phosphate do as well. Since the WHP specifications refer to a standard error, one would expect
that 68% of any experimental values would fall
within the specifications. Thus, we again see that
nutrient measurements in the one-time WHP
survey meet the overall requirements. It remains to
be seen whether oxygen comparisons follow this
trend since these have not yet been made for the
Pacific and Indian Oceans. As with salinity measurements, one should be able to identify particular sections with relatively high or low nutrient or
oxygen values. The WHP specifications for these
variables represent reproducibility constraints, since
no standard reference materials were available for
an absolute calibration of dissolved oxygen or
nutrients in the WHP, in contrast to salinity.
3.1.3 Current measurements in the
shipboard hydrographic programme
3.1.3.1 The nature of Acoustic Doppler
Current Profilers
An Acoustic Doppler Current Profiler (ADCP) is a
high-frequency multibeam sonar. A short pulse of
sound is transmitted along each beam, and the
Doppler shift of the sound scattered back is estimated as a function of time since the transmission,
or equivalently, range to the scatterers. The
Doppler shift is proportional to the relative velocity component along the beam between the transducer and the scatterers. There are typically four
beams, all at the same elevation (angle from the
vertical), and at 90° intervals of azimuth. If we
assume the scatterers are moving with the water,
and the water velocity is roughly uniform in any
horizontal plane intersected by the four ADCP
beams, then the four radial velocity components at
any given range yield estimates of both horizontal
velocity components, and two independent estimates of the vertical velocity component. Estimates from successive range cells provide a vertical
profile of the water velocity relative to the instrument. If the orientation and velocity of the instrument relative to the earth are known, the profile of
water velocity relative to the earth can be calculated. Range, resolution, and accuracy depend on
the sonar’s frequency and on other instrumental
and environmental factors; typical values for
common 150 kHz instruments are 300 m vertical
range, 8 m vertical resolution, and 11 cm s
91 standard deviation of single-ping horizontal velocity
component estimates. Ensembles of successive
single-ping profiles are normally vector-averaged
to reduce the variance of the velocity estimate.
ADCPs have been deployed in many ways and
locations: on the seafloor, looking upward; looking up or down from a mooring, or downward
from a moored buoy; on the bottom of a ship; and
lowered from a ship, usually as part of a standard
profiling instrument package including a CTD and
a water sampler. We are concerned here only with
the latter two: shipboard (SADCP) and lowered
(LADCP).
In calculating absolute (earth-relative) velocity
profiles from an ADCP mounted on a ship that is
underway, say at 5 m s
91
, the desired signal – the
ocean current – is the small difference between
two large numbers: the velocity of the ship relative
to the earth and the velocity of the ship relative
to the water. Both numbers therefore must be
measured with high accuracy, and with particular
attention to minimizing bias – any systematic
errors that persist longer than a few minutes. Two
navigational measurements are crucial: the position fixes that are first-differenced to calculate the
ship’s velocity over the ground, and the heading
measurements used to transform the ADCP velocity vectors from the instrument coordinate system
into the geographical coordinate system: east,
north and up. Prior to 1991, uncertainty in transducer heading was the primary factor limiting
shipboard ADCP accuracy for most applications.
The Global Positioning System (GPS) revolutionized shipboard ADCP profiling by providing not
only excellent position fixes, but unprecedented
heading accuracy. We will return to this topic.
Shipboard ADCPs measure upper ocean currents with effective horizontal resolution as fine
as 1–2 km, but are blind to currents below a few
hundred metres depth. Lowering a self-contained,
internally recording, ADCP on a wire extends the
profiling range to the ocean bottom, but with horizontal resolution limited to the station spacing.
The method is in many ways similar to shipboard
profiling, but there is a crucial difference: over
most of the depth range, the velocity of the instrument cannot be determined directly by navigation.
