Estimates of Surface Heat Fluxes
69
experiment. Eventually we all got in our ships and planes and onto our islands, and
commenced measuring from November 1992 through February 1993.
LEARNING TO MEASURE, AND CHECK, HEAT FLUXES
Especially for oceanographers, the centrepiece of the Coupled Ocean–Atmosphere
Response Experiment (COARE) was Bob Weller’s IMET mooring, located at 2
◦ S, east
of Papua New Guinea. Three ships—R/V’s Wecoma, Moana Wave, and our Australian
ship Franklin—worked in its neighbourhood. On the bow of each ship meteorologists
like Chris Fairall, Clayton Paulson, Steve Esbenson, and Frank Bradley undertook the
deeply laborious task of accurately measuring all four components of heat flux, the
two components of momentum flux, and precipitation. At the stern we oceanographers
performed the slightly easier job of keeping the SeaSoars and ADCPs going, to map
out the ocean temperature/salinity (T /S) and current structure in detail. All told, 22
instruments had to operate simultaneously at near-maximum precision without gaps
to close the heat budget, even to the seemingly modest accuracy of 10 W/m
2 (about
5% of the average input of solar radiation, the biggest term in the net flux).
I will not forget the sight of five research vessels steaming together past the
IMET buoy, to calibrate their meteorological measurements. Planes flew low overhead, to calibrate theirs. Two intercomparisons like this proved crucial to the eventual
success of this part of COARE, because the differences in raw data among the platforms were pretty sobering (due to difficult problems like salt on humidity wicks, birds
sitting on radiometers, and distortion of wind speeds by the ships’ superstructure).
By the efforts of Frank Bradley and others involved in trying to obtain the most
accurate air–sea fluxes possible, the COARE IMET mooring, the ships, and the aircraft
produced a new level of accuracy. Intensive pre- and postcalibrations, dedicated inthe-field comparisons, and new methods paid off. Special Teflon screens around the
humidity sensor on the IMET mooring let moisture in without allowing salt crystals
to grow. Intensive study of radiometer performance resolved problems with factory
calibrations; with sensitivity to solar heating; and with the possibility that sea birds
would land on one of the radiometers (to be detected by examining the minute-byminute data on the assumption that the bird can only sit on one radiometer at a time).
The results from the IMET buoy itself were described by Weller and Anderson (1996).
It took several years to sort these basic problems out; but when this was done,
the resulting heat fluxes from the bow of the ship matched the ocean heat uptake from
measurements at the stern of the ship to within the 10 W/m
2 we had set ourselves ( Feng
et al., 2000). The oceanographic part of the problem was nontrivial, since it involved
using the divergence of ADCP currents around small boxes to assess upwelling rates,
requiring very accurate knowledge of any misalignment of the ADCP. These issues
were reviewed by Godfrey et al. (1998). We had problems on Franklin, but we also
achieved this 10 W/m
2 accuracy using the same techniques and data from a later
Indian Ocean cruise (Godfrey et al., 1999).
69
experiment. Eventually we all got in our ships and planes and onto our islands, and
commenced measuring from November 1992 through February 1993.
LEARNING TO MEASURE, AND CHECK, HEAT FLUXES
Especially for oceanographers, the centrepiece of the Coupled Ocean–Atmosphere
Response Experiment (COARE) was Bob Weller’s IMET mooring, located at 2
◦ S, east
of Papua New Guinea. Three ships—R/V’s Wecoma, Moana Wave, and our Australian
ship Franklin—worked in its neighbourhood. On the bow of each ship meteorologists
like Chris Fairall, Clayton Paulson, Steve Esbenson, and Frank Bradley undertook the
deeply laborious task of accurately measuring all four components of heat flux, the
two components of momentum flux, and precipitation. At the stern we oceanographers
performed the slightly easier job of keeping the SeaSoars and ADCPs going, to map
out the ocean temperature/salinity (T /S) and current structure in detail. All told, 22
instruments had to operate simultaneously at near-maximum precision without gaps
to close the heat budget, even to the seemingly modest accuracy of 10 W/m
2 (about
5% of the average input of solar radiation, the biggest term in the net flux).
I will not forget the sight of five research vessels steaming together past the
IMET buoy, to calibrate their meteorological measurements. Planes flew low overhead, to calibrate theirs. Two intercomparisons like this proved crucial to the eventual
success of this part of COARE, because the differences in raw data among the platforms were pretty sobering (due to difficult problems like salt on humidity wicks, birds
sitting on radiometers, and distortion of wind speeds by the ships’ superstructure).
By the efforts of Frank Bradley and others involved in trying to obtain the most
accurate air–sea fluxes possible, the COARE IMET mooring, the ships, and the aircraft
produced a new level of accuracy. Intensive pre- and postcalibrations, dedicated inthe-field comparisons, and new methods paid off. Special Teflon screens around the
humidity sensor on the IMET mooring let moisture in without allowing salt crystals
to grow. Intensive study of radiometer performance resolved problems with factory
calibrations; with sensitivity to solar heating; and with the possibility that sea birds
would land on one of the radiometers (to be detected by examining the minute-byminute data on the assumption that the bird can only sit on one radiometer at a time).
The results from the IMET buoy itself were described by Weller and Anderson (1996).
It took several years to sort these basic problems out; but when this was done,
the resulting heat fluxes from the bow of the ship matched the ocean heat uptake from
measurements at the stern of the ship to within the 10 W/m
2 we had set ourselves ( Feng
et al., 2000). The oceanographic part of the problem was nontrivial, since it involved
using the divergence of ADCP currents around small boxes to assess upwelling rates,
requiring very accurate knowledge of any misalignment of the ADCP. These issues
were reviewed by Godfrey et al. (1998). We had problems on Franklin, but we also
achieved this 10 W/m
2 accuracy using the same techniques and data from a later
Indian Ocean cruise (Godfrey et al., 1999).