Sub-bottom profilers, which are usually single channel systems, are used for shallow seismic seabed profiling. There are a number of different systems which operate at different frequencies. The pinger , so called because of its high frequency acoustic pings, operate on a range of single frequencies between 3.5kHz and 7kHz. It can achieve seabed penetration from just a few meters to more than 50m depending on the sediment consolidation, and are capable of resolving soil layer stratigraphy to approximately 0.3m. The high frequency profilers are particularly useful for delineating shallow lithology features such as faults, gas accumulations and relict channels. A lower frequency system is the boomer. This instrument has a broader band acoustic source between 500Hz to 5kHz and typically can penetrate to between 30m and 100m with resolution of 0.3m to 1.0m and are excellent general-purpose tools. Even lower frequency is the sparker which is a very powerful instrument that can penetrate soils and rocks to 1,000m. The CHIRP sub-bottom profiler is a recent introduction to geophysical survey. Designed to replace the pingers and boomers, CHIRP systems operate around a central frequency that is swept electronically across a range of frequencies (i.e. a chirp.) between 3kHz to 40kHz and can improve resolution in suitable near-seabed sediments.
See also NOAA's summary table Summary view of single beam sounding technique (279 KB PDF).
NOTE: The content below is derived from Chapters 6 - 6.4 of Acoustic Techniques for Seabed Classification (2005) (11 MB PDF) by J D Penrose, P J W Siwabessy, A Gavrilov, I Parnum, L J Hamilton, A Bickers, B Brooke, D A Ryan and P Kennedy.
The geology of the seafloor and underlying strata represents key physical features of benthic habitats. Underwater acoustics have for many years been a fundamental tool for oceanography and marine geology because of the ability of these methods to determine physical properties of the seafloor, and to identify geological acoustic reflectors below the seafloor (McQuillin et al., 1984). In recent years, acoustic methods have also been used to measure small scale sedimentary structures and processes, with high temporal and spatial resolution, and they have been widely adopted by marine researchers because of their ability to rapidly and non-intrusively collect data (Davis et al., 2002, Walter et al., 2002, Kim et al., 2002). Although most acoustic sub-bottom profiling systems have been designed to acquire information about geological boundaries well below the seafloor, information relating to surficial and near-surficial sedimentary environments is also inherent in many of the commercially available systems (Davis et al., 2002). These data are valuable because seabed geomorphology can provide a good first approximation of different types of benthic environments and habitats (Kloser et al., 2001b).
Figure 1. Deployment of various shallow-water sub-bottom profiling systems. After Stoker et al. (1997).
Sub-bottom profiling systems comprise a sound source, either towed behind a vessel or firmly mounted to the hull, that produces an acoustic pulse of set frequency, power, and time duration. The acoustic pulses generated may be described as a ‘single-beam’ (see single beam sounders). The acoustic pulse travels through the water column (at a rate determined by water temperature, salinity and suspended material concentration), and penetrates the seafloor (Figure 1).
Some of the acoustic signal is reflected from the seafloor, whereas the remainder penetrates the seafloor and is reflected when it encounters boundaries between layers that have different acoustic impedance. The acoustic impedance of a material, Z, is dependant on the wet bulk density (of the sediment), ρ, and the compressional wave velocity, c, where:
The Rayleigh coefficient of reflection, R, is defined as the ratio of the reflected amplitude to the incident wave amplitude. For a normal incident acoustic wave this reflection coefficient is related to acoustic impedance by the relationship:
Subscripts 1 and 2 respectively refer to the impedance of the mediums above, and below, the reflecting interface. For reflection at the seabed, R will be positive when the acoustic impedance of the sediment is greater than that of the overlaying water (Davis et al., 2002, Walter et al., 2002).
The reflection of acoustic energy takes place at boundaries of differing acoustic impedance, and the reflection strength depends on the degree of impedance contrast. Typically, a portion of the incident energy is reflected from the sediment-water interface, whereas the remainder is transmitted deeper into the substrate (McQuillan et al., 1984, Stoker et al., 1997). The returning sound waves are recorded by an array of hydrophones (also usually towed further behind the vessel), or by a transducer/transceiver, depending on the type of system (Verbeek and McGee, 1995, McGee, 1995). The acoustic receiver resolves the various pulses of energy, with backscatter from shallower reflectors arriving first, forming a profile of information. The result is a continuous real-time displayed record of bathymetry (the first significant reflecting surface) and the boundaries between sub-bottom strata.
