Title: R/V Ka'imimoana Sonic Anemometer and CO2 (Picarro) Data

Author: Byron Blomquist

PIs: Byron Blomquist
University of Hawaii Oceanography
1000 Pope Rd, Honolulu, HI 96822
blomquis@hawaii.edu
808 956 5185

Chris Fairall
NOAA / ESRL / PSD3
325 Broadway
Boulder, CO 80305
Chris.Fairall@noaa.gov

1.0 Data Set Overview:

Data files are 10Hz CO2, U, V, W, Relative Wind Direction and Air Temperature (from the
sonic anemometer).  Wind speeds are corrected for ship motion contamination [Edson et al.,
1996].  A further decorrelation with respect to vertical acceleration and velocity was
applied to W, removing residual traces of motion contamination.  In addition, and N2 pulse
variable is included to facilitate synchronization of the CO2 and wind data.  The N2 pulse
square wave is recorded on the wind time base.  The corresponding pulse observed in the
CO2 signal is used to determine the time lag between wind and CO2 measurements.  For TORERO
the mean lag time was 15.2 +/- 0.07 sec.  [see Blomquist et al., 2013]

Data covers the period from 1/26/12 00:00 UTC to 2/28/12 22:00 UTC for the R/V
Ka'imimoana cruise track from Hawaii to Costa Rica.  Ship position may be determined from 
GPS data in the SAMOS underway dataset.

2.0 Instrument Description:

Wind and motion sensors and the CO2 inlet were on a 10m met tower on the bow of the ship.
The anemometer is a Gill R2 sonic, with motion corrections derived from 3-axis motion and
angular rate measurements from a Systron-Donner MotionPak.

The CO2 analyzer is a Picarro G1301 cavity ring-down spectrometer, modified as a CO2-only
instrument with upgraded pressure control valves, to minimize motion interference.  All
instruments ran at 10 Hz.  The CO2 sample stream was dried to a dew point of <-10C, so
CO2 mixing ratios are on a dry-air basis. [see Blomquist et al., 2013].

3.0 Data Collection and Processing:

Data were collected continuously for the duration of the cruise.

Wind speed is computed directly from the analog calibration of the Gill R2 sonic.  Sonic
temperature is computed from speed-of-sound, with corrections for the cross-wind
components (in MATLAB: sonT = (SpdSnd.*SpdSnd + (u.*u + W.*W)/2. + v.*v)/403-273.15;)
where SpdSnd is the speed of sound in m/s.  No corrections for humidity effects.

Plots of spectra and cospectra for raw and motion-corrected winds are presented in the
QuickTime movie file "RV_Kaimimoana_Wind_Spectra.mov".  These may be used to examine the
effectiveness of the motion corrections and the effects of flow obstruction contamination
of the fast wind data.  Note that in general, relative wind directions of +/- 120 degrees
from the bow are usually good, although for flux calculations it may be desirable to 
restrict the sector to +/- 60 deg.  The effects of severe flow distortion may be seen in
spectra plots when the relative wind direction is > +/-150 deg.

Note that in this processed dataset, a double angle rotation is applied to U, V and W to
bring U into the streamline - thus hourly mean V and W are zero.  U and V should not be
interpreted in geophysical coordinates without a proper rotation which takes into account
relative wind direction and heading.

For computing fluxes, CO2 and W data should be synchronized by analyzing the N2 pulse
response.  Fluxes for periods when relative wind direction is out of sector should be
discarded.  For other flux processing methods and selection criteria see Blomquist et
al., 2013.

4.0 Data Format:

Data files are tab-delimited ASCII with one header line of variable names.  File names are
of the type: CO2_2012_ddd_hhmm_merge.txt, where 2012 is the year, ddd is the day-of-year,
and hhmm is UTC.

One QuickTime movie file of quality control plots for wind motion corrections is also
included (RV_Kaimimoana_Wind_Spectra.mov).

List of parameters with units:
Timestamps		Date/Time string
Time_GPS_sec	Time in seconds from midnight 1/1/2012
CO2ppm			CO2 mixing ratio (dry) in ppm.
U				stream-wise wind speed (m/s)
V				cross-stream wind speed (m/s)
W				vertical wind speed (m/s)
RelWdir			Relative wind direction w/ respect to the ship's bow (deg)
N2_pulse		5 V pulse which drives an N2 injection into the CO2 inlet
SonicT			Sonic temperature (C)

This is data version G, Dec 2012.

5.0 Data Remarks:

Wind data have been corrected for motion effects.  CO2 data have been checked and are
precise to +/-5%.

6.0 References:

Edson, J. B., A. A. Hinton, K. E. Prada, J. E. Hare, and C. W. Fairall (1998),
Direct covariance flux estimates from mobile platforms at sea,
J. Atmos. Oceanic Technol., 15, 547–562.

Blomquist, B.W., L. Bariteau, J.W. Edson, C.W. Fairall, J.E. Hare, W.R. McGillis,
B.J. Huebert, S.D. Miller, E.S. Saltzman (2013), Advances in ship-based air-sea CO2 flux
measurement by eddy covariance, in preparation for JGR.