Skip to main content
School of Mathematical Sciences

PSD - Prof Silvia Liverani

Bayesian Spatio-Temporal Modelling for Mosquito-Borne Disease Surveillance and Outbreak Prediction

Supervisor: Professor Silvia Liverani

Project description:

Mosquito-borne diseases — including dengue, chikungunya, Zika, and malaria — represent one of the most pressing and dynamic public health challenges of our time. Climate change, rapid urbanisation, and shifting land-use patterns are expanding the geographic range of disease vectors into previously unaffected regions, creating urgent demand for sophisticated predictive models that can guide public health response. This PhD project will develop novel Bayesian statistical methodology to model the spatio-temporal dynamics of mosquito-borne diseases, with direct application to real-world outbreak surveillance and prediction. 

The project sits at the intersection of Bayesian statistics, spatial epidemiology, and environmental data science. The student will develop flexible, scalable models that integrate multiple data streams — including epidemiological case data, climate variables (temperature, precipitation, humidity), land use, and population density — to understand and predict disease dynamics across space and time. A key methodological challenge is the high dimensionality of such data, which demands computationally efficient approaches such as integrated nested Laplace approximation (INLA) alongside modern methods for scalable Bayesian inference. The student will also explore multivariate extensions capable of jointly modelling two or more co-circulating diseases, whose transmission shares common environmental drivers, exploiting this structure to sharpen predictions. 

Beyond prediction, the project will tackle fundamental inferential questions: How do climate and environmental variables mediate outbreak risk? How do diseases interact — through shared vectors, cross-immunity, or population dynamics — and how can these interactions be identified and quantified from observational data? This opens the door to causal questions about the environmental and behavioural drivers of disease spread, which the project may address using tools from causal inference within a Bayesian framework. 

The project will be developed in close collaboration with researchers at Fiocruz in Brazil, one of the world’s leading public health research institutions, providing access to rich, longitudinal Brazilian epidemiological datasets. Brazil is an ideal setting: it experiences large-scale, recurrent outbreaks of dengue, Zika, and chikungunya, with marked spatial heterogeneity driven by its diverse climate zones and urban landscape. The models developed will be directly relevant to real surveillance needs and have potential for deployment in support of public health decision-making. 

The student will develop strong technical skills in Bayesian computation, spatio-temporal statistics, and the analysis of complex health data, as well as experience of international research collaboration. There is scope to extend the methods to other geographic settings and disease systems, and to engage with the growing literature on climate-sensitive infectious disease modelling. 

This project is well suited to a student with a background in statistics, mathematics, or a quantitative discipline, and an interest in applying rigorous statistical methods to problems with real-world impact. Prior exposure to Bayesian statistics is desirable. The project offers an excellent platform for high-quality publications and for contributing to a field of growing global importance.

If you wish to inquire about the project, please contact the supervisor Professor Silvia Liverani s.liverani@qmul.ac.uk

Funding Notes:

This project is open to candidates applying for CSC / EPSRC Studentships and self-funded candidates.

 

Further information: 
How to apply 
Entry requirements 
Fees and funding

Back to top