Contributions to Global Ocean Observations
55
the desired transparent processing scheme, but the RDI frequency-tracking circuitry
did work. After several years of testing on Scripps ships, during which the demands
on installation and on the compasses available on ships were discovered, the RDI
shipboard ADCP was in operation. The first report of its use is Joyce, Wunsch, and
Pierce (1986) and Chereskin et al. (1989) explains the foibles of the RDI frequency
tracker.
The NORPAX surface-drifter effort turned out tragically. Drifters were attractive in NORPAX because the scientific focus was on the largest scales of variability
near the surface and because satellite tracking was becoming feasible, albeit at a
cost 10 times what it is today. Leaders in the NORPAX effort were Bill Richardson
(the Buoy Project pioneer who had moved to Nova University) and Gerald McNally
(SIO) who were tackling the problems of drogue and sensor survivability in satellitetracked buoys as well as how winds and waves affected accuracy of drifter-measured
surface currents. My own contribution was developing inexpensive drifters tracked
by, and reporting surface temperature measurements through, high-frequency radio transmissions received by military direction-finding stations around the Pacific.
Richardson’s fundamentally experimental approach was appropriate to the key questions of longevity and our groups planned a wintertime trial in the Gulf of Maine using
Richardson’s R/V Gulf Stream. Wayne Hill, a bright and vigorous young engineer,
and I were to join the trial in January 1975. For reasons now forgotten, at the last
minute I could not join the ship. The first report was that the Gulf Stream was missing.
Despite a massive Coast Guard search augmented by an aircraft and two tireless pilots
from Nova University, only one crewman’s body and some debris were ever found.
The R/V Gulf Stream was lost with all hands.
The investigation that followed the Gulf Stream’s loss put a new light on my
father’s adage on eyewitnesses. Reports of the ship being sighted in different places
were wildly inconsistent with each other. Then came the rumors, from Soviet submarines to suicide pacts, which grew without apparent basis. Evidently in a crisis
contradictory information is inevitable, rationality is an early victim, and conspiracy theories flourish. Fortunately, not everything of the NORPAX drifter program
was lost. Gerald McNally continued the quest for surface drifters to map large-scale
surface circulation and he succeeded (see McNally, 1981).
A direct outgrowth of the interest in air–sea interaction and the problems with
rotor-vane current meters was the development of the Vector Measuring Current Meter
(VMCM; Weller and Davis, 1980). It was bad luck that early current meters measured speed and direction rather than components of velocity or even the component
of velocity along the direction of a current-following vane. In either of these alternate methods high-frequency noise can be filtered without significantly biasing the
low-frequency velocity components. But if mooring-induced high-frequency effects
add to the observed speed, processing cannot reverse the error. As the name indicates,
the VMCM measures current components. The VMCM propellers, with a cosine response to the angle between its axis and the current, respond to vector components in a
quasi-linear fashion. Nevertheless, when the oscillatory component is large compared
55
the desired transparent processing scheme, but the RDI frequency-tracking circuitry
did work. After several years of testing on Scripps ships, during which the demands
on installation and on the compasses available on ships were discovered, the RDI
shipboard ADCP was in operation. The first report of its use is Joyce, Wunsch, and
Pierce (1986) and Chereskin et al. (1989) explains the foibles of the RDI frequency
tracker.
The NORPAX surface-drifter effort turned out tragically. Drifters were attractive in NORPAX because the scientific focus was on the largest scales of variability
near the surface and because satellite tracking was becoming feasible, albeit at a
cost 10 times what it is today. Leaders in the NORPAX effort were Bill Richardson
(the Buoy Project pioneer who had moved to Nova University) and Gerald McNally
(SIO) who were tackling the problems of drogue and sensor survivability in satellitetracked buoys as well as how winds and waves affected accuracy of drifter-measured
surface currents. My own contribution was developing inexpensive drifters tracked
by, and reporting surface temperature measurements through, high-frequency radio transmissions received by military direction-finding stations around the Pacific.
Richardson’s fundamentally experimental approach was appropriate to the key questions of longevity and our groups planned a wintertime trial in the Gulf of Maine using
Richardson’s R/V Gulf Stream. Wayne Hill, a bright and vigorous young engineer,
and I were to join the trial in January 1975. For reasons now forgotten, at the last
minute I could not join the ship. The first report was that the Gulf Stream was missing.
Despite a massive Coast Guard search augmented by an aircraft and two tireless pilots
from Nova University, only one crewman’s body and some debris were ever found.
The R/V Gulf Stream was lost with all hands.
The investigation that followed the Gulf Stream’s loss put a new light on my
father’s adage on eyewitnesses. Reports of the ship being sighted in different places
were wildly inconsistent with each other. Then came the rumors, from Soviet submarines to suicide pacts, which grew without apparent basis. Evidently in a crisis
contradictory information is inevitable, rationality is an early victim, and conspiracy theories flourish. Fortunately, not everything of the NORPAX drifter program
was lost. Gerald McNally continued the quest for surface drifters to map large-scale
surface circulation and he succeeded (see McNally, 1981).
A direct outgrowth of the interest in air–sea interaction and the problems with
rotor-vane current meters was the development of the Vector Measuring Current Meter
(VMCM; Weller and Davis, 1980). It was bad luck that early current meters measured speed and direction rather than components of velocity or even the component
of velocity along the direction of a current-following vane. In either of these alternate methods high-frequency noise can be filtered without significantly biasing the
low-frequency velocity components. But if mooring-induced high-frequency effects
add to the observed speed, processing cannot reverse the error. As the name indicates,
the VMCM measures current components. The VMCM propellers, with a cosine response to the angle between its axis and the current, respond to vector components in a
quasi-linear fashion. Nevertheless, when the oscillatory component is large compared