SIDE-SCAN SONAR : A PRACTICAL GLIDE

Transcription

1 nternational Hydrographie Review, Monaco, L (), January SDE-SCAN SONAR : A PRACTCAL GLDE by B.W. F i.km\ii\ <; National Research nstitute of Oceanology Council for Scientific and ndustrial Research Stellenbosch, South A frica A B S T R A C T The apparent lack in up-to-date literature of practical advice on operational aspects of side-scan sonar and the isom etric reproduction of records calls for a more specific approach. The author discusses step by step every m ajor feature that directly affects the com position o f the resulting sonograph. Special emphasis is given to distorting factors, which have to be elim inated when draughting isometric maps from sonograph mosaics. A number of practical aids are presented in the form of distortion ellipses and true distance nomograms, which sim plify the manual draughting process to a point beyond which only com puterization w ould provide substantial im provem ent. Since com puterization o f the draughting process will not be available for many users in the foreseeable future, these aids are considered vital. Finally, the various features discussed, as w ell as geological interpretations of selected exam ples from the South A frica n continental margin, are represented. Th e discussion is based on the EG&G Mark B side-scan sonar system. N T R O D U C T O N A m ajor shortcoming of most up-to-date literature on side-scan sonar is the lack o f practical inform ation and advice to solve the numerous technical and operational difficulties discussed at such length. t is the aim o f this essay to com pile such inform ation. M athem atical and other theoretical discussions are reduced to a minimum, this having been covered in detail by many authors and expertly summarized by L e e n h a r d t (974). Special emphasis is therefore given to all those aspects that directly concern the com position and interpretation of sonographs. From the start it should be noted that on two vita) aspects inform ation and experience cannot be fu lly passed on. Firstly, there is no fixed control

2 setting in order to achieve m axim um clarity and resolution; continual adjustments are necessary. And secondly, accurate interpretation o f the features recorded on the sonographs often remains a matter of argument. n both cases the operator w ill have to gain his personal first hand experience before achieving satisfactory results. Extensive test runs in well known areas and control dives by scientifically trained personnel have proved very successful. This discussion is based on the EG&G Mark B side-scan sonar system w hich has achieved considerable distribution. How ever, w ith the exception of the nom ograms (fig. 0 : -0) which are unique to this particular instrument, all other features discussed apply in general. H STO R CAL BACKGROUND Sound was first used to exam ine the sea-bed w hen W o o d and some colleagues developed the prototype echosounder in 929 ( W o o d et al., 935). Since then numerous adaptations have come into use, m ain ly in the form of bottom penetrating low frequency sound sources, today w idely applied in the various marine seismic techniques. t was not until 958 that sound was actually used to m ap geological surface features o f the sea floor utilizing the back-scattering effect o f high frequency sound waves (C h e s t e r m a n, C l y n ic k & St r i d e, 958). The first operable sideways-looking sonar was built by T u c k e r & S t u b b s (96) at the National nstitute o f Oceanography in England and since then the instrum ent has undergone rapid developm ent and im provem ent. This period is documented by m any publications discussing m ainly the technical problem s rather than practical applications. The articles by T u c k e r (966), C h e s t e r m a n et al. (967), S a n d e r s & C l a y (968) are good exam ples of this category. Howrever, side-scanning achieves its m axim um capability only when applied system atically! Sporadic recordings w ithin the lim its o f reconnaissance cruises can provide valuable background inform ation but w ill not suffice fo r detailed interpretations. U tilizing such inform ation, specific problems can be defined on a local or regional scale fo r a systematic approach w ithin the conventional survey procedure, thus optim izing and substantially im provin g costly scientific ventures. Exam ple : dredge stations conventionally selected from echo sounder profiles never achieved m ore than 60% success along the South A frican coast whereas those selected from side-scan sonar records proved 00% successful! Although the first system atic surveys were commenced as early as 964 by C l a y, E ss & W e is m a n it was not until the end o f the last decade that they became routine w herever the instrument wtas available. S a n d e r s et al. (969), W o n g et al. (970) and Ch e s t e r m a n et al. (97) provide interesting exam ples o f this nature. R u s b y (970) discusses a side-scan sonar system specially adapted fo r long range deep-water surveys commonly know n as the G.L.O.R..A. P roject. Th is instrument can cover a maximum range o f 22 km.

