Developing efficient statistical tools for problems arising in spatio-temporal modelling

Abstract

When analysing spatio-temporal data in applications such as in environmental, ecological, or epidemiological settings, it is useful to be able to efficiently parametrise the potentially high dimensional model in a way that still captures the spatial and temporal features of the data. Furthermore, detecting abrupt changes over space and time gives insight into the underlying mechanics of a system and can have a significant impact in the interpretation of the evolution of the process and its future states. The aims of this project include looking at methods that efficiently model spatio-temporal data by utilising the spatial context by parametrising this network structure as a constraint to better model the data and detect changes. We will expand on these approaches further where gaps in the literature exist by developing a set of robust statistical tools to interpret these datasets. In particular, we will consider how the space and time elements of the data interact and leverage them to reduce the dimension of the problem. Additionally, when we are able to detect change-points we will assess how easily can we identify them.

Date of Award	12 Nov 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Matthew Nunes (Supervisor) & Sandipan Roy (Supervisor)