Home

Offshore Wrecks

HMS Oswald

Seabed Imaging

Wey Chieftain

Join Us

Forum

News

Inshore Wrecks

Weymouth

Online Diary

FAQs

Links

Multibeam or Sidescan?

Just recently there has been a lot of coverage in the press regarding the merits of multibeam sonar.  Whilst multibeam has its place as a good tool for mapping known wreck sites or geographical features it is almost useless as a search tool.  Multibeam is a very useful tool for presenting good-looking data to the public and they might be forgiven for thinking it is the be all and end all.  With this in mind I asked my good friend Jerry James to try and explain the merits in lay man’s terms.  Make yourself a cup of tea read on.

Seabed imaging, object detection techniques and tools.

The seabed and related features such as wrecks, pipelines, navigation obstructions and general topography can be imaged or detected by several different techniques, each with their own merits. The following describes the basic theory of the various techniques, their merits and shortcomings and their application to the imaging and detection of seabed features.

Acoustic methods

Single beam Echosounder

The single-beam echosounder has been around for many years and is now a common sight on even very small craft.

Basically a short but powerful pulse of electrical energy is converted to sound/acoustic energy and is transmitted from the bottom of the vessel directly at the seabed. The frequency of these small boat sounders is usually in the region of 150kHz. The pulse is in the form of a narrow beam, which is defined by the way in which the sounder transducer fitted to the bottom of the boat is designed.

The sound travels through the water at approximately 1500 m/second, which is about 3 times the speed of sound in air. When the sound pulse hits the seabed, some of the energy will be absorbed by the seabed material and some will be reflected back up the way it came down. However a proportion will also be scattered in all directions as the seabed is not a flat perfect surface.

Another important factor is that the sound will spread as it travels through the water and thus the pulse energy will also be weakened as it travels.

A proportion of the original pulse, if we are lucky, will however eventually get back to where it originated from at the sounder transducer.

At this point the transducer works in reverse by converting the sound energy back to electrical energy. A bit of electronics works out the time it took from transmitting the pulse to receiving the “return”. Because we know the speed of sound in water and the time it took we can work out the distance the sound has travelled between transmit and return.

However don’t forget that the sound has done a round trip from the boat to the seabed and back to the boat, so we must divide by two to give us the depth of water under the boat.

Once we have received the return signal we can transmit another pulse and take another measurement. Hence the greater the depth, the longer it takes the return signal to get back and the fewer measurements we can take in a given time. If the boat is going too fast we may even miss the return signal, so we have to consider speed versus depth.

Acoustics is a funny thing and there are a number of factors that can affect the passage of sound through the water.

Fresh water and salt water tend not to mix very well and freshwater will hang around in pockets. When sound meets the “front” between fresh and salt water, it tends to act like a mirror and reflect the sound at the boundary of the two different density materials. The same can happen with different temperature layers in the water. Different temperatures tend not to mix just like different salinity (salt density) pockets. Sometimes only part of the pulse energy will get through and in extreme cases all of it will be reflected back, giving false depth readings.

The changing densities also have an affect on the speed of sound. The higher the density/salinity, the faster sound travels. Thus if our pulse has to travel through changing densities on its way down and back, the speed of sound will also vary. So far we have only used an estimated/average speed of sound (1500m/s) in our calculations, so you can see that errors can start to creep into our measurements.

When commercial echosounder surveys are carried out, a device called a bathythermograph is used at regular intervals to record temperature and salinity variations with depth. The results are used to fine tune and correct the echosounder data.

There are still two other affects that can mess up your day. As we said earlier, boundaries of salinity can cause partial or total reflection of the sound pulse. However sound passing through the boundary will be “bent” as it passes through the transition, much like light is bent by a prism. It’s not hard to see that this will cause a change in the distance travelled to the seabed and back by the sound pulse, again causing errors. This affect is called “ray bending”.

We are trying to get an accurate reading of the depth of water below the boat, by using a single narrow beam pointing directly at the seabed. What happens when the boat pitches or rolls? Yes you’ve got it, the beam moves around. The deeper the water, the more the beam waves around at the seabed, so we are not always measuring directly below the boat.

