7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

Advantages of parametric acoustics for the detection of the dredging level in areas with siltation Jens Wunderlich, Prof. Dr. Gert Wendt, Rostock University

Keywords sediment echo sounder, parametric acoustics, dredging level, siltation,

Abstract Dredging companies and local authorities are interested in exact information about the dredging area and the material conditions, like sediment structures, sediment types and sediment volumes especially in ship channels and harbours. To get these information several equipment for sediment echo sounding is used. But what are the differences? What are the advantages of non- linear echo sounders compared to linear ones, especially in dredging areas with siltation? These questions are briefly discussed in this contribution. After that some survey examples, using a parametric echo sounder system, are given.

Echo Sounding and Echo Sounder Parameters Sediment echo sounding means v = const transmitting pulses in direction ∆h surface to the bottom and receiving reflected y x signals from the bottom and sediment a x b layers, see fig. 1. z ∆Θ,∆Φ There are some important parameters noise reverberation to classify echo sounder systems: Θ • Directivity h • • Pulse length, pulse ringing ∆τ bottom echo • Pulse repetition rate • Beam steering and stabilizing bottom surface

• Heave compensation objects • Real time signal processing l

• Size, weight, mobility a,c = f(material) layer echo These parameters and their meanings for surveys shall be discussed briefly. layer borders

Figure 1: Echo sounding principle Directivity The directivity of the transmitted sound beam depends on the transducer dimensions related to the sound frequency. There is a main lobe with the beam width Θ and side lobes. The sound at the border area of the main beam has a longer travel time than the sound in the centre of the beam. This makes the reflected signal longer than the transmitted signal. Particularly in deep- water areas refraction due to changes of the sound velocity may enlarge the sounded bottom area and that makes the echo signal longer, too. Longer signals will mean less vertical resolution. The horizontal resolution depends on the area that is hit by the sound beam. A great beam width means bad horizontal resolution and diffraction hyperboles occurs at small structures. A great beam width will also mean greater surface and volume reverberation. Another problem are side lobes that cause echoes from outside of the desired direction. To get echo prints with high spatial resolution you need directivity with small beam width and without side lobes.

page 1 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

Frequency, pulse length, pulse ringing and pulse repetition rate The attenuation of sound inside the sediment layers is most important to the penetration depth that can be expected. The attenuation coefficient is proportional to the frequency and depends on type and structure of the sediment. For greater penetration depths lower should be used. On the other hand the vertical resolution depends on the pulse length of the transmitted signal. Shorter pulses will result in better resolution, which could mean using high frequencies. For different survey tasks different frequencies may be optimal, the echo sounding equipment therefore should be cover a wide frequency range. Short pulses can be transmitted only by systems with a high frequency bandwidth. The ringing effect at the end of the transmitted pulse depends on the transducer bandwidth. The ringing time should be as short as possible for echo sounding in shallow water. For safely detection of small sediment structures or buried objects the bottom area must be hit several times. Therefore the pulse repetition rate should be as high as possible.

Beam steering and stabilizing, heave compensation Caused by rough the transducer may move during the survey in 6 directions. The most important unwanted motions are roll, pitch and heave. They should be compensated, if high- resolution echo prints are required. Therefore beam stabilizing should be possible, especially at greater water depths. Heave compensation is useful at shallow water areas, too. Fig. 2 illustrates the difference between echo prints with and without heave compensation. At slopes it is useful to direct the beam perpendicular to the bottom surface to make sure to get the best penetration. For that reason beam steering should be possible. Beam steering may also be used for widening the search area without losses in the horizontal resolution.

Figure 2: Heave compensation example (left without; right with compensation)

Real time signal processing and size, weight, mobility The echo prints should be calculated in real-time to get first survey results immediately. Thus makes a powerful real-time signal processing necessary. If you need mobile equipment to use on small, maybe different ships, the size and weight of the echo sounder system and in particular of the transducer will be important.

