dc.description | This dataset consists of measurements and data products derived from a shipboard stereo video system, two styles of drifting wave buoys (SWIFT and Waverider), and the standard measurements made from the R/V Thomas G. Thompson. All data are stored as .mat files to be read into Matlab. The data from each platform are described below:
STEREO VIDEO
The stereo video data is found in two compressed directories: StereoXYZData_FiveMin and StereoXYZData_FullCapture. Each directory is broken down by date and then further by time (in UTC). These correspond to the start time of the stereo video segment. The duration of the full videos ranged from 20 to 60 minutes. In the StereoXYZData_FullCapture directory, the 10 best full video segments are processed. In StereoXYZData_FiveMin, the first five minutes from 52 video segments of intermediate or better data were processed.
Within each video, the main data is found in the subdirectory XYZResults. The data are stored as one-minute bursts. Files named ZGrid_01.mat and XYGrid_01.mat correspond to the first minute, ZGrid_02.mat and XYGrid_02.mat to the second, etc.
The XYGrid files give the x (xgrid, in meters) and y (ygrid, in meters) points of the gridded stereo data. For all the data given here, x and y range from -20 to +20 meters, with a resolution of 0.25 meters (size 161x161). The +x coordinate is the downwind direction.
The ZGrid files give the three-dimensional gridded elevation (zgrid, in meters) and radiance (imgrid, arbitrary units) data products. When the camera frame rate was set at 5 Hz (prior to January 1), the resulting arrays are size 161x161x300. When the frame rate was 7.5 Hz, the arrays are 161x161x450.
The other subdirectories give plots and information related to the processing of the raw stereo data into the gridded data products. DisparityPlots shows the rectified stereo images and their disparity image. XYZPlots shows the resulting elevation and radiance products. IGMVectors gives the time series of ship motion from the Novatel SPAN-IGM system, which were used to project the data into real-world coordinates. And StereoOptions.mat saves the options used in Matlab’s disparity function to process the stereo imagery.
SWIFT
The surface wave instrument float with tracking (SWIFT) buoy is described in Thomson (2012). All SWIFT data are found in a single file, SWIFTdata_TGT2015.mat. This file contains a single Matlab structure array (1x2522), named smallSWIFT. Each member of the struct array is derived from a burst of 512 seconds. The fields of the struct array are described below:
smallSWIFT.uplooking.tkedissipationrate: vertical profiles of turbulent dissipation rate in W/kg (= m^2 / s^3)
smallSWIFT.uplooking.z: depth bins, in meters, for the tke dissipation rate profiles. wave-following reference frame SWIFT.downlooking.velocityprofile: vertical profiles of horizontal velocity magnitude, in m/s, relative to the float (not corrected for drift) smallSWIFT.downlooking.z: depth bins, in meters, for the velocity profiles
smallSWIFT.winddirT: true wind direction, in degrees FROM North
smallSWIFT.winddirTstddev: standard deviation of true wind direction, in degrees
smallSWIFT.windspd: wind speed, in m/s, at 1 m height above the wave-following surface
smallSWIFT.windspdstddev: standard deviation, in m/s, of wind speed
smallSWIFT.time: UTC timestamp in MATLAB datenum format (serial days since 0 Jan 0000)
smallSWIFT.date: human readable date as day, month, year
smallSWIFT.airtemp: air temperature, in deg C, at 1 m height above the wave-following surface
smallSWIFT.airtempstddev: standard deviation of air temperature, in deg C
smallSWIFT.sigwaveheight: significant wave height, in meters
smallSWIFT.peakwaveperiod: peak of period orbital velocity spectra (note convention is usually wave height spectrum)
smallSWIFT.peakwavedirT: wind direction, in degrees FROM North
smallSWIFT.wavespectra.energy: wave energy spectral density, in m^2/Hz, as a function of frequency
smallSWIFT.wavespectra.freq: spectral frequencies, in Hz
smallSWIFT.wavespectra.a1: normalized spectral directional moments
smallSWIFT.wavespectra.b1: normalized spectral directional moment
smallSWIFT.wavespectra.a2: normalized spectral directional moment
smallSWIFT.wavespectra.b2: normalized spectral directional moment