3 A com prehensive com pilation of literature on the developm ent and application of side-scan sonar can be found in B e l d e r s o n et al. (972). A t the same tim e the book presents m any excellent exam ples of sonar recordings (sonographs) and their geological interpretations thus providing a very useful reference catalogue. More recent examples of detailed sidescan surveys are discussed by N e w t o n et al. (973), B e l d e r s o n et al. (974), M ck in n e y et al. (974), W e r n e r (975), W e r n e r & N e w t o n (975), L a u g h t o n & R u s b y (975) and B ry'a n t (975). First results of side-scan sonar investigations along the South A frica n coast were presented by G e n t l e (973). Side-scan sonar must be regarded as a m ajor breakthrough especially for m arine geology because, for the first time, it has become possible actually to map the surface features of the sea flo o r on a broad scale. SYSTEMS A N A L Y S S B asically the side-scan sonar system consists of three units : a transducer w hich form s the underw ater unit and is better known as the fis h a steel w ire reinforced cable acting as transm ission and tow cable sim u l taneously, and a dual channel recorder. The fish consists o f a streamlined, hydrodynam ically balanced body about m long containing tw o sets o f transducers that scan the sea-bed a A fish m i h cone angle F i g.. The sound lobes of the ultrasonic beam (exaggerated), a. Plan view; b. Vertical section.

4 68 i n t e r n a t i o n a l h y d r o g r a p h i c r e v i e w on either side. The ultrasonic beam is slightly depressed from the horizontal w ith the axis of the main lobe pointing 0 downwards (fig. lb ). The horizontal spread of the beam is narrowed down to.2 (fig. la ). This ensures a reasonable transverse resolution even at the largest range (0 in). The transverse resolution (R () is defined as the minimum distance between two objects parallel to the line of travel that w ill be recorded on paper as separate objects. This minim um distance is of course equal to the beam width at any particular point on the sea-bed, resulting in a steady decrease of the transverse resolution towtards the outer sweep ranges (Table ). This is a system-inherent effect that cannot be manipulated, and it is responsible for the apparent fading of recorded signals on the outer sweep ranges if individual features become smaller than the corresponding resolution. Exam ple. at the poinl on the sea-bed marked bv the 200 in range the transverse resolution will he R, = sin.2 X 200 = 4.9 m () The vertical spread of the main lobe is norm ally 20" which together w ith the vertical side lobes w ill cover the whole distance from a point vertically below' the fish to the lim it of the m aximum range, covering 0 in on each side (fig. lb ). A lternative settings are available for deep w ater surveys. Resulting from this coverage a certain vertical resolution is achieved depending largely on the range setting. The vertical resolution or range resolution (R,) is defined as the minim um distance between two objects perpendicular to the line of travel that w ill be recorded on paper as separate objects. Assuming that a m inim um spacing o f mm on the recording paper is needed to plot two objects separately, the resolution w ill be / of the range scale, the paper width per channel being mm (Table ). Exam ple : if the range setting is 200 m, then the range resolution w ill be : Rr = 200/ =.6 m (2) These lim its of resolution define the maximum scale of reproduction of sonographs. N orm ally a strong reduction becomes necessary anyway, and even a /0 000 reproduction o f a 200 m range record means reducing the w'idth of the record by the factor 0.6. The w hole process o f reproduction, however, is not quite as simple as that because other distortions, caused m ainly by varying ship speeds and heights of fish above bottom, involve a tedious process of correction before an isom etric map is produced. The sonar im age is constructed by signals received through the main lobe as well as through the side lobes (fig. lb ), and obviously the side lobes, being less intense, w ill have inferior qualities regarding resolution, especially in the case of the small central lobe. Although large objects are recorded without difficulty, smaller ones may be vaguely defined or not recorded at all. For the correct interpretation of sonographs it is important to knowr how large this area of reduced resolution is. Since we are dealing with a sound lobe producing a cone, the width of this cone area is a function of the fish height above sea-bed (table ). Furtherm ore, w'here the side lobes overlap w ith the main lobes, resolution is dram atically reduced, often resulting in a w hite gap close to the inner record m argin o f each channel.

