Low Wave Velocity Research Articles

Periodic finite element/boundary element modeling of capacitive micromachined ultrasonic transducers

The possibility to excite and detect acoustic waves in fluids using capacitive transducers built on silicon using surface micromachining offers attractive opportunities in the manufacturing of high quality low cost imaging probes. As in the case of standard probe transducers, simulation codes are required to accurately design such devices. The periodic structures extensively used for these capacitive transducers has to be accounted for. In this work, a two-dimensional finite element analysis of capacitive micromachined ultrasonic transducers (cMUT) is proposed, taking into account periodicity and radiation in fluids. The convergence of the calculation is verified using different computation approaches. It is then shown that the periodic computations provide a rapid and precise analysis of the cMUT compared to non periodic calculations. The mutual displacements are deduced from the periodic harmonic calculation, providing an efficient estimation of cross-talk phenomena arising for cMUT radiating in water. The capability of cMUT operating under such conditions to generate a low velocity wave guided at the fluid/silicon interface is theoretically pointed out.

The characteristics of photon and phonon standing waves in a periodic medium

The main characteristics of the spatial variation of the photon and phonon wave fields atthe band gap boundaries are analysed for a one-dimensional medium with periodic opticalor acoustic parameters. The derivations are based on symmetry considerations and onanalytical results derived from the basic differential equation for the wave field. A simplerelation is derived between the band gap width and the derivative of the field intensity atthe interface between the regions of high and low wave velocity. The general fieldcharacteristics are derived for some examples. Using the analysis a remarkable asymmetricbehaviour of the wave absorption near the Brillouin zone boundaries can be explained in astraightforward way.

Comparison of Teleseismically and Hydroacoustically Derived Earthquake Locations along the North-central Mid-Atlantic Ridge and Equatorial East Pacific Rise

This letter examines earthquake locations derived using independent teleseismic and hydroacoustic data sets for events along the north-central (15°-35°N) Mid-Atlantic Ridge (MAR) and equatorial (10°S-10°N) East Pacific Rise (EPR). It represents the first large-scale comparison of such data, providing ground-truth information for the teleseismic locations and insight into the process of T -wave generation. An assessment of location accuracy and catalog completeness in these remote ocean settings is critical for explosion monitoring efforts associated with the Comprehensive Test-Ban Treaty (CTBT), as well as seismotectonic studies at mid-ocean ridges (MOR's). During the last decade, the value of hydroacoustic studies in monitoring MOR seismicity has been demonstrated through work using the U.S. Navy's SOund SUrveillance System (SOSUS) ( e.g., Fox et al., 1994) and arrays of moored autonomous underwater hydrophones (AUH's) ( e.g., Fox et al., 2001; Smith et al., 2002). These studies utilize seismically generated tertiary (7) waves that propagate within the ocean's SOund Fixing And Ranging (SOFAR) channel (Tolstoy and Ewing, 1950). In locating submarine earthquakes, the main advantages of the hydroacoustic method stem from the efficiency of this low-velocity wave guide, which allows for the detection of much smaller events at longer ranges relative to that possible with waves that travel through the solid earth, as well as from the existence of a well defined ocean sound-speed model ( e.g., Teague et al., 1990). Although T waves may be produced within regions of shallow sloping bathymetry near oceanic islands, continental shelves, and subduction zones ( e.g., Shurbet and Ewing, 1957; Johnson and Norris, 1968; Talandier and Okal, 1998), here we examine their generation in association with shallow hypocenter events within a deeper-ocean or abyssal environment ( e.g., Johnson, et al., 1968; Fox et al., 1994). In this setting, the scattering of energy from a rough seafloor has emerged as the …

Observation of stimulated electron-acoustic-wave scattering.

A diffraction-limited laser interacts with a plasma whose conditions are uniform on the scale of the focused laser spot. Two distinct, narrow waves are observed in the backscattered spectrum with phase velocities of v(phi)/v(e) = 1.4+/-0.08 and 4.2+/-0.1, where v(e) is the electron thermal speed. The high-velocity wave is ordinary stimulated Raman scattering (SRS) from a Langmuir wave. The low-velocity wave corresponds to stimulated scattering from an electron-acoustic wave (SEAS), and implies strong electron trapping. Previous SRS data from low-density plasmas are reinterpreted in terms of SEAS.

