The Combined Sensor Program:
An Air-Sea Science Mission in the Central and Western Pacific Ocean
Madison J. Post, Christopher W. Fairall
NOAA Environmental Technology Laboratory
Jack B. Snider, Yong Han, Allen B. White, Warner L. Ecklund
Cooperative Institute for Research in the Environmental Sciences (CIRES)
University of Colorado, Boulder, Colorado
Klaus M. Weickmann
NOAA Climate Diagnostic Center
Daniel I. Cooper
Los Alamos National Research Laboratory
Los Alamos, New Mexico
University of Miami
Patricia K. Quinn
Pacific Marine Environmental Laboratory
Steven M. Sekelsky and Robert E. McIntosh
University of Massachusetts
Robert O. Knuteson
University of Wisconsin
Madison J. Post
NOAA/ERL R/E/ ET4
Boulder, Colorado 80303
(303) 497-6048 email@example.com
Accepted for publication on August 21, 1997.
This paper will appear in the December volume of the Bulletin of the American Meteorological Society.
Abstract: Thirteen national research organizations joined forces on a 30-day, 6800 nautical
mile survey of the Central and Tropical Western Pacific on NOAA's Research Vessel
Discoverer. The Combined Sensor Program (CSP), which began in American Samoa on 14
March 1996, visited Manus Island, Papua New Guinea, and ended in Hawaii on 13 April, used a
unique combination of in situ, satellite, and remote sensors to better understand relationships
between atmospheric and oceanic variables that affect radiative balance in this climatically
important region. Besides continuously measuring both short- and long-wave radiative fluxes,
CSP instruments also measured most other factors affecting the radiative balance, including
profiles of clouds (lidar and radar), aerosols (in situ and lidar), moisture (balloons, lidar, and
radiometers), and sea surface temperature (thermometers and Fourier Transform Infrared
Radiometers -- FTIRs). Surface fluxes of heat, momentum, and moisture were also measured
continuously. DOE's ARM Program used the mission to validate similar measurements made at
their CART site on Manus Island, and to investigate the effect (if any) of large nearby land
masses on the island-based measurements.
The most quantitative statement of our accumulated knowledge about the climate system is the
climate model or the general circulation model (GCM). The shortcomings of today's GCMs
have been recently evaluated by Browning (1994). One major shortcoming is our insufficient
understanding of the interaction between clouds and the Earth's radiation field. Clearly, there are
important but poorly understood feedbacks between the direct warming effects from an increased
greenhouse gas concentration and the indirect effects on global temperature from the resulting
changes in cloud distributions and types. Another related factor is the complex role of air-sea
interaction, in which clouds are a critical process. Marine clouds are dynamically, structurally,
and microphysically distinct from continental clouds. Marine boundary layer (MBL) clouds
strongly influence global climate because their relatively high albedos (compared with the ocean
background) give rise to large deficits in absorbed solar radiative flux at the top of the
atmosphere, while their low altitude prevents significant compensation in thermal emission
(Randall et al. 1984). Deep convective clouds common in the western Pacific create profound
reductions of the radiation reaching the surface; surface solar flux levels in the tropics can be
significantly lower than those in, say, Buffalo, New York, in December. Tropical deep
convection also generates huge areas of cirrus clouds whose effect on net cloud forcing is still
uncertain. For example, consider the recent debate about possible cloud radiative flux absorption
being significantly greater than that represented by present models (Cess et al. 1995; Chou et al.
1995; Evans et al. 1995; Ramanathan et al. 1995). This issue is particularly critical in the tropics
where clouds are thick and the solar flux is highest.
In the last decade, a number of research programs have focused on various aspects of this marine
problem. For example the NASA First International Satellite Cloud Climatology Program
(ISCCP) Regional Experiment (FIRE) has executed a series of experiments in midlatitudes to
investigate the role of cirrus and MBL stratus clouds in the climate system (Randall et al. 1984;
Albrecht et al. 1988, 1995). In the near future, FIRE will venture to the Arctic to study stratus
clouds of all types. The Tropical Ocean-Global Atmosphere (TOGA) program conducted an
extensive Coupled Ocean-Atmosphere Response Experiment (COARE) in the tropical western
Pacific Ocean in 1992-93 (Webster and Lukas 1992). COARE focused on air-sea fluxes and the
role of deep convection and larger-scale coupled interactions in short-term climate variability
(primarily El Niño). The pivotal role of the western Pacific "boiler box" in global climate has
been emphasized by recent debates about the importance of radiative flux and oceanic surface
temperature coupling, a phenomenon usually referred to as the thermostat hypothesis (Zhang et
al. 1995). This issue was examined (Zhang and Grossman 1996) in the Central Equatorial
Pacific Experiment (CEPEX), which followed immediately after COARE.
In recognition of the need for a different experimental approach to climate problems (both marine
and continental), the Department of Energy (DOE) initiated the Atmospheric Radiation
Measurement (ARM) program (Stokes and Schwartz 1994). The heart of the ARM program is
the development of several Cloud And Radiation Testbed (CART) remote-sensing facilities as
experimental analogs of a GCM grid cell. A second element of ARM involves participation in
intensive field studies to improve and test sensors and to investigate physical processes in more
detail. ARM has identified the Southern Great Plains (SGP), the North Slope of Alaska (NSA),
and the Tropical Western Pacific (TWP) climate regimes for installation of CART sites. The
SGP site is currently fully operational; the other two sites are not yet fully instrumented. The
ARM TWP effort has issues in common with those of COARE (the role of air-sea interaction
processes in interannual climate variability) and FIRE (the representation clouds in climate
models with an emphasis on satellite linkages), but its focus on local but long-term remote and in
situ measurements is quite different.
The development of the CART instrument systems for the TWP is under way; however, their
deployment in oceanic locations presents special problems. For instance, the harsh marine
environment typically takes a great toll on sensors; an island may impart unknown distortions to
measured profiles; and systems mounted on ships may require very demanding motion
corrections. Current plans for TWP CART sites call for several surface-based remote-sensing
packages, Atmospheric Radiation and Cloud Stations (ARCS), to be deployed on islands in the
western Pacific. The need and practicality of a purely ocean-based (ship or buoy) site is still
being debated. Clearly, an oceanic deployment will eliminate the uncertain complications from
island-generated perturbations to the flow and the resulting changes to the clouds, but the
logistics and costs may be prohibitive. The information required to make thoughtful choices is
To attack this problem, we proposed to draw on extensive experience at the NOAA
Environmental Technology Laboratory (ETL) in making marine atmospheric measurements. The
ETL effort began with ship-based direct measurements of air-sea fluxes in the COARE pilot
cruise in the equatorial Pacific in 1990 (Young et al. 1992) and with our first efforts at remote
sensing with a ceilometer and a stabilized Doppler wind profiling radar in the equatorial Eastern
Pacific in 1991 during the Tropical Instability Wave Experiment (TIWE; Chertock et al. 1993).
This same system was deployed in the Azores region in 1992 as part of the Atlantic
Stratocumulus Transition Experiment (ASTEX; White et al. 1995); land-based remote sensors
were also deployed on an island (Frisch et al. 1995). The complete system was deployed on three
cruise legs in the TWP during COARE (Young et al. 1995; Fairall et al. 1996a). Details on the
instruments and techniques used in these and subsequent efforts can be found in Fairall et al.
Besides ETL's interest in climate and forecasting issues in marine meteorology, both the NOAA
National Environmental Satellite, Data, and Information Service (NESDIS) and NOAA Corps
(uniformed organization established to operate NOAA's fleet of ships and aircraft) were
interested in developing capabilities to equip a few NOAA research vessels with a suite of in
situ and remote sensors. This would help NESDIS better calibrate and validate satellite data
products, help NOAA's National Weather Service (NWS) fill in data voids over the oceans to
improve numerical weather prediction, and provide valuable, quasi-operational research data to
the climate modeling community. With the support of these organizations, 30 days of ship time
for the first year of the plan, 1996, was requested and allocated on NOAA R/V Discoverer.