As we have seen the sound energy decreases with distance due to the beam spreading and absorption of the energy by the surrounding water. Thus if we want to measure large depths we will have to put more energy into our pulse in order to get sufficient “return” signal at our transducer. We will also have to try and keep our beam as narrow as possible to reduce the spreading losses. It is easier to manufacture high frequency transducers with narrow beam characteristics than it is to have narrow beam low frequency transducers. However, once again nature screws us up. High frequencies will incur greater losses than low frequencies as they travel through the water, so we need to choose an optimum frequency to use as well.

Considering all the above, its amazing that any echosounders ever work, but a compromise can be made and with modern electronics and processing power, very accurate depth measurements can be made.

Multibeam Echosounder

The Multibeam echo sounder does exactly what it says on the packet. The inherent problem when trying to survey the seabed with a single beam echosounder is that you affectively only get a set of readings along a narrow line. Unless you run your boat up and down a very narrowly spaced survey grid you could easily miss features. Some time ago someone clever thought why not have several sounders arranged as a fan and cover a wide swath of seabed with each transmit ping? Thus the multibeam echousounder was born.

The fan of the beam is again narrow in the along track view, but spreads out either side of the vessel and is made up of lots of overlapping single beams. Thus for every transmission “ping” a large number of depth readings are obtained.

There may be 90 or more narrow beams of about 1° wide giving a fan of 90° width. You do not need to be a geometry wizard to work out that with this fixed fan shape, the deeper the water the wider the “swath” covered at the seabed. However you only still have the same number of beams to give you a depth sounding so the number of soundings per metre of seabed reduces with depth. This obviously reduces the ability to decipher accurately what the seabed topography is and features could be missed. Also because the swath width changes with depth you have to vary the width of your survey tracks to make sure that you cover all the area adequately and don’t miss anything.

As with single beam echosounders all the acoustic problems remain, but are multiplied by the number of beams. Huge amounts of soundings are created over a short period of time and inevitably incorrect sounding will be taken due to acoustic problems. The outer beams, because of their steep angle, are more susceptible to errors.

A 3D image of the seabed can be created using different colours for different depths, thus highlighting features such as wrecks. You can also create a 3D mesh by joining the sounding points which gives you a 3D map effect

With multibeam data it is necessary to post process the data recorded very carefully in order to “clean” the bad soundings. This is somewhat problematical, especially when looking at complicated data such as wreck sites. It is very easy to over filter the data and throw away the baby with the bath water!

With care some very good “images” can be created of wrecks and other features using multibeam sounders, but it is always a trade off of depth against resolution of the data. In deep water only very course images can be created.

So far we have only considered measuring depth points using the returns from the echosounder. Within the returned signal there will also be what is known as “backscatter data. Each beam of the sounder spreads as it travels through the water and thus by the time it reaches the seabed it will “illuminate/insonify” an area of seabed which is known as the “footprint”. The “footprint” will get larger, the deeper the water gets. We can get the depth from the round trip time, but we can also measure the relative strength of each returned signal. This will tell us something about the reflectivity of the seabed. I.e. Is it soggy or hard? The backscatter data thus gives us seabed texture information. If we join up this data we can map boundaries between different sediment types etc.

Multibeam however gives only poor quality backscatter data, which for imaging purposes is not very good.

A multibeam fixed to a vessel can be used as a wreck hunting tool, but has several major drawbacks.

Swath width varies with depth.

Seabed imaging is not great and gets worse with increasing depth.

We need very accurate vessel motion sensing to enable data correction.

If the wreck does not stand up from the seabed, as in the case of older wooden wrecks, it will be difficult to detect as you will have to rely on poor quality backscatter data.

Adverse sonar conditions in deeper water mean that considerable post processing of the data may be required in order to make sense of the data.

On the plus side it is possible to do very good work in shallow water and some quite astounding images of wrecks can be obtained where there is plenty of relief.

Multibeams can also be mounted to ROV’s and used for some very close in work. This gets around the requirement for expensive vessel motion sensing equipment.

