Dennis Lindwall

Marine Seismic Surveys with Vector Acoustic Sensors

Abstract : Vector acoustic data will allow accurate three-dimensional imaging of a complex environment while corresponding pressure hydrophone data will fail. Newly developed sensors make vector acoustic-based surveys practical. This concept is demonstrated with data from an acoustic water tank. Using a simple but novel imaging algorithm, all of the main structures in the water tank were correctly located. This imaging algorithm uses none of the existing imaging or inversion methods common in exploration seismology.

3D underwater imaging using vector acoustic sensors

Marine surveys that use vector acoustic sensors may allow for 3D imaging of underwater environments with a much smaller amount of data than current 3D hydrophone surveys. Newly developed sensors make vector-acoustic-based surveys practical. This concept is demonstrated with data from a three-axis accelerometer and a collocated hydrophone in an acoustic water tank using a short-pulse source and passive scattering targets. One algorithm rectifies the vector data with scalar pressure data and another maps the vector data into a 3D volume, showing several slices of the volume images. The imaging algorithm maps the scattered energy using the direction and traveltime independently for each source-receiver pair rather than using the phase coherence methods common in exploration seismology. Imaging a more complex and realistic marine environment requires vector wavefield decomposition techniques and other theoretical developments but may allow for 3D vector-acoustic seismic surveys using logistics similar to 2D surveys that use conventional hydrophones.

Detection and tracking of quiet signals in noisy environments with vector sensors

We analyze the utility of vector sensors to detect and track underwater acoustic signals in noisy environments. High ambient noise levels below 300 Hz are often dominated by a few loud discrete ships that produce a complicated and dynamic noise covariance structure. Horizontal arrays of omni-directional hydrophones improve detection by forming (planewave) beams that “listen” between loud azimuthal directions with little regard to changing noise fields. The inherent 3-D directionality of vector sensors offers the opportunity to exploit detailed noise covariance structure at the element level. We present simulation performance results for vector sensors in simple and realistic environments using particle filters that can adapt to changing acoustic field structures. We demonstrate the ability of vector sensors to characterize and mitigate the deleterious effects of noise sources. We also demonstrate the relative value of vector vs. omni-directional sensing (and processing) for single sensors and compact arrays.

Undersea noise characterization using vector sensors and adaptive particle filters

Acoustic noise in the ocean is spatially complex and dynamic. It results from a composite of moving discrete and distributed sources with narrow and broadband signatures from near and distant locations over complex propagation paths. Whether the goal of noise characterization is to improve understanding of the sources and environmental aspects of noise or to find and describe weak signals, it can be better achieved with vector acoustic data and vector-based analysis rather than with a pure scalar-pressure approach. We will show analytically and with realistic acoustic noise simulations that the vector characterization of noise parameters such as directional distribution, frequency content, and time variation is more accurate and can be done with simpler instrumentation than what is commonly done with pressure data. We employ adaptive particle filters to model and estimate the temporal aspects of the noise fields. The filters state and observation vectors consist of signal directions and magnitudes, respe...

Optical bandwidth and focusing dynamics effects on an underwater laser acoustic source

Both femtosecond and nanosecond laser pulses can produce nonlinear effects in water, including filamentation and laser-induced breakdown resulting in acoustic generation. We examine the effects of GVD, varying wavelength, bandwidth, energy, and focusing configurations.

Observed anisotropy in low frequency acoustic signal loss at the sea floor.

Anisotropy of 20%–30% in the low‐frequency (10–120 Hz) acoustic bottom loss of marine sediments 1200‐m deep off the coast of Oregon is reported here. Two nearly perpendicular acoustic profiles were acquired with air guns towed at the sea surface, crossing at a point directly over a four‐component ocean bottom seismometer (OBS) on the seafloor. The area is a convergent margin, with a regional principal stress field perpendicular to broad bathymetric ridges and the shoreline, and numerous faults running parallel to the bathymetry. Amplitudes of sediment reflections are consistently 20%–30% less on the profile parallel to the bathymetry than on the profile perpendicular to the bathymetry. This is the opposite of what would be expected if the dominant attenuation mechanism were scattering from the fault planes. The anisotropy exists at angles for which acoustic energy penetrates sediments only a few hundred meters from the crossing point, where layering in both directions is laterally consistent. The direct arrival amplitudes, from the source to the OBS, are identical on both transects, suggesting that neither the acquisition technique nor the water column is responsible for the anisotropy. Possible explanations include seafloor bed forms and wavelength‐scale roughness at the buried sediment interfaces.

Underwater acoustic generation with narrow and broadband lasers

Underwater laser acoustic sources, generated by intense pulsed lasers on above‐water and underwater platforms, are under investigation. In a novel configuration, a tailored intense broadband laser pulse can be designed to propagate many meters underwater and compress at a predetermined remote location. Controlled compression of these optical pulses is governed by a combination of optical group velocity dispersion (GVD) and nonlinear Kerr self‐focusing, resulting in photoionization, localized heating, and shock generation. Recent and ongoing experiments include near‐field acoustic source characterization using lens‐focused pulses of a broadband 400 nm Ti:sapphire laser, as well as 1064 nm and 532 nm narrowband YAG laser pulses. Also, the nonlinear optical Kerr index of water at 400 nm was precisely measured. Acoustic source characterization includes measurements of photoacoustic energy conversion efficiency, acoustic power spectrum, and directivity. Experimental results will be presented, and laser sources...

3D Underwater Imaging Using Vector Acoustics

Vector acoustic data will allow accurate 3-D imaging of an underwater environment while corresponding pressure hydrophone data using the same acquisition geometry can only produce 2-D images. Newly developed sensors make vector acoustic based surveys practical. This concept is demonstrated with data from an acoustic water tank. Using a simple but novel imaging algorithm, all of the main structures in the water tank were correctly located. This imaging algorithm uses none of the existing imaging or inversion methods common in exploration seismology.

High‐resolution seismic investigations of shallow flow site in the Gulf of Mexico

Results from a very recent (Feb 2005) deep-tow multichannel seismic investigation of a shallow flow site near the MARS platform were taken along a track line where, previously, industry standard high-resolution seismic data had been acquired. The increased resolution obtained with data from the deep-tow seismic instrument (Gettrust, Ross and Rowe (1991) Wood and Gettrust (2001)) reveal details of the geologic structure not resolved using the conventional technique. We observe structural differences between the north end of a high-resolution seismic track line and the southern end (near the MARS platform). We compare these differences together with other deep-tow data taken to the west of the shallow flow site (which we assume represents an area where no shallow-flow potential exists). We conclude that there are features in the deep-tow seismic data that are consistent with fluid flow or migration in the area where shallow-flow has been observed. However, it is not yet clear from our preliminary analyses if the deep-tow MCS technique can be used to predict areas where shallow-flow is likely.

Fast Thinlayer Inversion (Theory)

Summary Conventional analysis of seismic data for layered medium properties breaks down when the layers are thinner than the seismic wavelength. We describe a fast and accurate method to calculate the reflection response of a thin layer and then use this method to invert seismic data for the thickness and physical properties of the thin layer. A reflectivity forward model is used because it includes all the spectral effects of the thin layer interference and can be used for any elastic for fluid case. The data are prepared for the inversion by time windowing the thin layer reflection which is then transformed to the frequency rayparameter (ω�p) domain. The error function calculation for the inversion is done in the ω�p domain so that many fewer loops are used in the reflectivity calculation. By isolating the layer and doing all the calculation in the ω�p domain, the speed of the inversion calculation is increased by a factor of approximately 1000.