Several physical parameters of the acoustic signal emitted, such as output power, signal frequency, and pulse length affect the performance of the instrument and influence its usefulness in various marine environments. Increased output power allows greater penetration into the substrate, however, in the case of harder seabeds (for example gravels or highly compacted sands), or very shallow water, will result in multiple reflections and more noise in the data (McQuillan et al., 1984). The resolution of acoustic systems is defined as their ability to differentiate closely spaced objects, or resolve discrete echoes returning from closely spaced reflectors. In general, higher frequency broadband signals are more discriminating, although higher frequencies are preferentially filtered out in the environment. Higher frequency systems (2 to 20 kHz) produce high definition data of sediment layers immediately below the seafloor. These higher frequency signals have shorter wavelengths, and they are able to discriminate between layers that are close together (e.g. 10’s of cms). Lower frequency systems give greater substrate penetration, but at a lower resolution. Longer pulse length transmissions (or ‘pings’) yield more energy, and result in greater substrate penetration. On the other hand, longer pulse lengths decrease the receivers ability to discriminate between adjacent reflectors, thus decreasing the system resolution. However, the penetration depth depends on the hardness of the overlying layers and the presence of gas deposits, such as methane (Davis et al., 2002). The presence of sub-surface gas deposits can significantly degrade an acoustic signal because it may result in a negative R impedance value (see Equation 2 above). In addition to the frequency and bandwidth, other factors affecting system resolution are beam width (or area of seabed insonified), depth of water below the transmit/receive array, signal to noise ratio, and electronic signal processing.
The resolution obtained in a sub-bottom profile is related to the frequency of the acoustic source, higher frequencies providing greater detail. However the attenuation of sound and therefore bottom penetration is inversely related to frequency, necessitating a variety of sub-bottom profiling tools specific to different marine environments (Stoker et al., 1997). Single-beam acoustic reflection systems operating within the low kHz range are useful for high resolution assessment of the top 100 m of sedimentary material below the sea floor, with penetration depth inversely related to transmit frequency. The correlation between signal frequency and penetration is not linear – at frequencies below approximately 800 Hz the penetration increases dramatically (Figure 2).
Figure 2. Frequency, depth of penetration, and approximate system vertical resolution (R), with typical sonar system ranges (depicted by horizontal bars). After Stoker et al. (1997).
Sub-bottom profiling systems are limited by a narrow swath width (or small area of seabed insonified), so comprehensive coverage of the seafloor is time-consuming and expensive to obtain. As with other single-beam acoustic methods, the footprint is relatively small and dependent on depth. Acquisition lines are often spaced closely together or in transects across the area of interest (depending on the purpose and constraints of the survey), allowing sub-surface reflectors to be traced in three dimensions from line to line. Shallow water surveys allow faster shot rates and vessel speed, although speeds greater than 10 knots may result in interference due to high degrees of ship noise and water turbulence (Stoker et al., 1997).
In the years since the first crude sub-bottom profiling systems were developed (Knott and Hersey, 1956), a multitude of processes and techniques have been utilised to produce acoustic pulses, detect their returning echos, and subsequently process this information into a meaningful representation of acoustically reflective surfaces. As discussed above, limitations related to frequency, resolution and depth of penetration have led to a diversification in system design (Lean and Pratt, 1991). Many types of sub-bottom profilers are currently available, and there is a great range of variation between types of systems and also between various equipment manufacturers.
High frequency (> 2-3 kHz) boomer, pinger and chirper systems are tailored to give very detailed information about near surface features, with bottom penetration in the order of 100 m. Medium frequency (~1 kHz) profilers such as the sparker system can penetrate to a depth of approximately 500 m, and maintain relatively good resolution. Very low frequency and high energy systems (<1 kHz), such as airgun systems may penetrate a number of kilometres into the seabed, with a corresponding reduction in resolution (Stoker et al., 1997).
Table 1. System characteristics of various classes of sub-bottom profilers.
|System||Operating Frequency||Acoustic Source||Receive Array||Typical Resolution||Depth of Penetration||Source Mounting Style|
|Sparker||50Hz - 4 kHz||Electrical spark in water||Towed streamer||> 2 m||500 m||Towed|
|Chirper||Swept 1-10kHz||Swept frequency transducer (1-10kHz)||Transducer||~0.05 m||< 100 m||Hull mounted or towed|
|Boomer||300 Hz to 3 kHz||Boomer plate||Towed streamer||0.5 to 2 m||< 200 m||Towed sled|
|Pinger||Tuned between 2-12 kHz (eg 3.5kHz)||Combined piezo- transducer/ transceiver||Combined piezo-transducer/ transceiver||0.2 m||10 - 50 m||Hull mounted|
This review will focus on the high to medium frequency systems as these are most relevant for the study of modern and recent seabed geomorphology in shallow marine environments. Table 1 summarises some of the basic details about various classes of sub-bottom profiling systems.
