Surveying

technical

Multibeam echosounders:
trends in the hydrographic industry
by John Fraser and Rich Lear, RESON

Decreasing quantity of experienced multibeam users juxtaposed with the increasing demand for qualitative and
quantitative multi-layered (i.e. bathymetry and backscatter information of seabed topography and seabed material)
multibeam sonar data results in growing demand for a comprehensive and efficient system for acoustic sonar
mapping. This article explores, through discussions with multibeam sonar industry experts, the increasing demands
for acquisition of "more/faster/better" sonar data through status quo or limited hands-on operator experience. Future
generation multibeam sonars must be geared for high efficiency; a mantra for robust common-goal solutions.

uge leaps in technology have

occurred since the first primitive
forays into depth sounding
but technology also brings with it
challenges. The dramatic increase
in data quantities necessitate large
investment in software and hardware
and of course highly skilled personnel
not only to operate but to support it.
The question has to be posed: Is
todays technology genuinely increasing
a surveyor's data gathering capability
in less time, using less equipment?
The investment in capabilities that
translate directly into efficiency gains
which of course in turn bring cost
savings to operators. In this article, we
will look at some of these features and
what benefits they bring.

Fig. 1: In shallow water, object detection is normally paramount because of issues such as ship
safety in harbours. In this case a combination of ultra-high resolution and coverage are critical.

History
Through hydrographic survey history
from the lead line; single beam; sweep,
to ultimately multibeam. The hardware
systems have always required skilled
operators, that have been guided
by education and experience. Since
the theory has remained, how has
hardware actually evolved?
With the introduction of the now
familiar PC interface, and reduction of
hand plotted charts, concurrently came
multibeam systems available for all
users not just deep water military.
Certain sonar manufacturers in the
late 1980s pioneered the shallow water
multibeam industry with the release of
products such as the RESON SeaBat
9001. This system truly revolutionised
this market and made multibeam
accessible to a multitude of operators
who, until then, had used single beams
and sidescan almost exclusively in

PositionIT October 2010

Fig. 2: The plot shows theoretical equi-angle (blue) and equi-distant (green) spacing on a flat
sea floor.

shallow water. The success of such

systems may be judged by the fact that
they were in continuous production for
over 17 years and many are still in use
worldwide today.
Hydrographers then demanded a
greater range of tools with which to
collect high resolution bathymetry.
Greater range; greater coverage;
higher resolution and faster ping
rate were all perquisites for previous
generation multibeam systems. Still
today some attributes can not be met

by all systems in the market place,

however yet again next generation
systems are raising the standards to be
judged from. Significant development
has taken place specifically on features
and capabilities to maximise the
considerable investment required
to own and operate one of these
cutting-edge systems.

Standards and shallow water

The hydrographic community wish
to work to recognisable standards.

19

SURVEYING

technical
which to the typical surveyor mean
very little and provide little practical
benefit.
However, will other requirements
change the equation? As a case study
let us assume a survey was conducted
in the following area:
l

Fig. 3: An extract from data collected in both equi-angle and equi-distant mode near Santa
Barbara harbour in January 2007.
l

50 m depth, 1000 x 1000 m area,

10% overlap client specification,
5 kts survey speed,
3 hits in a "1 m box" client
specification
System ping rate within 10% of
theoretical maximum.

In reality for example using RESON

systems
l

RESON SeaBat 7125 2,7 hits per

box need to reduce speed to 4
kts

RESON SeaBat 7101 1,9 hits per

box need to reduce speed to 2 kts

RESON SeaBat 7125 1hr 30 min

(narrower swath less time)

RESON SeaBat 7101 1hr 40 min

(wider swath more time)

So how are systems more efficient?

"Hands off" online

Fig. 4: Roll stabilisation allows the swath width on the sea floor to be maximised thus
increasing line spacing and reducing vessel time. The usable swath is that portion between
the red lines.

The publication, Standards for

dynamic as systems change. In

Hydrographic Surveys (S-44), is one

shallow water, object detection is

of the series of standards developed

normally paramount because of issues

by the International Hydrographic

such as ship safety in harbours. In

Organisation (IHO) to help improve

this case a combination of ultra-high

the safety of navigation. From

resolution and coverage are critical. A

the 1st Edition of S-44 entitled

system providing the combination of

Accuracy Standards Recommended

a swath width in excess of

for Hydrographic Surveys published

4 x depth (a swath path of some

in January 1968 to the current 5th

100 m in 25 m water depth) and the

Edition Standards for Hydrographic

ability to robustly detect objects down

Surveys changes have occurred

to approximately 0,5 m in

in operational capability. As a

25 m and down to less than the size

result, the latest standard is not a

of a soccer ball in 5 m water depth

specification, as that responsibility

may be judged as favourable and

is given to individual hydrographic

offer greatest user benefit.

offices/organisations to prepare
specifications based on the

Survey efficiency

standards.

