Opportunistic Analysis of Acoustic Data from EX1305
NOAA Ship Okeanos Explorer

hurried draft version - Friday 5th March 2021
John E. Hughes Clarke
Professor, CCOM, UNH
 Indra Prasetyawan
PhD student CCOM

  This web page has been generated to illustrate some interesting aspects of archived acoustic data collected as part of an old (2013) NMFS mission on board the Ocean Exploration vessel Okeanos Explorer. All the data were downloaded from NCEI. At this time, the intended audience is just Shannon and Meme from OER.

Interesting Examples are presented of combined EK60- 18kHz and EM302 water column imaging.

A - Huge Internal Wave Packet propagating from the Shelfbreak onto the Bank.

This sample was taken from EX1305 transit ....   On a tidal period, Internal Solitary Waves (ISWs) are generated at the shelf break and propagate onto the bank. This example is an extreme case with a leading wave of ~ 60m height in only ~ 80m of water.

location map
map illustrating the section investigated (white rectangle)

section
combined plot of :

This packet had a major influence on the multibeam bottom tracking due to pronounced refraction distortion driven by the depressed and tilted velocline. The zoomed-in area shown below focuses on the first 5 undulations of the ISW and illustrates the specific impact on the bottom tracking.

Note that the conventional vertical sections do not give any indication of the orientation of the undulations. Only by cutting a horizontal section through the EM302 pie (bottom figure) can one infer their full 3D structure. The first undulation on the left, which initially looked like a broader feature from the 2D section, is actually just a more oblique section. It actually generates the sharpest bathymetric artifact that ends  up being about at 45 degrees to the shiptrack. The relationship between wave azimuth and projected seafloor orientation is described in this USHC2017 paper


zoom ISW probs
of all those interesting apparent morphologic features on the seabed - nothing is actually real.

On the left hand side, the SSP was just replaced and one can see the gross refraction artefact was removed (vessel steaming from left to right). This would be fine if the watermass was still represented by that last SSP which measures the thermocline BEFORE the arrival of the ISW packet. 

The rapid outer beam fluctuations vary between +/-2 and up to +/-4m at the outer edge of the swath (65 deg). In only 80m of water this is up to +/-5% of depth, well in excess of any typical bathymetric survey accuracy requirement. But a cool oceanographic phenomena.

A - Shear between the in-flowing(?) Scotian Slope water and the overlying Scotian Shelf water.

This sample was taken from the EX1305 transit through North East Channel which defines the eastern end of Georges Bank. It is a glacial scoured trough with a depth of ~ 280m. It is often described as a passage way for deep water exchange from the open ocean into the Gulf of Maine. As far as I know (?) the ADCP was not running so we can't prove whether there really was shear. Nevertheless, the water column imaging makes a pretty compelling case.

location map
map illustrating the section investigated (white rectangle)

section
combined plot of :
The section illustrated above was chosen as there are two oceanographic phenomena - the near surface internal wave packet and the mid water Kelvin Helmholtz wave packet.

There are three EM302 configuration ("mode") changes involved here:

The EM302 is switching back and forth  from Shallow to Medium on the floor of the NorthEast Channel and it is very apparent that, for the purposes of water column imaging - the Medium mode is to be preferred. This would contrast with the bathymetric resolution requirement which would favour the shorter pulse in the shallower water. As the vessel runs off the shelf break, it switched to the longer pulse.


Detailed Analysis of the KH wave packet.

There are three ways we can look at that short window:



plan view

As can be seen, you now are looking at an almost geographically correct plan view of the wave patterns (the waves are also moving while the ship steams over so there is still a little distortion). From that you can tell their wavelength and azimuth. Note that while generally orthogonal to the shiptrack (as they were steaming sub parallel to the orientation of the NE Channel), the azimuth and wavelength do vary in the different regions.

Classic Pie Diagrams

This is the only display that is standardly visible in real time for most multibeams.  A still frame tends to be noisy and of little use. By scanning through a series of sequential frames though, faint patterns, lost in the speckle, become apparent and the relative translations are now visible.

Ultimately the visibility depends on the relative intensity of the echoes from the coherent structure (due to the displaced scatterers, whether zooplankton or micro-turbulence we don't honestly know) relative to the sparsely distributed backgrounds scatterers. One way to suppress uncorrelated ping to ping background noise is to do averaging. This is an along track average and therefore represents a slight spatial filtering. But, remembering that this EM302 is operating in dual swath mode and the swaths are only optimally spaced at the seabed, the closer you are to the transducer, the more the swath pair are sampling the same volume. Thus paired averaging is really no loss of resolution. The right hand image shows the same data with pairs averaged.


each swath
 pair averaged
individual swaths without along track averaging
 two swath averaging.

Offset Vertical Sections

This presentation mode allows one to look at the across track variations. If the wave crest is truly orthogonal to the ship track, the only change you will see is a resolution issue. But if the azimuth is slightly oblique one gets a sense of the wave "rolling". Often extra structure becomes visible.
Note that we are  seeing sections both through the KH waves at ~120m depth, as well as the near surface internal waves on the thermocline at ~ 20m depth.

offset vertical


One technical aspect is worthy of note. There is an optional gain that one can apply to the logged 8 bit water column data (the "x_log_offset" term). Back in 2013, this was by default set to zero. Nowadays this is default set to 30. The result is that without an offset, the lowest scattering strength logged is -64 dB (about half way into weak scatterers). With 30 dB offset, this drops to -94 dB and you can see those weaker echoes. In the 2013 case, it was set to zero so the centre sectors are thresholding out and not seeing the weaker echoes. Note for example that the above thermocline scattering is  are almost invisible in the centre two sectors. This is not recoverable from the archived data.



page developed by JEHC, Jan-Feb 2021