S12A-01
The Receiver Function Method for Multicomponent Ocean Bottom Cable Seismic Data
The Receiver function method has been originally developed to analyze earthquake data recorded by multicomponent (3C) sensors and consists of deconvolving the horizontal component by the vertical component for a selected phase. The deconvolution process removes travel path effects from the source to the base of the target as well as the earthquake source signature. In addition, it provides the possibility of separating the emergent P and PS waves based on adaptive subtraction between recorded components if plane waves of constant ray parameters are considered. The resulting RF signal is the local PS wave impulse response generated at impedance contrasts below the 3C receiver. We have extended this method to the wide-angle multicomponent seismic reflection acquisition geometry that has source above the interfaces instead of below as in earthquake seismology and contains many interfering primary (reflected) arrivals. The method has been developed in the (τ-p) domain that helps to decompose the wavefield in to P and S-waves. An iterative deconvolution is used to remove the source effects, and results in pure P and S-wave fields, and leads to P and S-wave receiver function. The presence of pressure wavefield recorded by 4C ocean bottom cable data helps in the reconstruction of the both P and S-wave receiver function. The method has been tested on synthetic data example and then applied to multicomponent ocean bottom cable data. Our method provides much-improved P and S-wave image of the sub-surface. The resulting P and S-wave receiver function could be further used for amplitude versus offset analysis or for full waveform inversion.
S12A-02
Time-lapse Measurements of Scholte Wave Velocity Over a Compacting Oil Field
Acquisition of time-lapse seismic data over producing oil and gas fields is a proven method for optimizing hydrocarbon production. Most current data have been acquired using towed-streamer seismic vessels but new systems incorporating permanent Ocean Bottom Cable (OBC) systems are gaining in popularity, both as a way to achieve better repeatability and also to reduce the cost of acquiring many time-lapse repeats of the baseline survey. Over the last three years, more than seven repeat data sets have been acquired at the permanent OBC system installed (by the operator, BP) over the Valhall oil field located offshore Norway. This system contains ~2400 four-component receiver stations that are recorded using a dense areal shot grid ("carpet" shoot) that provides high fold and has delivered excellent time-lapse signals starting from the first repeat occurring just three months after the baseline. Time-lapse OBC data are conventionally used to measure amplitude and velocity changes of body wave reflections (PP and PS) but other measurements are also possible. In particular, Scholte waves are strongly visible on records acquired everywhere in the field on appropriately processed data and, given the high fold (because of the dense shots), Scholte wave velocity and anisotropy time-lapse changes obtained with both hydrophone and geophone sensors are accurately and robustly estimated. The resulting shallow velocity maps are very sensitive to the seabed strains and show large velocity changes overlying deep production. Also, reconstruction of compressional "head wave" velocity difference measurements and vertically propagating shear wave shallow time-lapse statics produce maps that resemble the Scholte wave maps, with differences that reflect the physics of the propagation modes and effective fold. A reservoir model that includes deep reservoir volume changes together with appropriate geomechanical properties in the overburden and a shallow conversion of strain to velocity is used to successfully predict the measured velocity changes. The strain/velocity conversion requires asymmetry between crack opening and closing as well as velocity hysteresis and, in fact, the measurements provide an excellent laboratory for testing fracture-model/velocity conversion on in-situ rocks. After calibration, the model together with the data can constrain both volume changes in the reservoir, for making drilling decisions as well as the overburden geomechanical rock properties model, which itself is used for well- path selection and facilities decisions. Scholte wave velocity measurements can also be made using an oil platform as a "passive" source, removing the need for a conventional source near the seafloor. Finally, these measurements might be applicable on time- lapse controlled source measurements of greater generality in a wider geophysical context wherever an accurate measurement of a time-varying surface strain is desired.