The Combined Sensor Program (CSP), a joint NOAA-DOE effort, was thus established to
promote seagoing climate research and as a continuation of the ongoing island and ship-based
studies. We then developed a plan to conduct a cruise to the TWP aboard the NOAA ship
Discoverer, which was instrumented with the most advanced in situ and remote sensors in the
ETL inventory, supplemented by sensors from a number of universities and government
laboratories active in the ARM program. Several investigators from the Aerosol
Characterization Experiment (ACE) and the World Ocean Circulation Experiment (WOCE) that
preceded CSP were invited to leave some of their instrumentation on board the Discoverer and
join CSP, providing aerosol and air/ocean chemistry measurements and direct satellite reception.
Thus, the inventory of measurements for this cruise far exceeded any of ETL's previous efforts.
Besides the obvious opportunity to address the numerous ARM scientific issues discussed above,
this field program was also intended to provide an opportunity for ARM to gain further
experience with shipboard remote sensing of clouds that might be useful in the development of
TWP CART facilities. The scientific goal of the CSP mission thus evolved to "obtain an
unprecedented data set on marine clouds and radiation" in the central and western tropical Pacific
by deploying a suite of ship-based sensors on a 30-day mission. It was also to develop
"improved analysis techniques and improved cloud-radiative interaction parameterizations and
cloud dynamical models, and to evaluate cloud effects on the surface energy budget of the
tropical western Pacific (TWP)." Implicitly, another goal was to help establish whether the
island-based measurements could be considered representative of the region, including the open
ocean, or if they were influenced by nearby land masses. Thus, our intention was to deploy a
ship-based system functionally equivalent to various components of a prototype ARCS system
and to compare our shipboard measurements with those from similar island-based sensors in the
prototype ARCS system operated by Pennsylvania State University and the ARM TWP team at
Manus Island in Papua New Guinea.
To satisfy these goals, we planned to gather data to 1) parameterize cloud-aerosol-radiation
interactions in a tropical marine environment; 2) improve parameterizations of surface fluxes of
heat, momentum, and moisture; 3) determine cloud cover and cloud layering statistics in the
TWP; 4) evaluate marine boundary layer height and internal characterization; 5) characterize
TWP stratus and cumulus clouds in terms of radiative effects, ice and water mass, and effective
radii; 6) validate aerosol and radiance data from polar-orbiting satellites; 7) characterize the
effects of the MBL on measurements of aerosols, clouds, and radiative balance near Manus
Island; 8) evaluate the effects of island topography on measurements of aerosols, clouds, and
radiative balance; 9) examine details of the surface properties of the ocean and the dynamics of
cool-skin and warm-layer effects (Fairall et al. 1996b); 10) obtain precipitation distributions and
statistics in conjunction with simultaneous radar data; 11) investigate a host of remote sensor
measurement and algorithm development issues; and 12) cross-calibrate measurements with
similar DOE/CART instrumentation on Manus Island.
2. Cruise strategy and execution
Because of the importance of obtaining meaningful, simultaneous measurements with the ARM
instrumentation at the TWP CART site, we needed to spend as much ship time as possible in the
vicinity of Manus Island, Papua New Guinea. This required the Discoverer to steam at
maximum speed from our starting point at Pango Pango, American Samoa (14.28S, 170.68W),
to Manus Island (2.0S, 147.3E), and to then steam to our final destination of Honolulu, Hawaii
(21.2N, 157.8W), at maximum speed. Rather than taking direct routes to these three ports, we
decided to sacrifice a small amount of time and undertake a more meaningful survey of the
central Pacific's warm pool by spending significant time cruising parallel to the equator between
1S and 2S, a region relatively free of landmasses yet representative of equatorial conditions.
We therefore planned a nominal track that progressed northwest from Samoa to 2S, 180W
(dateline), then directly west to a point northeast of the TWP CART site.
Although winds are highly variable in this part of the Pacific, a perusal of climatological winds
prior to the mission indicated that southeasterly flow is likely over Manus Island in March and
April. An approach to the island from the northeast would permit the same airmasses to be
sampled over Manus and by Discoverer and in the open ocean downwind of Manus, albeit not
simultaneously. Therefore we planned to steam southwest towards Manus Island until we were
at a point 100 km from the island, and hold station for 2 days. We would then steam quickly to a
point 30 km from the island and hold station for another 2 days, before proceeding to the island
and holding station there for 2 days. These stations would permit us, within synoptic scale
variability, to discern statistical differences in island-based and ship-based measurements as a
function of separation, and thereby to begin to assess the representativeness of island-based
measurements with respect to similar measurements over the open ocean.
Outbound from Manus Island to the northeast, we planned to reoccupy the same 30- and 100-km stations for another 2 days apiece, before repeating the meridional survey of the central Pacific by once again steaming east just south of the equator. This plan would carry us back to the point where we originally intercepted the 2S parallel, before diverting northeast to Hawaii. All significant data-taking would therefore be space and time symmetric. Differences between measurements inbound to Manus Island and outbound would help quantify variability and representativeness in the region. The plan would give us 10 days in the vicinity of Manus Island, accomplish a 3200-km, 11-day survey of conditions in the west and central equatorial Pacific, and get the ship to Honolulu within our allotted mission time of 30 days.
The actual cruise track is shown in Fig. 1.
To a large extent we were able to follow the nominal
plan outlined above. However, while en route to Manus Island we were asked by NOAA to
rescue one Tropical Atmosphere-Ocean (TAO) buoy that had broken loose from its mooring and
was drifting near our track, and to service another nearby TAO buoy. We decided to honor these
requests, because the primary mission would not be compromised, and because the meridional
survey would not be affected significantly. The only other important departure from the plan was
the launch of a small, drifting heat-flux buoy from the first 100-km station. Instead of
proceeding directly to the 30-km station 2 days later, we diverted the ship to recover the heat-flux
buoy, which had been carried by currents to the west, considerably off our intended track. The
Battelle National Laboratory (BNL) radiometer package was off-loaded at Manus Island when we
arrived there on March 27. It was subsequently installed and operated at the CART site for the
remainder of the CSP mission.
3. Instrument descriptions and performance
A list of onboard investigators, affiliations, and instruments is given in Table 1. Contributing
organizations not listed in Table 1 were the NOAA Corps (ship operations and data), the NPGS
(AVHRR receiver), NASA (disdrometers), and the University of Washington (aerosol
measurements). Thus 13 organizations were involved in the mission. Two of the instrument
systems, the Pacific Marine Environment Laboratory (PMEL) aerosol sensors and the Naval
Postgraduate School (NPGS) Advanced Very High Resolution Radiometer (AVHRR) satellite
receiver, had participated in prior FY96 Discoverer missions and had only to be reactivated
and/or reconfigured to participate in CSP.