The ideal echo sounder The ideal echo sounder for surveying should have • Small beam width and no side lobes • Wide frequency range • Short sound pulses without ringing • High pulse repetition rate • Beam steering and stabilizing as well as heave compensation possible • Echo prints calculated in real time • High mobility and therefore small and light equipment

Some of these desires are only to fulfil using non-linear acoustics.

page 2 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

Linear and Non-linear Acoustics Linear echo sounders generate the sound pulse of the desired frequency directly. The directi- vity depends on the ratio of the transducer dimension and the signal frequency. Therefore a good directivity at low frequencies requires large transducers. But such transducers are heavy and expensive. Parametric echo sounders transmit two signals of slightly different high frequencies at high sound pressures (primary frequencies f1 and f2). Because of non-linearities in the sound propagation at high pressures both signals interact and new frequencies are arising, see fig. 3.

The difference frequency (∆f = f2 - f1) is low and penetrates the sea bottom. The primary frequencies may be used for exact deter- mination of water depth even in difficult situations, e.g. soft sediments. Figure 3: Non-linear acoustics principle

Figure 4: Linear (left) and parametric transducer array (right) Fig. 4 shows a linear and a parametric transducer array. The linear transducer has an active sound area of approximately 0.7 m × 0.7 m and weights approx. 350 kg. The parametric transducer with comparable directivity has an active sound area of approximately 0.2 m × 0.2 m and weights approx. 15 kg. 1 1 4 kHz 100 kHz 0.9 8 kHz 0.9 4 kHz 0.8 12 kHz 0.8 8 kHz 0.7 0.7 12 kHz 0.6 0.6 0.5 0.5 0.4 0.4 Amplitude 0.3 Amplitude 0.3 0.2 0.2 0.1 0.1 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 Angle [°] Angle [°] Figure 5: Directivity pattern for linear (left; computed) and parametric transducer (right) Using parametric sound generation, the directivity for the difference frequency is similar to the primary frequency directivity and is nearly the same for different secondary frequencies. This is important if echo prints from different frequencies should be comparable, so for the calculation of sediment attenuation using multi-frequency signals.

page 3 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

Fig. 5 shows some experimental data from a parametric transducer array with an active sound area of about 0.2 m × 0.2 m. All difference frequencies between 4 and 12 kHz (ratio 1:3) have nearly the same half power beam width as the primary frequency of about 100 kHz. The left side of fig. 5 shows the computed directivity of a linear transducer that is 10 times greater. In this case different radiated frequencies have different half power beam widths. Therefore the sounded area will be not the same at different frequencies and echo prints cannot be compared.

2 primary frequency secondary frequency 1.5

1 Amplitude 0.5

0 0 1 2 3 4 5 6 7 8 9 10 z / z0 Figure 6: Sound pressure at the beam axis (computed; circular piston transducer) The general sound pressure distribution using non-linear acoustics along the acoustical axis of a circular piston transducer is shown in fig. 6. There are no local minima or maxima in the near- field area for the difference frequency. The distance between the near field and the far field is longer for the parametric signal (approx. 5-10 times near-field length) than for the linear signal (approx. 3 times near-field length). The maximum sound pressure and the sound propagation depends on the near-field length of the primary frequency and the primary to difference frequency ratio. Due to the interaction length of the parametric transmitter, the decrease of the sound caused by geometrical losses and physical attenuation will be lower compared to linear echo sounders.

