Modifications of mode water properties by sub-mesoscales in a bio-physical model of the Northeast Atlantic

Northeast Atlantic mode waters (NEAMW) are formed by subduction in a region which is a strong sink of atmospheric CO2. The mechanisms underlying this sink were thoroughly examined in the frame of the POMME experiment, which covered the 2001 seasonal cycle with particular focus on meso to sub-mesoscales. The biological and physical data collected during POMME were used to initialize, constrain and evaluate a regional, 4.5&\#xa0;km-resolution, bio-physical model simulation of the 2001 seasonal cycle in the region of NEAMW formation. We used this model to examine the contribution of sub-mesoscales on the annual budget of carbon export below the mixed-layer and on NEAMW biogeochemical characteristics. This is done by comparing a high-resolution (eddy) simulation at 4.5&\#xa0;km resolution with a high-diffusivity (non-eddy) simulation where the only change is a stronger eddy diffusivity. We found that the model mixed-layer depth is more stratified and closer to reality in the eddy simulation. This result confirms with observational data the proposed mechanism of restratification of the upper ocean driven by sub-mesoscales. We found that the phytoplankton bloom and the subduction of NEAMW display strong contrasts at the sub-mesoscale. Nevertheless, the mean intensity of the bloom and of the subsequent biological pump are only marginally modified by sub-mesoscales (less than 5\%), while the intensity of the physical pump (subduction of carbon) is moderately reduced (−10\%) in the non-eddy experiment. Moreover, the biogeochemical and thermodynamical characteristics of NEAMW are substantially affected, with, in particular, a wider range of densities and biogeochemical characteristics biased toward winter conditions in the eddy simulation. These differences ensued essentially from lateral induction across sub-mesoscale filaments, which is found to contribute to subduction before seasonal stratification, i.e. before the effective subduction period.