top of page


We have constructed a world in which the potential for high-tech catastrophe is embedded in the fabric of day-to-day life.  -   Malcom Gladwell
EPSRC UKRI Innovation Fellowship - Multivariate Max-stable Processes with Application to the Forecasting of Multiple Hazards


July 2018 - Summary

Over the next three years, this Fellowship aims to champion innovation in the statistical modelling of the structural components that capture how extremes from multiple phenomena are likely to manifest themselves jointly across a certain region and over time. Real life applications abound in the multivariate infinite-dimensional max-stable processes frameworks: the Fukushima nuclear disaster in 2011 was ignited by the combination of a huge earthquake followed by a tsunami; the sting jet phenomenon often unleashes very extreme local wind speeds, heavy rainfall and extreme temperatures on a nuclear plant; hurricanes like Harvey, Irma, Maria and Nate, which ravaged part of the US and Central America last year, are not so much surge events any longer, they are big but they are getting slower.


The research programme will have predominant focus on risk assessment aspects and will be carried out in the context of the Fellow’s ongoing collaborations with the Energy sector, in particular with EDF Energy. The statistical methodology to be developed as part of this EPSRC UKRI Innovation Fellowship will contribute to ensure safety standards and reliable operation of different stakeholders spanning the energy sector, which will ultimately give people access to more affordable energy, provide more interaction and safety and thus more choice.

May 2022 - Culmination event

Energy Forecasting Innovation Conference - Building capacity from Modern Statistical Methodology

24-26 May 2022, King's College London

July 2022 - Acknowledgments

Mentors: Prof. Sarah Dance (University of Reading), Prof. Sue Todd (University of Reading)

Further UKRI support in here

FCT Exploratory Project DEXTE – Development of Extremes in Time and Space, EXPL/MAT-STA/0622/2013



Rainfall at nearby locations is strongly correlated. The mutual dependence can be quite different during heavy rainfall. The theory of max-stable processes offers a framework for modeling dependence at higher levels of the process. Using this framework questions about spatial rainfall, like extreme rainfall in a certain area, can be addressed. Max-stable processes are useful for – for example – problems on the amount of toxic elements in the air or water but also for problems about geological formations. The basic theory is well established (see [1]). We propose to develop the theory further in a useful way and to develop models for applying the theory. We will also develop an application of the theory to trends in rainfall in time and space (see output paper [2]).

May 2014 - Culmination event: Course on Empirical Processes and their Applications to Extreme Value Statistics


[1] A. Ferreira and L. de Haan (2014). The Generalized Pareto process; with a view towards application and simulation. Bernoulli.

[2] L. de Haan, A. Klein Tank and C. Neves (2015). On tail trend detection: modeling relative risk. Extremes.

bottom of page