5 T ab le nter-relationships between various parameters. H eig h t o f fish ab o ve sea-bed (m ) C o n e area p er channel ( m ) Suitable sw eep ranges ( m ) X X X X X X n O X X X X X X X X X X X X X X X > Transverse re solu tion ( m ) R an ge resolu tion ( m ) O p tim u m h eigh t o f fish ( m ) The dual channel recorder contains most of the electronics as w ell as the graphic mechanism. The transducer sends out short sonar pulses of 0. ms duration and the returning signals are am plified in the recorder and fed, in form of variable currents, to the helix electrodes which sweep out from the centre of the recording drum. The current passes through the recording paper to the printer blade-electrode and from there to earth. The current passing through the paper produces m arks of varied intensity which are proportional to the strength of the incom ing signals, thus reflecting the nature o f the sea-bed (fig. 3). The recorded image is called the sonograph and, in a sense, it is rem otely similar to a continuous high altitude aerial photograph. Each returning signal is plotted on the paper at the position corresponding to the time it was received after the outgoing pulse. This position w ill shift across the paper proportionally to the selected sweep range. Each range is divided into equal time intervals which are autom atically plotted as lines on the recording paper. These time intervals, defined as slant ranges, correspond to m travel distances of the outgoing pulse. Th e 00 m range, for example, w ill show 4 such m lines, and the 200 m range 8 lines. Doubling the range therefore means increasing the scan area roughly by the factor 2. This results in a reduction of all features recorded on the paper by the factor 0.5. The instrument is calibrated such that by doubling the range the pulse rate is reduced by one half, and in order to avoid gaps on the recording paper the paper feed rate is also reduced by one half. n this w ay the longer travel times o f the sonar signals are compensated for.

6 The stronger the returning signal, the darker w ill be the m ark on the paper. Th is intensity o f the return signal is a function o f m aterial properties as w ell as topography. Large objects (e.g. boulders, rock pinnacles, ridges and sandwavesj are not only good reflectors but also produce an acoustic shadow zone behind them w here nothing is recorded, thus leaving w hile patches 0 the paper. Th e width of this shadow and the position o f the object relative to the fish are utilized to calculate the height of such objects (fig. 2). The height of an object (H,) is equal to the product of the acoustic shadow length (L g) and the height o f the fish above sea-bed (H ;), divided by the sum o f the acoustic shadow length and the range distance at which it is recorded (R.). Thus : L, x H/ H. = - L. + FL (3) L, x Hf L. + R. 36 x 20 -= 7.5 m The previous discussion has already indicated that the slant range (R,) is not identical to the recorded distance along the sea-bed (fig. 2). n order to determ ine the exact position o f an object or feature relative to the line o f travel, the horizontal range has to be calculated using the theorem o f Pythagoras. T h e horizontal range (R *) or better the true distance over ground is equal to the square root of the differen ce between the squares o f the slant range (R ) and the height o f the fish above sea-bed (H f). Thus : R = V / RJ2 - h; (4) T o speed up this operation o f calculating true positions, tables, graphs or, even better, straightforw ard nom ogram s can be utilized. The author has draughted such aids, and their application is discussed later in this article. As long as the process o f accurate reproduction of sonographs in m ap form is not com puterized these aids are the quickest and most reliable means available. A nother typical feature o f sonographs are the identical depth profiles recorded on both channels along the inner portion o f the paper. Since the recorder plots time intervals o f returning signals, it w ill also record the slant range corresponding to the vertical distance between the fish

7 and the sea-bed at its respective range position on the paper. Furtherm ore, the upward facing lobes o f the sonar beam also record the sen-surface, which form s a reflecting horizon and appears as a thin undulating line at its respective range position on the paper. This line can easily be distinguished, and by adding the range distance from fish to sea-bed and fish to sea-surface the total w ater depth can be calculated (fig. 3). F ig. 3. Schematic figure illustrating various recording principles height of fish above sea-bed (a) = 20 m ; depth of fish below sea-surface (b) to ta l w a t e r d e p th (a + b ) = 30 m.

8 The depth profile therefore does not necessarily record true changes in topography only, blit w ill also record a change in the position of the fish relative to the sea-bed. Howrever, since depth profile and sea-surface line are com plem entary, any change in the position of the fish w ill always be reflected by a shift of the sea-surface line, whereas true topographic changes w ill not affect the sea-surface line. This feature is very im portant because it gives complete control over the position o f the fish, a factor that plays such a vital part in the correction o f distortions. Tow methods O PE R A TO N A L PROCEDURES The fish can be towed o ff the stern of a «hip, o ff either sidg or svsn from o ff the bowr. The method adopted depends largely on the objective of the survey and on the environm ental conditions such as w ater depth, sea flo o r topography and sw'ell conditions. During deep-water surveys the fish is best tow ed astern, because there it w ill be far below the surface well aw^av from in terferin g propeller turbulence. n shallow water the fish is norm ally towed off port or starboard, where it w ill also be away from the propeller. How'ever, care must be taken when turning the ship in order to avoid the fish slipping beneath the hull and possibly being damaged by the propeller. Th e cable should always be secured in a w ay that w ill allow quick recovery o f the fish. Tow procedure Th e most effective tow procedure can be sum m arized as follow's : Lay out on deck as much cable as w ill be necessary to effectively scan the deepest parts o f the survey area. This becomes unnecessary once a slip-contact cable winch is available; Reduce ship speed to at most 5 knots before low ering fish into the w ater; Connect fish and recorder, and low er fish into w ater; Switch on first the recorder and then the trigger; Adopt survey speed, and low er the fish w hile observing its position relative to the sea-bed on the recorder; Stop low ering once the fish has reached optim um height above sea-bed; usually about /0 of the selected scanning range. Short ranges m ay necessitate raising the fish higher than the optimum, especially in rugged areas. A lw ays m aintain a good safety m argin; Proceed to survey, annotating event marks at selected time intervals. A dju st fish height above bottom as m ay become necessary. n em ergency situations a sudden increase of ship speed w ill raise the fish very fast; ncrease ship speed when turning on profile in order to avoid the fish sinking to the bottom ; A fter com pletion of survey, pull in the fish until it becomes visible, then switch o ff first the trigger and then the recorder, and carefully hoist the fish aboard.

