Abstract
Two coarse-resolution model experiments are carried out on an OGCM to examine the effects of North Atlantic Deep Water (NADW) formation on the thermohaline circulation (THC) and ventilation timescales of the abyssal ocean. An idealized age tracer is included to gauge the ventilation in the model. One experiment is forced with the present-day climatology, the other has a negative salinity anomaly imposed on the North Atlantic surface to eliminate the formation of NADW. The Atlantic branch of the THC is reversed and the ventilation of the deep Atlantic basin is severely reduced when NADW formation is prevented. The Southern Ocean forms bottom water in both experiments, but downwelling and upwelling in the Southern Ocean are both reduced when NADW is included due to increased stratification of the water column. The Indian and Pacific basins are upwelling regions in both experiments and upper-level upwelling is stronger there when NADW is included; this change leads to cooler temperatures and reduced ventilation of the upper ocean.
There is a need for the distinction to be drawn between upwelling, the vertical movement of water parcels, and positive buoyancy forcing, which converts denser water masses into lighter ones. A density-regime analysis reveals that most of the positive buoyancy forcing associated with the THC occurs equatorward of 30° latitude in both hemispheres. The THC in the model ocean seems more modular than the “conveyor belt” metaphor implies; each high-latitude region functions quasi-independently, responding to its own heat and freshwater forcing. The newly ventilated subsurface water masses formed in each region compete for space in the water column. The production of NADW increases the upper-level stratification throughout the World Ocean. The idealized age tracer reveals that the deep Pacific and Indian basins are primarily ventilated by southern bottom water and the inclusion of NADW causes only minor changes. The qualitative pattern of ventilation in the model is determined by the number and location of subsurface water mass formation regions; the actual ages in the deep ocean are very sensitive to the vertical diffusion prescribed in the model.
Corresponding author address: Paul J. Goodman, JISAO, University of Washington, Box 354235, Seattle, WA 98195-4235.