S12A-03
A New Method for MCS Refraction Data Analysis of the Uppermost Section at a Mid-Atlantic Ridge Core Complex
The first refraction arrival or moveout of reflection arrivals are generally used to obtain velocity structure of the sub- surface. However, in deep-water environments and in the absence of near seafloor reflections, it is not possible to determine uniquely the velocity structure just below the seafloor from conventional methods. Here a new approach to analyzing seismic refractions recorded on a multi-channel streamer is tested with a subset of R/V Ewing data obtained over the central dome of Atlantis Massif in order to determine the high-resolution P-wave velocity of the uppermost ~1 km. This oceanic core complex on the Mid-Atlantic Ridge 30°N provides access to intrusive crust exposed at the seafloor via detachment faulting, thus eliminating the imaging problems due rough seafloor and heterogeneous basaltic carapace that can mask lower crustal structure at most mid-ocean ridges. In addition, IODP drilling results in the vicinity provide ground truth to depths of 1.4 km. Multi-channel seismic (MCS) data were downward continued to a new datum 1.5 km below the sea surface using a prestack phase shift approach. A filtered receiver wavefield from a single shot point was first extrapolated to depth, and this step was repeated for all shots. Next, the data were sorted into common-receiver location space and the 160 shots per common-receiver gather were downward continued to 1.5 km below the sea surface, producing a dataset that is equivalent to a reflection profile collected at a water depth of 1.5 km. Because the trace spacing in these two domains differ by a factor of 3 (i.e., 12.5 m versus 37.5 m), filtering parameters were changed prior to invoking a phase shift in the F-K domain, to minimize aliasing (with the common receiver domain trace spacing being most problematic at the 37.5 m distance corresponding to the shot spacing). The requirement of a second extrapolation in common-receiver location space reduces the full-fold line length, within the downward continued gathers, by an amount equivalent to a streamer length (6 km) from each end of the original record section. For most of the shots, clear refracted arrivals are present at ranges 0.2-4.5 km as compared to 1.5-6 km on original gathers, which should provide velocity starting from just below the seafloor down to about 0.5-1.5 km depth. Refracted arrivals have been picked for every shot trace and these data will allow us to invert for detailed velocity structure in the uppermost section. The data density and continuity is over an order of magnitude greater than a previous near-bottom-source OBS refraction experiment that obtained ambiguous results in this area. The presentation will include illustrations of this new methodology and the tomography results as well as comparison to results obtained using conventional methods in the same area (Canales et al., session T19 Fall AGU07).
S12A-04
Wide Angle Converted Shear Wave Analysis of North Atlantic Volcanic Rifted Continental Margins
High-quality, wide-angle, ocean bottom seismometer (OBS) data have been acquired with a low frequency (9 Hz) seismic source across the Faroes and Hatton Bank volcanic rifted continental margins in the North Atlantic. In these regions thick Tertiary flood basalt sequences provide a challenge to deep seismic imaging. S-wave arrivals, which are dominantly converted from P- to S-waves at the sediment-top basalt interface, were recorded at 170 4-component OBS locations. Variation in the conversion efficiency was observed along the profiles. Tomographic inversion of over 70,000 converted S-wave crustal diving waves and Moho reflections was performed to produce S-wave velocity models and hence, when combined with pre-existing P-wave velocity models, a measure of the Vp/Vs ratio structure of the crust. Resolution testing shows the structure of the oceanic crust and continent-ocean transition is generally well resolved on both profiles. Lateral and vertical changes in Vp/Vs resolves changing crustal composition within, and between, oceanic and continental crust, including regions in the lower crust at the continent-ocean transition with high P-wave velocities of up to 7.5 km/s and low Vp/Vs ratios of ~ 1.75 associated with intense high-temperature intrusion at the time of break-up. Vp/Vs ratios of 1.75-1.80 at the base of the thickened oceanic crust are also lower than generally reported in normal oceanic crust. The P-wave travel-time tomography revealed a low velocity zone (LVZ) beneath the basalt on the Faroes margin and additional constraint on the Vp/Vs of the LVZ beneath the Fugloy Ridge has been gained by analysing the relative travel-time delays between basalt and basement refractions for P- and S-waves. This approach is less subject to the velocity-depth ambiguity associated with velocity inversions than is the determination of P- or S- wave velocity alone. Comparison of the calculated Vp/Vs ratio and P-wave velocity with measurements from relevant lithologies reveals that the LVZ is likely to contain sill-intruded Paleocene sedimentary rock rather than igneous hyaloclastites similar to those found beneath the basalt in a nearby well. Immediately beneath the LVZ, a unit with Vp/Vs ratios of 1.80-1.85 and P-wave velocities of 5.5-6.0 km/s is interpreted as sill-intruded sedimentary rock of a pre-breakup Mesozoic basin. We thank C.J. Parkin, A.W. Roberts and L.K. Smith for their contributions.
