Extreme value theory provides an efficient tool to solve relevant problems in various disciplines, including risk management, meteorology and many other fields. The objective of our project is to go beyond the stationary framework. We propose the development of new probabilistic and statistical tools to deal with non stationary data. Although the present proposal addresses theoretical challenges, upstream of the societal challenges, it will be most of the time driven by climatological and hydrological applications, or more generally by environmental applications.

Our project is subdivided into two sub-themes:

• Non stationarity tests for extremes,

• Modeling non stationarity in the dependence structure of extremes.



Clémentine PRIEUR at LJK

Anne-Catherine Favre at LTHE (IGE)


Related collaborations:

Patricia TENCALIEC (PhD student funded by AGIR, defense on February, the 1st, 2017)

Philippe NAVEAU at LSCE (Paris)

Gianfausto SALVADORI at Universita del Salento (Italy)