Scripps Institution of Oceanography
J.Geophys.Res., submitted
Preliminary Results from Directly Measured Mid-Depth Circulation in
the Tropical and South Pacific
by Russ E. Davis
Introduction
It is uncertain that with achievable densities of hydrographic observations
even the most sophisticated inverse models would be capable of determining both
absolute velocity and irreversible mixing processes. For this reason WOCE fielded
an ambitious program of direct velocity observations including various types of
neutrally buoyant floats deployed in sufficient number to measure the average
absolute velocity and its eddy variations near 1000 m depth. There exists no pre-
cise definition of "the general circulation velocity" to use when comparing hydro-
graphic inferences with these direct velocity measurements, but for WOCE the work-
ing definition was "that part of the decade-average velocity field with scales of
200 km or more." The one-kilometer design depth for floats reflects the objective
of determining a level of known motion for referencing mean geostrophic shear in
the face of time variability. The 1000 m depth was a compromise between the gen-
eral decrease of eddy variability with depth, the desire to maximize the area over
which a reference velocity could be determined without bottom interference, and
the increased energy required for autonomous floats to cycle to the surface from
greater depths. A target 500-km resolution for the mean was selected as an achie-
vable resolution well matched to the density of hydrographic observations that
would be available from historical and WOCE cruises at the end of that experiment.
The South Pacific WOCE float program involved 300 Autonomous Lagrangian Cir-
culations Explorer (ALACE) floats deployed to a depth near 900 m primarily from
WOCE Hydrographic Program cruises. ALACEs are small neutrally buoyant floats
(Davis et al., 1992) that, by adjusting their buoyancy, cycle from depth where
velocity is tracked to the surface where the float is located by System Argos sat-
ellites and relays observations like submerged depth and temperature through those
satellites. In the South Pacific survey most ALACEs remained at depth for 25.5
days, rose to the surface in about one hour, remained on the surface for 24 hrs
while transmitting to Argos, and returned to depth, achieving within 100 m of
their target depth in about two hours. The position of surfacing and diving were
estimated, with typical precision of 1 to 3 km, by extrapolating the Argos fixes
to the ascent and descent times using an objective estimation procedure.
South Pacific deployments began in mid 1991 and continued through early 1996.
At their typical cycle time, these floats have battery energy to operate for over
70 cycles or five years and at this writing 65% are still operating. [] The
total number of observations from the South Pacific survey is expected to even-
tually be twice what is available now.
The Data
All available observations up to the end of 1996 were [released]. In the
region between 120E and 60W this consisted of 10,290 submerged displacements
over durations between 10 and 100 days, summing to 702 float-years of observation.
Since these data are novel, it is worthwhile to discuss their character and the
procedures used to edit and correct them.
All floats report averages of their submerged temperature & pressure (depth).
The temperature records are used here only to verify the pressure record or to
estimate submerged depth when the pressure record is either suspect or the pressure
reading exceeds the range of the float's analog to digital converter. When depth
readings are suspect the time series of temperature, depth and submerged velocity
were examined to classify each uncertain displacement into one of three categories
(a) apparently submerged at or below nominal depth, b) apparently not submerged or
submerged at a depth too shallow to be reported, or (c) depth unknown. Early de-
signs were prone to losing their wave damping disk after some time and then, being
excessively buoyant, did not submerge. Since the floats can ground themselves, the
entire set of positions was searched for instances where the ocean depth at the
float positions was possibly as shallow as the float depth. If bottom contact was
suspected and the submerged speed was lower than in neighboring displacements where
grounding was not indicated, [the record has been marked.]
Some displacements extended over more than one programmed submerge-and-surface
cycle. Based on laboratory experience with the ALACE hydraulic system & the obser-
vation that the number of transmissions received from the surfacings immediately
before and after these missed cycles was rarely below average, it is believed more
likely that these floats failed to surface rather than that they surfaced but
failed to signal Argos. Consequently these data were retained.
ALACE data from 1991 deployments on WHP lines P17C, P16C and SR3A.
ALACE data from 1992 deployments on WHP lines P6, P13C, P14C, P31, P16S & P17S.
ALACE data from 1993 deployments on WHP lines P19, SR3, P14N and P10.
ALACE data from 1994 deployments on WHP P18, P31, P21, and Drake94 & Tasman95.
ALACE data from 1995 Pacific deployments from Tasman95.
[Later deployments and other oceans have also been included.]
ALACE trajectories consist of submerged displacements and on-surface drifts.
Averaged submerged pressure (dbar) and temperature are P,T.
Missing and out-of-range cold & warm Ts are marked as -9.999, -8.888 & 88.888.
Missing depths are marked -777 if believed to be submerged or -999 if believed
to be surfaced. Out-of-range shallow & deep pressures are marked -888 and 8888.
[The trajectories consist of a sinking point and a surfacing point, extrapolated
from the surface Argos locations using objective interpolation. Since the floats
can ground themselves, the set of positions was searched for instances where the
ocean depth at the float position was possibly as shallow as the float depth. If
bottom contact was suspected, the record was marked.]