Multibeam features have been

It is therefore important for those

local persons defining standards, to
ensure that the latest equipment
capability be considered for existing
or indeed new hydrographic purposes.

20

a priority amongst certain

manufacturers, and they have
put considerable effort into
developing real-world features to
aid the surveyor. For a long time
manufacturers have been extolling

Specifications can be more system

the virtues of their systems based

specific and as such will be quite

on detailed technical specifications

In the past, some sonar systems have

required a great deal of operator
intervention, in some cases needing
a dedicated operator to monitor the
system throughout all phases of
operation. This obviously does not
make sense and so modern systems
developed in the last few years often
now employ an autopilot type
capability.
This feature means the operator
can spend the majority of their
time performing critical tasks such
as data quality monitoring and less
time adjusting the sonar system. The
autopilot must be very reliable and
robust even under difficult conditions.
Autopilot should assist the operator
by automatically setting certain
sonar parameters from a pre-defined
look-up table. Quite often multiple
tables may be stored and loaded
as required. In use, the autopilot
system analyses bottom detect data
from a number of pings. If a sufficient
number of those pings are judged
to have good bottom detect quality,
then the average bottom depth and
maximum usable swath angle are
therefore determined.
Once the depth and maximum swath
angle are determined, an appropriate
range setting is determined and
commanded to the sonar. The process

PositionIT October 2010

SURVEYING

Operating on 100 m range scale

Ping rae 5Hz

technical

Operating on 50 m range scale

Ping rate 10 Hz

Fig. 5a and b: The operator may reduce the swath coverage sector from maximum to a predefined minimum angle using set increments.
Often two modes high density and reduced beams may exist.

then repeats for the next number of

pings and so on. If too few of these
pings are judged to be good (for
example, if the water depth exceeds
the current range), then autopilot
will command the next greater
range setting to the sonar before
analysing the next set of pings. In
this way autopilot mode can be reset
to the minimum range and it will
incrementally try greater ranges until
the bottom is found.
In addition to controlling the range,
autopilot mode also automatically
adjusts sonar parameters such as
power, gain, and pulse length. A user
editable table is made up of rows,
each corresponding to a minimum and
maximum range and including a set of
parameters. Depending on which row
the current range setting falls within,
the parameters for that row will be
commanded to the sonar. In this way,
critical sonar settings as well as range
are all automatically controlled when
in autopilot mode. Typical variables
include: range against power, pulse
length, gain, spreading and absorption
variables.

Beamforming
A majority of current multibeam
systems allow equi-distant (ED)
beamformed footprints which
maximise productivity, although the
number of beams vary according to
the specific system geometry.
Traditional (and some contemporary)
multibeam systems operate only in
equi-angle (EA) mode. As its name
suggests, in this mode the distance
between the centreline of each beam
is an equal angle. When viewed from
above it becomes obvious that the

PositionIT October 2010

Fig. 6: The sonar head may be physically tilted (typically up to 30) to one side

to illuminate the desired area, a quay wall in this case, and the user interface is
adjusted to match.

density of the beam centres on the

sea floor decreases towards the edge
of the swath.
The plot in Fig. 2 shows theoretical
equi-angle (blue) and equi-distant
(green) spacing on a flat sea floor.
It may be seen that in equi-angle
mode (blue), sounding spacing is
most dense at nadir and decreases
towards the edge of the swath
whereas in equi-distant mode (green)
nadir density is maintained across the
entire swath.
This has a critical impact on survey
operations especially where object
detection is a mandatory requirement.
The plot in Fig. 3 is an extract from
data collected in both equi-angle and
equi-distant mode near Santa Barbara
harbour in January 2007.

In this example, equi-angle nadir

spacing (red data) is approximately
6 cm (in 9 m water depth)
However, at the edges of the swath,
equi-angle spacing has increased
to approximately 28 cm while the
equi-distant (blue dataset) remains at
6 cm across the entire swath
(see Fig. 3).
In a survey operation where a
requirement exists for a specific
number of soundings on an object, it
may be seen that equi-distant spacing
allows the entire swath to be used
whereas equi-angle spacing quickly
exceeds the specification.

Comparison of equi-angle and

equi-distant footprints
The availability of equi-distant
beamforming has increased

21

SURVEYING

technical
This means that the line spacing has
to be reduced in order to maintain
100% coverage and avoid data gaps.
The amount of overlap will depend
on the amount of roll experienced
on a particular day and so it is
difficult to predict or plan line spacing
accordingly.

