Lateral mixing and advection of reactive isotope tracers in ocean basins: observations and mechanisms

Abstract

The global mapping of dissolved non-conservative isotopic tracers (Be, Pb, Nd) in the world's oceans has revealed relative homogeneities in isotope ratios that by far exceed those of the tracer's isotopic input ranges. For tracers with continental sources such as Pb and Nd, a first possible mechanism to account for such an isotopic homogenisation is the mixing within their source areas, such as rivers, deltas, or continental shelves. However, a further mixing mechanism within ocean water is likely, because in the Pacific Ocean the two isotopes of Be are mixed to within only 13% dispersion, despite their entirely different sources (cosmogenic 10Be entering the oceans interior by rain, stable 9Be entering the ocean margins from the continents). A second possibility is that well-mixed tracers are imported from the Antarctic Circumpolar Current (ACC) into the Pacific by Antarctic Bottom Water (AABW). However, since the area of the ACC is also one of intensive tracer scavenging, for Pb and most of the Pacific's Be imported sources are not dominant by mass balance. Hence a third mechanism, lateral intra-basin mixing between coastal and central gyre waters is suggested to be the most important mixing mechanism. We perform a numerical assessment of the feasibility of tracer mixing by lateral advection and eddy mixing, in two dimensions in a basin-width subtropical ocean gyre with laterally varying scavenging and input intensity. For Be, strong cross-streamline diffusion is required to achieve homogenisation of the 10Be derived from the interior with the margin-sourced 9Be. This is favoured by long scavenging residence times ( τ) of Be, which is the case in both thermocline waters and deep waters. For Pb, most of which is entirely margin-sourced, isotopic homogenisation is favoured by high velocities (Péclet Numbers) in the gyre and τ of at least 20 y. This is the case in thermocline waters. If short-residence time tracers are dispersed in such an efficient way over large sections of an ocean basin important implications for the use of these isotopes in paleoceanography are: (1) sites of tracer input (e.g. dust) do not necessarily spatially correlate with the distribution of such input in either seawater or marine sediments; (2) gyre-wide, or even basin-wide responses might arise from relatively local changes in the provenance of isotopes, and need not necessarily result from global climatic changes, or variations of the deep thermohaline circulation.