Dissolved Oxygen Response on the Climate Variability of the Mediterranean Sea

Dissolved oxygen concentration is an important oceanic parameter for the functioning of the marine ecosystem. The distribution of dissolved oxygen in the ocean interior is controlled by the air-sea interaction processes, circulation changes, and biological effects. Climate-driven changes in these processes should, therefore, regulate the oceanic dissolved oxygen variability. In this study, we investigate the temporal and spatial variability of the dissolved oxygen for the period 1960-2011. We focus on the Mediterranean Sea that is characterized by a combination of long-term trends and climatic shifts, known in the literature as "transients". In such an oligotrophic area the atmospheric deposition, which can supply bioactive trace nutrients to the ocean surface, could have a significant impact on the primary productivity and the consumption of the dissolved oxygen. The analysis of a reconstructed database interpolated into 1/8^ο x 1/8^ο grid by means of Data-Interpolating Variational Analysis (DIVA) revealed a combination of processes related to its temporal evolution. At the surface layer the solubility-driven changes determine the dissolved oxygen concentration. In deeper layers the interannual variability is more related to dynamical processes that may involve dense-water convection, biological consumption or mixing, rather than temperature trends. Clear trends cannot be observed in the vertical profile of the dissolved oxygen even though they are evident on a global scale. In the Mediterranean Sea, the observed changes of minimum/maximum oxygen zones are mostly related to abrupt shifts. The attribution of the observed variability involves complex physical and biogeochemical processes and requires further analysis by a coupled model implementation. In this context, a high-resolution configuration of NEMO at 1/36^o resolution is coupled with the biogeochemical model PISCES to carry out long-term hindcast simulations and help to identify each process separately.