My main research interest is extreme value theory, especially the concept of extremal dependence. Application areas of interest include droughts, heatwaves and other environmental phenomena. My research is in collaboration with the Met Office.

Statistical Methodology for the Spatial Extent, Duration and Severity Aspects of Meteorological Extremes

Much of basic extreme value theory (EVT) is concerned with modelling rare events at a given location. This one-dimensional (univariate) problem is of great interest, but may not be sufficient to model many applications. For example, if a large rainfall event has occurred in one location we may expect other locations nearby to be having a similarly large event. It is clear in this situation that the location are not independent and thus should not be modelled as such.

The concept of extremal dependence can be used to model how likely we are to observe extreme events at different locations and time points. The aim of my thesis is to look at the problem of modelling the duration, severity and spatial extent of environmental phenomena such as droughts and heatwaves. Unifying the different aspects of extremal dependence into a coherent framework will provide a useful tool for analysing current patterns and predicting any changes in future patterns of droughts and heatwaves.

Jonathan Tawn (Lancaster University) and Simon Brown (Met Office)