Throughout the cruise we launched at least four balloons carrying Vaisala radiosondes per day,
but near Manus Island we increased the frequency to eight per day. Both the ship's balloons and
CSP balloons were launched, and often both the ship's receiver (A.I.R.) and CSP receiver
(CLASS) tracked the sondes. Twice daily, for standard synoptic reporting times, sonde data were
transmitted through the Geostationary Operational Environmental Satellite (GOES) satellite into
the Global Telecommunication System (GTS) database. In windy conditions (> 20 knots relative
wind) launches were difficult and many sondes were lost, resulting in data gaps. Typical
pressure altitudes reached by the sondes were 75 and 45 hPa for 100 and 200 g balloons,
There is some evidence that standard radiosondes may be subject to moisture biases in the
tropics. Systematic dry biases with Vaisala sondes on the order of 1 g kg-1 were observed in the
boundary layer for groups of launches from a number of different platforms in TOGA COARE
(Bradley and Weller, 1997). Because of incomplete records, it is not known if this dry bias is
associated with specific manufacturers, batches, or the handling of the sondes; some evidence
suggests that storage in air-conditioned spaces just prior to launch may be a factor. For the CSP
cruise, several steps were taken to minimize the thermal and humidity shock to which the sondes
were subjected prior to launch. The sondes were stored in the shade at ambient temperature and
humidity. During setup up and the Omega navigation lock period (about 20-30 minutes), sondes
were hung outdoors in a shaded space that was well ventilated with marine air. Sondes were
never allowed to bake in the direct sun prior to launch. A preliminary check of one sonde per
day on ten consecutive days yielded an average of 1.5 g kg-1 difference between the humidity at
20 m height (i.e., our surface observation) and a representative value for the mixed-layer, close to
the nominal difference of 1.25 g kg-1 expected on the basis of similarity theory and near-surface
aircraft observations (Bradley and Weller 1997). Thus, our preliminary opinion is that dry biases
were not a significant source of error for the CSP cruise; however, this issue will be examined in
more detail. If there is a dry bias, it would cause a similar bias in radiometer retrievals of total
integrated water vapor that were calibrated using radiosonde data. A bias of 1 g kg-1 over a depth
of 1 km would represent a bias of about 0.1 cm, versus a typical total of 5 cm, in the amount of
precipitable water vapor measured by the microwave radiometer.
b. Radar wind profiler
This mechanically-stabilized 915-MHz system (Fairall et al. 1997) performed well, providing
continuous coverage of horizontal and vertical components of wind in the lower atmosphere
throughout the cruise. It often detected clouds and precipitation overhead. Its data acquisition
parameters were matched to those of the Manus Island profiler, to facilitate comparisons. Lowest
level winds (< 500 m) measured by this profiler may be contaminated by sea clutter, and may be
biased low (up to 1 m s-1) .
c. Flux system
The high-speed (20-Hz) temperature, wind, and humidity sensors mounted on the bow tower
worked well throughout the cruise, but the sensors' data-logging system experienced some
startup problems during the first 3 days of the mission. The flux system sensed and compensated
for instantaneous ship motion in real time. Broadband radiative fluxes and conventional bulk
meteorological variables were also obtained as part of this system.
d. Cloud radar
This two-frequency, polarimetric system (Sekelsky and McIntosh 1996) experienced problems
throughout the cruise, but through innovative repairs still acquired a significant dataset. It
occupied a position far aft on the starboard fantail, and was subjected to large-amplitude, low-frequency vibrations, especially when the ship accelerated. One of its air conditioners failed on
the first day of the mission, necessitating replacement. Only one frequency (95 GHz) was
operative en route to Manus Island; the radar began taking data at 95 GHz the second day of the
meridional survey (18 March). At Manus Island we received an emergency shipment of repair
parts and got the 33-GHz channel working (28 March), but the 95-GHz channel failed nearly
simultaneously. Thus radar data obtained near Manus Island after 28 March and during the
second (easterly) leg of the meridional survey were with the 33-GHz channel. However, on 7
April this channel became problematic again; hardware components and data taking were
switched to the 95-GHz channel for the remainder of the cruise. Still, a significant new dataset
for central and western Pacific clouds was obtained, especially when one considers
simultaneously obtained supporting data. Images of cloud radar data are available at
e. Marine-atmospheric emitted radiance interferometer (M-AERI)
This prototype instrument measured upwelling and downwelling infrared spectra at high spectral
resolution (0.5 cm-1) from near visible to the thermal infrared wavenumbers (550 to 3000 cm-1),
both at zenith and at a number of nadir-viewing angles (Smith et al. 1996; Knuteson et al. 1997).
This permits one to infer sea surface temperature (SST) of the top millimeter of the ocean (the
"skin") to a high degree of accuracy (±0.2 K absolute) and precision (±0.1 K, 3 sigma), and to
calculate profiles of atmospheric temperature and moisture in the marine boundary layer. It
operated well throughout the cruise, acquiring good data 82% of the time. Periods of rain
accounted for 12 % of the down time (optics were covered), while maintenance (2%) and loss of
detector coolant (3%) accounted for most of the remaining down time.
f. Fourier transform infrared radiometer (FTIR)
This instrument is similar to the M-AERI instrument, but it obtained data only while pointing
vertically. Its resolution is 1.0 cm-1 with a spectral range from 500 to 2000 cm-1 (Shaw et al.
1995). This FTIR operated throughout the cruise, monitoring downwelling zenith radiation more
continuously than the M-AERI did. During station-keeping near Manus its blackbody references
were realigned, after which time comparisons with the M-AERI were remarkably good. It is vital
for proper analysis of FTIR and M-AERI data to know when clouds are overhead; information
from the ceilometer, lidar, all-sky camera, microwave radiometers, and cloud radar will therefore
be used heavily in analyses of these data.
g. Microwave radiometers
This two-channel (23.87- and 31.65-GHz), zenith-viewing system continuously and
simultaneously monitored integrated cloud liquid and gaseous water substance overhead. The
system employed a spinning reflector (Jacobson and Nunnelee 1997) to direct down-welling
radiation from the atmosphere into the radiometers' antenna and prevent accumulation of
precipitation, thus permitting observations in nearly all weather conditions. Satisfactory tip-
calibrations (Hogg, et al. 1983) were not possible due to a small antenna misalignment that could
not be corrected underway, because of ship motion. Instead, data from 19 radiosonde profiles
acquired under clear conditions during the early part of the mission were used to calibrate the
system, which performed continuously and solidly throughout the mission. Any bias in
radiosonde-derived integrated water vapor will cause a bias of similar magnitude in the
radiometer data (see discussion on biases in subsection (a) above).
h. Elastic and Raman lidar
Two lidar systems were used during the mission--solid-state Nd:YAG system that produces
pulses of light at 1.06 µm (near infrared) and 0.53 µm (green), and an excimer gas (fluorine)
system that produces pulses of light at 0.24 µm (ultraviolet). The former system measures light
backscattered from aerosols, cloud particles, and air molecules which is nearly unchanged in
frequency, except for possible Doppler shifts; hence it is called an elastic lidar. The ultraviolet
lidar measures weak signals that have been re emitted by lidar-excited molecules at a frequency
typically below the exciting frequency. The amount of the shift represents the vibrational or
rotational state to which the molecule was excited, as discovered by Raman; the amount of shift
is species specific, enabling one to measure the concentrations of different gaseous components
of the atmosphere. By measuring the amount of Raman-shifted light from both water vapor and
nitrogen molecules as a function of range, the Raman lidar can remotely measure the mixing ratio
profiles. By scanning the lidar beam, one can map areal or vertical distributions of the gasses
(Cooper et al. 1997).
Both lidars began collecting data the second night of the mission (16 March), including a series
of moisture and aerosol profiles up to cloud base level, and at times higher. However, within 8
hours it was discovered that smoke, oil, and soot from the ship's engine exhaust had deposited on
the lidars' scanning mirrors. When these deposits were illuminated by the intense laser light,
they vaporized, oxidizing the mirrors' coatings as well. The result was nearly disastrous (total
loss of both lidars), but the lidar operators developed ingenious temporary solutions to allow
partial data taking. Even with frequent cleaning of the remaining optics, all exposed mirrors
were completely ruined by 3 April. About this same time the high-voltage supply for the
Nd:YAG laser failed irreparably. The combination of events curtailed all lidar data acquisition
for the remainder of the mission (10 days). Despite the setbacks, exciting new results were
apparent in the data, and the lidar observations will be vital for interpreting several of the other
datasets (e.g., from the M-AERI, FTIR, and aerosol samplers).