S12A-05
Dip Corrections for Seismic Reflection Velocity Analysis
We use structural dip information to mitigate noise in time processing of active-source seismic reflection data, especially in land or low-fold marine datasets characterized by low signal-to-noise ratios (SNR). We estimate dip fields in stacked sections using robust iterative semblance scans. In two-dimensional data, the dip fields consist of single time-dip components defined at each image point; in three-dimensional data, the fields consist of simultaneously constrained in-line and cross-line dip components. The dip fields are used as input for a Common Reflection Surface (CRS) stack that sums along move-out curves in both the midpoint and offset dimensions to create sections with improved SNR. We show that this process is equivalent to a post-stack local slant stack that trades along-dip resolution for noise suppression. Dip fields are also used for velocity analysis in super-gathers, groupings of adjacent common midpoint (CMP) gathers commonly used for velocity analysis. We use the CRS travel-time formulae to correct for differences in zero-offset travel-times for traces with different midpoints within the super-gather. The corrections supplement dip move-out (DMO), which corrects for multiple conflicting dips in traces within a single CMP gather. While DMO collapses semblance maxima from conflicting dips into a single maximum, our dip corrections increase the sharpness of the peaks and make the semblance panels easier to interpret. As a result, larger super-gathers with more traces can be used to increase SNR without blurring the semblance maxima. We then apply the same technique to residual move-out (RMO) analysis using dip fields computed on migrated sections. Super-gathers are also commonly used for computing for RMO semblance panels, so the dip corrections similarly lead to sharper maxima and facilitate the use of larger gathers. Examples are shown for both 2D and 3D datasets.
S12A-06
A new integrated method for the crustal structure analysis using OBSs and control sources
Recently, in order to determine crustal structure in the oceanic region, a huge amount of seismic data has been collected using a set of digital OBSs and a large volume (e.g., ~8,040 cubic inches) tuned airgun-array. Such survey can provide full-waveform data of wide-angle reflected and refracted arrivals. Most of previous analyses using OBS data have commonly been carried out only by the travel-time inversion method using first arrivals only. However, it is well known that the travel time inversion strongly depends on a starting model. To avoid this problem, it is necessary to make a reasonable crustal initial model to satisfy full-waveforms, MCS reflection data and geological and geophysical backgrounds. In our analysis we are adopting a combination of several datasets such as MCS data, Pg, Pn first and later arrivals, PmP and crustal reflected phases, P-S and S-P converted phases, and the amplitude of each phase. Seismic records are interactively processed using gPastuph software developed by Fujie et al. (2007). Those dataset are analyzed by the forward modeling, the ray tracing, the first-arrival and reflection travel-time tomography, and the calculation of synthetic seismograms. We have applied the above analysis for the crustal structure studies in the western Pacific Ocean near the Japanese archipelago. Crustal structures obtained by the above analysis stream satisfactory fit to the original waveforms and MCS reflection sections. The O-Cfs of between observed travel-times and theoretical travel-times are ~30-40 ms. Most of analysis has been carried out by colleagues in Japan Continental Shelf Survey Co., JGI Inc., and Kawasaki Geological Engineering, Co. Ltd.