The horizontal position of the instrument relative to the ship is known only at the times of
launch and recovery. As the instrument descends,
the ADCP records a large number of overlapping velocity profiles, each with a range of only
3.1 Shipboard Observations during WOCE
111
King, Firing and Joyce
that 68% of any experimental values would fall
within the specifications. Thus, we again see that
nutrient measurements in the one-time WHP
survey meet the overall requirements. It remains to
be seen whether oxygen comparisons follow this
trend since these have not yet been made for the
Pacific and Indian Oceans. As with salinity measurements, one should be able to identify particular sections with relatively high or low nutrient or
oxygen values. The WHP specifications for these
variables represent reproducibility constraints, since
no standard reference materials were available for
an absolute calibration of dissolved oxygen or
nutrients in the WHP, in contrast to salinity.
3.1.3 Current measurements in the
shipboard hydrographic programme
3.1.3.1 The nature of Acoustic Doppler
Current Profilers
An Acoustic Doppler Current Profiler (ADCP) is a
high-frequency multibeam sonar. A short pulse of
sound is transmitted along each beam, and the
Doppler shift of the sound scattered back is estimated as a function of time since the transmission,
or equivalently, range to the scatterers. The
Doppler shift is proportional to the relative velocity component along the beam between the transducer and the scatterers. There are typically four
beams, all at the same elevation (angle from the
vertical), and at 90° intervals of azimuth. If we
assume the scatterers are moving with the water,
and the water velocity is roughly uniform in any
horizontal plane intersected by the four ADCP
beams, then the four radial velocity components at
any given range yield estimates of both horizontal
velocity components, and two independent estimates of the vertical velocity component. Estimates from successive range cells provide a vertical
profile of the water velocity relative to the instrument. If the orientation and velocity of the instrument relative to the earth are known, the profile of
water velocity relative to the earth can be calculated. Range, resolution, and accuracy depend on
the sonar’s frequency and on other instrumental
and environmental factors; typical values for
common 150 kHz instruments are 300 m vertical
range, 8 m vertical resolution, and 11 cm s
91 standard deviation of single-ping horizontal velocity
component estimates. Ensembles of successive
single-ping profiles are normally vector-averaged
to reduce the variance of the velocity estimate.
ADCPs have been deployed in many ways and
locations: on the seafloor, looking upward; looking up or down from a mooring, or downward
from a moored buoy; on the bottom of a ship; and
lowered from a ship, usually as part of a standard
profiling instrument package including a CTD and
a water sampler. We are concerned here only with
the latter two: shipboard (SADCP) and lowered
(LADCP).
In calculating absolute (earth-relative) velocity
profiles from an ADCP mounted on a ship that is
underway, say at 5 m s
91
, the desired signal – the
ocean current – is the small difference between
two large numbers: the velocity of the ship relative
to the earth and the velocity of the ship relative
to the water. Both numbers therefore must be
measured with high accuracy, and with particular
attention to minimizing bias – any systematic
errors that persist longer than a few minutes. Two
navigational measurements are crucial: the position fixes that are first-differenced to calculate the
ship’s velocity over the ground, and the heading
measurements used to transform the ADCP velocity vectors from the instrument coordinate system
into the geographical coordinate system: east,
north and up. Prior to 1991, uncertainty in transducer heading was the primary factor limiting
shipboard ADCP accuracy for most applications.
The Global Positioning System (GPS) revolutionized shipboard ADCP profiling by providing not
only excellent position fixes, but unprecedented
heading accuracy. We will return to this topic.
Shipboard ADCPs measure upper ocean currents with effective horizontal resolution as fine
as 1–2 km, but are blind to currents below a few
hundred metres depth. Lowering a self-contained,
internally recording, ADCP on a wire extends the
profiling range to the ocean bottom, but with horizontal resolution limited to the station spacing.
The method is in many ways similar to shipboard
profiling, but there is a crucial difference: over
most of the depth range, the velocity of the instrument cannot be determined directly by navigation.
The horizontal position of the instrument relative to the ship is known only at the times of
launch and recovery. As the instrument descends,
the ADCP records a large number of overlapping velocity profiles, each with a range of only
3.1 Shipboard Observations during WOCE
111
King, Firing and Joyce