Lateral variations of the upper mantle structure in Eurasia from group velocities of surface waves

Group velocities of both Rayleigh and Love waves in the 20–70 s period range along paths across Asia and the Eastern Europe are used to obtain group velocity maps of the region. For this purpose, we developed a method for tomographic inversion on spherical surface, which is practically free of a priori assumptions. Checker-board test shows that velocity anomalies of 10°×10° size can be resolved. The ‘local’ dispersion curves are inverted to velocity-depth curves up to 160 km depth along three profiles: (1) Tibet, the South China Platform and the southern part of the Japan Sea; (2) the West Siberian Platform, the Tienshan and the Pamir; (3) the East European Platform and the Black Sea Basin. Lateral S-wave velocity variations are presented at the depths of 60, 90 and 120 km. A low velocity zone in asthenosphere under Japan Sea is extended under the South China and the North China–Korean Platforms. The upper mantle structure of the Russian (East European) and the West Siberian Platforms is different: a low-velocity zone at 120–140 km depths is observed under the West Siberian Platform, whereas the velocity under the Russian Platform increases monotonously with depth. A low velocity layer appears under the Black Sea indicating an oceanic structure of the Basin. Low surface wave velocities in Tibet are caused not only by a thick crust, but also by low S-wave velocities in the upper mantle. The Precambrian West Siberian, North Indian and North China–Korean Platforms are characterized by high velocities at the depth of 120 km unlike the younger tectonic units.

Effect of support conditions on seismic response of cable-stayed bridges

The seismic response of a cable-stayed bridge is the result of the interaction between the input earthquake characteristics and the bridge properties like mass, stiffness, and support conditions. The support type used for the deck at the towers of cable-stayed bridges is one of the important parameters that greatly influences the seismic response of the bridge. This is demonstrated in the present paper by study of the change in modal characteristics and seismic response of two cable-stayed bridges for different types of the support conditions. The first bridge has concrete towers and a steel deck with a 371 m main span, and the second bridge also has concrete towers, but a composite deck with a 465 m main span. The study shows that elastic supports for the deck at the towers give very low deck displacements and deck bending moments compared to pinned or fixed connections. Roller supports produce low tower base shear and base moment but not necessarily a low response for the bridge as a whole. Roller supports also cause the bridge to have very low first longitudinal direction mode frequency; this indicates that the bridge is very flexible in the longitudinal direction. A study of the effect of variation of the longitudinal stiffness of the elastic deck supports shows that it strongly influences the first few longitudinal direction modes, but not the symmetric or higher modes. The study also shows that the seismic response of the bridge is highly influenced by the variation of the support stiffness for some earthquakes like 1940 El Centro earthquake, but to a lesser degree for others. Wave propagation effects were also studied for different support conditions of the bridge. It is observed that the seismic response shows a higher degree of variation with the changing wave velocities when the bridge deck has pinned or fixed supports, and that the change in the bridge response with the wave velocities is not very pronounced when the bridge deck has elastic or roller supports. For the first bridge, the critical responses were observed at very low wave velocities, and for the second bridge, the uniform motion case caused the critical response of the bridge at many locations.Key words: cable-stayed bridges, support conditions, seismic response, wave propagations.

Effect of support conditions on seismic response of cable-stayed bridges

The seismic response of a cable-stayed bridge is the result of the interaction between the input earthquake characteristics and the bridge properties like mass, stiffness, and support conditions. The support type used for the deck at the towers of cable-stayed bridges is one of the important parameters that greatly influences the seismic response of the bridge. This is demonstrated in the present paper by study of the change in modal characteristics and seismic response of two cable-stayed bridges for different types of the support conditions. The first bridge has concrete towers and a steel deck with a 371 m main span, and the second bridge also has concrete towers, but a composite deck with a 465 m main span. The study shows that elastic supports for the deck at the towers give very low deck displacements and deck bending moments compared to pinned or fixed connections. Roller supports produce low tower base shear and base moment but not necessarily a low response for the bridge as a whole. Roller supports also cause the bridge to have very low first longitudinal direction mode frequency; this indicates that the bridge is very flexible in the longitudinal direction. A study of the effect of variation of the longitudinal stiffness of the elastic deck supports shows that it strongly influences the first few longitudinal direction modes, but not the symmetric or higher modes. The study also shows that the seismic response of the bridge is highly influenced by the variation of the support stiffness for some earthquakes like 1940 El Centro earthquake, but to a lesser degree for others. Wave propagation effects were also studied for different support conditions of the bridge. It is observed that the seismic response shows a higher degree of variation with the changing wave velocities when the bridge deck has pinned or fixed supports, and that the change in the bridge response with the wave velocities is not very pronounced when the bridge deck has elastic or roller supports. For the first bridge, the critical responses were observed at very low wave velocities, and for the second bridge, the uniform motion case caused the critical response of the bridge at many locations.Key words: cable-stayed bridges, support conditions, seismic response, wave propagations.

