Comparison of shear measurements and mixing predictions with a direct observation of diapycnal mixing in the Atlantic thermocline

Author Posting. © American Geophysical Union, 1995. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research 100 no.C7 (1995): 13481–13498, doi:10.1029/95JC01023.Four sets of velocity and density profiles have been measured with an autonomous profiler during an upper ocean intentional‐tracer (SF6) diapycnal diffusivity measurement, the North Atlantic Tracer Release Experiment (NATRE). The tracer was injected near 310 m depth in the Canary Basin. Two profile sets were collected 6 months after tracer release, and two were collected 1 year after release, all within the horizontal boundaries of the SF6 patch. Shear and strain can be combined with turbulent kinetic energy dissipation and diffusivity measurements (published elsewhere) to test existing expressions for dissipation and diffusivity due to shear‐induced turbulence. These expressions arise from internal‐wave decay modeling. One expression of dissipation parameterized in terms of shear, based on stochastic nonlinear internal‐wave interaction, has fared well empirically; its extension to estimate diffusivity is evaluated. Shear variance of the first two data sets was about 1.6 times GM76, and 2.5 to 3.0 times GM76 in the later sets. The average parameterized mixing estimate computed using all of the temporally limited shear measurements overestimates annual mean NATRE diffusivity, 1.5 × 10−5 m2 s−1, by a factor of 1.2. A modified parameterization gives an underestimate. To first order, this supports the present understanding of open‐ocean diffusivity in terms of fine‐scale shear and internal‐wave decay, that is, the slow diapycnal mixing was not a consequence of unusually low shear. Adjustment of the shear‐induced mixing models to better fit the data is not warranted because of the lack of direct comparability between the various measurements, the expected natural variability of the shear, and sampling errors.This work was supported by National Science Foundation grant OCE-9216204