Fig. 7: Installing a MBES on a ROV, ensuring alignment with Inertial heading sensor.

productivity dramatically especially in

survey operations where a number of
soundings per grid cell is specified.
The increased density in the outer
beams maintains a high number of
soundings across the swath allowing
wider line spacing, less overlap, fewer
survey lines, higher productivity,
higher efficiency, and of course cost
savings.
As an example, let us assume a
RESON SeaBat 7125 multibeam
system in 20 m water depth with
a requirement to meet three
soundings across and three soundings
along-track to detect a 1 m "cube".
The along-track soundings are
determined purely by water depth and
vessel speed so remain constant for
both modes.
256 Equi-angle mode: In EA mode,
256 beams are formed at a spacing
of 0,5 This translates to a spacing
on the sea floor between soundings
of approximately 18 cm which
means that at nadir (directly under
the transducer) the 1 m cube will
be defined by some six soundings
across-track and would meet the
detection criteria. If one now
performs the same analysis across
the swath, it becomes apparent that
it is only possible to place three
hits on the cube until one reaches a
swath angle of approximately 45.
This means that out of the available
128 swath from the 7125 only 90 is
usable. 128 swath provides a swept

22

path of 4x water depth (80 m in 20 m

depth) but of this 80 m width only
50% is usable.
512 Equi-distant mode: In ED mode,
512 beams are formed at a variable
spacing to maintain the nadir footprint
across the entire swath. This means
that the 18 cm footprint and the
ability to place six soundings on the
1 m cube is maintained across the
entire swath meaning that a usable
swept path of 80 m is generated. This
translates to a huge efficiency gain,
allowing line spacing to be reduced
by 50% and completing the survey
is around half the time it would have
taken with EA footprints.

Roll stabilisation
Roll stabilisation allows the swath
width on the sea floor to be
maximised thus increasing line spacing
and reducing vessel time. Real-time
input from a motion sensor is used to
steer each receive beam dynamically
for every sample thus maintaining the
swath vertical irrespective of vessel
motion. Under zero roll conditions,
the swath is vertical and the centre of
the swath is directly below the vessel.
When the vessel rolls, the swath is
rotated and the projected swath on the
sea floor is laterally displaced.
This has the effect of reducing
the usable swath width due to the
distortions on the edge. In the
example in Fig. 4 the usable swath is
that portion between the red lines.

With roll stabilisation active the

swath stays vertical irrespective
of vessel motion thus maximising
swath. Roll stabilisation requires the
input from a motion sensor and the
operator is able to configure sensor
type and communications protocol.
A motion monitoring display is
provided in order to quality assure
the data being collected. Needless to
say, implementing roll stabilisation
requires extremely tight control over
the system timing, not only of the
multibeam but also the motion sensor.

Variable swath
A variable swath capability allows the
swath width to be varied according
to the environment, or operational
requirements, for example pipeline
inspection.
The operator may reduce the swath
coverage sector from maximum to a
predefined minimum angle using set
increments. Often two modes high
density and reduced beams may
exist.
In high density mode, all beams are
formed and are compressed into the
reduced sector providing increased
sounding density. In reduced beams
mode only the natural number of
beams are formed to populate the
selected sector. It is important to note
that such a feature does not affect
the beam width which is a physical
characteristic of the receiver array
and operating frequency.
Reducing the sector while maintaining
the same number of beams has the
effect of decreasing the spacing
between beams thus increasing
sounding density. Variable swath
should be available in all beam
modes, equi-angle or equi-distant and
may also be used in conjunction with
roll stabilisation.
The benefits of this feature are
twofold:
l

Increased sounding density in

a small sector may be used
when tracking pipes or other

PositionIT October 2010

SURVEYING

technical

Figs. 8a, b, c: Sounding chart; sounding chart with surface bathymetric shading; and aerial photograph with bathymetry.

linear features where the user

is not interested in data outside
a defined sector. The increased
sounding density helps to define
small targets.
l

A reduced sector allows an

increase in ping rate for any given
depth. This has immediate effects
when one thinks of requirements
for a specific number of soundings
across and along-track. It means
that in any given water depth
and vessel speed, the ability to
meet these criteria is increased
significantly by reducing the swath
angle.

Data quality
An automated indicator of data quality
for each sounding will assist the
operator/data processor in manually
determining the amount of resurvey
or processing required. Online there
are generally two different methods
of flagging data, one is a rather
coarse flag which defines whether a
particular sounding has passed the
two main criteria in the bottom detect
assessment process brightness and
colinearity. This method is very robust
and when processing data statistically
one sees that if the sonar is set
correctly, only around 0,5% of the
soundings do not pass both criteria.

Total propagated error (TPE)

The calculation of total propagated
error (TPE) is used to aid the surveyor
in determination of sounding quality
by statistical methods and is typically
calculated in the data acquisition
software. One of the inputs into
the TPE calculation is the sounding
measurement error as computed in
the sonar system. Other error sources
which are considered are position,
draft, squat, load, tide, geodetic
calculations, lever arm calculations,
timing offsets, speed, heading,
attitude, sound velocity and others.