<i>E</i>-region mid-latitude decametre irregularities observed at four radar frequencies. Experiment and first results

Abstract. In this paper an experiment designed for multifrequency azimuthal Doppler-spectrum investigations of decametre-scale plasma irregularities in the mid-latitude E region is introduced and some preliminary results are presented. The observations were made with the high-frequency Valensole radar in the south of France. The radar operated in a multifrequency mode that allowed simultaneous measurements of Doppler spectra at the four frequencies of 9.23, 11.03, 12.71 and 16.09 MHz, which correspond to scatter from field-aligned irregularities with wavelengths of 16.2, 13.6, 11.8 and 9.3 m, respectively. In addition, a digital ionosonde was operating beneath a small part of the radar viewing region. The data show that lower-frequency echoes are stronger, more frequent and more spatially extended than higher-frequency ones, in general agreement with theory and rocket measurements. On the other hand, the preliminary analysis shows no pronounced differences of the Doppler spectrum with radar frequency. Some trends in the spectral moments do exist, however, which can be understood as the result of temporal and/or spatial mixing of backscatter from directly generated (primary) plasma waves by the gradient drift instability and secondary low-velocity waves. Finally, a close relation of mid-latitude coherent backscatter with patchy sporadic E-layers is present in the data, but the details of this relation remain unresolved.

Vibrations of Bridge Structure under Kinematic Wave Excitations

Random vibrations of a bridge structure under propagating, seismic excitations are considered and joint effects of pseudostatic and dynamic vibrations are studied. Respective matrix equations of motion are derived. Detailed numerical analysis of a four-span bridge modeled as a three-dimensional (3D) frame under kinematic wave excitations with an oblique angle of propagation is carried out. Key parameters of seismic loads are identified and their effect on both displacement and force response is analyzed. The root-mean-square response is normalized with respect to uniform excitations. The normalized displacement response stays between 0 and 1 while the normalized force responses may exceed 1, indicating additional spatial effects. These effects were identified to result from pseudostatic response, mostly for low apparent wave velocity with respect to support distances. A sensitivity, first passage, analysis is carried out to study the joint effect of propagation velocity and angle on the response.

Fault-zone guided waves from explosions in the San Andreas fault at Parkfield and Cienega Valley, California

AbstractFault-zone guided waves were successfully excited by near-surface explosions in the San Andreas fault zone both at Parkfield and Cienega Valley, central California. The guided waves were observed on linear, three-component seismic arrays deployed across the fault trace. These waves were not excited by explosions located outside the fault zone. The amplitude spectra of guided waves show a maximum peak at 2 Hz at Parkfield and 3 Hz at Cienega Valley. The guided wave amplitude decays sharply with observation distance from the fault trace. The explosion-excited fault-zone guided waves are similar to those generated by earthquakes at Parkfield but have lower frequencies and travel more slowly. These observations suggest that the fault-zone wave guide has lower seismic velocities as it approaches the surface at Parkfield. We have modeled the waveforms as S waves trapped in a low-velocity wave guide sandwiched between high-velocity wall rocks, resulting in Love-type fault-zone guided waves. While the results are nonunique, the Parkfield data are adequately fit by a shallow wave guide 170 m wide with an S velocity 0.85 km/sec and an apparent Q ∼ 30 to 40. At Cienega Valley, the fault-zone wave guide appears to be about 120 m wide with an S velocity 0.7 km/sec and a Q ∼ 30.

Spectral Relative Motion of Two Structures due to Seismic Travel Waves

This paper proposes a spectral difference method to estimate the maximum relative displacement between adjacent elastic structures subjected to seismic traveling waves. The method, based on random vibration theory, seeks the cross-correlation between the responses of adjacent structures subjected to differential earthquake excitation with time lag. The results of computer simulation with 15 earthquake records show a good agreement between the proposed spectral analysis and time-history analysis. The relations between the maximum relative displacement of adjacent structures and their vibration periods, damping, as well as the time lag of earthquake excitations are clarified using the closed form solution obtained. Time lag of the excitations among adjacent structures alters the correlation and phase of vibration of the adjacent structures. The cross-correlation coefficient explained herein captures this aspect. Its magnitude under nonzero time lag can be very different from that under zero time lag. In some instances, these coefficients can have different signs. This leads to a significant change in the peak relative displacement of adjacent structures. These are especially true when the structures are subjected to surface waves; they rest on soft soils where low wave velocities develop, and their supports are remotely located.

