reduce instrument noise). Results for an estuarine deployment (Fig. 14.24) show
correlation of quite large fluctuations compared to the deep sea. You are left to think
about why the flux (TOU) over a deployment of many minutes is the average of w′
[O 2 ]′, the product of the fluctuating components (after subtracting the means from all
measures of w and [O 2 ] to get w′ and [O 2 ]′). A comparison of this technique to
canister and electrode profile results at 1450 m in Sagami Bay (eastern Honshu)
showed excellent agreement (Berg et al. 2009). Eddy correlation equipment is being
commercially manufactured, and widespread data will be forthcoming (stay tuned).
Fig. 14.24 Simultaneous 30-second series of [O 2 ] fluctuations and vertical velocities
(+ is up) recorded at 25 Hz at 15 cm above the sediment in Aarhus Bay, Denmark.
The horizontal lines are means. Thick lines are from low-pass filtering.
(^) (After Berg et al. 2003.)
Measurement Results
(^) An assembly of the global data-set (Glud 2008), eliminating sites with very low [O 2 ],
shows an exponential drop-off of both TOU and DOU with depth (Fig. 14.25a). For
TOU, the change is 85-fold from 10 m to 4000 m. The scatter does not overwhelm the
definite relationships, but it implies strong differences in organic-matter supply
among sites. This corresponds to an exponential increase in the depth of oxygen