Border Backscatter Calibration Site

US-Canada Border Backscatter Calibration Site
EM712 Reference Data Collection
CSL Heron - April 10th 2022

John E. Hughes Clarke, Miguel Candido, Brandon Maingot
Center for Coastal and Ocean Mapping
University of New Hampshire

version - 13th April 2022

Overview

There is a growing need to provide a means of absolutely calibrating the seabed backscatter strength measurements collected by multibeam sonars operating within US (and Canadian, Portuguese and Brazilian) waters. In the US, there is a new impetus, driven by the 2019 Presidential Memorandum on Characterizing the US EEZ.

This preliminary web pages outlines details of a new absolute backscatter calibration site, ideal for an EM712, that lies along the US/Canada border in protected waters which would be a suitable test location for any US or Canadian vessel to compare against on an opportunity basis.

This builds on a 2019 experiment using calibrated split beam sonars, that took place from the CSL Heron within shallower adjacent Canadian waters off Sidney BC. The aim is to transfer that calibration to a deeper water location along the border. The short term drive is to allow the NOAA Ship Fairweather (FA), which should be transiting past in May, to inherit that calibration for her newly installed EM712.

Showing location of 2019 EK-80 calibration work
and location of 2022 transfer of that calibration

The CSL Heron backscatter calibration experiment builds on a long-standing collaborative relationship between four partners:

University of New Brunswick
University of New Hampshire
Canadian Hydrographic Service, (DFO)
Office of Coast Survey (NOAA)

Site Location

To avoid administrative overhead, the site was chosen to lie exactly along the US-Canada border in the eastern end of the Strait of Juan de Fuca, between Victoria and Port Angeles. The site was chosen to deliberately avoid the heavy commercial vessel traffic that is directed through specific shipping lanes in the area.

location map showing section along US-Canada Border outside shipping lanes
Red rectangle was the original reconnaissance section
Black rectangle is the focus calibration site.

To establish the best site, the full SW-NE section of the border was run as a reconnaissance. 30-50m deep shoal areas exist at either end (perhaps suitable for EM2040 calibration in the future?). To be most amenable to EM712 calibration (in which the shallowest mode is routinely used to 100m) the deepest, lowest relief and most homogeneous backscatter section was selected.

full 9300m long border reconnaissance transect (SW to left, NE to right) showing:
- EM712 water column
- EM712 backscatter
- EM712 bathymetry
- chart identifying legal border

From the full 9 km section a ~3.5 km long sub-section (20 minutes at 6 knots) was selected. The morphology in the area is relatively smooth (varying only 10m within the nearly 2 nm transect). And the backscatter strength is relative uniform and, most importantly perhaps, anomalous high (averaging ~ -5dB) and without any discernible specular component. This suggests that the seabed consists of washed cobbles and gravels which respond in a typically Lambertian manner. Such seafloor sediments are common in the tidally swept area here.

To illustrate the homogeneity, a representative ~600m wide bathymetric grid and backscatter swath through the reference calibration area is presented below:

chosen section - 3450m long - acquired by CSL Heron (EM712, 70-100 kHz, Very Shallow Mode)
bathymetry - colour scale from 102m to 112m
backscatter - grey scale from -12.5 dB to +(!)2.5 dB
(top , before sector-balancing , middle, after)

Previous Work

In 2019, Ivan Guimaraes collected EK-80 data using a calibrated set of split beam transducers (ES-70, ES-120, ES-200 and ES333) which, by chirping each through their full bandwidth provides backscatter measurements over the range 45 to 450 kHz.. This is used to derive the backscatter strength "truth" (see his MSc thesis, UNH 2020) for details. This was collected over 5 sites in the Sidney area in depths ranging from 15 to 55m.

Showing 2019 experimental method - 4 chirped EKs covering 45-450 kHz
suspended below CSL Heron, rotated in elevation and azimuth.
Resulting "truth" at 5 locations

At the time, data from the multibeams then installed on the CSL Heron including EM710 (SIS-4 .all) and EM2040P (SIS-5 .kmall) were collected to try and also calibrate them. That EM data together with a new 2022 dataset including the new EM712 (SIS-5 .kmall) and EM2040C (SIS-5 .kmall) was just collected by Miguel Candido (MSc in progress, UNH).

