Home > News > Degree Defenses > Jérôme Servonnat's PhD Defense

PhD Defense

Jérôme Servonnat (LSCE)

Title : Climate variability in the North Atlantic domain during the last millennium: evaluation of the influence of solar forcing with the IPSLCM4 model

Date and time : The 16-12-2010 at 14h00

Type : thèse

Université qui délivre le diplôme :

Location : Amphitéâtre Bloch, Bat 774, CEA Saclay, Orme des Merisiers, 91191 Gif sur Yvette
Members of jury :

Hugues Goosse (rapporteur)

Jean-Louis Dufresne (rapporteur)

Jesus-Fidel Gonzalez Rouco

Joël Guiot

Serge Planton

Pascal Yiou (Directeur de thèse)

Myriam Khodri (Co-directrice de thèse)

Summary :

Studying the climate of the last millennium gives the possibility to deal with a relatively well-documented climate essentially driven by natural forcings. Concerning the forced secular variability of the climate during the preindustrial last millennium, important uncertainties exist on the amplitude of the multidecadal to centennial variability of the Total Solar Irradiance. Two main hypotheses emerge from the literature on the amplitude of the TSI secular variability, which are identified by an increase between the Maunder Minimum and today of 0.25% or of 0.1% of the contemporaneous TSI value. At the same time, the timing of the main oscillations of the TSI is almost the same in the existing TSI reconstructions.

My work during my PhD has been the evaluation of the influence of solar forcing on climate variability during the last millennium with the IPSLCM4_v2 coupled model. I achieved two numerical simulations of the climate of the last millennium for this purpose. The first one consists in a millennium-long control integration (CTRL), and the second one (SGI) was forced by reconstructions of the Total Solar Irradiance (TSI), CO2 concentrations and orbital parameters during the last millennium. The chosen TSI reconstruction follows the strong amplitude hypothesis, i.e. 0.25%, to favour a strong but realistic forced signal.


As a first step, we have shown that the variance of the forced simulation is statistically different from the control integration for a spatial scale wider than 5.106km² (spatial extent of Europe). The Northern Hemisphere temperature in SGI is in agreement with previous similar numerical exercises, and the simulated variability is realistic in terms of amplitude compared with reconstructions. A mismatch has been evidenced between the reconstructions and the simulation between 1000 and 1200 AD during the so-called Medieval Warm Period. Statistical temperature decomposition shows that this mismatch is linked with solar forcing in the simulation. The statistical temperature decomposition applied by grid point supports the idea that the studied forcings weakly affect local temperature reconstructions.


A particular interest has been drawn on the North Atlantic domain. A model-data comparison in Europe has been done with a temperature reconstruction and numerical simulations from the ANR ESCARSEL project which includes my PhD. The amplitude of the variability of the temperature mean over Europe is similar in the reconstruction and the simulations. However, the comparison between the forcing signatures has shown that although solar forcing appears clearly in the simulations, it is not striking in the temperature reconstruction. We also analysed the North Atlantic Oscillation but it is not affected by solar forcing in the SGI simulation.

The conclusion of my work is that the solar forcing used in the simulation can not explain all aspects of the secular temperature evolution between 1000 and 1850 AD, neither on the Northern Hemisphere nor in Europe.

University :


Contact :
Return to list of degree defenses