This small, eyesafe, commercial laser radar system performed continuously throughout the
cruise, measuring cloud bases occurring below 3.66 km (12,000 ft) altitude. A similar instrument
on Manus Island was set up to record cloud bases to higher altitudes of 7.62 km (25,000 ft).
j. AVHRR and GMS satellite sensors
Except for 23 and 24 March, when the satellite receiving system had to be realigned after a
computer power failure, AVHRR satellite data were received continuously throughout the
mission. Images of 1.1-km resolution were obtained in five channels of wavelength 0.63, 0.86,
3.7, 11.0 and 12.0 µm. Geostationary Meteorological Satellite (GMS) water vapor images of
0.1 (62 km) resolution in the 6.5-7.0 µm band were also archived for postanalysis. Degraded
images of 0.25 (156 km) resolution were made into video "loops" in S-VHS format, with each
loop showing synoptic-scale features and the ship's track. The higher resolution GMS images
are available at http://www.cdc.noaa.gov/~climsat/products.html.
k. Aerosol samplers
The in situ samplers for characterizing boundary layer aerosols performed admirably throughout
the cruise. Both aerosol chemistry and size distribution were measured continuously and results
were plotted in near-real time throughout the cruise (Quinn et al. 1996). Sample air for all
measurements was drawn through a 6-m long heated sample inlet and dried to 50% relative
humidity. The top of the inlet was 18 m above sea level and 10 m forward of the ship's stack.
Measurements were made only when the concentration of particle > 15 nm diameter was small
enough to indicate no contamination, the relative wind was > 3 m s-1, and the relative wind
direction was forward of the inlet.
The number size distribution between 0.02 and 0.6 µm was measured every 10 min with a
differential mobility particle sizer at a relative humidity of 30%. The number density between
0.6 and 9.6 µm was measured at the same time at a relative humidity of 50% by an aerodynamic
particle sizer. Seven-stage multijet cascade impactors were used to determine mass size
distributions of Cl-, Br-, NO3-, SO4=, MSA-, Na+, NH4+, K+, Mg2+, and Ca2+, with time resolution
or 12 - 24 h. For these measurements, the 50% aerodynamic cutoff diameters were 0.27, 0.37,
0.64, 1.2, 2.3, 4.7, and 12.0 µm. Two-stage impactors measured total organic and elemental
carbon on a 3-day time scale. Aerosol total scattering and hemispheric backscattering
coefficients were determined with an integrating nephelometer at wavelengths of 0.45, 0.55, and
0.70 µm. The aerosol absorption coefficient at a wavelength of 0.55 µm was also measured.
Interpretation of these detailed aerosol size and chemistry data give strong indications of surface
air mass origin and age, and they will be valuable in better understanding sea-air-cloud
interactions and the shortwave radiative balance.
l. SST sensors
A number of in situ and remote sensors measured the skin, surface, and bulk temperatures of the
upper ocean throughout the cruise. These included the ship's measurements on water drawn
from a depth of 5 m near the bow, two infrared thermometers, two floating sensors, and M-AERI. The floating sensors were deployed only while the ship was holding station or moving
slowly. The complete set of readings permits us to quantify and understand relationships
between SSTs measured at depths of 1 mm, 10 cm, and 5 m.
While the ship was stationkeeping, it was pointed approximately into the wind with sufficient
power to maintain about 1 m s-1 forward motion. This allowed us to maintain steerage and to
minimize any potential thermal effects of the ship's hull on the near-surface sea temperature
measurements. Both near-surface sea temperature sensors were trailed in the water from
outrigger booms near the bow to keep them 2-3 meters away from the hull. Turbulence shed
from the ship's hull may have generated some local vertical mixing, but this modest forward
speed should have been sufficient to eliminate significant warming of the water at the sensors.
m. All-sky camera
This system began recording hemispherical images of the sky and clouds shortly after we began
the first (westward) half of the meridional survey, but only during daylight hours. On the second
(eastward) half of the survey it also recorded images at night when there was sufficient
moonlight. Time-lapse animation of the entire mission's image sequence is available in VHS
n. Sun photometer
This multichannel (filter wheel) instrument was on its first field deployment and experienced
considerable problems. It acquired direct solar flux data at several visible and infrared
wavelengths intermittently before Discoverer arrived at Manus Island, but only when it was
pointed and held on the sun manually. During station-keeping near Manus Island it began to
track the sun automatically, but not reliably. Therefore more data were acquired during the
second half of the mission, but they must be scrutinized for accuracy, as misalignments can easily
be misinterpreted as increased atmospheric attenuation.
o. Portable radiation package
This system of several radiometers performed nearly flawlessly during the first half of the
mission, providing important information on direct and diffuse solar (shortwave) radiation, and
downwelling infrared radiation (Reynolds and Smith 1997). For the last half of the CSP mission
it was taken off the ship, then set up and operated on Manus Island, where it performed equally
p. Rain sensors
Two STI optical rain gauges (Sheppard and Joe, 1994; Wang et al., 1978) measured rainfall rates
on both sides of the ship. The sensor on the windward side appeared to provide better readings
than the one on the leeward side. Rain rates up to 142 mm hr-1 were recorded. Disdrometers
measured the raindrop size spectrum, and appeared to perform well throughout the cruise.
However, after the cruise, serious concerns about calibration and linearity raised doubts that any
of the rain data can be used quantitatively.
q. Surface meteorological sensors
In addition to the ship's meteorological sensors for surface air temperature, humidity, and
pressure (approximately 33, 12, and 10 m above sea level, respectively), there were several other
independent sets of sensors measuring the same variables. Spot intercomparisons were not
always satisfying. Hand-held standards for temperature and relative humidity were taken to each
sensor and readings were taken simultaneously to help remove biases when data are
r. Heat flux buoy
A small drifting-buoy system (Soumi et al. 1996) was deployed when Discoverer arrived
northeast of Manus Island at the first 100-km station. A floating sensor for measuring surface
heat flux was tethered to it. The buoy regularly transmitted its position and sensor data to the
University of Wisconsin via the ARGOS satellite. This experimental system was retrieved 2
days later en route to the first 30-km station, but it was not deployed again because the sensor
was faulty. However, some reliable data were obtained by the system.
4. Meteorological context
An important post-cruise activity was to review and summarize the meteorological conditions
throughout the cruise, and to compare them with mean conditions for the month. An in-depth
report, summarized below, was produced and distributed to principal investigators (PIs). Mean
conditions were derived from reanalysis of an historical 40-year dataset (Kalnay et al., 1996).
Knowledge of the synoptic and regional meteorology is needed to better interpret and analyze the
Three phenomena provide insight into the time-mean and time-varying large scale atmospheric
circulation observed during the CSP. Two of these, the annual cycle and El Niño/Southern
Oscillation (ENSO), helped determined the time-mean CSP conditions, while the third, Madden-Julian Oscillations (Madden and Julian 1972), dominated the large-scale, time-varying part. As
is typical for the coupled ocean-atmosphere system (Bjerknes et al., 1969; Graham and Barnett,
1987), we experienced a close relationship between the pattern of SST and tropical convection.
the occurrence of large-scale, deep convection increased substantially as the ship moved into
waters having temperatures > 28 C. To a good approximation, the atmospheric circulation could
be understood as a response to the divergent flow associated with this convection (Gill 1980).
The time-mean SST and its departure from a 15-year climatology are shown in
Figs. 2a and 2b,
respectively. Normally during northern spring the center of the west Pacific Ocean warm pool is
in the Southern Hemisphere (Fig. 2a), and equatorial SST
gradients are weak in the region
traversed by the Discoverer. In a typical year convection occasionally reaches eastward to the
dateline, the equatorial cold tongue is weak, and SSTs approach 28C there. This tendency was
disrupted during the CSP by a weak to moderate cold event within the ENSO cycle, which
moved equatorial convection to regions west of 160E. The observed SST anomalies
produced a stronger zonal SST gradient along the equator, with 27C water at the dateline
becoming 29C water at 160E. As a result, strong gradients in convective activity and in the
atmospheric zonal circulation were also located near 160E.
