A global comparison of subtropical underwater formation rates

Subduction-related formation rates for Subtropical Underwater (STUW) are estimated for the North Atlantic and South Indian Oceans and compared with earlier work for the North and South Pacific. Subduction rates are calculated using drifter (1988-1998) and tracer (1987-1995) data. The drifter method involves calculating lateral induction (LI) from the mixed layer depth gradient, and vertical pumping (VP) from Ekman convergence. The tracer method involves calculating pCFC-12 ages on density surfaces to estimate the inverse age gradient. The two methods are independent except for the calculation of the formation area. The tracer and drifter methods agree within the error estimates, except in the North Pacific, and can be used to put bounds on the STUW subduction process: North Atlantic (44-36 m/yr, 2 Sv), North Pacific (26-17 m/yr, 4 Sv), South Pacific (32-33 m/yr, 7 Sv), and South Indian (22-25 m/yr, <1 Sv). In the drifter method, analysis of the terms contributing to subduction of STUW shows, LI is negligible in the North Atlantic and North Pacific, and positive in the South Pacific. In the South Indian Ocean LI is large and negative due to the anomalous poleward location of the STUW. Ekman pumping (EP) is highest in the South Indian Ocean followed by the North Atlantic, consistent with the strength of the wind stress curl. The North Pacific has smaller EP than the South Pacific, even though the wind stress curl is higher in the North. This effect is mainly due to the location of South Pacific STUW closer to the equator than the North Pacific STUW, and more buoyancy forcing in the South Pacific Ocean. The conclusion is that there is considerable variation in thermocline ventilation between ocean gyres.