Good multibeam equipment is expensive and coupled with the need for motion correction equipment, generally puts it out of reach of individual ownership.

Sidescan Sonar

Sidescan sonar has been around longer than multibeam and is still the main tool for seabed imaging. It is used widely in the offshore oil and gas industry, but cheaper more basic systems are within the reach of wealthier individuals, institutions such as Universities and Diving Clubs.

It was developed by a guy called Harold Edgerton back in the seventies, when he noticed that if you pointed a single beam echo sounder at a shallow angle across the seabed you could get an image of a thin strip. If you then move forward and “ping” again you can image another strip. If you add the strips together you get a continuous picture of the seabed. Stick another echo sounder pointing the other way and you can build up an image in two directions. I.e. either side of your vessel track.

If you then put this arrangement in a towed body and tow it behind a boat, you help to get away from the boat motion problems and you can maintain the swath width by generally towing the towed body/fish at a constant height above the seabed.

By careful adjustment of the spreading angle of the beam through transducer design, he gradually refined the system and improved performance.

Sidescan sonar uses the backscatter/ intensity of the “return” signal and will not give you depth information in its basic form. However if an object is sticking up from the seabed it will cause an acoustic shadow behind it, where no sound energy is reflected.

If we know the height of the towfish above the seabed we can use simple geometry to measure the height of the object by measuring the length of the acoustic shadow.

The height of the towfish above the seabed is calculated from the “first return”. This is the point on the seabed almost directly beneath the towfish. We can use this first return in the same way as a single beam echosounder works to give us the towfish height. The height of the towfish above the seabed is known as the “water column”.

Modern PC based data acquisition systems can track this height automatically from the first return and it can be used to automatically determine the height above the seabed of wrecks and obstructions.

Again a sidescan uses time as its basis for working out and presenting the seabed image.

A burst of high frequency sound is transmitted simultaneously from transducers mounted either side of the towfish. The transducers are designed to form a beam that is wide in the vertical plane, (typically 30 to 40 degrees) but very narrow in the horizontal plane (typically 1 to 1.5 degrees). This spreading burst of energy will insonify a narrow strip of seabed either side of the towfish track. If we want to view a swath width of say 100 metres total, we first calculate how long it will take for the sound to travel out and back for just one side (Port or Starboard). A total swath of 100m will mean a distance of 50m either side. We have to allow for the sound to do a round trip, so returns from the extreme range will have travelled 100m before it gets back to the transducer.

If we assume a speed of sound of 1500 m/s, it will take our round trip just 1/15th of a second to get back. The fist return will take a lot less and is dependent on the towfish height. If we record all the return data received at the transducers between “transmit” time and for 1/15th of a second afterwards, we will have data representing 50 meters of seabed either side of the towfish.

If we are moving forward we can then transmit again to insonify another strip, add the data to the previous strip and build up a 100 m wide image of the seabed.

Thus the swath width defines the rate at which we can transmit. This is known as the “ping rate”.

To be perfectly accurate we must also realise that the distance the sound travels is not the true plan range, as it is actually the hypotenuse of a right angle triangle (slant range) and we must take this into account when calculating the position of objects and features in relation to the towfish track. We can use the towfish height in conjunction with the slant range to calculate the true plan range. Again, modern sonar processing system will do this automatically if required.

The most common frequencies used are either 100 kHz or 300 kHz. The higher the frequency the better the detail (resolution) of the image, but the higher frequencies will not be capable of such a large swath as the lower frequencies due to higher absorption losses. A good dual frequency sonar can give swaths of up to 800m at 100 kHz and 300m plus at 300 kHz. For most survey and wreck search activities it is best to use the higher frequency settings so that small features are not missed.

However if you are looking to re- locate something like the wreck of the Britannic, as I did a few years ago for a diving expedition, the longer ranges at lower frequencies saved a lot of time. Once located you can have a detailed look at the higher frequency short range settings.

Some sidescans operate at 600 kHz or above and give very detailed images, but at the expense of operating range.