Regional seismic phases across the Ligurian Sea: Lg blockage and oceanic propagation

The Ligurian Sea (West Mediterranean) is an example of a narrow oceanic basin which stops Lg wave propagation. Locating the exact point of Lg extinction was one of the goals of the SISBALIG II seismological experiment, which deployed both inland stations in Provence and Corsica and Ocean Bottom Seismometers (OBS) in the Ligurian Sea. A precise analysis of all available regional seismological data using energy diagrams reveals the role of the Provence margin in the Lg phase extinction for seismic events occurring in the southwestern Alps. The extinction is not a progressive phenomenon along oceanic travel path but occurs in a narrow zone of about 20 km width in the vicinity of the Provence margin. The extinction is accompanied by the generation of two types of diffracted waves: the mantle phase Sn ( V s > 3.5 km s −1), and very low velocity S waves appearing both near the Provence margin and in the middle of the Ligurian basin. We performed numerical simulation of SH wave propagation in two-dimensional (2D) media to investigate the mechanisms of Lg extinction. The simulations confirm that simple crustal thinning at the continent-ocean boundary alone cannot cause the observed extinction of Lg waves, although the change in Moho depth does induce the diffraction of a part of the incident Lg wavetrain into mantle waves, giving rise to the Sn phase on the other side of the basin. The existence of soft oceanic sediments is a very important factor in explaining the extinction, as a large part of the Lg energy is transferred into slow S waves trapped in the sediments. Moreover, comparison of synthetic seismograms and observed data shows that the observed two packets of low-velocity waves are an indication of the asymmetry of the Ligurian basin, with the existence of two shallow basins with very low S velocity (1.0 km s −1) near the Provence margin and in the middle of the Ligurian Sea. However, the attenuation of Lg waves in the vicinity of the Provence margin is more abrupt in the data than in the simulations. This discrepancy could result from the complex geometry of the continent-ocean transition zone and its possibly very diffractive nature, two factors that could add to Lg extinction at the margin and are impossible to take into account in models.

Numerical modeling of fault-zone trapped waves: Acoustic case

Abstract We have obtained numerical results for an acoustic wave propagating in a 3D medium with mixed boundary conditions. The specific model involves a vertical low-velocity fault zone of varying thickness embedded in an otherwise homogeneous half-space. At the fault-zone boundary, the stress and displacement are continuous; on the free surface, the stress is zero. A pseudo-spectrum method is employed to achieve sufficient resolution with reasonable computation time on a CRAY. Results show the development of guided waves trapped in the fault zone. These guided waves display large amplitudes and lengthening waveforms; they propagate at lower velocity with amplitudes that drop off mainly due to energy leakage out of the fault zone. As the width of the fault zone varies, the wave energy tends to funnel into the new low-velocity wave guide. At large distance, these guided waves become the dominant arrivals on seismograms. The waveforms are useful to recover the geometry and properties of the fault zone.

Detonation of liquid explosive mixtures in powdered media

The propagation of detonation in liquid explosive mixtures (LEM) (nitrobenzene solutions in tetranitromethane) in inert and active powdered media of normal loose packing has been experimentally studied. It is shown that for small distances the general wave process is determined only by the propagation of detonation in LEM through narrow channels formed by separate particles of the powdered material independently of its acoustic stiffness. As the distance increases, in heterogeneous explosive mixtures with HMX a comparatively low-velocity wave smoothly changes to stationary detonation of the composition as a whole.