High resolution, tuned-frequency profilers typically operate at frequencies that range between 1 to 30 kHz. This achieves relatively high depth resolution, however the bottom penetration is significantly less than that obtained by lower frequency systems. Signal penetration is further limited in coarse sediment or highly compacted sands, due to scattering (Damuth, 1975, Whitmore and Belton, 1997). Tuned-frequency profilers typically use the same transducer for both transmitting and receiving the signal. The source signal frequency is also highly consistent between pulses in order to provide better resolution and tracking of thin subsurface layers.
Chirper systems are so named due to their emission of a chirp sound (rather than a ping) for a single frequency unit. A chirp sonar is a wide-band, frequency modulated (FM) sub-bottom profiler that is capable of producing very high resolution profiles in soft sediments. The system generates an FM pulse from a resonant source that is phase and amplitude compensated, which helps to suppress noise (Schock et al., 1989, McGee 1995). These units obtain very good subsurface images due to their ability to sweep through a range of frequencies (by varying the amplitude and frequency of the emitted pulse in a predetermined pattern), usually between 1.5 to 11.5 kHz for shallow water applications, or as low as 0.4 to 8 kHz for deep seismic reflection (McGee 1995). A complex signal processing algorithm correlates returns, and estimates the attenuation of sub-bottom reflections by waveform matching with a theoretically attenuated waveform. Chirp sonars are typically able to achieve vertical resolutions down to ~ 5 cm, and can provide relatively artefact free sub-bottom profiles attenuating to 100 m depth (Schock et al., 1989). Longer chirp pulses can be used for deeper penetration. Noise is suppressed, and resolution is improved in chirp systems due to lower peak input power to the transducer, and transducer voltage is controlled to prevent source ‘ringing’ (decaying oscillations in the transmitted acoustic pulse). The chirp system transducer may be towed or hull mounted, and can operate in water as shallow as 30 cm (Schock et al., 1989), or can be used in the deep ocean if mounted within a towed vehicle (Parent and O’Brien 1993).
Single-beam acoustic sub-bottom profilers and bottom-detection units are often referred to as ‘pingers’. These systems are similar to the depth-sounding units used for navigation on most vessels, however are tuned to a lower frequency for penetration into the seabed (Luskin et al., 1954). Single-beam systems are described more fully in the single beam section. Pinger systems comprise a piezoelectrically resonated ceramic element within the transducer/transceiver, to produce (and receive) a controlled pulse length, narrow-frequency acoustic signal (McGee, 1995). Pinger-type profilers are specifically tuned to a particular frequency. For simple bottom detection in shallow waters this frequency may be as high as 200 kHz, however for significant water depth and substrate penetration frequencies of 12 or most commonly 3.5 kHz are employed (Damuth, 1980). Operation of pinger sub-bottom profilers at 3.5 kHz typically results in 10 - 50 metres of substrate penetration, at a resolution down to 0.2 metres, depending upon sediment type.
Many low frequency systems are regarded as low-resolution profilers – for example water and air guns, sparkers, sleeve exploders, bubble pursers, and boomers. In these systems, the energy source transmits a signal of broad spectral content, and requires separately towed hydrophone arrays for receiving the return signal.
The sparker is a relatively higher powered sound source, dependant on an electrical arc which momentarily vaporizes water between positive and negative electrodes (Trabant 1984). The collapsing bubbles produce a broad band (50 Hz - 4 kHz) omni-directional acoustic pulse. Sparkers typically yield better penetration, but poorer resolution than boomer systems, with depth of penetration up to 500 metres (Stoker et al., 1997), and vertical resolution usually greater than 2 metres (depending on energy settings). Sparker sources are commonly used in regions where compacted sands and other coarse semiconsolidated sediments are found. Some sparker seismic systems have been developed in which the electrical discharges take place at the focus of a paraboloidal reflecting surface, in order to obtain a downward-oriented, approximately plane acoustic wavefront. This system enables a greater substrate penetration (Gasperini et al., 1993).
