Statistical AI Approaches for Optimal Health Warning Systems. Department of Mathematics and Statistics,UQ-Exeter Institute PhD Studentship (Funded) for January 2027 Entry.

Project Description
Temperature extremes and poor air quality are two of the main hazards humans are exposed to, particularly in urban environments where over 55% of population reside and rising. Although individually both temperature extremes (e.g. heat-stress) and air quality (e.g. Nitrogen Oxide from car fumes) are known to severely impact human health, it is unclear how these stressors interact although it is expected that their effect is compounded. The hazard to human health is therefore not well-understood, where individual assessments of either temperature or air quality are likely to underestimate the health risk.

It is therefore not surprising that most health early warning systems are designed for either temperature extremes or air pollution but not their synergy. At the same time, there is a lack of a coherent and mathematically rigorous methodology for how health, environmental and population exposure and vulnerability data can be combined to optimally issue warnings in order to minimise health risk.

The aim for this project is to investigate the use of statistical AI methods for a) estimating the synergistic effect of temperature, humidity and air quality on human health (mortality and morbidity); b) to propose a prescriptive framework for using these estimates to optimally issue health warnings and c) to investigate the implications of climate change to the compound risk from environmental extremes.

The project will use state-of-the-art statistical AI approaches in environmental epidemiology (a unique strength of the Exeter-Queensland team), to quantify the degree to which the effects from temperature, humidity and air quality on human health are compounded, while also allowing for population characteristics (age, sex, socio-economic background) and exposure. The project will also use ideas from decision theory to investigate approaches for optimally issuing warnings with a view to minimise health risk to the respective sections of the population. Moreover, the project will utilise the latest data from climate models which contain information for how the climate, the population and its vulnerability/exposure are projected to change under various socio-economic scenarios. The information in these data sets and the estimated health risks as a function of environmental extremes will be combined to produce predictions of future health risk under different climate scenarios.

The project is expected to produce new knowledge into the synergistic effects of weather and air quality extremes, the way the estimates health risks might change in warming planet and present a health warning framework with which this knowledge can be used to mitigate health risk. The project partners, Public Health Scotland and the Met Office, will offer unique knowhow in terms of health surveillance, operational risk management and data access as well as insights regarding societal relevance and potential adoption and piloting of the resulting warning system.

Contact
Questions about this project should be directed to Associate Professor Theo Economou at T.Economou@exeter.ac.uk

Advert information