Ba k a e -A aly e
HN104 — 06/2016
15
The modelling proved quite di cult given the
poor quality of the real time backscatter measure
merits A much more successful approach, and
one which is currently being investigated at IFRE
MER is to use a BS model obtained from another
calibrated echo sounder (single beam or split
beam) mechanically steered at several transmis
sion angles in order to obtain a more realistic BS
angular response curve. Nevertheless, the calibra
tion survey and subsequent analysis allowed for a
substantial improvement in the precision of the
mean BSvalue of the BSmosaics
Cbmpared to other multibeam echo sounders
previously investigated by the author, the EM
installed on OOGS»Amundsen« shows very strong
transmission sector patterns modulations Several
attempts were made to nd the causes of these
modulations During this investigation, it was dis
covered that some of the transmit channelsof the
EM are in adegraded state. However, the most
plausible explanation is the fact that the EM ’s
transducers are encased behind a titanium en
forced polymer protective window which is likely
to a ect the transmitted source level and sector
patternssigni cantly more than for normal vessels
For operations in ice, this window is however es
sential to the protection of the transducers
Cbnclusion
This project has focused on the application of a
unique calibration procedure dedicated to the
compensation of transmission sector patterns of
multi sector, multi swath multibeam echo sound
ers Uncalibrated transmission sector patterns
modulate the backscatter reqconse and are visible
as intra line artefacts The calibration procedure
was applied on backscatter data collected by 003S
»Amundsen«. The procedure included the success
ful conduct of a calibration survey in the Canadian
Arctic and subsequent analysis of the backscatter
data usng the software SbnarSCope. The precision
of the mean BSvalue of virtually all Transmission
Modesisimproved in the to degreesangular
range. Ftesidual sector boundary steps remained
problematic and visible in the calibrated images,
and overall, the calibration could be improved
with better estimation of the parameters of the
BS model. Indeed, due to the multibeam echo
sounder’s protective window deteriorating the raw
backscatter measurements, the modelling of the
BSfrom empirical data proved di cult.
Several investigations were carried out in order to
identify the cause of a remaining inter sector bias
This lead to the discovery of tranamisaon channel
problems However, the cause of the bias was most
likely due to ahardware software incompatibility. The
fact that this problem did not lead to any discernible
artefact on the bathymetry data highlights the fact
that backscatter measurementsare more senativeto
changesin system dependent parameters
Multibeam backscatter o ersseveral applications
in seabed dassi cation, habitat mapping and sea
bed monitoring. These applications can potentially
be very bene dal in fundamental sciencessuch as
geology and benthic biology and ecology. Sden
tists wishing to mate use of multibeam backscatter
must however be fully aware of the limitations and
pit fallsof multibeam echo sounder measurements
Most notably, the status of the instruments should
be well understood, the context under which the
datawascollected and the processing applied prior
to delivery should be well controlled."
Fig. 7: Angular compensated
BSmosaic following
the transmission sector
pattern calibration (a left)
and following a statistical
compensation to remove the
remaining OdBInter line bias
(b, right)
Acknowledgements
The author would like
to add ress thanks to the
following Individuals:
Jean Marie Augustin
(IFFBVIB^ for hlsguldance
and for providing a loan of
SonarSbope for t h Is p roj ect ;
Rof. Harald Sernberg (HCU)
for his supervision;
Rof. Rrtrlck Lajeunesse
(Université Laval) for providing
the opportunity to participate
on an expedition onboard
(DOGS »Amundsen«;
Gabriel Jeyal (Université
Laval) for assisting In the
data col lection.