Boomer sub-bottom profilers comprise an insulated metal plate adjacent to an electrical coil, typically mounted on a towed catamaran. This electro-mechanical transducer is known as a ‘boomer plate’. A powerful electrical pulse, generated by a shipboard power supply and capacitor banks discharges to the electrical coil, causing a magnetic field to explosively repel the metal plate. This energetic motion generates a broad band, high amplitude impulsive acoustic signal in the water column (Trabant 1984, McGee 1995). The frequency of the acoustic pulse is in the range 300 Hz to 3 kHz (or more) with the majority of the energy being directed vertically downward (Verbeek and McGee, 1995). Most boomer systems rely on a (potentially dangerous) high voltage power supply and capacitor banks for the generation of the high voltage electrical pulses required. More recent developments in efficient, low voltage boomer systems have circumvented some of the problems inherent with the high voltage systems (Davis et al., 2002). Resolution of boomer systems ranges from 0.5 to 2 m, and penetration typically ranges from 50 to 200 metres, depending on sediment type. A limitation of these systems is that the transmitted acoustic pulses do not have the repeatability necessary to provide high accuracy measurements of seafloor properties.
Other Sub-Bottom Profiling Systems
A large number of lower frequency sub-bottom profiling systems are currently in use, including low frequency sparkers, parametric echo-sounders, airgun/sleevegun systems (of various volumes), watergun systems, and multi-channel receive techniques. These systems are not treated here as they are more relevant for deeper water geological investigations (Stoker et al., 1997, Lean and Pratt, 1991, Verbeek and McGee, 1995). Some shallow-water work has been undertaken using single, small-volume airgun systems coupled with multi-channel receive arrays, achieving resolution approaching that of standard 3.5 kHz (pinger) sub-bottom profiles, although obtaining much greater substrate penetration (Lee et al., 2004).
The physical properties of the seabed (eg sediment bulk density, grainsize) can be approximated from analysis of the acoustic reflection response, with sediment depth dependant on the resolution of the profiling system. Much information can be gained about benthic habitats through the analysis of sub-surface reflectors, and the thickness of surficial sedimentary units. Analysis of both newly acquired (and the large volumes of previously archived) sub-bottom profiling information is a relatively under-utilised source of information about seafloor sedimentary environments. A number of studies have attempted to quantitatively determine or discriminate sedimentary features of the seafloor from profiling data (de Moustier and Matsumoto, 1993). These have involved the direct measurement of a pressure coefficient of the seafloor (Kim et al., 2002), statistical approaches utilising the entire acoustic echo signal, such as the Karhunen-Loeve transform (Milligan et al., 1978), and highly quantitative approaches such as the inversion of boomer acoustic impedance values (Davis et al., 2002), and use of swept-frequency chirp system impedance and attenuation coefficients (LeBlanc et al., 1992a, Kim et al., 2002).
Depositional and post-depositional processes play a major role in determining the nature and spatial distribution of seafloor sediments, integrating factors such as geological setting, sediment supply, oceanographic conditions, and sea-level change (Davis et al., 2002). Acoustic reflection systems can provide important insights into the physical character of the seabed and sub-surface that help in the interpretation of other remotely sensed benthic data (such as sidescan, single and multibeam echosounder data, video, satellite and airborne scanner imagery) and indicate the physical processes responsible for present-day form and distribution of benthic habitats. Sub-bottom reflectors provide information relevant to an assessment of present day benthic habitats at a range of scales. A basic measure provided is the depth to the bedrock reflector that underlies any seabed sediment, which indicates whether an area is a site of sediment accumulation or erosion. Also, former reefs now buried by sediment can be identified in sub-bottom profiles, providing a useful record of major environmental changes that have occurred (Figure 3).
Similar insights can be provided into the spatial distribution of benthic habitats, for example bedrock reflectors may extend to the surfaces and provide habitats that may significantly contrast with the surrounding seabed (Figure 3). These types of acoustic data can also allow the discrimination of relatively hard sedimentary bottoms from bedrock reefs, which may have similar echosounder characteristics or may not be visually discernible when covered by marine organisms. The dynamics of benthic environments can also be indicated by surface and preserved subsurface bedforms. For example, dune and sand-wave structures in sandy deposits indicate they are highly mobile, while planar bedforms in certain areas reflect significant sediment accumulation (Figure 3).
Figure 3. Boomer sub-bottom profiles of the seafloor around the Whitsunday Islands, Great Barrier Reef platform, Australia (after Heap, 2000). The system used was an EG & G TM Uniboom sounder, triggered every 0.5 s at 200 J, and towed 0.3 m below the surface 11 m behind an 8 element hydrophone array. (A) The reflectors reveal a range of recent, Holocene, and pre-Holocene features, showing an exposure of bedrock surrounded by recent sand accumulations. (B) Steeply NE dipping bedding structures (clinoforms) and surficial dune bedforms record the accumulation and present-day movement of sand into a depocentre.
