, £765,538.

## Supervision

**Boonsuriyatham**, Panthakan

Forecasting Local Net-electricity Demand at Scale**Foster**, Hamish

Understanding the interplay of lifestyle and deprivation to support policy and intervention development: a mixed methods study**Halliday**, Alba

Bayesian Hierarchical Modelling Frameworks for Correcting Reporting Delays in Disease Surveillance**Muegge**, Robin

Spatiotemporal areal data modelling: COVID-19 applications and boundary detection for big data**Zhu**, Qiangqiang

Spatio-temporal modelling of population-level disease risk when the populations at risk have partially unknown spatial locations**Zou**, Zhaoyuan

Methodological developments in Data Fusion for Environmental Statistics

**Completed / Existing research staff**

- Alastair Rushworth - January 2013 - May 2015, working on the EPSRC funded project 'A rigorous statistical framework for estimating the long-term health effects of air pollution'.
- Francesca Pano - June 2015 - December 2015, working on the EPSRC funded project 'A rigorous statistical framework for estimating the long-term health effects of air pollution'.
- Gary Napier - March 2015 - August 2018, working on the MRC funded project 'A flexible class of Bayesian spatio-temporal models for cluster detection, trend estimation and forecasting of disease risk'.

**Completed / Existing PhD students**

- Helen Powell - 'Statistical methods in air pollution and health studies', October 2008 until July 2012.
- Serdar Neslihanoglu - 'Financial time series modelling', March 2011 until June 2014, jointly with Prof McColl.
- Craig Anderson - 'Bayesian space-time modelling of disease risk', October 2011 until February 2015, jointly with Dr Dean.
- Guowen Huang - 'Space-time air pollution and health modelling', August 2013 until July 2016, jointly with Prof Scott.
- Francesca Pannullo - 'Space-time air pollution and health modelling', October 2013 until August 2017, jointly with Prof Leyland and Dr Waclawski.
- Eilidh Jack - `Space-time modelling of disease risk', October 2015 - July 2019, jointly with Dr Dean.
- Yoana Borisova - `Space-time modelling of air pollution', October 2016 until current day, jointly with Prof Scott.
- Kamol Sanittham - `Spatial modelling of disease risk', October 2016 until March 2021, jointly with Dr Anderson.
- Xueqing Yin - `Spatio-temporal modelling of disease risk', January 2019 until July 2022, jointly with Dr Anderson and Dr Napier.
- George Gerogiannis - `Multiple membership multiple classification models', October 2018 - current day, jointly with Prof Tranmer and Prof Moore.
- Zhaoyuan Zou - `Spatial misalignment in environmental data', October 2019 until current day, jointly with Dr O'Donnell and Prof Miller.
- Enoch Havyarimana - `Modelling inflamatory and rheumatic diseases', October 2019 until current day, jointly with Dr Basu and Dr Cullen.
- Robin Muegge - `Space-time modelling of disease risk', October 2020 - current day, jointly with Dr Jack and Dr Dean.
- Hamish Foster - `Factors contributing to ill health in UK Biobank', October 2020 - current day, jointly with Prof Mair.
- Qiangqiang Zhu - `Fusing machine learning and spatial statistics to estimation air pollution and its health effects', October 2022 - current day, jointly with Dr Stoner.
- Panthakan Boonsuriyatham - `Spatio-temporal prediction of energy consumption', November 2022 - current day, jointly with Dr Browell.

**Completed / Existing MSc (by research) students**

- Abita Bhaskar - MSc Statistics - 'Environmental exposures and cardiovascular morbidity in Scotland', October 2007 until February 2009, jointly with Prof Scott.
- Neil Murphy - MSc Statistics - 'Modelling aids incidence in South Africa', October 2009 until September 2010, jointly with Prof McColl.
- Greg Halbert - MSc Statistics - 'Estimating the changing effects of air pollution on health in Scotland', October 2010 until September 2012.
- Katie Stewart - MSc Statistics - 'How much spatial autocorrelation is too much?', October 2013 until January 2015.
- Cillian Docherty - MSc Statistics - 'Environmental indicators', October 2015 until September 2017, jointly with Prof Scott.
- Cara MacBride - MSc Statistics - 'Spatial predictive modelling of average property prices', October 2022 until September 2023, jointy with Dr Davies.

## Research datasets

## Additional information

**Twitter**

@DuncanPLee

**Computer software**

**CARBayes** - A package for the statistical software R to implement spatial generalised linear mixed models for areal unit data, with inference in a Bayesian setting using Markov chain Monte Carlo (MCMC) simulation.The spatial autocorrelation is modelled by a set of random effects, which are assigned a conditional autoregressive (CAR) prior distribution. A number of different CAR priors are available for the random effects, and full details are given in the vignette accompanying the package. The software is available on the Comprehensive R Archive Network (CRAN) here. Over 131,600 downloads since 2013.

**CARBayesST** - A package for the statistical software R to implement spatio-temporal generalised linear mixed models for areal unit data, with inference in a Bayesian setting using Markov chain Monte Carlo (MCMC) simulation.The spatio-temporal autocorrelation is modelled by random effects, which are assigned conditional autoregressive (CAR) prior distributions. A number of different random effects structures are available, and full details are given in the vignette accompanying the package. The software is available on the Comprehensive R Archive Network (CRAN) here. Over 51,800 downloads since 2014.

**Professional activities**

**Member of the UK government advisory committee on the medical effects of air pollutants (COMEAP)**between March 2020 and the current day.**Conference chair**for the biennial GEOMED conference, which was held in Glasgow in August 2019.**External examiner**of BSc and MSci programmes at Queens University Belfast from 2018 to 2023.**Associate editorships**at Journal of the Royal Statistical Society Series C, Biometrical Journal, Spatial and Spatio-temporal Epidemiology, and Biometrics.**Editorial board member**at Spatial Statistics.**Guest edited**three special issues of Statistical Methods in Medical research and one special issue of Spatial and Spatio-temporal Epidemiology.**External PhD examiner**at the Universities of Bath; Exeter, Lancaster; Leeds; Oxford; Queen Mary University London, Queensland Technology; Southampton; Technology Sydney.**Secretary of the environmental statistics section (ESS)**of the Royal Statistical Society between 2013 and 2016.**Honorary appointment with Public Health Scotland**to work on Covid-19 surveillance modelling from April 2020 to current day.**Invited member of the Health effects working group**that is advising the Cleaner Air for Scotland (CAFS) strategy, versions 1 and 2, January 2019 to 2026.