Modeling the distribution of Nd isotopes in the oceans using an ocean general circulation model

Abstract

Nd isotopic compositions of authigenic phases in marine archives are increasingly used as a tracer of past changes in ocean circulation. Such applications assume that water masses are “tagged” with distinct Nd isotope compositions in source regions and that away from these regions the Nd isotope ratios observed in deep waters mainly reflect water mass mixing. However, the degree to which the simple assumption of “quasi-conservative mixing” reflects the complicated reality of general circulation in the oceans is not well understood. Here, we present numerical simulations of the Nd isotope distribution in seawater using an offline ocean general circulation model (OGCM), treating the Nd isotopic composition of seawater as a conservative tracer. Nd isotope data from modern surface waters are used to generate an interpolated map of Nd isotope compositions for the entire surface ocean. This map is treated as a fixed surface boundary condition that is propagated into the interior ocean by the advective–diffusive transport of the ocean model until a steady state is reached. This treatment of Nd isotopes was chosen because it mimics the behavior of well-known and well-understood conservative tracers, such as potential temperature and salinity. The model results for all non-surface grid points are then compared to published data. This simple approach produces Nd isotope estimates for North Atlantic Deep Water that are consistent with observations, but produces values lower than observed in the deep Pacific and Southern Oceans, indicating the need for a source of radiogenic Nd directly to the deep Pacific Ocean. Sensitivity tests show this to be a robust result. By fixing the return flow of the Pacific Ocean to the Southern Ocean with Nd isotope ratios that are consistent with modern observations, the model results are brought into close agreement with a majority of the available data. With the addition of a deep Pacific source of radiogenic Nd, more than half of the simulated values from the interior ocean fall within one ε-unit of the corresponding observations. Our results thus indicate that the deep Pacific Ocean requires a source of radiogenic Nd in addition to inputs to the surface ocean, and that when the deep waters of the North Atlantic and North Pacific are set with realistic values, a majority of intermediate areas are consistent with quasi-conservative mixing of these two end-members.