9 Navigation Accurate navigation is o f param ount im portance when p erform in g side-scan sonar surveys, because only then can the position of any recorded feature be accurately determined. E vent marks annotated w ith time, position and control settings should be carefully entered into a special side-scan log. Event m arks should not be spaced too fa r apart. f very accurate surveys are to be carried out then event m arks spaced only 2 minutes apart m ay become necessary, especially when scanning at sm all ranges where paper feed rates are high. W h en operating at 200, 0 or 0 m ranges, 5, 0 or 5 m inute event spacing's w ill be sufficient. t should be noted that each event m ark notes the position o f the ship and not the position o f the fish. Th e distance between fish and ship has to be subtracted from the on-line p rofile when plotting survey sheets. n areas w here automated navigation is not available a local system m ay have to be set up. Close inshore this is no problem, fu rth er out at sea how ever, new systems have to be introduced. Underw ater acoustic tria n gulation utilizing transponder navigation seems very prom ising and has reached an advanced stage of developm ent. Various systems and p rocedures are discussed by S piess et al. (969), T y r r e l l (969), M u d ie et al. (970) and K e l l a n d (973). NTE R PR E TATO N OF SONOGRAPHS Principles A sonograph consists basically of a sheet of paper m arked by shades of varyin g intensity and resolution. Features w ith sharp outlines alternate w ith vaguely defined areas in w hich subtle changes o f tone m ay occur. T o interpret these various appearances correctly w e must be aware o f the factors that can cause changes in tone or intensity on the recording paper. Th ere are two m ain sources o f influence that m ay cause darkening of the recording paper. One is purely electronical, caused by m anipulation o f the control settings on the recorder. For each ehanncl there are three manipulators. The general gain control w ill intensify all incom ing signals evenly across the paper, whereas the tw o tim e-varied gain controls initial and slope gain selectively a m p lify certain parts o f the record m ore than others in order to compensate fo r decrease o f intensity due to frequency dispersion o f the outer range signals. Th e second m ain source is the incom ing signals. H ere we must distinguish between two types. One type of signal is caused by material properties depending on the relative reflectivity o f the various m aterials on the sea-bed. Rock and gravel, for example, are better reflectors than sand and w ill therefore record darker. Sand, in turn, is a better reflector than mud, etc. The other type o f incom ing signal is caused by topographic features. Slopes facing the transducer reflect sound w aves better than surfaces lying oblique to the sound beam and w ill consequently plot darker.

10 M aterial reflectors and topographic reflectors can therefore produce exactly the same effect on the recording paper, and it is often up to the operator to decide which he is dealing with. T o do this accurately, he needs experience and preferably first hand know ledge about the particular reflectin g patterns caused by various rock types, gravels, sands, and characteristic form s associated w ith them such as bedding, jointing, ripple marks, sand waves etc. Exam ples of these w ill be discussed later by interpretation o f selected sonographs. Th is delicate interaction betw een the various signals must constantly be com prehended. f the recorder controls w ere set such that a smooth sand surface w ould produce an evon, light tone then all other features would record ligh ter or darker, depending on their relative reflectivity. This would elim inate a m isinterpretation o f electronically produced effects. U nfortunately- preselected control settings apply only in very smooth areas. n most other areas a continual adjustm ent o f control settings is necessary in order lo achieve m axim um resolution by com pensating for losses or gains in intensity caused by a change in the relative position of the fish above the sea-bed. Th is can hardly be elim inated w ithout other complications. A careful note should therefore be made o f all m ajor changes o f control settings by annotating them in the side-scan logbook. Distortions H avin g discussed in fa ir detail the m ajor system-inherent and operational aspects o f side-scan sonar, w^e can now proceed to factors causing distortions in the records. Sonographs do not norm ally represent isom etric maps o f the sea-bed, and various distorting factors have to be observed when reproducing sonograph mosaics in m ap form. Some practical aids w ill sim p lify this process. 9 0 ' 80' F ig. 4. Construction figure of distortion ellipses with which the compression effect parallel to the line o f travel can be corrected.