Since the 1960’s, 3.5 kHz systems have been used to gain seafloor penetration of 100 m or greater, and Damuth (1975) developed a surface sediment classification system for 3.5 kHz echograms based on an earlier, lower penetration 12 kHz system data (Hollister 1967). The simple echo characteristics of the seafloor may be classified into three main categories, based on parameters that include the clarity and continuity of echoes, which show a qualitative relationship with seafloor morphology. These include distinct echoes, prolonged indistinct echoes, and hyperbolic indistinct echoes (Table 2).
Table 2. Description and examples of echo character types(after Damuth, 1980, Whitmore and Belton, 1997, Rollet et al., 2001).
|Distinct||IA||Sharp continuous with no sub-bottom reflectors|
|IB||Sharp continuous with numerous parallel sub-bottom reflectors|
|IC||Sharp continuous with non-conformable sub-bottom reflectors|
|Indistinct||Prolonged||IIA||Semi-prolonged with intermittent parallel sub-bottom reflectors|
|IIB||Prolonged with no sub-bottom reflectors|
|Hyperbolae||IIIA||Large, irregular hyperbolae with varying vertex elevation (>100m)|
|IIIB||Regular Single hyperbolae with varying vertices and conformable sub-bottom reflectors|
|IIIC||Regular overlapping hyperbolae with varying vertex elevation (<100 m)|
|IIID||Regular overlapping hyperbolae with vertices tangental to the seafloor|
|IIIE||Type IIID hyperbolae with intermittent zones of distinct (IB) echoes.|
|IIIF||Irregular Single hyperbolae with non-conformable sub-bottoms|
These parameters were further sub-divided based on presence or absence of a sub-bottom reflectors, and the hyperbolae characteristics which relate to the morphology of the seafloor. Different types of echoes form through the interaction between the seabed and the echo-pulse, and sediments affect the echo return depending on their density, layering structure, and topography (Flood, 1980).
To construct an echo-character map based on 3.5 kHz data, the bottom returns of all available survey lines across a region need to be examined to develop a classification system that is suitable and specific to that area. Importantly, the number and spacing of survey lines and the variability of echo types will determine whether it is reasonable to extrapolate between them, and thereby produce meaningful maps of the seabed (Damuth, 1980).
However, comprehensive ground-truthing information is required to validate these interpretations. These should include a large number of bottom samples, see section on fine scale classification, such as sediment grabs, cores, or dredges, from varied geographic and bathymetric locations as well as from each distinct facies type encountered (Whitmore and Belton, 1997). Although relatively qualitative, this method has been effectively used to classify both deep and shallow water sediments in numerous studies (Hollister 1967, Damuth, 1975, Damuth, 1980, Blum and Okamura, 1992, Whitmore and Belton, 1997, Rollet et al., 2001).
The remote classification of shallow marine sediments can also be carried out in a more quantitative manner with a high resolution chirp sonar. LeBlanc et al. (1992a) first used this equipment to characterise depositional environments using acoustically derived density, sediment compressibility, rigidity, and acoustic impedance parameters (based upon a function of the Rayleigh reflection coefficient, given in equation 2). From these measured values, models allow the prediction of sediment sound velocity, porosity, and wet bulk density (Figure 4).
This method relies on the construction of a database of quantitative reflection parameters for each depositional environment, necessitating significant ground truthing (e.g. sediment cores). Further work by LeBlanc et al. (1992b) focussed on the determination of mean sediment grainsize based upon modelling of acoustic attenuation in sediments. Based upon a large compilation of historical attenuation data, an empirical equation was developed to predict sediment type in real time from chirp sonar data. More specifically, a relationship between sediment type and ‘relaxation time’ is discussed, which equates to a measure of the time needed to change the density by application of a sudden pressure (Figure 5).
Further work into classification techniques for chirp sonar data has resulted in the development of techniques that utilise the entire acoustic echo signature, rather than just the reflection coefficient (Kim et al., 2002). These have involved cluster analyses of whole datasets (Milligan et al., 1978), and statistical properties of the bottom type based on time-frequency analysis (Andrieux et al., 1995). More recently, application of a similarity index to the uncorrelated principal components of the echo signature (which are derived from the Karhunen-Loeve transform) have been used to classify the seafloor (Kim et al., 2002). This index is based upon the first principal component of the reflected echo (see Kim et al., 2002), and has been shown to correlate with sediment facies based upon grain size, particle sorting, hardness, and homogeneity of the substrate.
Figure 4. Acoustic impedance as a function of both sediment porosity and wet bulk density (after LeBlanc et al., 1992a).
Figure 5. Sediment acoustic attenuation measurements, or relaxation time values plotted as a function of mean grain size in phi units (after LeBlanc et al., 1992b).