The time-mean atmospheric wind field, decomposed into streamfunction (psi) and velocity potential (chi) fields, is shown in Fig. 3 and 4, respectively. The (psi) and (chi) fields depict the nondivergent (Fig. 3) and divergent (Fig. 4) components of the flow. The two levels shown in each figure were chosen to be near the layers of maximum inflow (surface) and outflow (150 hPa or ~15 km) associated with deep tropical convection. In both figures, the pattern of deep convection is illustrated by shading of outgoing longwave radiation (OLR) values <230 W m-2. OLR is frequently used as a proxy for deep tropical convection (Waliser et al., 1993).
The 150-hPa (psi) field (Fig. 3a)
shows a pattern of stationary waves consisting of twin anticyclones
straddling the equator along 130E and cyclones on either side of the equator centered well east
of the dateline. Along the equator easterly flow between the two anticyclones becomes westerly
flow east of ~165E. The surface (psi) field (Fig. 3b)
shows the general easterly tradewind flow
giving way to weak cyclonic features in the western Pacific beneath the upper-level anticyclones.
Equatorial convection is located in the upper-level easterly flow between the twin anticyclones
and in weak westerly flow at the surface. The Fig. 4
(chi) fields show upper-level divergence and
low-level convergence in the regions of time-mean deep convection (shading), especially over
the western Pacific Ocean. Axes of 150-hPa divergence and surface convergence then extend
eastward in two convergence zones on either side of the equator. These are approximately
coincident with a similar structure in the time-mean convective pattern, which, however, is more
intense in the Southern Hemisphere. Enhanced subsidence occurs along the equator between the
two convergence zones in the region of cooler than normal SST.
The time-mean conditions were largely controlled by ENSO cold-event conditions that spread westward from the east Pacific in spring 1995 reaching the west Pacific during spring 1996. The other large-scale phenomenon controlling convection in and to the east of the warm-pool region is the so-called Madden-Julian Oscillation (MJO; Madden and Julian 1972). The MJO is characterized by large-scale convective flareups that develop over the Indian Ocean and then move east to the Pacific Ocean (Lau and Chan 1985; Knutson and Weickmann 1987). Individual MJOs represent the convective envelope for a wide range of convective activity that evolves and decays on many time and space scales (Nakazawa 1988; Hendon and Liebmann 1994). For daily-mean data, the most prominent of these multiscale features is the so-called supercluster that seems especially well-defined in the spring and fall seasons when the seasonal-mean convective activity occurs near the equator.
A detailed picture of the time evolution of deep convection during the CSP can be seen in the
time-longitude diagram in Fig. 5. Both the total (Fig. 5a) and anomalous (Fig. 5b) OLR depict
the MJO that affected the western Pacific region during the CSP mission (late March to early
April 1996). The MJO consisted of two superclusters both of which developed over the Indian
Ocean and then moved eastward at a speed of ~8 m s-1. The first supercluster reached the area of
Manus Island while the Discoverer was on station and gave the most intense and prolonged
convection experienced during the cruise (at least in equatorial regions). The second supercluster
was somewhat stronger, but its convection set up slightly farther to the west (~140E) as the ship
left Manus Island. However, as a result of this convective development, the atmospheric
circulation in the cold-tongue region during the Discoverer's eastward return journey in early
April had changed compared with the westward track in mid-March, especially in upper levels.
In summary, the atmospheric circulation over the western and central equatorial Pacific during
spring 1996 was somewhat atypical; it had stronger than usual zonal gradients in SST,
convection, and atmospheric zonal winds. These features primarily reflected a weak to moderate
cold event in the ENSO cycle. Although the center of persistent, large-scale convection was
confined to the far western Pacific Ocean region, a Madden-Julian Oscillation and its associated
synoptic variability provided an episode of deep convection somewhat farther east over Manus
Island while the Discoverer was on station.
5. Preliminary results
About 6 months after the mission, on 28-29 October 1996, the CSP science team held a
workshop in Boulder, Colorado, to examine preliminary data. Prior to the workshop, many
datasets were contributed to a web-accessible data management system at ETL, and more
datasets have been contributed subsequently. The team decided to password-protect CSP
datasets until PIs ensured their quality, had the first chance to use them, and published findings.
The data management system makes it convenient to access others' data in a common format
(netCDF) and to use datasets synergistically. These data are available to the entire DOE/ARM
community; soon the public will have access. CSP and Manus Island datasets that have been
contributed to ETL's data management system are listed in Table 2. Several PIs have created
Graphics Interchange Format (GIF) images of their data and made them available via the
worldwide web. Visit http://www4.etl.noaa.gov to learn about both data and image sites. What
follows in this section is a pot pouri of glimpses at CSP data:
Among the immediate payoffs of combining and comparing datasets was a better understanding
of interrelationships between various types of SST measurements.
Figure 6 depicts a 24-h period
at the initial 30-km station northeast of Manus Island, when winds were calm or light. During
daylight hours SST in the upper layer of the sea (0-5 m) increases due to insolation, but relatively
large temperature differences occur between the ship's sensors sampling water at a depth of
approximately 5 m, other sensors at 0.1 m depth, and the M-AERI instrument, which measures
the temperature of the radiating skin, which is nominally 1 mm deep. The skin temperature is
generally a few tenths of a kelvin lower than the temperature of the water immediately below it,
due to evaporative cooling. Surface layer overturning in the absence of insolation at night tends
to equalize all three measurements. Combined measurements such as these help better
understand the radiation budget and air-sea interactions (Webster et al. 1996; Soloviev and
Schlussel 1996; Fairall et al. 1996b).
Under these same calm conditions the Raman lidar detected a pattern of water vapor cells that were tens of meters in diameter (Fig. 7), a pattern never before seen or hypothesized (Cooper et al. 1997). This discovery can be corroborated by statistical intercomparisons with data from other CSP sensors, such as radiometers, flux systems, and the wind profiler, possibly leading to a better understanding of moisture flux at the surface. These fine-scale observations will help us better understand the mechanics by which moisture and latent heat are carried aloft from the sea surface under nearly calm conditions.
Figure 8 depicts the time histories of both integrated water vapor (red) and integrated liquid water in clouds (blue) for the periods of station-keeping near Manus Island, at 100 km (top two panels), 30 km (third and fourth panels from the top), and close to the island (bottom panel). Downward pointing arrow indicate that rain was detected at the surface. These data corroborate other data (e.g., ceilometer data) that indicate it was relatively dry and less cloudy far removed from the island (100 km stations) than at the 30 km or nearby stations. Whether or not this is a general trend, or one induced by transient synoptic meteorology, remains to be determined by further inspection of the meteorological analyses mentioned in the previous section.
Figure 9 depicts the radiance difference between two independently-operated FTIRs, one from NOAA/ETL and the other from the University of Wisconsin. Absolute radiances for both instruments were in the range 40-100 mw m-2 sr-1 cm for the 700-1300 cm-1 band. The r.m.s. difference shown in Fig. 8 is well below 1 mw m-2 sr-1 cm and is very small on a relative basis. This is a typical comparison, and it corroborates the empirical adjustment to an ARM-standard, line-by-line-radiative-transfer model (LBLRTM) made by S.A. Clough, based on ETL FTIR spectra taken during TOGA COARE and CSP (Shaw et al. 1995; Han et al. 1997). Figures 10 shows the radiance differences between the model and measurements (a) before and (b) after the adjustment. The Air Force Geophysical Laboratory (AFGL) subsequently adopted this adjustment for LOWTRAN, MODTRAN, and FASCOD, improving these highly-used algorithms. Accurate, high resolution FTIR measurements such as these can be used to retrieve profiles of temperature and atmospheric constituents in the boundary layer, thereby more thoroughly characterizing it, and to measure SST. Of course, by scanning from nadir to zenith they also provided a very detailed assessment of infrared upwelling and downwelling flux at the surface, to better understand radiative balance. For these devices it is critical to know when clouds are in the field of view, because clouds are strong infrared radiators. Therefore simultaneous measurements by the microwave radiometer, cloud radar, ceilometer, lidar, all-sky camera, and other CSP sensors help considerably in data analysis.
