Statistical modelling changes in climate extremes

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Mathematics

Abstract

Intense meteorological events such as extreme precipitation and windstorms can cause havoc for the people affected and typically, result in huge financial losses. There is a clear need to have better information attributing causes to change in frequency and intensity of extremes in observations, and improving predictions. This PHD project aims at developing flexible spatial extreme value models with primary focus on the detection and attribution of human contribution to the intensification of extreme precipitation. The methodology used will be a combination of recent developments in spatial extreme value theory and climatological modelling expertise. Key research questions include: What are the spatio-temporal trends in intense precipitation, given noise in the data and spatial information? How are rainfall extremes changing compared to mean rainfall and how does this link to warming in large regional scales?

Publications

10 25 50
publication icon
Auld G (2023) Changes in the distribution of annual maximum temperatures in Europe in Advances in Statistical Climatology, Meteorology and Oceanography

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509644/1 01/10/2016 30/09/2021
1935526 Studentship EP/N509644/1 01/09/2017 30/09/2021 Graeme Auld
 
Description We have derived results that allow us to quantify the degree of clustering of extreme values may occur in non-stationary time series. Extremal clustering refers to the fact that in a time series of dependent data, several large values may be observed in quick succession and in many applications it is important to understand the extent to which this may occur.
Our results generalize and extend previously known results for stationary series.
Exploitation Route Our results are theoretical in nature and although we have begun the process of finding applications, we believe that others will help us to turn these results in to new statistical methodologies for the extremes of non-stationary time series.
Sectors Environment,Financial Services, and Management Consultancy