Ocean Deoxygenation Conference | Kiel 2018

Increased observational efforts have revealed a multi-decadal decrease of oxygen concentrations with superimposed interannual to decadal variability in the oxygen minimum zone (OMZ) of the eastern tropical North Atlantic (ETNA). Recent studies have linked this variability to long-term changes in the ventilation by the latitudinally alternating zonal jets (LAZJs).
In this study a 1.5 layer non-linear shallow water model coupled to an advection-diffusion model is employed in basins with either rectangular or Atlantic geometry to obtain a conceptual understanding of the influence of the LAZJs on the ventilation of the ETNA OMZ. Using an equatorial annual period forcing, westward propagating off-equatorial Rossby waves are generated that subsequently break up into non-linear eddies. The responsible non-linear triad instability mechanism thereby sets the amplitude and size of the generated eddies, which rectify to LAZJs when temporally averaged. An oxygen-mimicking tracer is transported by the resulting velocity field, forming a region with minimum tracer concentration whose location is in general agreement with the observed ETNA OMZ. The thickness-weighted tracer budget reveals that the Eulerian mean advective flux convergence outweighs the eddy advective flux convergence in balancing the strongly simplified tracer consumption. Thickness-weighted averaging the advective flux convergence also yields an eddy mixing term which is shown to play an important role in the budget and is analysed in more detail. Despite the purely annual period forcing, interannual to decadal and longer tracer variability is excited in the basin, including the region of the ETNA OMZ. A comparison between modelled and observed oxygen trends in the lower OMZ does not lead to a rejection of the null hypothesis that the observed decadal oxygen trends are part of the system's intrinsic variability. However, the observed pronounced decadal oxygen decrease in the upper OMZ during 2006-2013 is not reproduced by the model. On a multi-decadal time scale, the picture is reversed. In contrast to the upper OMZ, the multi-decadal oxygen decrease in the lower OMZ is not reproduced by the idealised model. While this would support the idea of an anthropogenically driven long term deoxygenation of the lower OMZ, it is important to bear the simplicity and shortcomings of the model in mind. Further, the sparsity in measured oxygen data before the recently increased observational efforts complicates the reliable estimation of multi-decadal trends.