88
Michael J. McPhaden
July–August 1987 exposed a problem in which the temperature sensors failed progressively from the deepest to the shallowest depths. The manufacturer made design
modifications to fix this problem, but sensors below 120 m systematically failed on
a second trial deployment of four reconditioned drifters in the western tropical Pacific in February 1988. After another round of design modifications and upgrades
to our remaining inventory, we staged a third field test in the California Current in
July–September 1988 in which two drifting thermistor chains transmitted reasonably good data for 3 months. Encouraged by these results, we purchased 13 more
drifters (3 with wind sensors) for a large-scale deployment in the western equatorial
Pacific.
Nearly 2 years behind our original schedule, we launched 19 drifting buoys
between 140
◦ and 165
◦ E in October–November 1989 with the intention of capturing
the upper ocean response to a westerly wind burst. Our timing was perfect and within
days of deployment we caught the onset of a strong wind burst (McPhaden et al.,
1992). However, the thermistor chains failed as they had in the first two field tests,
with 50% of the sensors out of commission after only about 2 months (McPhaden
et al., 1991). Like the Neanderthal, drifting buoys with thermistor chains became an
evolutionary dead end in the development of the TOGA Observing System.
ORIGINS OF THE TAO ARRAY
While the drifting thermistor chain project was frustrated by failure, Stan Hayes
methodically advanced the development of the Autonomous Temperature Line Acquisition System (ATLAS) thermistor chain mooring at PMEL. Like Dave Halpern
before him, Stan had a distinct advantage over the competition, namely, Hugh Milburn
and his engineering division at PMEL. The ATLAS design incorporated many proven
concepts from PMEL current meter moorings used in previous equatorial ocean studies (Milburn and McLain, 1986). In addition, significant cost savings were achieved
by eliminating current meters in favor of temperature as the primary measurement.
Also, elimination of current meters with their movable parts (rotors and vanes, or
propellers) meant that mechanical wear and biofouling in the biologically productive
upper equatorial ocean became less of a constraint on deployment duration. Thus, the
ATLAS mooring design lifetime was 1 year, or twice that of a typical surface current
meter mooring. Real-time data telemetry through Service Argos was also a standard
ATLAS feature.
At the start of TOGA in January 1985 there was one ATLAS mooring in the
equatorial Pacific at 2
◦ S, 110
◦ W transmitting surface air temperature and ocean temperatures in the upper 500 m. Stan added surface winds to the ATLAS measurement
suite in 1986, adapting earlier design concepts developed for real-time wind measurements from current meter moorings (Halpern et al., 1984). Initial ATLAS mooring deployments were concentrated between 110
◦ W and 140
◦ W as part of EPOCS (Hayes et
al., 1989). To extend the array into the western Pacific, Stan established collaborations
Michael J. McPhaden
July–August 1987 exposed a problem in which the temperature sensors failed progressively from the deepest to the shallowest depths. The manufacturer made design
modifications to fix this problem, but sensors below 120 m systematically failed on
a second trial deployment of four reconditioned drifters in the western tropical Pacific in February 1988. After another round of design modifications and upgrades
to our remaining inventory, we staged a third field test in the California Current in
July–September 1988 in which two drifting thermistor chains transmitted reasonably good data for 3 months. Encouraged by these results, we purchased 13 more
drifters (3 with wind sensors) for a large-scale deployment in the western equatorial
Pacific.
Nearly 2 years behind our original schedule, we launched 19 drifting buoys
between 140
◦ and 165
◦ E in October–November 1989 with the intention of capturing
the upper ocean response to a westerly wind burst. Our timing was perfect and within
days of deployment we caught the onset of a strong wind burst (McPhaden et al.,
1992). However, the thermistor chains failed as they had in the first two field tests,
with 50% of the sensors out of commission after only about 2 months (McPhaden
et al., 1991). Like the Neanderthal, drifting buoys with thermistor chains became an
evolutionary dead end in the development of the TOGA Observing System.
ORIGINS OF THE TAO ARRAY
While the drifting thermistor chain project was frustrated by failure, Stan Hayes
methodically advanced the development of the Autonomous Temperature Line Acquisition System (ATLAS) thermistor chain mooring at PMEL. Like Dave Halpern
before him, Stan had a distinct advantage over the competition, namely, Hugh Milburn
and his engineering division at PMEL. The ATLAS design incorporated many proven
concepts from PMEL current meter moorings used in previous equatorial ocean studies (Milburn and McLain, 1986). In addition, significant cost savings were achieved
by eliminating current meters in favor of temperature as the primary measurement.
Also, elimination of current meters with their movable parts (rotors and vanes, or
propellers) meant that mechanical wear and biofouling in the biologically productive
upper equatorial ocean became less of a constraint on deployment duration. Thus, the
ATLAS mooring design lifetime was 1 year, or twice that of a typical surface current
meter mooring. Real-time data telemetry through Service Argos was also a standard
ATLAS feature.
At the start of TOGA in January 1985 there was one ATLAS mooring in the
equatorial Pacific at 2
◦ S, 110
◦ W transmitting surface air temperature and ocean temperatures in the upper 500 m. Stan added surface winds to the ATLAS measurement
suite in 1986, adapting earlier design concepts developed for real-time wind measurements from current meter moorings (Halpern et al., 1984). Initial ATLAS mooring deployments were concentrated between 110
◦ W and 140
◦ W as part of EPOCS (Hayes et
al., 1989). To extend the array into the western Pacific, Stan established collaborations