Advantages of Parametric Systems for Surveying The properties of parametric systems described above causes a lot of advantages in surveying compared to linear echo sounders. Parametric systems have small beam width for the transmitted low frequent signal in spite of small transducer dimensions. The beam width only depends on the primary frequency related to the transducer aperture, even for the secondary frequency. There are no significant side lobes for the difference frequency and you will get a constant directivity for different secondary frequencies (fig. 5). Therefore the footprint is the same for different frequencies and you will get comparable results. Caused by the high system bandwidth of a parametric system really short signals can be transmitted without ringing, for example 1 cycle of 12 kHz. This makes parametric systems useful in particular in shallow water areas. Due to the small beam width and the high frequency bandwidth the bottom echoes at parametric transmitting have a steeper slope than echoes from linear echo sounders. These steeper signals are better to detect at low signal to noise ratios, so in areas with dredging activities. Therefore a high resolution of layers and detection of small changes in the acoustic impedance becomes possible. It will be produced a more realistic and more accurate picture from the bottom layer and the sediment structures beneath the bottom. Short pulses, narrow beams and the absence of side lobes results in less volume reverberation and less reverberation from the bottom surface compared to linear systems. This results into a better signal to noise ratio, especially in areas with siltation.

page 4 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

The Parametric Echo Sounder System SES-96 Based on components developed by the Research Group of Rostock University, INNOMAR Technology GmbH Rostock developed a product line of new parametric echo sounders. Sound pulses are generated by a small piezoceramic phased transducer array with up to 16 separately controlled elements. Electronic beam stabilizing and steering is possible. Parametric echo sounding transmits focused low frequent signals in spite of small transducer dimensions. The high frequent signal from the primary transmitter frequency is processed separately to detect the bottom surface. Real time signal processing produces colour echo prints to detect sediment layers and buried objects with high resolution. High repetition rates are used to improve the signal to noise ratio and to raise the degree of probability to find small single objects and small bottom structures. The echo sounder system SES-96 consists of a main device and a transducer array. The main device contains transmitters, receivers and modules for analogue and digital real time signal processing. Analogue to digital converters (ADC) are used for digitising the receiver signal with 16-bit resolution at sampling rates up to 200 kHz depending on signal bandwidth. All received data are stored digitally on hard disk including GPS data and other important system parameters. The echo sounder data are stored in an optimised binary format but may be converted into the standard SEG-Y format for post processing using other equipment. A motion sensor measures the ship’s movements including the heave. Sensors from Seatex, TSS and Octans are supported. The roll and pitch values are used for beam stabilizing. Echo prints are heave compensated by the sensor’s heave value. All the signal and image processing is done in real time.

Water depth range 0.5 … 500 m Vertical resolution <6 cm Penetration depth up to 50 m Accuracy of the depth measurement 0.02 m +0.02% of the water depth Primary transmitter frequency ca. 100 kHz Secondary transmitter frequency 4, 5, 6, 8, 10, 12 kHz Transmitter pulse length 0.08 … 1 ms Repetition rate up to 100 s-1 Beam width ±1.8° @ 4 … 12 kHz Beam steering range ±16° Transducer dimensions ca. 20 × 20 cm Transducer weight incl. cable (in air) ca. 32 kg

Table 1: Main Parameters SES-96

Figure 7: SES-96 main device (left) and mounted transducer (right)

page 5 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

SES-96 Survey Example A lot of surveys under different conditions were done using the parametric echo sounder system SES-96 with good results. Compared to linear echo sounders the surveys were more efficient and volume or mass calculations were more accurate. The data from the exemplary described survey were provided by INNOMAR Technology GmbH Rostock [3, 4]. During the selected survey the signal to noise ratio was sometimes bad due to a lot of particles in the , high ship traffic and therefore noise from vessel engines and bubble creations. Another source of noise were the dredging activities of the hopper dredgers. The siltation causes a high attenuation of the sound in the water column. Therefore the high- frequency depth information got lost sometimes. Very good results were received at the 10 kHz difference frequency. The main task was to monitoring the siltation, which had sometimes a thickness of up to 6 m. Even under these conditions it was possible to detect a clear bottom line below the siltation - the reached dredging level. The different colours in fig. 9 at the siltation areas (the nearly flat sediment parts) are indicating different acoustical impedances (and therefore a different density and/or sound velocity in the sediment material). The high signal dynamic of the SES-96 system and the small half power beam width allow the measurement of differences in the acoustical impedance in the range of one percent. The change of the acoustical impedance in the right part of the plot is less than in the left part and indicates a softer siltation (probably not as old as on the left side). The penetrations on the left and right of the figure show layers much deeper than the depth of the multiple signals. During the processing the water depth (red line, fig. 8 and fig. 9) was digitised from the HF channel. The dredging level was digitised too (black line, fig. 9), also below the siltation.