smallSWIFT.lat: latitude in decimal degrees
smallSWIFT.lon: longitude in decimal degrees
smallSWIFT.watertemp: water temperature, in deg C, at 0.5 m below the surface
smallSWIFT.salinity: water salinity, in PSU, at 0.5 m below the surface
smallSWIFT.puck: three color channels of a WetLabs puck flourometer
smallSWIFT.driftdirT: drift direction TOWARDS, in degrees True (equivalent to "course over ground”)
smallSWIFT.driftspd: drift speed in m/s (equivalent to "speed over ground")
WAVERIDER
Wave measurements were also made from drifting Datawell DWR-G4 Waverider buoys. These data were processed using Datawell’s w@ves21 software. The results are given in the file WaveriderSpectra_TGT2015.mat. This file contains all arrays of dimension mxnx263, corresponding to the 263 waverider points, each corresponding to 30 minutes of measurements. The variables are described below:
buoyNum: Unique identifier for each buoy (in this case, either 3 or 4)
f: Frequency vector, for use with Sf, in Hz
fMat: Frequency array, for use with Sftheta, in Hz
kurtDir: Kurtosis of the directional spectra (dimensionless)
lat: Average latitude, in decimal degrees
lon: Average longitude, in decimal degrees
meanDir: Spectral mean direction, in degrees
Sf: Wave power spectral density, in m^2/Hz
Sftheta: Wave directional spectral density, in m^2/(Hz*degrees)
skewDir: Skewness of the directional spectra (dimensionless)
spreadDir: Spectral mean directional spread, in degrees
thetaMat: Direction array, for use with Sftheta, in degrees (clockwise from true North)
time: UTC time at end of 30-minute measurement, in Matlab datenum format
R/V THOMPSON
Meteorological measurements were made from the R/V Thompson’s standard met package of instruments. The resulting variables are:
airtemp: Air temperature, in degrees C
airtemptime: UTC time vector corresponding to airtemp, in Matlab datenum format
cog: Ship course-over-ground, in degrees, clockwise from true North
heading: Ship heading, in degrees, clockwise from true North
headingtime: UTC time vector corresponding heading, in Matlab datenum format
lat: Ship latitude, in decimal degrees
lattime: UTC time vector corresponding to lat and lon, in Matlab datenum format
lon: Ship longitude, in decimal degrees
sog: Ship speed-over-ground, in m/s
sogtime: UTC time vector corresponding to sog and cog, in Matlab datenum format
sst: Water temperature, in degrees C
ssttime: UTC time vector corresponding to sst, in Matlab datenum format
winddir: True wind direction, in degrees, clockwise from true North
winddirtime: UTC time vector corresponding to winddir, in Matlab datenum format
windspd: True wind speed, in knots (measured from roughly 20 meter height)
windspdtime: UTC time vector corresponding to windspd, in Matlab datenum format | en_US |
dc.description.abstract | These data accompany the article “Sharp-crested Breaking Surface Waves Observed from a Ship-Based Stereo Video System,” by Schwendeman and Thomson in the Journal of Physical Oceanography (DOI: http://dx.doi.org/10.1175/JPO-D-16-0187.1). The data were collected during a cruise in December 2014 and January 2015 in the North Pacific Ocean onboard the R/V Thomas G. Thompson. Measurements were made from a shipboard stereo video system, two styles of drifting wave buoys (SWIFT and Waverider), as well as the standard R/V Thompson measurement systems. Ship, SWIFT, and Waverider data are provided as standalone files (“ShipData_TGT2015.mat”, “SWIFTdata_TGT2015.mat”, and “WaveriderSpectra_TGT2015.mat”). Stereo video data is provided as gridded sea surface elevation and surface radiance data products. The raw image frames are not provided, but are available upon request from the authors. The gridded data products are stored in one-minute segments, and are contained in two directories. One (“StereoXYZData_FiveMin”) contains five minutes of data from 52 vidoes of moderate-to-high quality. The other (“StereoXYZData_FullCapture”), contains all data (20-60 minutes) from only the highest quality videos. Finally, two directories are provided to replicate the analysis from Schwendeman and Thomson (2017). The analysis codes (written in Matlab) are provided in “StereoAnalysisCodes_JPO.” When run correctly, these codes produce the results and figures found in “ResultsAndPlot_JPO.” | en_US |