Quantifying the Role of Mixed Layer Entrainment for Water Mass Transformation in the North Atlantic

Abstract

Forcing by wind stress and air?sea buoyancy flux climatologies between σt = 23.5 and 26.5 results in differing water mass transformations in the North Atlantic, reflecting the opposing tendencies of wind stress and air?sea fluxes. This difference needs to be reconciled in terms of various processes that lead to diapycnal advection and mixing. This study attempts to quantify the contribution of one such process to water mass transformation?the small-scale but ubiquitous process of mixed layer entrainment and deepening. An estimate is computed using formulas developed earlier, the Levitus hydrography, and a mixed layer model forced by observed fluxes. The mixed layer and forcing data are taken from the Marine Light Mixed Layer Experiment mooring, which includes both spring and fall mixed layer transitions. The sensitivity to averaging of synoptic events is also explored. Calculations presented here indicate that, if hourly winds are used, the water mass transformation due to mixed layer entrainment has annual peak contributions of about O(4) Sv for σt = 24.0 (Sv ≡ 106 m3 s?1). This is comparable to the annual transformation attained by diapycnal mixing in the upper-ocean water masses. However, with daily averaged winds and without diurnal variation in buoyancy forcing, this contribution is up to an order of magnitude smaller. Another set of mixed layer simulations includes an annual cycle with a shallow and strong summer thermocline. Inclusion of synoptic summer forcing for this scenario leads to transformation values several times larger than above, about O(14) Sv at σt = 24.0. The peak contribution in this case is almost two orders of magnitude smaller if the synoptic forcing is averaged daily and the diurnal cycle is not resolved. These results suggest that the numerical diagnostics using general circulation models may significantly underestimate entrainment mixing if the combination of diurnal variation and synoptic wind events is not resolved or explicitly parameterized.