Gutenberg velocity section of the earth and its possible geomechanical explanation. I. Zonal disintegration and the hierarchical series of geoblocks

Introduction. In the last ten years throughout the whole world there has been a steady tendency towards mining economic minerals at greater depths which is accompanied both by deterioration in the mining and technical conditions for working deposits and by active development of mine pressure. Of particular importance and significance in this aspect is the large-scale scientific and technical problem of creating safe technologies for working deposits of economic minerals at considerable depths, including unmanned working. It is generally accepted that scientific bases of mining technology are a fundamental knowledge of the geomechanical and physical processes in blocks of Earth around underground workings which occur due to disturbance of the original condition of rock masses by the cavities formed. As numerous full-scale experiments show, the block structure of rock masses, geotectonics, and the seismicity of regions have a considerable effect on the nature and features of the geomechanical and physical processes around underground mine workings. In the light of past achievements in geomechanics, geotectonics, and seismology some connections for these disciplines are clearly developed. Among key points in particular is the discovery of zonal disintegration of rocks around underground workings [1, 2], the variable sign of the reaction of rocks to dynamic effects [3, 4], and also a number of other physical factors [5-9] pointing to the particularly nonlinear behavior of rocks. Mainly these are the existence of an "invariant" ratio of the linear dimensions of rock blocks to the amount of crack opening in the structural hierarchy of rock masses/z~(6) [8],* the process of self-blocking of "composite" materials leading to formation of a cellular structure in the form of a passive core and an active supporting shell [5], and also detection of a new type of low-velocity waves whose elementary carriers are geoblocks of a different hierarchical level and exhibiting rapid dispersal [6, 7]. It is possible to show that there is a profound interconnection between all of these effects which is caused by the existence of the phenomenon of zonal disintegration of rock around underground workings [1] and the alternating sign of rock reactions to dynamic effects [4]. Apart from a desire to illustrate, although partially, this intercormection, a stimulus for this work is the requirement for attributing the curious, in our view, results of comparing the effects which are obtained on the basis of these two discoveries with a number of basic factors from the field of Earth physics and geotectonics recently accumulated by geologists and geophysicists, and which have not so far found any satisfactory explanation from a physical viewpoint. As geological and geophysical factors the authors have in mind mainly the following: 1) the structure of the Gutenberg velocity section of Earth for longitudinal and transverse waves; 2) subdivision of the geological history of Earth into five sections of approximately the same duration (1 billion years): Catarchean, Archean, Aphebian, Riphean, and Phaneroan; 3) separation of 19 tectomagnetic epochs in Earth starting from an age of 3.54 billion years and a final age of 0.093 billion years, with an average period of 170-200 million years; 4) paleoclimatic recurrence; 5) cyclic changes of the overall area of modern continent territories covered by seas;

Heat flow constraints on the South Pacific Superswell

The South Pacific superswell has been defined as a large region of anomalously shallow bathymetry, low Love wave velocities, and low effective elastic thicknesses relative to those predicted for its age. These phenomena have been interpreted as reflecting a combination of lithospheric reheating and thinning, and dynamic uplift due to mantle flow. We use heat flow data to better constrain the thermal structure of this region and examine the predictions of various possible models. The average heat flow for the superswell region does not differ significantly from that for lithosphere of similar ages elsewhere on the Pacific plate. Given their uncertainties, the heat flow data imply that thermal lithospheric thickness exceeds 60 km, but cannot discriminate between greater thicknesses. The lack of observed high heat flow appears not to be explained by biases due to water circulation in the thin sediment cover, since the superswell heat flow is not higher than for sites elsewhere with similar sedimentary environments. The Darwin Rise has been proposed as a fossil superswell in the Cretaceous, on the basis of the many similar characteristics to the South Pacific superswell. We find that the Darwin Rise heat flow values do not exceed those for similar ages elsewhere in the Pacific and Atlantic. This observation suggests that any thermal effects associated with the formation of the Darwin Rise are no longer present, and it is consistent with the idea of a fossil superswell. The surface heat flow data thus provide no evidence that the temperatures in the uppermost portion of the lithosphere; are significantly higher in the entire superswell region than in other areas of comparable age. This observation is intriguing given the suggestion that the thin effective elastic thicknesses inferred from seamount loading may reflect reheating of the lithosphere. Models in which the plate thicknesses and/or the basal temperatures are increased to yield temperatures high enough to explain the low effective elastic thicknesses predict surface heat flow much higher than observed. Reheating the lithosphere, as is proposed for hot spots, yields temperatures adequate to explain the effective elastic thicknesses only if reheating occurs at very shallow depths, and again implies a surface heat flow much greater than observed. Hence, unless shallow reheating is somehow localized beneath the seamounts, the thinner elastic thicknesses may reflect mechanical, rather than thermal, weakening of the lithosphere.

Wave propagation in laminar models of cracked media

Wave propagation in laminar models of cracked media is investigated. Particularly noteworthy is the low-velocity wave, which cannot be explained within the framework of a single elasticity theory.

