Munk [1966] used vertical
advection/diffusion balances for the
central Pacific to deduce an upwelling
rate of 10
m/sec and a vertical
(diapycnal, i.e., crossing isopycnals)
eddy diffusivity of 10
m
/sec. Numerical models have
required a general value of this size in
order that the basic circulation match
the observed circulation. Basin-wide
calculations of required eddy
diffusivity for the bottommost layers,
based on how much deep water flowing
northward through a channel must be
converted to less dense water somewhere
to the north, also suggest this
magnitude of eddy diffusivity [ Hogg,
1987, for the Vema Channel in the South
Atlantic; Roemmich, personal
communication, for Samoan Passage in the
South Pacific]. However, microstructure
measurements in the upper ocean have
indicated that the vertical eddy
diffusivity is more like 10
m
/sec, an order of magnitude
smaller [ Osborn and Cox, 1972]. Recent
abyssal microstructure measurements in
the northeastern Pacific also show this
value [ Toole et al., 1994]. Deliberate
dye release experiments in the Santa
Barbara basin and in the North
Atlantic's subtropical gyre also show a
vertical eddy diffusivity of 10
m
/sec, in remarkable
confirmation of microstructure
measurements made at the same locale
[ Ledwell et al., 1993]. Thus the
observed vertical eddy diffusivity of
the ocean is an order of magnitude
smaller than required by models and the
vertical profiles, suggesting that
boundary mixing might be much more
important than interior mixing [ Toole et
al., 1994; Davis, 1994].