Series chronologiques non stationnaires non lineaires. Le cas des series de temperatures en Europe
Thi Thu Huong HoangEDF R&D & Departement de Mathematique d'Orsay
Friday, March 5th, CERFACS Conference Room - 14h00
Abstract.
This is a thesis of Mathematics, which deals with advanced statistics.The work is focused on analyses of the recent evolutions of some characteristics of the temperature, like mean, variance and extremes. A particular attention is given to the link between the evolutions of these quantities. These analyses are conducted using mainly the observed temperature series provided in the framework of the ECA&D project. Then, two main goals are defined:
- to propose a methodology to estimate future temperature return levels in a non stationary context
- to build a realistic stochastical model to simulate the temperature evolution, with a special interest in the reproduction of extreme events.
Cerfacs' Conferences Home Page
