Coastal counter-currents setup patterns in the Gulf of Cadiz

Alongshore coastal counter-currents (CCC) are frequent features of Eastern Boundary Upwelling Systems, where they temporally alternate with upwelling driven jets of opposite direction. Along the northern margin of the Gulf of Cadiz inner shelf, these CCCs are oriented poleward (eastward) and responsible for sharp temperature increases during the upwelling season, along with potential decline in water quality at the coast. This research is based on a multi-year ADCP velocity time-series (2008-2017), recorded at a single location (23 m water depth) over 13 deployments up to 3 months-long. The analysis focuses on the water column alongshore velocities during current inversions (i.e., the transition from equatorward upwelling jets to poleward CCCs). A set of parameters were derived from the flow structure to identify distinct types of inversions and to hypothesize about their driving mechanisms. Results show that 77% of the inversions start near the bed, propagating then to the upper layers. The bottom layer also changes direction before the surface layer for most events (71%). The vertical shear in this case is one order of magnitude greater than in the (less frequent) opposite situation. No seasonal variability is observed in the CCC occurrences. However, the parameters analysed in this study suggest different types of inversion between winter and summer. In winter, inversions are well defined (low variability), with similar patterns near the surface and bed layers as a result of a strong barotropic component. In summer the inversion patterns are more variable. In particular, the upper and bed layers are often importantly decoupled during inversions, indicating the strengthening of baroclinicity. A categorization of inversions events is proposed based on cross-correlation and multi-variable analyses of the developed parameters. Various types of inversion are obtained, suggesting that CCCs are driven by different forcings that may act separately or jointly.