PositionIT October 2010

The uncertainty value of each

sounding is reported as part of the
output record and allows the surveyor
to continuously monitor the overall
system performance during all phases
of operations. For a direct display, the
main user display may be coloured
to display the uncertainty value. The
data is also generally available in
quality data acquisition software for
use during processing.

Changing orientation
It is becoming more and more
common to fuse different datasets
such as bathymetry with laser or
photography to further enhance the
final data products. The ability to
robustly define vertical structures is
a key feature in all good premium
multibeam systems and one that is
used frequently.
In this type of application the
sonar head may be physically tilted
(typically up to 30) to one side to
illuminate the desired area for
example quay wall in Fig. 6 and the
user interface is adjusted to match. A
sonar feature may then be employed
to enable the use of modes such as
equi-angle or equi-distant and indeed
roll stabilisation.
Advanced multibeam systems may
also employ a pipeline mode which
maximises the concentration of
beams in the centre section and also
provides undisturbed sea bed towards
the outer edges of the sector, again
maximising efficiency in these specific
applications.

Improved receivers
The old adage of poor quality in, poor
quality out remains true in this as in
most fields in our world. The purpose
of a receiver sub-system to detect the
minutest signal in the water and pick
this out from a background cacophony
of vessel noise (own vessel and
other traffic); wave noise and other

sources. Improving receiver design

is a major thrust and one which will
help to improve the all-important data
quality resulting in the end in more
accurate soundings.

Portability and upgradability

Moving away from hardware and
processing wizardry, traditionally,
there have been two different
configurations of shallow water MBES
systems; one in which as much
electronics as possible are located
in the sonar head which is located
underwater and the second where
the electronics in the transducers are
minimised and some sort of junction
box used to house electronics. Both
these approaches have advantages
and disadvantages.
The highly integrated "surface heavy"
approach reduces shipping weight and
cost, installation and integration time,
operator workload, vessel size and
power requirements and provides a
one-man operation suitable for rapid
deployment on small vessels as well
as a tidy installation each time.
The rapid improvements in computer
hardware and processing power
makes it obvious to place those items
where they are easily upgradable
that is in the topside processor.
Having the heart of a system in the
wet end where severe constraints
are placed on the designers simply
does not make sense as upgrades will
be expensive, time consuming and
will more than likely require major
engineering effort. Being able to
simply replace a plug-in card in a PC
provides longevity and a future-proof
system. Less serviceable parts in the
water arguably make a system more
reliable with less potential down time due
to failures.

Installations
As a general statement, most portable
installations be the temporary or semi

23

SURVEYING

technical

permanent are now installed under a

best practice criteria. The practice of
using fabric tapes and broomsticks to
measure offsets is thankfully redundant.
With the high resolution MBES system
combined with RTK positioning, there
is no doubt professional procedures
for installation all involve land survey
techniques using a total station or
similar measuring device for both angles
and distances. This is a test for a true
certified hydrographic surveyor.

Deliverable product
With hardware capable of producing
larger datasets, imaging complex
structures, end clients also need
to consider what is a suitable and
appropriate data deliverable. Will a
hardcopy paper chart suffice, or will an
interactive dataset to be viewed in stereo
using a freely distributed data viewing
software package downloaded off the
internet be preferable?

24

How will the data be used? Such a

question falls beyond the scope of this
paper, however when users consider
hardware procurement, they must bear
in mind how they are going to use the
acquired data.

Conclusion
Today, industry users and clients of
multibeam echosounder manufacturers
are presented with an array of systems
each promising unique capabilities. It
is important for users not to lose focus
on what features can actually be used
by them for the benefit of operations.
It is also critical that the system
fundamentals such as bottom detection
algorithms, beam forming ability etc.
are not over looked. Clients should also
consider reliability and service from the
respective manufacturers.
However, users who procure systems
must also consider not only survey
standard compliance, but also maritime
safety hardware compliance standards

such as CE certification and also

IEC60945 (Maritime Navigation and
Radiocommunication Equipment and
Systems).
Irrespective of hardware capability,
the user (hydrographic surveyor) must
be educated, trained and experienced
in conducting operations. End data is
still only as good as the quality of the
installation, integration and acquisition.
Irrespective of data processing software
technology, it is still a critical quality
to have a qualified online surveyor,
however to achieve the data in the first
place requires a quality piece of sonar
hardware and sonar processing ability.

Acknowledgements
Christian Blinkenberg, and Mairi Forrest,
RESON Offshore Ltd
This paper was presented at Hydro 09
and is published here with permission.
Contact John Fraser, RESON,
john.fraser@reson.com

PositionIT October 2010