Does Stommel's Mixed Layer “Demon” Work?

Abstract

Stommel argued that the seasonal cycle leads to a bias in the coupling between the surface mixed layer and the main thermocline of the ocean. He suggested that a ?demon? operated that effectively only allowed fluid at the end of winter to pass from the mixed layer into the main thermocline. In this study, Stommel's hypothesis is examined using diagnostics from a time-dependent coupled mixed layer-primitive equation model of the North Atlantic (CME). The influence of the seasonal cycle on the properties of the main thermocline is investigated using two methods. In the first, the rate and timing of subduction into the main thermocline is diagnosed using kinematic methods from the 1° resolution CME fields. In the second, tracer diagnostics of the CME and idealized experiments using a ?date? tracer identifying the timing of subduction are performed. Over the subtropical gyre, both approaches generally support Stommel's hypothesis that fluid is only transferred from the mixed layer into the main thermocline over a short period, ?1 month, in late winter/early spring. Tracer date experiments are also conducted using the eddy-resolving ?° CME fields. Eddy stirring is found to enhance the rate at which the tracer spreads into unventilated regions, but does not alter the seasonal bias of the Stommel demon mechanism.