The measured aerosol number size distribution as a function of meteorological regime is shown in Figure 11. Within the South Pacific Convergence Zone (SPCZ) and the Intertropical Convergence Zone (ITCZ) deep tropical cumuliform clouds persist and precipitation occurs frequently; here the Atiken mode (particle diameters between 20 and 70 nm)
dominated the size distribution. Sporadically an ultra-fine mode (diameters < 20 nm) appears, possibly transported by air from the upper troposphere into the boundary layer by subsidence. There was no detectable accumulation mode (diameters between 70 and 500 nm) because of aerosol scavenging by precipitation. In the wedge-shaped region between the SPCZ and ITCZ divergent easterly winds prevailed and precipitation was much less frequent. Here the size distribution is distinctly bimodal, showing both Aitken and accumulation modes. Near Manus Island, three meteorological situations were encountered over the 10-day period. During "Manus
dry" and "Manus wet" the accumulation mode was depleted, again indicative of an air mass
scavenged by precipitation. During "Manus old" the size distribution was bimodal, indicating
less frequent precipitation. Here a larger mean diameter of Aitken particles implies that the
aerosol is well-aged and may have been transported from the region of divergent easterlies.
Inbound to Hawaii we experienced northeast trade winds and size distributions similar to those
found in the divergent easterlies. However, the air mass here appears to have been more
anthropogenically influenced because the concentration of elemental carbon, which results from
combustion, was higher. Also, the concentration of radon, which serves as a tracer for
continentally-derived air, was higher. See Quinn et al. (1996) for common features between this
CSP summary and data from earlier RITS and ACE campaigns in the Pacific.
Comparisons between the ceilometer on board Discoverer, a device with a maximum altitude extent of 3.66 km (12,000 ft), and the ceilometer at the Manus Island CART site, a 7.62-km (25,000 ft) device, were revealing. Table 3 summarizes the cloud fraction and cloud base statistics as functions of platform (ship or island) and ship location (by zone). It also shows correlation statistics between ship- and island-based ceilometers. As expected, correlation decreases with increasing separation, but is still significant at 30 km separation. Figure 12 depicts the detailed track of the ship for the 10 days it was within 100 km of the island, and shows the three zones cited in Table 3 associated with station-keeping.
Figure 13 indicates that the cloud fraction over the island was always higher than over the ship, even if the island ceilometer data were processed only to the same altitude as the ship's data (3.66 km). The island ceilometer detected clouds above 3.66-km altitude that increased in the late afternoon, persisted through the evening, and finally dissipated in the early morning hours. These midlevel clouds may have been the result or remnants of daytime convective activity over mainland Papua New Guinea. Probability distributions for the bases of low clouds (< 3.66 km) over both the ship and island were quite similar, as shown by Fig. 14. Such ceilometer data, when compared to appropriately averaged radiative flux and SST measurements, will help assess the effect of clouds and cloud base on these variables.
Figure 15 shows a 24-h segment of simultaneously-acquired reflectivities from radars at three different frequencies: (a) 915-MHz wind profiler on Discoverer, (b) 3-GHz precipitation radar on Manus Island (Ecklund et al. 1995), and (c) 33-GHz cloud radar on Discoverer. UTC date is 30 March 1996. Data were acquired while Discoverer was at the 30-km station, downwind from Manus Island. Thus precipitation events in the middle panel (Manus Island data) appear slightly earlier than corresponding events in the top and bottom panels (Discoverer data). Relationships between the reflectivities seen in these panels can be used to determine microphysical properties of the clouds and precipitation, such as rain rate, size distributions, effective radii, etc., important quantities for modelers of clouds, precipitation, boundary layer processess, and radiative balance.
The CSP mission brought together and operated a unique combination of shipborne in situ and
remote sensors in the central and western equatorial Pacific ocean for a 30-day period beginning
14 March 1996. The combination of sensors characterized the marine boundary layer, the
radiation budget, air-sea interactions, and clouds more completely than did any previous
campaign. The unique combinations of measurements should lead to better understanding of
fundamental climate-related processes and feedbacks, interferences to satellite data products, and
the representativeness of CART measurements on Manus Island.
The operations plan for CSP was executed nearly flawlessly, and all major goals of the mission
were accomplished. The CSP dataset has already resulted in new insights to fundamental
processes important to the modeling of climate (e.g., see discussion in section 5), and more are
anticipated. The SST measurements, and their interrelationships under a variety of
meteorological conditions, will be used to improve satellite SST retrievals and better understand
and parameterize energy and moisture fluxes across the air-sea interface. Other investigations
remain. For example, enhancements in aerosol backscatter just below cloud base can be
compared with lifting/growth models to verify the models and the assumption of a well-mixed
marine boundary layer. On the eastbound return leg we unexpectedly experienced small regions
of dry downwelling air, where ozone and the concentration of small (new) particles rose
dramatically. These regions may have been on the boundary between different air masses and
may be important in the downward mixing of important species generated at higher altitudes.
The multitude of upwelling and downwelling radiometric measurements, and the accompanying
aerosol and water vapor data, will undoubtedly help improve AVHRR products.
Because we experienced both typical and atypical atmospheric conditions, from clear and dry
with light winds, to disturbed, unstable, and wet during the genesis of a tropical cyclone, we will
be able to better parameterize fluxes, aerosols, and clouds across a broad range of conditions in
climate models. We also have a small sample of data (10 days) to help determine if large land
masses near Manus Island, or the island itself, affect CART measurements.
While the overall campaign was a large success, there were nonetheless some disappointments.
The sun photometer did not perform as well as anticipated, and the lidar systems failed
prematurely. The cloud radar seldom operated with both frequencies simultaneously. Thus the
quantity of data from several of the centerpiece instruments was not as large as hoped, reducing
the number of hours available for multisensor synergism. Also, the instrument complement on
Manus Island was considerably smaller than originally planned, because of delays in deploying
the first ARCS container of standard DOE/ARM instruments.
Plans are for PIs to continue to analyze data and to hold a special session at the 1997 AGU Fall
Meeting to present new findings. A similar mission for 1999 is being considered, to build on
CSP findings and to better compare shipboard measurements with a complete set of ARM/ARCS
instrumentation that will be on Nauru Island in 1999, in contrast to the incomplete set that was
on Manus Island during CSP. The Nauru mission will be more comprehensive than CSP, with
plans to include two ships and research aircraft. Also, Nauru Island is much smaller than Manus
Island, and we expect significantly different island effects.
It is impractical to list as co-authors or to acknowledge everyone who made significant
contributions to the CSP mission. John Herring and many others in NOAA Corps helped plan
and coordinate the mission many months before any equipment was shipped, and Steve
Piotrowicz of NOAA/OAR helped secure ship time for it. The entire ship's crew was
unbelievably supportive during the mission, especially considering that the Discoverer was near
the end of an extended deployment and was on its next to last mission, scheduled to be
decommissioned after nearly 30 years of stalwart service to NOAA and our Nation. The crew
had to deal with a plethora of odd requests from many "rookie" atmospheric scientists on board.