11 A n obvious distortion w ill occur parallel to the line of travel due to variable ship speeds, resulting in a com pression o f all sonographs in this direction. The effect is best dem onstrated by distortion ellipses (N e w t o n et al., 973) w hich show in sim plest form the progressive shortening of the on-line component w ith increasing speed. Fig. 4 illustrates how7 the various ellipses are constructed. At about 2 knots virtu ally no distortion occurs and the ellipse represents more or less a circle. T h e distorting effect on some comm on shapes o f increasing speed is dem onstrated schem atically in Fig. 5. The distances covered by the ship in line o f travel at various speeds per unit tim e are listed in Table 2. 2 knots V 4 knots 8 knot: knots 6 knots A? A / o ' i 7 C ~ D ^ 7 F i g. 5. Distortion effects on some common shapes parallel to the line of travel caused hj' various ship speeds. T a b l e 2 Distance over ground covered in line of travel at various speeds per unit time knots m/sec m/min

12 Fici. (i. Distortion ellipses for various ship speeds.

13 A nother aspect connected w ith the com pressional effect is the distortion o f all linear displays. A true angle o f e.g. 45 off the line o f travel at 2 knots reads in fact 63 at 5 knots and 74" at 9 knots. Th is effect too is dem onstrated in Fig. 4. n order to elim inate these distortions when plotting isom etric maps, the author has provided a series o f distortion range distance Fig. 7. Converting a sonograph (A ) into an isometric chart (B). Ship speed = 6 knots; Range = 00 m ; Fish height above bottom = 20 m.

14 78 i n t e r n a t i o n a l h y d r o g r a p h i c r e v i e w ellipses w ith their respective true angles for ship speeds between 3 and 9 knots (fig. 6). Th e result of such a correction is dem onstrated in fig. 7. T h e height o f the fish above sea bed causes a lateral distortion perpendicular to the line of travel. Range distance and true distance over ground coincide only on a line horizontally abeam o f the fish. Since the fish has to be towed some distance above the sea-bed, a certain amount of the recording paper is lost to the depth p ro file which plots at its respective range distance. The higher the fish is towed above the sea-bed the m ore paper is lost for actual recording (fig. 8). Th is can proceed until the smaller ranges are totally engulfed by the depth p ro file; e.g., the m range w ill cease to record the sea-bed at a fish height of m and more. The true distance therefore has to be fitted into the rem aining part o f the paper, beginning w ith 0 m w here the depth p ro file begins to record the sea-bed. fish F ig. 8. Lateral distortion caused by fish height above sea-bed. a. Fish height = m ; b. Fish height = m. Furtherm ore, due to the obliqu ity o f the sonar beam, equal truedistance intervals w ill not fo llo w a linear scale on the sonograph. The short-range intervals are compressed and the far-range intervals slightly stretched. n order to correct this distortion, a linear correction has to be applied w7hich w ill vary w ith fish height above sea-bed; ship speed does not a ffect this lateral component. Correction factors can be calculated and presented in table form or in graph form. Th e simplest form, however, is a direct conversion o f these values into nom ograms which can be used like rulers. For a certain fish height above sea-bed and the particular recording range, the respective nom ogram is selected and superimposed on the record, thus allow ing the true lateral distance to be read o ff directly. Th e author has prepared such nom ogram s for fish heights between 0 and 00 m above sea-bed and fo r all respective ranges (fig. 0 : -0).

15 upslope fish d o w n s lo p e F ig. 9. Lateral distortion caused b y slope in sea-bed. Upslope : K», = R + a; Downslope : KhJ = H* b. These corrections apply strictly for plane sea-beds only, and other nom ogram s would have to be calculated and drawn for various slopes of sea-bed in relationship to fish height and range. Fig. 9 illustrates the com plexity of distortions caused by sloping sea-beds; e.g. a 0 slope would increase the upslope component by ca. 3%, whereas the downslope component would be reduced by ca. 2%, thereby compressing and stretching the lateral distance on the respective channel. Com m only, sea-bed slopes rarely exceed -2, and correction o f the resulting small distortion would hardly be w orth the effort. Sim ilarly, distortions caused by density gradients in the w ater column are norm ally small, and correction would become necessary only w here large gradients are observed. n svich cases detailed density surveys would have to precede the side-scan operation. nterferences There are three main sources of acoustic interferences which w ill produce characteristic patterns on the sonograph. A continuous interference pattern, overlying the sonograph is caused by sefsmic instruments run simultaneously w ith the side-scan sonar. The com bination of surface m apping and sub-bottom profiling, however, yields valuable inform ation for the m arine geologist besides being m ore economic, and interference can be reduced considerably by running both instrum ents w ell apart. A second type of interference is caused by dense particle suspension in the w ater column. W henever suspension becomes dense enough the sonar beam is partially dispersed and partially reflected before it reaches the sea-bed, resulting in w hite gaps through all ranges, whereas the area w ithin the depth profile is blackened. Sim ilar effects, but in form of scattered dark patches w ith sharp outlines, are produced by shoals of fish.

