UMR CNRS 5805 EPOC
Environnements et Paléoenvironnements
Océaniques et Continentaux

        

Consignes de Sécurité et Numéros d'Urgence
(à l'usage des personnels du laboratoire)

le 10-12-2013 à 10:45

Lieu Salle Stendhal, DGO, Bâtiment B18 Intervenant(s) Giovanni SGUBIN, Postdoc, UMR EPOC

The human contribution in greenhouse gases emission is a potential significant component of the causes of recent global climate changes, and this makes the discussion on how and to what extent this can alter the future climate particularly debated. In order to tackle this topic, the Intergovernmental Panel on Climate Change have proposed different scenarios, i.e. CMIP5 protocol, representing possible evolutions of the future greenhouse gas concentration. Here I review a set of model outputs within the CMIP5 framework.
The first part of the discussion is aimed to the evaluation, model by model, of any hints of abrupt changes under RCPs and historical scenarios forcing. In particular, besides the general warming trend shown by the 32 analyzed models at global scale, many sub-regions where abrupt changes occur are identified. Although these rapid changes are not a common feature of all of the models, their identification is an important statement as well in terms of 'low-probability high-potentiality' impact.
The second part is mainly focused on the response of the Atlantic Meridional Overturning Circulation (AMOC), one of the key phenomenon in climate changes discussion. In particular its reversibility is studied by means of 4 different global climate models (EC-EARTH, HadGEM2-ES, IPSL-CM5A-LR and MPI-ESM-LR) under an increase of radiative forcing (ramp-up) and a subsequent symmetrical reversal of radiative forcing (ramp-down). All the models show a weakening of the AMOC during the ramp-up process whereas, once the external forcing is reversed, the results are model dependent: for IPSL-CM5A-LR, the AMOC holds over a weakening trend and no clear hints of recovering are visible; for HadGEM2-ES, the AMOC trend reverses once the external forcing also reverses, without nevertheless recovers its initial value; for EC-EARTH and MPI-ESM-LR, on the other hand, the recovering is anomalously strong yielding an AMOC overshooting. Once tested a robust linear dependence between AMOC strength and meridional density gradient between North Atlantic Deep Water (NADW) formation region and South Atlantic, the stability properties of the AMOC in terms of freshwater feedbacks in an interhemispheric 2-boxes conceptual model is examined: the IPSL-CM5-LR results to lie in a bistable regime where the AMOC can potentially collapse, the HadGEM2-ES show a monostable regime, while multiple states are found for EC-EARTH and MPI-ESM-LR during the period of integration of the experiments. This suggests that, to some extent, the stability regime of the AMOC, beyond pointing out the potentiality for its abrupt collapse, can also play a role in the recovery process. Finally, the origin of meridional density gradient anomaly is detected in order to evidence the mechanisms underlying the reversibility of the AMOC. It will be shown that different behaviors found in the models depend on the way in which the salinity anomaly built-up during the ramp-up in the tropical Atlantic is released elsewhere during the ramp-down. For those models in which an overshooting of the AMOC is evidenced indeed, such a salinity anomaly is mainly advected northward reinforcing the density in NADW formation region faster than in the South Atlantic. This increases the north-south density contrast and, as a consequence, the AMOC strength.