Advection-Diffusion Process Inference via Statistical Oceanographic Methods in the North Atlantic and Southern Oceans
Quantities pertaining to ventilation, the process by which water is exchanged between the surface mixed layer and the interior of the ocean, are estimated here: To do this, the utility of the thin-plate spline (TPS) as an interpolation procedure is described as a way to estimate mapping errors. Second, using the concept of transit-time distribution (TTDs), or mass fractions of water getting from the surface to an interior location over some time interval, the volume transports of Labrador Sea Water (LSW), Antarctic Intermediate Water (AAIW), and the LSW component of North Atlantic Deep Water (NADW) are estimated through three different repeated cross-sections of the ocean. Evidence is found that both the single and mixture inverse Gaussian representations of a TTD are insufficient, that at least some of the AAIW seen south of Tasmania forms from Subantarctic Mode Water (SAMW) that is entrained and detrained in the southeast Indian Ocean, and that at least some of the LSW seen south of Iceland is formed in the Irminger Basin. Third, estimates for rates at which the North Atlantic thermocline is ventilated advectively are given using profiler, scatterometer satellite, and altimetric satellite observations from several consecutive years. Ekman pumping is a secondary effect relative to the horizontal and temporal structure of the mixed layer depths (MLDs) for controlling ventilation. Fourth, the trends of mode water subduction rates and the extent to which along-isopycnal diffusion contributes to subduction in the North Atlantic and Southern Oceans are estimated by utilizing TTDs. Using a statistical technique called Bayesian model averaging (BMA), observations of CFC-11 are in- corporated into a model simulation-derived TTD-based estimates of ventilation rates, which are compared with diapycnal velocity-based estimates, and uncertainties are quantified. It is found that along-isopycnal diffusion has increasingly contributed to SAMW ventilation over time and that both ventilated North Atlantic Subpolar Mode Water (SPMW) and, to a lesser extent, ventilated North Atlantic Subtropical Mode Water (STMW) erode by processes with at least two distinguishable time scales. Last, using BMA at each grid point, a density dependence of the model parameter for the along-isopycnal diffusivity, which has yet to be determined to be the path-averaged along-isopycnal diffusivity of the real ocean, is inferred. The model parameter for the along-isopycnal diffusivity is suggested to be smaller for water masses that form by deep convective processes than for water masses that form by turbulent processes that mix distinguishable water masses at intermediate depths.
- Oceanography