Mae Chu of InterConex and her agents David McNeil (Samoa) and Dick Pearse (Manus Island)
provided indispensable aid in shipping and handling a great deal of sophisticated equipment
throughout the mission. Mark Fiscus, Chuck Long and Paul Johnston helped coordinate
activities on Manus Island with those of the ship's, and hosted a short but very helpful tour of
island facilities. Kurt Nielson was injured while setting up the AVHRR receiver in Samao, but
still managed to get it operating before the ship departed. Dave Covert and Tim Bates helped
bring the ACE aerosol instrumentation back online, and Gene Furness helped install rain gauges
and disdrometers. Jim Churnside, Joe Shaw, Scott Abbott, and Jesse Leach helped install much
of ETL's equipment in Samoa. Bill Porch provided S-band radar data and ceilometer data,
respectively, from Manus Island. John Bates and Wesley Berg acquired GMS data and produced
both images and video from it. Joan Hart was instrumental in providing analyses for the
meteorological summary. David Welsh, Tom Glaess, Sandy King, and LingLing Zhang
developed many of the capabilities for data management. Of course, all the CSP participants in
Table 1 who are not listed as co-authors deserve special acknowledgment for their strong
contributions, especially Duane Hazen who assisted the chief scientist greatly. Finally, we must
thank Tom Ackerman and the rest of the ARM Science Team Executive Committee for having
the vision to promote and support the ASTEX and CSP missions under contract DE-AI02-92ER61366.
CSP participants, affiliations, and instruments. Name Affiliation Instrument M.J. Post NOAA/ETL Chief Scientist Chris Fairall NOAA/ETL Flux system, SST sensors, wind profiler, ceilometer Yong Han NOAA/ETL FTIR Jeff Hare NOAA/ETL Flux system, radiosondes, solar radiation sensors Duane Hazen NOAA/ETL Radiometers (microwave) Jack Snider NOAA/ETL Radiometers (microwave) Dan Cooper DOE/LANL Lidars (elastic and Raman) Bill Eichinger DOE/LANL Lidars (elastic and Raman) Larry Tellier DOE/LANL Lidars (elastic and Raman) Mike Osborn DOE/LANL Lidars (elastic and Raman) Scott Smith DOE/BNL Radiometer package (visible & IR) Derek Coffman NOAA/PMEL Aerosol samplers, satellite receiver Vladimir Kapustin NOAA/PMEL Aerosol samplers, satellite receiver Patricia Quinn NOAA/PMEL Aerosol samplers, ocean chemistry sensors Steve Sekelsky U. Massachusetts Cloud radar Lihua Li U. Massachusetts Cloud radar Tom Saxen Colorado St. Univ. Rain gauges, disdrometer Jorge Valero Penn. St. Univ. Sun photometer Peter Minnett Univ. Miami All-sky camera, SST sensors Bob Knuteson Univ. Wisconsin M-AERI Nick Nalli Univ. Wisconsin M-AERI Brian Osborne Univ. Wisconsin M-AERI John Short Univ. Wisconsin M-AERI
CSP and Manus Island Data Sets
(in ETL Data Management System)
CSP Manus Island
Aerosol Size Distributions Pyranometer Flux Aethalometer Absorption 915 MHz Winds (Moments) w/RASS Anemometer Relative Wind (1 min avg) 915 MHz Winds (Spectra) w/RASS Anemometer Relative Wind (30 min avg) Rotating Shadowband Radiometer Anemometer Wind (30 min avg) Surface Meteorology Rawindsonde (CLASS) Pyrgeometer Ceilometer Ceilometer Fluxes, Meteorological FTIR Spectra MAERI Spectra Nephelometer, submicrometer Nephelometer, total Rain Rates, Optical (CSU/NASA) Meteorology, PMEL Rain Rates, Optical (PMEL) Radon Pyranometer Flux (BNL) Pyranometer Flux (PMEL) Pyranometer Flux (ETL) 915 MHz Winds (Moments) 915 MHz Winds (Spectra) Water Vapor and Liquid (microwave radiometer) Rotating Shadowband Radiometer Sea Water Chemistry Ship's Meteorology and Ocean Data Ship's navigation Data Observers' Meteorological Notes
Table 3. Statistics from data collected by a ceilometer on board Discoverer
in zones 1-3 (Fig. 10) and by a ceilometer at the Manus Island CART site.
Ncf: Number of data points in cloud fraction correlation.
rcf: Cloud fraction correlation coefficient.
Nch: Number of data points in cloud base height correlation (restricted to cloud fractions > 0.1)
for hourly data and cloud fractions > 0.5 for 10-min data).
rch: Cloudbase height correlation coefficient.
Albrecht, B. A., D. A. Randall, and S. Nicholls, 1988: Observations of marine stratocumulus
clouds during FIRE. Bull. Amer. Meteor. Soc., 69, 618-626.
Albrecht, B. A., C. S. Bretherton, D. Johnson, W. H. Schubert, and A. S. Frisch, 1995: The
Atlantic Stratocumulus Transition Experiment - ASTEX. Bull. Amer. Meteor. Soc., 76, 889-904.
Bjerknes, J., L.J. Allison, E.R. Kteins, F.A. Godshall, and G. Warnecke, 1969: Satellite mapping
of the Pacific tropical cloudiness. Bull. Amer. Meteor. Soc., 50, 313-322.
Bradley, E.F. and R.A. Weller (Editors), 1997: Fourth Workshop pf the TOGA COARE Air-Sea
Interaction (FLUX) Working Group. Report, TCIPO, UCAR, 9-11 October 1996, Woods Hole
Oceanographic Institute, Massachusetts, USA, 61 pp. Available from TOGA COARE
International Project Office, UCAR, P.O. Box 3000, Boulder, CO 80307-3000, USA.
Browning, K.A., 1994: Survey of perceived priority issues in the parameterizations of cloud-related processes in GCMs. Quart. J. Roy. Meteor. Soc., 129, 483-487.
Cess, R. D., M. H. Zhang, P. Minnis, L. Corsetti, E. G. Dutton, B. W. Forgan, D. P. Garber, W.
L. Gates, J. J. Hack, E. F. Harrison, X. Jing, J. T. Kiehl, C. N. Long, J.-J. Morecrette, G. L.
Potter, V. Ramanathan, B. Subasilar, C. H. Whitlock, D. F. Young, and Y. Zhou, 1995:
Absorption of solar radiation by clouds: Observations versus models. Science, 267, 496-499.
Chertock, B., C.W. Fairall, and A.B. White, 1993: Surface-based measurements and satellite
retrievals of broken cloud properties in the Equatorial Pacific. J. Geophys. Res., 98, 18,489-18,500.
Chou, M.-D., A. Arking, J. Otterman, and W.L. Ridgeway, 1995: The effect of clouds on
atmospheric absorption of solar radiation. Geophys. Res. Lett., 22, 1885-1888.
Cooper, D.I., W. Eichinger, R. Ecke, J. Kao, J. Reisner, and L.Tellier, 1997: Initial investigations
of microscale cellular convection in an equatorial marine atmospheric boundary layer revealed by
lidar. Geophys. Res. Lett., 24, 45-48.
Ecklund, W.L., P.E. Johnston, J.M. Warnock, W.L. Clark, and K.S. Gage, 1995: An S-band
profiler for tropical precipitation studies, Preprint Vol., 27th AMS Conf. on Radar Meteorology,
Vail, Colo., 335-336.
Evans, W.F.J., C. Reinhart, and E. Purckrin, 1995: A ground-based measurement of the
anomalous cloud absorption effect. Geophys. Res. Lett., 22, 2135-2138.
Fairall, C.W., E.F. Bradley, D.P. Rogers, J.B. Edson, and G.S Young, 1996a: Bulk
parameterization of air-sea fluxes in TOGA COARE. J. Geophys. Res., 101, 3747-3767.
