Time–depth plots of temperature, salinity, and density to
300 m (Fig. 3A–C) were generated from monthly median
profiles. Percentage saturation of dissolved oxygen, Chl a
concentration, and particulate optical backscatter coefficients
were scaled to 150 m to better visualize their patterns
(Fig. 3D–F) and because the optical sensors were turned
off below this depth to conserve batteries. Variations in the
optical signal were restricted largely to the upper 100 m
(with the exception of occasional encounters with detached
nepheloid layers over the slope and some indications of
particle settling below 100 m at certain times of the year),
justifying this power-saving procedure. Satellite measurements
of SST spatially averaged over the area within the
box (Fig. 1; data not shown) were similar to annual
patterns of Seaglider temperatures averaged over the upper
10 m. The maximal summer surface temperature recorded
by Seaglider was 17.85uC; minimal winter surface temperature
was 8.49uC. Although all temperatures were within
the ranges recorded at National Oceanic and Atmospheric
Administration buoy No. 460055 since 1976, they were on
the lower end of that range, most likely because of buoy
location 3u further offshore at 46uN, 131uW ,http://www.
ndbc.noaa.gov/images/climplot/46005_st.jpg..
The depth of the seasonal mixed layer, assessed as the
depth of the pycnocline, varied both seasonally and
interannually (Fig. 4A). Because of the approximately
monthly spacing between Seaglider samplings, it was not
possible to resolve seasonal changes more finely. Maximal
depth of the winter pycnocline varied from approximately
80 m in 2004 and 2005 to .110 m in 2007, with the deepest
mixed layers in late winter and early spring. The transition
between deep winter and shallow summer mixed layers,
typically occurring between May and June in 2004 and
2006 or between June and July in 2005 and 2007, was rapid
enough to elude capture by Seaglider’s monthly sampling.