Mass and volume transport variability in an eddy-filled ocean

The possibility that the oceanic general circulation is undergoing changes as part, or the cause, of major climate shifts is being intensely discussed, with some published results relying on data from moorings spanning the North Atlantic Ocean. The circulation is, however, extremely noisy. Here, I use existing estimates of the frequency and wavenumber content of geostrophic eddies in the ocean to show that variations in ocean-wide integrated transport must appear even in the absence of a true long-term trend. Expected fluctuations exceed ±20 × 10 kg s (or ±20 × 10 m s) and exhibit multi-year timescales. Existing knowledge of the eddy field allows predictions of observed variability and produces lower bounds on the (multi-decadal) timescale required to detect true trends of a large magnitude. Detecting and understanding the effect of climate change on the ocean circulation requires observations in three dimensions over long periods of time. Until comparatively recently, the ocean circulation was viewed as consisting primarily of a large-scale, very slowly changing, flow. The discovery in the 1970s (ref. 5) of an intense field of variability, with mid-latitude spatial scales of 100 km and larger, and timescales of months and longer, greatly complicates efforts to discern long-timescale shifts in the basin-wide circulation. In particular, moorings deployed across the North Atlantic, as in the Rapid Climate Change Program (RAPID) at 25N, will produce apparent basin-scale mass transport trend-like variability, whose magnitude can be estimated a priori—the purpose here. The ocean circulation is in near-geostrophic balance, meaning that there is an equilibrium between the Coriolis and pressure forces. An important peculiarity of this balance, heavily relied on by oceanographers, is that the total mass of fluid moving in the meridional overturning circulation (MOC) above any depth z0, ocean-wide, depends only on the pressure difference across the basin—as long as bottom topography does not intervene above z0. Thus, we need only observe pressure changes, for example, adjacent to the Bahamas on the west, and near the coast of Africa on the east, to determine fluctuations in the ocean circulation between those two places above the crest of the Mid-Atlantic Ridge (at about 2,000m). Pressure data located between these two points are readily shown to ‘drop-out’ when computing the total meridional movement of water. More generally, and independent of method, zonally integrated meridional transport in geostrophic flows is extremely sensitive to end effects, whether the measurements are of pressure, or of velocity directly. REPRESENTING THE EDDY FIELD