Decrease of internal free calcium and human sperm movement

In order to elucidate the effects of calcium on the movement of human spermatozoa, studies were conducted using motile cells selected by swim-up migration at 37 degrees C in 5% CO2 in air in a synthetic BWW medium containing 1.7 x 10(-3) M CaCl2 or BWW without added calcium (BWW-Ca). Preliminary experiments have confirmed that the addition of EGTA (5 x 10(-3); 10(-2) M) to BWW medium decreased the intracellular calcium concentration ((Ca++)i) of spermatozoa, as measured in cells loaded with a fluorescent Ca++ indicator, Quin-2. Concomitant measurements of (Ca++)i and sperm movement (analysed by videomicrography at 200 f/s at room temperature) were carried out on Quin-2 loaded cells incubated in BWW-Ca medium plus EGTA (10(-5) M; 10(-4) M; 10(-3) M). Under these conditions a decrease in (Ca++)i was observed and associated with a decrease in mean amplitude of lateral head displacement (ALH). Analysis using an automatic analyser (Hamilton Thorn at 37 degrees C) confirmed these results: the percentage of spermatozoa swimming with ALH greater than or equal to 6 microns is decreased when the external free calcium in BWW-Ca is decreased by the addition of 10(-5) M, 10(-4) M, or 10(-3) M EGTA. Flagellar analysis of the sperm population characterized by ALH greater than or equal to 6 microns showed a large proximal curvature of the tail associated with a low propagation wave velocity and a low beat frequency as compared to the spermatozoa with ALH less than 6 microns with similar progressive velocities. These characteristics result in a high flagellar beat efficiency (in terms of head displacement per beat). The disappearance of this pattern of movement when intracellular calcium is lowered indicates that calcium plays a complex role in the relationship between curvature and wave propagation. The ability of spermatozoa to modulate their movement in response to an alteration in the intracellular calcium level confirms the role of calcium in controlling flagellar movement in intact cells.

Acoustic properties of transparent polysiloxanes

AbstractIt is well known that the glass transition temperature of polydimethylsiloxane elastomers lies close to −120°C. According to the time temperature superposition principle, we may state that these materials keep their caoutchoutic state when they are excited by a 10 MHz mechanical wave. This is an important point particularly for acoustooptic applications. Thus, in the range of 1–10 MHz, materials with good efficiency are missed; low acoustic wave velocity is needed to give the best conditions, which means to be in a caoutchoutic state for polymeric materials. With this work, we show that three polysiloxanes, polymerized under different conditions, do not fulfill this condition. Thus the shift of Tg with frequency is very large and these polymers reach their glassy state at relatively low frequency. The acoustic wave velocity is in good agreement with mechanical predictions provided by master curves of the Young modulus.

Seismic reflection results from an ice island over the Arctic continental Shelf

In 1982 an Arctic ice island, 3 km by 7 km and 41 m thick, broke away from the Ward-Hunt Ice Shelf on Ellesmere Island and began drifting southwest with the ocean currents. This island provided a stable platform in the polar pack ice, for the seismic reflection exploration of the Arctic Continental Shelf north of Axel Heiberg Island. A seismic reflection array, consisting of a 107 station, 3.53 km, main arm and a twelve station, 2.4 km, cross arm, was constructed on the island. A set of 6 s common midpoint (CMP) data was recorded from the thirteen shotpoints for each kilometre the island drifted. Each charge, 1 kg of 60% Geogel, was detonated 110 m beneath the ice surface. Over July and August, 1985, a total of 87 sets of shots were recorded. The raw record sections were dominated by very strong low-velocity waves and water layer reverberations. Up to fourteen distinct water bottom multiples are seen in both the raw shot gathers and the CMP-stacked sections. Median, ( f– k) and optimum array filters were used along with CMP stacking, deconvolution and common shot gather stacking to reduce the water-bound noise. These processes were tested on both simulation data and the real data. A P-wave travelling through the ice layer dominated most of the first arrivals. First break refraction analysis, where possible, revealed 600–800 m of locally varying shallow sedimentary rock (4–5 km s −1) overlying a higher velocity layer (5.5–6 km s −1). The CMP sections contain numerous short discontinuous reflections indicating a lack of simple structure in the high-velocity rock. The larger (107) fold of the common shot gather stack increased the signal to noise ratio and revealed an extensive, roughly horizontal reflection at a two-way time of 1.6 s. This corresponds to a depth of about 4.2 km and is interpreted as a reflection from the top of the Precambrian basement.