Prof. Glenn Marion

With colleagues and collaborators in BioSS and beyond I aim to show and extend the power of mathematical modelling to impact on societal challenges through real world application and development of improved methods.

Complex systems: Modelling interacting processes enables understanding of emergent properties critical to addressing the ‘law’ of unintended consequences. This work aims to create tools and insights for better management, for example how culling and surveillance for wildlife disease control can fail. Ongoing and future work will look at how pesticide application counterintuitively increases crop losses and how trade between farms could be used to control endemic disease in livestock.

Statistical inference for process models uses field or operational data to, estimate quantities too costly to measure directly, make predictions, and test scientific hypotheses. This work has led to development of methods for model choice, model assessment and faster mixing in data augmentation Markov Chain Monte Carlo for Markov and semi-Markov processes. This research improved inference for: phylodynamics; spatial disease risk in small or ongoing epidemics; host genetic effects from epidemic data, and; demography and disease dynamics from capture mark recapture data. Ongoimg work is creating software applications; improving particle filtering, and creating novel epidemic inference methods for when the population at risk unknown.