Quantifying pore water exchange across the sediment-water interface in the deep sea with in situ tracer studies

A numerical model of NaBr tracer transport in benthic flux chamber incubations has been developed to simulate temporal changes in chamber water tracer concentrations and applied to four deep sea locations: the California and North Carolina margins, the Ceara rise, and the Cape Verde plateau. Model variables include the chamber volume (which determines the initial tracer concentration), and rates of diffusive and nonlocal exchange across the sediment-water interface. Chamber volume and solute exchange rate estimates were obtained by optimizing the fit between observations and model results using a chi super(2) statistical scheme. We observe enhanced solute exchange in regions with high organic carbon rain rates, over a wide range of bottom water oxygen concentrations. Furthermore, our results demonstrate that solute exchange rates cannot be assumed constant, even within relatively short deployments. In six of 22 cases, chamber volumes and corresponding benthic solute fluxes derived from model fits were more than 20% greater than values calculated using linear extrapolations of chamber tracer concentration-time relationships. To avoid this discrepancy, future benthic chamber studies should seek to minimize the time interval between tracer injection and sampling while maintaining sufficient time to achieve complete homogenization of the tracer in the chamber waters.