Miguel is currently using the archived "truth" to attempt to absolutely calibrate the CSL Heron EM712 (a 1x 2 degree system) backscatter strength observations. As a result, in theory (....) the Heron now represents a "golden" calibrated system. Thus it can now move to different areas and collect a reference dataset, against which other EM712s can compare.

This is the intent of this program in which a new 40-100 kHz backscatter reference site is being developed. It would be administratively complicated for the FA to steam directly into Canadian waters to occupy the sites off Sidney (and additional complications include the fact that it is an area with restricted maneuvering, relatively heavy ferry and recreational boating traffic). As an alternate approach, on April 10th, the CSL Heron EM712 system was deployed to the US-Canada border line in an area both outside the shipping lanes and away from coastal restrictions.

The site chosen (illustrated in the chart extracts above) is directly in the transit path of the FA as it relocates from Newport/Seattle to Ketchikan, Alaska (assuming it uses the inside passage). All data collection can be performed within US waters (simply by steaming 1/2 a boat's width to the SE of the border).

CSL Heron EM712 Data Collected

To provide a reference backscatter data set using the typical pulse lengths and frequency ranges of a standard EM712, the CSL Heron collected data at ~6 knots along the chosen border transect on April the 10th 2022. The following permutations were collected:

70-100 kHz

Very Shallow Mode
Shallow Mode
Medium Mode

"40" kHz (40-70 kHz)

Very Shallow Mode
Shallow Mode
Medium Mode

Additionally EM2040C data were collected using 300 kHz for all modes from : Shallow/Medium/Deep/VeryDeep/ExtraDeep/ExtremeDeep.

A preliminary relative (inter-sector, inter-swath, inter-pulse-length, inter-frequency) backscatter calibration was performed. Typically +/-5 dB of correction needs to be applied to remove the gross inter-sector offsets (up to 10 dB for the "40" kHz mode outer sectors). The plot below illustrates the sector-specific corrections estimated and applied.

Showing sector-specific sonar-relative beam pattern corrections for the CSL Heron EM712

To demonstrate the effectiveness of the method, the plot below illustrates that, when these relative corrections were applied to the CSL Heron EM712 data collected in the 6 modes, the inter-sector, inter-swath artefacts were very effectively removed.

Showing the impact of applying the relative sector-specific calibration to CSL Heron EM712 data
acquired using 6 unique frequency/pulse length combinations.
- top - corrections
- centre - before application
- bottom - after application

Note that the example here is only a RELATIVE calibration. Through the magic of Miguel's new algorithms, and working with Ivan's reference data, the idea is to do an ABSOLUTE calibration (the "Holy Grail").

Survey Line Recommendations for FA

Assuming that the FA is indeed passing through the Inside Passage on its way to Alaska, this site is conveniently away from all major shipping lanes. The coordinates for the 3450 m long reference base line are provided (the one in which the reference data was acquired on 10th April) and plotted on the chart below:

48.31657591 -123.21753805 SW end of Line
48.34261491 -123.19237257 NE end of Line

location map of 3450 m long transect on US-Canada border
(backdrop is NOAA Chart 18465 (depths in FATHOMS)

I believe... the border separates the responsibility for vessel traffic reporting. On the Canadian side, vessel reporting is handled by CCG Victoria Traffic. On the US side, it is handled by USGS Puget Sound Vessel Traffic.

Ideally, after a quick CTD, the FA would collect a pair of reciprocal lines along this transect for each operational mode of the EM712. Including turns, this translates to ~ 1 hour of shiptime per mode (2 x 20 minutes lines at 6 knots and probably about 10 minutes per line to turn around and do a run-in.).

Which Modes?.

At a minimum, I would expect the 70-100kHz modes to be tested in dual swath (standard OCS/CHS hydrographic survey settings). This would include at least the continental shelf modes (Very Shallow, Shallow and Medium) and, given NMFS interest in the upper slope, perhaps also Deep and Very Deep (Extra Deep implies you need a lower F sonar). And if single swath is being used routinely, then perhaps those permutations. Ideally, all data should be collected at +/-70 degrees.

For my (JHC) interest, I would like to see the "40" kHz modes as the centre frequencies are almost an octave lower. For hydrography, however, this compromises the beamwidths (~ 2x fatter). But for some seafloors this allows improved characterization. Note, however, that as currently configured, the EM712 cannot hop between the two frequencies ranges, so a second survey is currently required.

page generated, JEHC, Apr 13th 2022