Figure 8: HF-echo plot with water depth (red line)

Figure 9: LF Echo plot (10 kHz) with overlaid HF depth (red) and dredging level (black)

page 6 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

The whole area contained a lot of high spots, with a height of more than 5 m. These spots are the result of the dredging activities. The SES-96 detects these high spots not only in the high frequency channel but also in the low frequency channel. In the 15 kHz channel of the standard echo sounder plots (not shown in this paper) these high spots are often not visible caused by the steep slopes of the spots. This results in a wrong digitisation, because these spots are skipped. The volume calculation then becomes inaccurate. This lost of information in echo prints from a linear echo sounder compared to the parametric SES-96 is shown in fig. 10 to fig. 12. These echo prints are generated in parallel at the same time.

5 m

Figure 10: HF echo print SES-96 (100 kHz)

5 m

Figure 11: LF echo print SES-96 (12 kHz)

210 kHz

5 m 15 kHz

Figure 12: Echo print from a linear echo sounder (210 kHz / 15 kHz)

page 7 of 8 7th Workshop on Dredging and Surveying, Scheveningen (The Haag), June 07th-08th 2001

Survey Example: Cruise SO147 "Peru " The parametric echo sounder system SES-96 shows good results not only in shallow water areas but also at water depths up to about 1500 m. During June 2000, staff members of the Underwater Acoustics research group of Rostock University took part at cruise SO147 [5]. This journey along the coastal upwelling zone off the Peruvian coast was organized by the German Federal Institute for Geosciences and Natural Resources (BGR) Hannover. For high-resolution echo sounding the linear SEL-96 and a special deep-sea variant of the parametric SES-2000 sediment echo sounder systems were used, both developed at the Rostock University. During SO147 echo-sounding profiles with a total length of about 6000 km at water depths between 50 and 1600 m with sediment penetrations up to 60 m were acquired. Fig. 13 shows an echo print example from this survey with very weak sediments on top of the bottom, so called mud waves, in water depths around 300 m. These mud waves were not visible at older echo prints from the same area generated by other devices.

10 m

Figure 13: Echo print from Peru; SES-2000DS 10 kHz (Range 270 ... 320 m) References [1] Barnick, W.; Wendt, G.; Jakowlew, A.N.; Kablow, G.P.: Hydroortungssysteme zur vertikalen Bodensondierung; Nowosibirsk; 1992 [2] Hamilton, M.F.; Blackstock, D.T.: Nonlinear Acoustics; Academic Press Ltd.; San Diego; 1997 [3] Lowag, J.: Advanced Techniques for Dredging Applications; Hydro International 09/2000; pp. 55ff [4] Müller, S.; Lowag, J.: Parametric Sediment Echo Sounder SES-96 for Dredging Applications; Workshop of the Hydrographic Society; Rotterdam 02/2001; pp. 27ff [5] Wunderlich, J.; Wendt, G.: High Resolution Echosounding at SO147 Sonne-Statusseminar; Hannover; 03/2001; pp. 35-38 [6] Wendt, G.; Wunderlich, J.: Sediment- und Objektortung mit parametrischen Sendeverfahren; DAGA 2001; Hamburg 03/2001

Contact Address Prof. Dr.-Ing. habil. Gert Wendt; Dipl.-Ing. Jens Wunderlich Rostock University Institute for Communications and Information Electronics R.-Wagner-Str. 31 D-18119 Rostock-Warnemünde Germany e-mail: [email protected] WWW: http://www-nt.e-technik.uni-rostock.de/ntie/fg_hydro.html

page 8 of 8