Fairall, C.W., E.F. Bradley, J.S. Godfrey, G.A. Wick, J.B. Edson, and G.S. Young, 1996b: Cool
skin and warm layer effects on the sea surface temperature. J. Geophys. Res., 101, 1295-1308.
Fairall, C.W., A.B. White, J.B. Edson, and J.E. Hare, 1997: Integrated shipboard measurements
of the marine boundary layer. J. Atmos. Oceanic Tech., 14, 338-359.
Frisch, A.S., C.W. Fairall, and J.B. Snider, 1995: On the measurement of stratus cloud and
drizzle parameters with a K-band Doppler radar and a microwave radiometer. J. Atmos. Sci.,
Gill, A.E., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy.
Meteor. Soc., 106, 447-462.
Graham, N.E. and T.P. Barnett, 1987: Sea surface temperature, surface wind divergence, and
convection over tropical oceans. Science, 238, 657-659.
Han, Y., J.A. Shaw, J.H. Churnside, P.D. Brown, and S.A. Clough, 1997: Infrared spectral
radiance measurements in the tropical Pacific, J. Geophys. Res., 102 (D4), 4353-4356.
Hendon, H.H., and B. Liebmann, 1994: Organization of convection within the Madden-Julian
Oscillation. J. Geophys. Res., 99, D4, 8073-8083.
Hogg., D.C., F.O. Guiraud, J.B. Snider, M.T. Decker, and E.R. Westwater, 1983: A steerable
dual-channel microwave radiometer for measurement of water vapor and liquid in the
troposphere. J. Climate Appl. Meteor., 22, 789-806.
Jacobson, M.D. and W.N. Nunnelee, 1997: Design and performance of a spinning flat reflector
for millimeter-wave radiometry. IEEE Trans. Geoscience Remote Sensing, 35, 464-468.
Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G.
White, J. Woolen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K.C. Mo, C.
Ropelewski, J. Wang, A. Leetmaa, R. reynolds, R. Jenne, and D. Joseph, 1996: The
NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc.,77, 437-471.
Knuteson, R.O., F.A. Best, P.Minnett, H.E. Revercomb, and W.L. Smith, 1997: High spectral
resolution infrared observations in the tropical western Pacific using a marine atmospheric
emitted radiance interferometer (MAERI). Proc., Ninth Conf. on Atmospheric Radiation, Long
Beach, CA, Amer. Meteor. Soc., 180-183.
Knutson, T.R., and K.M. Weickmann, 1987: 30-60 day atmospheric oscillations: composite life cycles of convection and circulation anomalies. Mon. Wea. Rev., 115, 1407-1436.
Lau, K,-M., and P.H. Chan, 1985: Aspects of the 40-50 day oscillation during northern winter as
inferred from outgoing longwave radiation. Mon. Wea. Rev., 113, 1889-1909.
Madden, R.A., and P.R. Julian, 1972: Description of global-scale circulation cells in the tropics
with a 40-50 day period. J. Atmos. Sci., 29, 1109-1123.
Nakazawa, T., 1988: Tropical super clusters within intraseasonal variations over the western
Pacific. J. Meteor. Soc. Japan, 66, 823-839.
Quinn, P.K., V.N. Kapustin, T.S. Bates, and D.S. Covert: 1996: Chemical and optical properties
of marine boundary layer aerosol particles of the mid-Pacific in relation to sources and
meteorological transport. J. Geophys. Res., 101, 6931-6951.
Ramanathan, V., B. Subasilar, G. J. Zhang, W. Conant, R. D. Cess, J. T. Kiehl, H. Grassl, and L.
Shi, 1995: Warm pool heat budget and shortwave cloud forcing: A missing physics? Science,
Randall, D. A., J. A. Coakley, C. W. Fairall, R. A. Kropfli, and D. H. Lenschow, 1984: Outlook
for research on subtropical marine stratiform clouds. Bull. Amer. Meteor. Soc., 64, 1290-1301.
Reynolds, R.M. and S. Smith, 1997: The R/V Discoverer cruise to Manus Island -- The BNL
Portable Radiation Package (PRP) Evaluation, Report #64333, Brookhaven National Laboratory,
Upton, NY 11973, 45 pp.
Sekelsky, S.M. and R.E. McIntosh, 1996: Cloud observations with a polarimetric 33-GHz and
95-GHz radar. J. Meteor. and Atmos. Phys., 58, 123-140.
Shaw, J.A., J.B. Snider, J.H. Churnside, and M.D. Jacobson, 1995: Comparison of infrared
atmospheric brightness temperatures measured by a Fourier transform spectrometer and a filter
radiometer. J. Atmos. Ocean. Tech., 12, 1124-1128.
Sheppard, B.E and Joe, P.I, 1994: Comparison of raindrop size distribution measurements by a
Joss-Waldvogel disdrometer, a PMS 2DG spectrometer, and a POSS Doppler radar, J. Atmos.
and Oceanic Tech, 11, 874-887.
Smith, W. L., R. O. Knuteson, H. E. Revercomb, W. Feltz, H. B. Howell, W. P. Menzel, N. R.
Nalli, O. Brown, J. Brown, P. Minnett, and W. McKeown, 1996: Observations of the infrared
radiative properties of the ocean: Implications for the measurement of sea surface temperature
via satellite remote sensing. Bull. Amer. Meteor. Soc., 77, 41-51.
Soloviev, A. v., and P. Schlussel, 1996: Evolution of cool skin and direct air-sea gas transfer coefficient
during daytime. Bound.-Layer Meteor., 77, 45-68.
Soumi, V.E., L.A. Sromovsky, and J.R. Anderson, 1996: Measuring ocean-atmosphere heat flux
with a few in-situ sensors. Proc., 8th Conf. on Air-Sea Interaction, Atlanta, GA, Amer. Meteor.
Stokes, G.M. and S.E. Schwartz, 1994: The Atmospheric Radiation Measurment (ARM)
program: Programmatic background and design of cloud and radiation testbed. Bull. Amer.
Meteor. Soc., 75, 1201-1221.
Waliser, D.E., N.E. Graham, and C. Gautier, 1993: Comparison of the highly reflective cloud and
outgoing longwave radiation datasets for use in estimating tropical deep convection. J. Climate,
Wang, T.I., K.B. Earnshaw, and R.S. Lawrence, 1978: Simplified optical path-averaged rain
gauge. Appl. Opt., 17, 384-390.
Webster, P. J., and R. Lukas, 1992: TOGA COARE: The coupled ocean atmosphere response
experiment. Bull. Amer. Meteor. Soc., 73, 1377-1416.
Webster, P. J., C. A. Clayson, and J. A. Curry, 1996: Clouds, radiation, and the diurnal cycle of sea
surface temperature in the tropical western Pacific. J. Clim., 9, 1712-1730.
White, A. B., C. W. Fairall, and J. B. Snider, 1995: Surface-based remote sensing of marine
boundary-layer cloud properties. J. Atmos. Sci., 52, 2827-2838.
Young, G.S., D.R. Ledvina, and C.W. Fairall, 1992: Influence of precipitating convection on the
surface energy budget observed during a Tropical Ocean Global Atmosphere pilot cruise in the
tropical western Pacific Ocean. J. Geophys. Res. 97, 9595-9603.
Young, G.S., S.M. Perugini, and C.W. Fairall, 1995: Convective wakes in the equatorial Pacific
during TOGA COARE. Mon. Wea. Rev., 123, 110-123.
Zhang, G. J., V. Ramanathan, and M. J. McPhaden, 1995: Convective-evaporative feedback in
the equatorial Pacific. J. Climate, 8, 3040-3051.
Zhang, G. J., and R. L. Grossman, 1996: Surface evaporation during the Central Equatorial Pacific Experiment: A climate scale perspective. J. Climate, in press.