16 A third type of frequent interference is caused by ultrasonic waves generated by passing ships. t form s a discrete pattern and can easily be distinguished from other interferences. t is sim ilar to seismic interference without, however, being continuous. Exam ples of these main interferences are presented in the appendix. CONCLUSON Producing isom etric maps from sonograph mosaics is a tedious and tim e-consum ing undertaking, even when all the practical aids presented and discussed here are applied. B e r k s o n & C l a y (973), H o p k in s (972a) and S p o t t i s w o o d e & D o r s o n (975) discuss methods that w ill accelerate considerably the manual draughting process. However, when much detailed w ork is envisaged then com puterization o f the isom etric reproduction process w ill become inevitable. p recent years a number o f attempts towards this solution have been undertaken, and the most prom ising results so far have been achieved by w orkers at Bath U niversity, England (H o p k i n s 972b, K e l l a n d & H o p k in s 972, K e l l a n d 972). F or practical reasons it is advisable to photograph all sonographs that display m eaningful features and to store the negatives in a file system. Th is w ill ensure that records rem ain in telligible even after long storage, w hich has a clear detrim ental effect on the quality of the original records. SONO G RAPH EXAM PLES T o assist in the interpretation o f the aspects discussed in this article, a num ber o f selected sonographs are presented below in an appendix. Since all these sonographs are original recordings, hitherto unpublished, from the South A fric a n continental m argin they contain valuable in form a tion about the geology and sedim entology o f this area ACKNOW LEDGEM ENTS T h e author wishes to thank the Geological Survey of South A frica for providin g the side-scan sonar instrument, and the National Research nstitute o f Oceanology (C.S..R.) for their cooperation. The Master and the crew o f the R. V. Thomas B. Davie deserve, as always, special acknowledgem ent for their effort in the rough waters o ff Southern Africa. Mr. J. W illia m s at the D epartm ent of Geology, U niversity of Cape Tow n, is thanked fo r the photographic reproduction o f the sonographs.

17 R E F E R E N C E S B e l d e r s o n, R.H., K e n y o n, N.H., S t r i d e, A.H., S t u b b s, A.R. (972) : Sonographs o f the Sea Floor. Elsevier Scientific Publ. Co., Am sterdam. B e l d e r s o n, R.H., K e n y o n, N.H., St r i d e, A.H. (974) : Features of submarine volcanoes shown on ons ranee sonoeraphs. J. Geol. Soc. Lond., 30, pp B e r k s o n, F.M., C l a y, C.S. (973) : Transform ation of side-scan records to a linear display. nt. Hydrogr. Rev., (2), pp B r y a n t, R.S. (975) : Side Scan Sonar for Hydrography. nt. Hydrogr. Rev., 5 2 ( ), pp C h e s t e r m a n, W.D., C l y n ic k, P.R., S t r i d e, A.H. (958) : A n acoustic aid to sea bed survey. Acustica, 8, pp C h e s t e r m a n, W.D., St. Q u in t o n, J.M.P., C h a n, Y., M a t t h e w s, H.R. (967) : Acoustic surveys of the sea floor near Hong Kong. nt. Hydrogr. Rev., 4 4 ( ), pp C h e s t e r m a n, W.D., W o n g, H.K. (97) : Bottom sediment distribution in certain inshore waters of Hong Kong. nt. Hydrogr. Rev., 48 (2), pp C l a y, C.S., Ess, J., W e is m a n,. (964) : Lateral echo sounding of the ocean bottom on the Continental Rise. J. Geoph. Res., 69 (8), pp G e n t l e, R.. (973) : Sidescan sonar in m arine geological investigations o f the South A frican continental m argin. S. Afr. J. Sci., 69 (Dec.), pp H o p k in s, J.C. (972a) : A note on methods o f producing corrected side scan sonar displays. nt. Hydrogr. Rev., 49 (2), pp H o p k in s, J.C. (972Z>) : Tape recording of side scanning sonar signals. nt. Hydrogr. Rev., 4 9 ( ), pp K e l l a n d, N.C., H o p k in s, J.C. (972) : Mosaics from high resolution side scan sonar. Offshore Services, Oct. 972, pp K e l l a n d, N.C. (972) : Preparation of isom etric records from high resolution side scanning sonar data using the Bath U niversity playback equipm ent. UCS nternal Report 72/23. K e l l a n d, N.C. (973) : Assessment trials o f underwater acoustic triangulation equipment. nt. J. Nnut. Archaeol. Underwater Expl., 2 (), pp L a u g h t o n, A.S., R u s b y, J.S.M. (975) : Long-range sonar and photographic studies of the median valley in the FAM O U S area o f the M id-atlantic Ridge near 37 N. Deep-Sea Research, 22, pp L e e n h a r d t, O. (974) : Side scanning sonar - A theoretical study. nt. Hydrogr. Rev., 5 (), pp M c K in n e y, R.F., St u b b l e f i e l d, W.L., S w i f t, D.T.P. (974) : Large-scale current lineations on the central N ew Jersey shelf : investigations by side-scan sonar. Marine Geology, 7, pp