As we said earlier the Sidescan Sonar has been around for a long time, and it is still the best tool for searching the seabed for wrecks and features. A more recent development known as interferometric sidescan uses additional transducers to resolve the phase relationship of returns and uses this data to produce very accurate bathymetry (topography) data. As the bathymetry data is created from the sidescan data it means that it is now possible to create 3D maps of the seabed with co-registered texture data, making it a very powerful tool for seabed mapping.

This overlaying of sidescan data onto bathymetry data is possible by merging multibeam and sidescan data, but because one sensor is generally boat mounted and the other is towed behind the boat it is fraught with data alignment problems.

Sidescan sonar is capable of imaging quite small features.

Often on the harder to find wooden wrecks, there will only be small items on the seabed such as iron knees, strapping and maybe cannons and anchors. Because sidescan displays return intensity, these hard objects reflect better than the surrounding seabed and “stand out” in the displayed data even though they have no appreciable height. A mutibeam would be hard pressed to resolve such small and low lying objects. These artefacts tend to lie in what is known as a debris field of other objects associated with the wreck. When searching for wrecks we tend to look for features such as straight lines or right angles that do not occur naturally. When we find a group of these in an otherwise natural looking seabed area we can start to use other tools to investigate further.

Scanning Sonar

Scanning sonar’s use basically the same principals as sidescan. A single narrow beam is scanned over a seabed area and the returns are used to create a radar like image. It is possible to scan complete circles wit some systems. They are a useful tool for close in inspection work and are often used on ROV’s for obstacle avoidance etc. They are not particularly good as a search tool and suffer from the same problems regarding platform motion as a multibeam. They do have the advantage of not having to move the vessel and can be used in a stationary position.

Sub-bottom profilers

A sub-bottom profiler is basically a thumping great echo sounder. Very high energy pules of sound of about 3-4 KHz, formed in a narrow beam, are directed straight at the sea bed. Some of the energy will be reflected back from the seabed and some will penetrate it. As the sound travels through the seabed it meets different density layers such as sand ,mud and different types of rock. At each change of density, some sound will be reflected and some will carry on to the next change. The returns from each layer can be plotted in time, in the same way as sidescan sonar data and an image of the strata below the seabed can be made.

The sub-bottom profiler is not a good search tool as it only has a single narrow beam like an echo sounder, but it comes into its own when investigating possible wreck sites for buried objects such as wooden ribs and keels that have been preserved in the silt or sand.

The depth of penetration depends on the material build up of the seabed, but 10 to 15 meters is fairly common in softer seabed types.

Other methods

Magnetometer

The magnetometer is basically a metal detector. It can be towed behind a vessel or hand held by a diver. It is designed to measure very small magnetic field changes. It must first be tuned to null out the earths magnet field well away from any other magnetic materials, preferably in the middle of a field somewhere. Once the system has been nulled it can be towed behind a boat and the output trace observed for significant changes. Large steel wrecks will cause a significant local disturbance to the earth’s magnetic field and will be easily seen. However smaller items are harder to pick up and tend to be lost in the general clutter of changes due to natural affects. Outcrops of iron ore near the seabed surface will upset the system and make it difficult to see the smaller stuff. It also has the draw back of only detecting ferrous metals.

The magnetometer is useful for further investigating possible buried sites, but is not generally useful as a large area search tool unless looking for large steel wrecks.

Used in conjunction with a sidescan and sub bottom profiler it helps complete the picture and will increase the confidence that a wreck site has been found before undertaking diving activities.

Non ferrous detectors.

These are metal detectors that will pick up non ferrous metals like gold, silver copper and aluminium. A pulsed electrical field causes electrical currents to be induced in the metal objects. When the currents collapse after the energising pulse has gone they in turn generate a small electrical field that can be detected. You have to be very close to the objects, so they are no good as a towed system. However they can be mounted to an ROV or even a diver.

Summary

As can be seen there are a number of tools that can be used to help locate and identify possible wreck sites. None are perfect but the following observations are a round up of their merits and drawbacks.

Webmaster - DGWebs

© Crown Copyright 2008 - 2018 all rights reserved.