For the web
ACES- Atlantic Coral Ecosystem Study
Andrew Wheeler, Maxim Kozachenko & Gerry Sutton
Introduction
University College Cork (Coastal Resources Centre) has successfully carried out
the acquisition of 100kHz and 410kHz side-scan sonar surveys in the Rockall
Trough and Porcupine Seabight. The acquisition of these images was technically
challenging involving the deployment of items of equipment at the limit of their
operational effectiveness. Not only were we successful in this respect but we
also managed to survey larger areas than anticipated due to the smoothness of
the operation. As a result, these images represent the best at present) views of
these deep-water habitat collected using side-scan sonar. Furthermore, as
predicted, these images allow us to observe these deep-water coral bioherms at
a spatial resolution appropriate for the evaluation morphological features and
environmental controls. These images are also at the appropriate spatial scale to
allow the integration of regional side-scan and multibeam mapping with video
and sample data. They therefore provide fundamental data in their own right as
well as facilitating further scientific investigations.
The surveys were performed on the Darwin Mounds, northern Rockall Trough
(100 & 410kHz coverage) and the Belgica Mounds (410kHz coverage) (see
Figure 1). Images from this survey are viewable on the AGU poster (Wheeler et
al., 2000) <LINK>
Figure 1. Location of side-scan survey
Methods
All data was acquired onboard cruise RRS Discovery 248 using a dual frequency
GeoAcoustic side-scan sonar system depth rated to 2000m. Swathe width was
600m at 100kHz and 200m at 410kHz. A Hytech cable counter was used to
assist in layback calculation. A CODA DA100 sonar processing system supplied
a real-time image recorded to DAT tapes. A hardcopy output of the side-scan
record was also obtained directly from the GeoAcoustics transceiver using an
Ultraelectronics Wideline 200 series 12” thermographic recorder with manual
fixes.
To obtain optimum results the side-scan sonar towfish was flown at 25m off the
seabed at 100kHz and 10m off the seabed at 410kHz at 3.5 knots. This usually
required between 2000 and 3700m of cable layout in 1000m of water.
Side-scan sonar data was processed using the PRISM software packages
developed at the Southampton Oceanography Centre (Le Bas & Huhnerbach,
1999). Data was originally sub-sampled to ensure comparable across-track and
long-track resolution and remove any data redundancy. Data was processed by
initially removing the water column. The remaining seafloor backscatter was then
process to convert the return period to across track distance. Images were then
navigated and mosaiced based on the ship-position and layback. Layback
calculation take into account wire-out, wire catenation effects and inertia
movement around corners. Initial parameters were fine-tuned to provide the
optimum mosaic. Rubber-sheeting was not performed.
Darwin Mounds
The Darwin mounds measure approximately 70m in diameter and 5m in height.
All of the deep-water coral mounds possessed discrete boundaries with a rubbly
surface texture. The rubbly texture probably reflects small accumulations of coral,
coral debris and other organisms with positive relief acting as nuclei for further
growth. Live coral is represented by spots of very high backscatter and tends to
be slightly more dominant at the edges of the mound. Internal structuring of the
mound surface is irregular with a slight concentric arrangement of coral colonies.
Distinctions can be made between undisturbed mounds (with well developed
coral patches) and other mounds that have become covered by influxing
sediment with no obvious coral stands. These may have been impacted by deepwater trawling activities.
Tails to mounds were visible as higher backscatter suggesting a change in
sediment composition or surface roughness that may be caused by a number of
processes. Some of the tails also have a positive relief and always occurred in
the lee of the mound. When this was the case, the mounds were often scoured
on the opposing side of the mound. Background sediment around the mounds
was usually featureless although some mounds occurred in or at the edge of
sediment wave fields. Sediment within the mounds is sander than the outlying
sediments. It is speculated that these sandy patches may be original (possibly
related to water escape structures imaged in deep water) offering a viable
substrate for coral colonisation (Masson et al., 2001)
There was abundant evidence of benthic trawling over mounds and in the
intervening seabed. Some areas were 100% trawled. Areas of “speckly”
backscatter on the TOBI imagery in the east of the area were probably caused by
a combination of geological and biological processes. The high backscatter
patches may represent coral and/or patches of more reflective seabed (e.g.
gravel). In some areas, 50% of the seafloor was covered with these high
backscatter patches. Inter-patch arrangement is random although in some
places there was a tendency towards a current-parallel fabric. Trawling this area
is also intense.
In the south of the area, pockmarks were imaged. These appeared as shallow
depressions which sometimes have smaller, positive conical features (?sand
volcanoes) at their centre and/or rim. Linear sediment waves were also imaged
mainly in the north of the area of varying wavelengths and crest heights.
Belgica mounds
Side-scan imagery of mounds reveals that they are largely covered in sediment
waves. The topography of the mounds (often up to 100m high) provides an
obstacle to basal currents causing acceleration in current speed and facilitating
sediment transport in the form of migratory waves. A grading of the sediment
wave styles exists in relation to mound proximity and position up the flank of the
mounds. At 410kHz frequency, the side-scan was able to identify mounds with
coral accumulations. These appear as several metre-size accumulations of high
backscatter. Accumulations occur on all parts of the sediment wave although a
tendency for preferential growth on crests and upper flanks is possible. Further
away from the foot of the mound, coral accumulation density decreases and they
occur on the wave crests and upper flanks and not in the troughs.
On the flanks of the easternmost ridges, current velocities were sufficient to
produce barchan dunes forms feeding off the proceeding dune tails. These were
often associated with inferred gravel sediment waves.
Inter-ridge areas are characterised by a homogenous seafloor with isolated
sediment wave fields. On their peripheries of some of these sediment wave fields
and in isolated areas between the carbonate mounds, small coral mounds 50m in
diameter infrequently occur (named Moira Mounds).
References
Le Bas, T. & Huhnerbach, V. (1991). P.R.I.S.M. Processing of Remotely-sensed
Imagery for Seafloor Mapping Operators Manual Version 3.1. Southampton
Oceanography Centre, UK.
Masson, D.G., Bett, B.J., Billett, D.S.M. & Wheeler, A.J. (2001) The Darwin
mounds - possible fluid escape features in the northern Rockall Trough.
Abstract, European Union of Geosciences EUGXI, Strasbourg, 8-12 April
2001.
Wheeler, A.J., Bett, B.J., Billett, D.S.M., Masson, D.G. & Scientic Party, Officers
and Crew of Discovery 248 (2000). Very high resolution side-scan mapping of
deep-water coral mounds: surface morphology and processes affecting growth.
Abstract: AGU Fall Meeting 2000, San Francisco, EOS, 81(48), F638.