S12A-07
Upper Crustal Evolution Along the Juan de Fuca Ridge Flanks from Travel Time Tomography of Seismic Layer 2
We are performing 2-D travel time tomography using data from long (150-300 km) multi-channel seismic lines, collected in 2002 across the Endeavor and Cleft segments of the Juan de Fuca Ridge (JdFR), to examine upper crustal evolution along the ridge flanks. 1-D velocity models have previously been computed at ~3 km intervals on super CMP gathers and are here used to define the initial velocity models for inversion. The 1-D models, which provide good regional constraints of the upper crustal velocity, show little increase in velocity with crustal age on the unsedimented west flank of the JdFR whereas velocities close to double on the heavily sedimented east flank by ~4-5 Ma. Superimposed on these large-scale regional changes are shorter wavelength (~5-10 km) variations of seismic layer 2A, possibly associated with basement structure. Continuous 2-D tomography with higher lateral resolution, from the full streamer data, will enable detailed analysis of the upper crustal structure and evolution. For the 2-D travel time tomography, we picked the first arriving refraction in shot gathers (2B arrival). In our data, this phase is generally visible in the last ~2 km of the 6 km streamer. We are using a modified version of C. Zelt's FAST tomography code to invert for the minimum-structure 2-D velocity models. The dense sampling of sources and receivers allow us to image velocity variations at lateral scales of 1 km or less. We anticipate that our 2-D models will validate the previous 1-D models on a regional scale as well as yield more detailed information about layer 2A and upper layer 2B. In particular, we hope to determine if the there are detectable fine scale variations in upper crustal structure with basement relief, if there is any local effect of propagator wakes on upper crustal velocities and if the alteration observed in layer 2A extends into 2B.
S12A-08
The Agulhas Plateau: Structure, Formation and Environmental Impact of a Large Igneous Province
The Agulhas Plateau (AP) is an oceanic plateau in the Indian Ocean south of the African continent. We investigated its structure and evolution using seismic refraction, seismic reflection and potential field data. Our P- wave velocity-depth model of the AP along profile AWI-20050200 shows 20 km thick oceanic crust on average with velocities well above 7 km/s in the lower 10 km of the plateau. Velocity and density structure strengthen the hypothesis for an evolution as a Large Igneous Province (LIP). Like other LIPs, such as the Ontong Java Plateau, the AP consists of three major structural units: an extruded cover, an intruded middle crust and a lower crustal body. Plate-tectonic reconstructions suggest a coeval formation of the AP together with Maud Rise and Northeast Georgia Rise. Between these three parts a triple junction was situated which caused the fragmentation. Using the structural information from this seismic refraction profile and another crossing one, we could estimate the total volume of material accreted to the plateau to about 4 × 106 km3. We differentiated between intruded and extruded parts of igneous material and used the extruded basaltic volumes for estimation of carbon dioxide and sulphur dioxide emission. The volume of this extruded layer with velocities between 3 and 5 km/s and a thickness of 1.8 km on average was estimated to be about 4 × 105 km3. In this layer, volcanic flows could be identified in coincident seismic reflection data. Our calculations lead to implications on the possible impact of the AP formation to the Cretaceous environment. We conclude that the emitted carbon dioxide volume was too small to have a direct impact on the climate, but feedback mechanisms possibly enhanced the effect. Sulphur dioxide has not such a long resistance time in the atmosphere as carbon dioxide, therefore had a small impact on the climate, but probably a more important one on the ocean environment by increasing the acidity of the surrounding water masses. Due to heat generation during the AP formation, solved oxygen from the ocean could have been released into the atmosphere, leading to anoxic conditions in the ocean. In summary, we gained new information about the structure and evolution of the AP, which are essential to estimate the influence of its formation on environmental conditions during its formation in the Cretaceous.