Physical processes governing ocean heat content variability

This thesis summarizes the research program of the XXXIV PhD cycle of Uni versity of Rome Tor Vergata, falling within the European Union Project Copernicus Climate Change Service C3S_511: Quality Assessment of Essential Climate Variable (ECV) Products. The general objectives of this project regards the validation of all the data, both ob served or simulated by climate models available through the C3S Climate Data Store (CDS). To pursue this aim, the project should have applied analysis methodologies already partly developed, such as the Earth System Model Validation Tool (ESM ValTool), to evaluate statistical averages, climate variability, uncertainty evaluation, ability to capture extreme events. In one word to assess the suitability of available data to identify changes and variability of the Earth climate system, through the production of IPCC-style Assessment Reports (IPCC stands for Intergovernmental Panel on Climate Change) 1 . In the first part I will go through the definition, description, current understanding and evaluation of the Atlantic Meridional Overturning Circulation in state-of-art ocean reanalyses. The focus will be on describing and interpreting an ocean circulation regime change characterizing one of the data sets, proposing a mechanism to explain the observed variability2 . In particular, we find out that state-of-the-art reconstructions show signs of a reduction in the northward transport of watermasses, reflecting the possible existence a stronger circulation regime prior to mid-1990s, and a weaker one after wards. We explored the effects of these circulation changes on key components of the ocean climate, finding a tilt of the Gulf Stream Path toward lower latitudes passing from one period to the other, consistently with diminished Deep Water Formation in high latitude seas. Finally, observing the correlation with the North Atlantic Oscillation index we proposed a mechanism by which it is possible to interpret this ocean variability as a response to a persistent atmospheric perturbation. We also argue that this response appear overamplified in two of these datasets, and probably the flux-adjustment carried out in order to reduce Sea Surface Temperature biases along the Gulf Stream front is among the causes for this3 . Indeed, given the limitations in computational power of current High Performance Computing facilities, the resolution of global general circulation models (GCM) is still limited to about 25 km for the ocean component, which implies that all physical processes below the grid spacing must be parameterized, in order to account for their average effect on the whole grid cell - this is a primary cause for biases in circulation features. Given this non negligible aspect, vast uncertainties still remain about the effects of small scale processes in large scale dynamical features of the ocean, and this is an open research topic in the scientific community. There are two ways to tackle this problem: one is to deal directly with a general circulation model and study with sensitivity experiments the effect of choosing a certain scheme for sub grid physics; an other approach could be to study the effect of smaller scales in a numerical setup which is able to focus directly those scale of motions which have to be parameterized in an ocean GCM. In a nutshell, this second approach motivate the last part of the thesis, in which I will describe a more general and theoretically oriented study on rotating stratified turbulence in an ideal, thin, tri-periodical box (which should be close to represent an open-ocean domain), without including the effect of topography. Namely the aim is to disentangle the relative role of waves with respect to vor tices, and one possible way to do it is to employ a linear eigenmode decomposition introduced by Peter Bartello in 1996. We studied a recent version of Bartello’s approach, using it to selectively forcing only waves, or only vortices in the system under investigation from the beginning. Aiming to understand how the energy is partitioned amongst these modes, especially when different forcing projections (which will excite only waves or vortices) are used, this approach would help to characterize respectively the importance of rotation or stratification to nonlinear terms while varying their corresponding dimensionless parameters4 ; then, in a future development it will be possible to explore where and why one particular setup gets close to real ocean conditions. Conclusions and a trace of my future line of scientific investigation will follow.