18 M u d ie, J.D., N o r m a r k, W.R., C r a y, E.J. (970) : Direct m apping o f the sea flo o r using side-scanning sonar and transponder navigation. Geol. Soc. Am. B ull, 8, pp N e w t o n, R.S., S e ib o l d, E., W e r n e r, F. (973) : Facies distribution patterns on the Spanish Sahara continental shelf mapped w ith side-scan sonar. M eteor" Forsch.-Ergebnisse, Reihe C, pp R u s b y, S. (970) : A long range side-scan sonar fo r use in the deep sea. nt. Hydrogr. Rev., 47 (2), pp S a n d e r s, J.E., C l a y, C.S. (968) : nvestigating the ocean bottom w ith sidescanning sonar. n M o r g a n, P a r k e r (Eds.) : Proc. 5th Symp. rem ote sensing of environm ent. A n n Arbor, Michigan, nst. Sci. and Techn., W illo w Run Lab., Univ. M ichigan. S a n d e r s, J.E., E m e r y, K.O., U c h u p i, E. (969) : M icrotopography of five small areas o f the continental shelf by side-scanning sonar. Geol. Soc. Am. Bull., 80, pp S p ie s s, F.N., L o u g h r id g e, M.S., M cg e h e e, M.S., B o e g e m a n, D.E. (969) : A n acoustic transponder system. nt. Hydrogr. Rev., 46 (2), pp S p o t t i s w o o d e, J., D o b s o n, M. (975) : A fibre optic recording oscilloscope fo r the display o f side-scan sonar and seismic data. Marine Geology, 8 (5 ), pp. M73-M85. T u c k e r, M.J., St u b b s, A.R. (96) : A narrow-beam echo-ranger for fishery and geological investigations. Brit. J. Appl. Phys., 2, pp T u c k e r, M.J. (966) : Sideways look in g sonar fo r m arine geology. Marine Technology, 2 (9), pp Geo- T y r r e l l, W.A. (969) : U nderw ater acoustic positioning - Principles and problems. nt. Hydrogr. Rev., 4 6 (2 ), pp W e r n e r, F. (975) : Principios de interpretaciôn del sonar lateral y ejem plos de aplicaciôn en la plataform a continental Argentina. (n press, A rgen tin a). W e r n e r, F., N e w t o n, R.S. (975) : T h e pattern of large-scale bed form s in the Langeland Belt (Baltic Sea). Marine Geology, 9 (), pp W o n g, U.K., C h e s t e r m a n, W.D., B r o m h a l l, J.D. (970) : Comparative sidescan sonar and photographic survey of a coral bank. nt. Hydrogr. R e v., 47 (2), pp W o o d, A.B., S m it h, F.D., M cg e a c h y, J.A. (935) : A m agnetostriction echo depth recorder. T..E.E., 76, pp

19 true distan ce i i range distan ce 40 5 FSH HEGHT ABOVE SEA-BED: 0 m RANGES RANGES , r n ,, t l l r ! 2 300! 400 r i l l true d istan ce 9 ^P ^P ^P L J l l range distance FSH HEGHT ABOVE SEA-BED 20 m u - r - J T - J h r i i i i i i i i j i u ^, i i i lr _ l i T i i l i jj i. i ( l., L J L J L l J L -----L T i r r r n r i i i il i h - h f i i i i i i i i i tw *. l), -0. Nomograms for conversion " raunc disiaiiccs into true distances over ground (Seale /). Mo

20 RANGES RANGES 00 true d istan ce range distance FSH HEGHT ABOVE SEA-BED = 40 m i 0 h 0 20 l , 0C 200 i l i il l l l l 200! l ,, c'n!

