2014 GSA Annual Meeting in Vancouver, British Columbia (19–22 October 2014)

Paper No. 194-7
Presentation Time: 9:40 AM

USING HEAT AS A TRACER TO QUANTIFY TRANSIENT FLUSHING BELOW THE SEDIMENT-WATER INTERFACE IN A SANDY CONTINENTAL SHELF


WILSON, Alicia, School of the Earth, Ocean & Environment, Univ of South Carolina, 701 Sumter St, Columbia, SC 29208, WOODWARD, Gwendolyn, Environment and Sustainability Program, University of South Carolina, 701 Sumter St, Columbia, SC 29208 and SAVIDGE, William, Skidaway Institute of Oceanography, 10 Ocean Science Circle, Savannah, GA 31411

Ocean currents and waves drive porewater flushing through sandy seafloor sediments, analogous to hyporheic flow in streams. This flow contributes to particle trapping and nutrient cycling in benthic environments, but field observations of the depth and timing of flushing are extremely sparse. We used heat as a tracer for flushing below the seafloor and developed a numerical model to invert the resulting thermal dataset to determine the timing, magnitude and depth of flushing.

Temperature was recorded at four depths in the upper 12 cm of the sediment column at a site 50 km offshore of St. Catherine’s Island, GA, during two periods in the summer of 2008. The numerical model simulated transient 1-D (vertical) heat conduction and advection. The thermal effects of rapid 3-D flushing near the seafloor were included via an additional thermal dispersion term, based on similar approaches in the hyporheic flow literature. The dispersion term decreased exponentially with depth and was therefore parameterized by an apparent dispersivity and a half-depth (analogous to a radioactive half-life). An optimization routine was used to estimate the apparent dispersivity and half-depth at 15-minute intervals. The thermal dataset did not extend deep enough to allow estimation of the rate of long-term porewater advection, but the code can be altered to estimate advection rates for datasets that reach below the zone of rapid flushing.

Multiple combinations of apparent dispersivity and half-depth matched the observed thermal signal to within error, but optimal solutions also minimized the total added dispersion. All solutions indicated similar periodicity. Particularly during the first observation period, rapid flushing of the upper 10 cm of the sediment column occurred as the tide came in. Flushing occurred despite the lack of local seafloor topography (e.g. ripples). Together, this suggests that flushing was driven by thermal convection, as flood tides brought cooler water over the site, rather than by physical flushing associated with tidal currents. Thermal convection has the potential to be very common in sandy continental shelf environments over a range of seasons.