"Underway" versus "On Station" velocity measurements
 in the R/V Revelle HDSS 50 kHz Doppler Sonar system
during P16S

These plots explore the difference in measured velocity in earth coordinates, obtained while the ship was moving versus stationary. Comparing underway and on-station data is a good way to check for problems with Doppler velocity data, since the boundary between these data collection conditions should be invisible in the final ocean velocities. For this particular collection of figures, all measured velocities had a 150m-250m reference layer subtracted: therefore navigation (GGA messages or calculated ship speed) are not present in any calculations.

These data were all collected during a southbound CLIVAR cruise, which stopped every 1/2 degree between 17S and 71S. "On station" periods were typically for 3-4 hours (long enought for a full-depth CTD cast); "underway periods" took 2-3 hours as the ship steamed south to the next station. The weather was calm and scattering was low during the first week. The next few weeks saw an increase scatterers, and a subsequent increase in wind and sea state. After we passed 60S, the weather improved, but data penetration was poor.

In an earlier calculation, we averaged a 520m-620m layer (of similarly dereferenced velocities) and looked at the underway-on-station velocity difference.  We observed "there is a 4 cm/s difference , averaged over the cruise, in the meridional component of this vertically-integrated shear". The plots below are an attempt to address the question of whether the on-of station shear difference we saw was (1) from the effect of steaming over scattering layers or (2) a systematic shear bias of some kind and if so, how it was distributed in the vertical and in time.

Comparisons were made between pairs of 40-minute segments, determined by
  1. the first 40 and second 40 minutes on station
  2. underway-onstation transition:
  3. the first 40 and second 40 minutes underway

It seems that for the on-off station shear question, the effect we saw was in part due to the "S" induced in a profile by steaming over scattering layers, because of the visual correlation between regions of high scattering and changes in the vertical structure of the underway-onstation velocity difference.   However, the profiles show a variable but persistent negative underway-onstation velocity difference below about 300m for much of the valid vertical extent of the data.  These velocities were referenced to a 150m-250m velocity, so a nonzero underway-onstation velocity differences away from the reference layer do indicate a shear relative to the reference layer.  Because the ship was heading south, the form of the "S" is to have a negative (southward) lobe above the scattering layer, and a positive lobe below the scattering layer.  It is possible that the "S" shapes are sufficiently asymmetric to account for the remaining nagative difference observed below. 

data context
 stopped
- lower stddev (gray)
- pink and blue lines both on station
 on-station versus underway
- middle stddev (gray)
- red lines are STOPPED
-green lines are UNDERWAY
 underway
- highest stddev (gray)
- pink and blue lines both underway
section "a"

decimal
day 10-15
section "b"

decimal
day 15-20
section "c"

decimal
day 20-25
section "d"

decimal
day 26-32
(bad weather, underway data trashed)
section "e"

decimal
day ~38-42


For more detail, the underway-onstation velocity differences (red and green, above) are plotted as vertical profiles with dereferenced signal/noise (to see scattering layers and the dirunal cycle).  Links below brin gup these plots:


data context underway-onstation velocity differences
color-coded, as a function of time 		underway-onstation velocity differences
waterfall plot, over dereferenced "amp"
section "a"

decimal
day 10-15

 color-coded waterfall
section "b"

decimal
day 15-20 		color-coded waterfall
section "c"

decimal
day 20-25 		color-coded waterfall
section "d"

decimal
day 26-32
(bad weather, underway data trashed)
 color-coded waterfall
section "e"

decimal
day ~38-42

 color-coded waterfall


Jules Hummon
April 10, 2005