21 true distan ce range distan ce FSH HEGHT ABOVE SEA-BED : m RANGES RANGES i s'o l l l l l 00 l i i " l l ~----- r S 00 l i 200 i i i i V L i 75 2 H H 0 00, r - l 400 i i! true distance range distan ce FSH HEGHT ABOVE SEA- BED = 60 m i 0 20 il P, f l f, T, , 5 200,l ( r b il,

22 RANGES RANGES 00 true distan ce range distan ce FSH HEGHT ABOVE SEA-BED 80 m i l SO f , l llll M ) l l l l l l l l. i 3 4

23 range d istan ce FSH HEGHT ABOVE SEA-BED = 90m RANGES RANGES true ' L - J " Ul 'l ' , ,----- U U.Jp l.l j l! L ,, L J ' t. i i i r * ! - l ' l ' * \ *...] T " l!!!!! distan ce range distan ce FSH HEGHT ABOVE SEA-BED = 00m -i r HH '.T l l l i ,------,------, f in i i i, i i i i, i i i, i i i, i i l i i,i T! " f **~ \ 'l '» *» ' h l 0 3 4

24 APPENDX Fie».. The continuous pattern of parallel dark dashes (a) are caused by a seismic instrument (boomer) run simultaneously with the side-scan sonar. Fig. 2. A typical interference pattern is caused by ultrasonic waves generated by a passing ship.

25 F ig. 3. The sharply outlined, dark patches (a) are caused by shoals of fish. The light and dark sea-bed features (b = sand, c = gravel) are due to material reflectors and not topography, as indicated by the smooth depth profile At lower right (d) isolated rock outcrops produce topographic reflectors associated with acoustic shadows. F ig. 4. Dense particle suspension in the water column (a) forms a diffuse reflector which partially or totally obliterates the sea-bed. Where suspension is densest (b) it results in an acoustic shadow. F ig. 5. The turbulent wake of a small boat forms a reflecting trail (a). The pattern on the sea-bed is mainly due to material reflectors. A low calcrete abrasion platform (b) is partially covered by sand (e).

26 Fie,. Hi. Material reflectors are here weakly emphasized by low topography (a = calcrete, b = sand). Very subtle changes in topography are often clearly recorded if they form a continuous linear pattern such as the traces of a suction dredger on the left fc). Fig. 7. Similar materials are here strongly emphasized by high topography and structural differences (a = granite, b = shale, c = contact zone). Note the artificial slope in sea-bed caused by a change in ship speed which is indicated by a shift in the seasurfaee line (d) and a corresponding shift in the opposite direction by the depth profile (e). F i g. 8. The strike direction of bedded rock is emphasized by corresponding acoustic shadows (a). Between outcrops extensive areas are covered by gravels displaying wave ripple marks (b). The wavelength is about.5 m. Note the bifurcation ot individual ripples.

27 F Hi. 9. The dark reflectors are produced by low outcrops of bedded rock (a) which is partially covered by irregular algal reef biohernis lb;. Both are partially blanketed by sand (c), indicating the reefs to be relict features. (Continental shelf off Natal, South Africa F Kl. 2). A fault associated with a dyke (a) truncates a bedrock bench (b). This feature on the African east coast shelf suggests that the bedrock may be a member of the Karroo System. Fir,. 2. The east coast shelf of Southern Africa is characterized by large areas of underwater dunes, which are recorded as topographic reflectors (a). These dunes, moving south with the Moçambique Current, are up to (i m high. Climbing megaripples (b) indicate fluctuations in the current velocities which result in smaller bed forms.

28 Fir D ark re fle c to rs are h ere produ ced by m a te r ia l (a) as w e ll as to p o g ra p h y (b) L a rg e sand w a v e s b rea k up in the tu rb u len t flo w p rod u ce d by the rou gh bed rock surface F ig. 23. M a teria l re fle c to rs (a) and to p o g ra p h ic re fle c to rs (b) fo r m an in te rferen ce p a tte rn du e to d iffe re n t strik e d irec tio n s o f b e d d ed rock and lo w sand w a v es. F ig. 24. S an d rib b o n s w ith m e g a rip p le s (a) o v e rlie g ra v e l (b ). T h e a rra n ge m en t and stru ctu re o f the sand rib b on s are in d ic a tiv e o f stron g n e a r-b o tto m cu rren ts o f ca. 2.5 kn ots (.,} m /sec).