Research

Geospatial statistics for epidemiology, methods and applications.

Research interests

My research interests lie in the interface of statistics, epidemiology and health informatics with a particular interest in spatial statistics, public health and environmental epidemiology. I am interested in developing novel model-based geostatistical methods for tropical disease mapping.

My methodological themes include spatial and spatio-temporal statistics; joint modelling of multiple outcomes; geostatistical methods for spatial misalignment and Machine Learning.

My application themes include real-time health surveillance; tropical disease mapping; environmental epidemiology.

PhD Opportunities

I welcome enquiries from motivated students interested in pursuing a PhD in spatial statistics and its applications to global public health. Possible research directions include:

  1. Mixture models for spatially clustered disease data — developing Bayesian mixture approaches to detect and characterise spatial clustering in geo-referenced health surveys, with applications to NTD mapping.

  2. Multivariate spatio-temporal modelling — extending geostatistical frameworks to jointly model multiple correlated outcomes (e.g. co-endemic diseases) across space and time, including scalable inference methods for large datasets.

  3. Hybrid geostatistical and machine learning methods — investigating principled ways to combine the predictive power of machine learning with the uncertainty quantification and interpretability of model-based geostatistics for health risk mapping.

  4. Geostatistical methods for spatially misaligned data — addressing the challenge of combining data collected at incompatible spatial scales (e.g. individual surveys, areal counts, and remote-sensing covariates) within a unified inferential framework.

Prospective students are encouraged to contact me with a CV and a brief statement of their research interests. Funding opportunities (including EPSRC and university scholarships) are available for strong candidates. Please email olatunji.johnson@manchester.ac.uk to discuss.

Current projects

  • Climate and Health — Causal Inference methods for estimating the indirect effect of climate change on health.
  • Hybrid Machine learning and Geostatistical models — Investigating how best to combine these approaches.
  • COVID-19 - Developing spatio-temporal models and mathematical models for COVID-19.

Software

I develop open-source R packages and Shiny applications for spatial statistics and global health, including SDALGCP, ESPENAPI, MBGapp, and variogramApp. See the Software page for the full list with documentation and live demos.