- Home
- Conferences
- Conference Proceedings
- Conferences
56th EAEG Meeting
- Conference date: 06 Jun 1994 - 10 Jun 1994
- Location: Vienna, Austria
- ISBN: 978-90-73781-05-4
- Published: 10 June 1994
61 - 80 of 537 results
-
-
Migration and multi-parameter inversion in transversely isotropic elastic media
More LessTransverse isotropy is a common form of anisotropy encountered in many physical applications. It can be intrinsic to a material or rock type, occurring due to thin layering or also by alignment of pores and fractures due to the stress conditions affecting a medium. When such a medium is encountered, conventional isotropic inversion methods must be abandoned for anisotropic imaging and multi-parameter inversion methods. This requires development of new imaging algorithms and gaining an understanding as to the nature of the inverse problem.
-
-
-
Confidence measures for AVO inversion
Authors W. Beydoun, S. Jin and C. HanitzschAmplitude Versus Offset (AVO) inversion attempts to estimate multiparameter elastic information from prestack P-P reflection data (e.g., P- and S-wave impedance relative changes) prior to petrophysical/lithological characterization. Much more information can be conveyed regarding these parameters by specifying confidence measures in addition to parameter estimates. The term confidence is purposefully used to reflect our inability to provide "true" uncertainties. However, the intent is to indicate in some relative sense when or where criteria measuring processing quality are violated. This can be achieved in three ways: (1) the reflectivity versus angle domain, (2) the depth image (parameter) domain, and (3) the data (seismogram) domain. The emphasis is to demonstrate the usefulness of confidence measures by using consistently the three domains.
-
-
-
From seismic surface measurements to pseudo VSP data
Authors R. Ala‘i and C. P. A. WapenaarVertical seismic profile (VSP) has been accepted as an important tool for the calibration of seismic surface data processing and for the improvement of geological interpretation in the direct vicinity of the borehole . In this paper we describe a new technique for the transformation of surface data into VSP data, being referred to as Pseudo VSP (containing the same information but presented in a different format) . The nucleus of pseudo VSP generation is downward extrapolation of a. wave field from the surface into the subsurface . The pseudo VSP will improve the integration of surface data with real VSP. The generation of pseudo VSPs along a line , where well information is not available, will allow us to extend the geological knowledge in all lateral directions . The pseudo VSP generation technique requires high quality shot records, a description of the source properties and a macro model of the subsurface . For extensive discussion of different extrapolation operators we refer to Berkhout (1982) and Wapenaar and Berkhout (1989) .
-
-
-
Source and pseudo-logs estimation from VSP inversion
Authors J. L. Petit, V. Richard and M. CuerThe well seismic profiling has been widely used for a long time, first contributing to the definition of a velocity depth model, second producing a reference trace for accurate stratigraphic calibration of surface seismic events. For zero-offset vertical seismic profiles (VSP), a processing technique by inversion of the acoustic wave equation has been proposed by Lefebvre (1985), Macé and Lailly (1986). This VSP inversion method carries out in one step what would have required three steps in standard processing (up - and downgoing wavefield separation, deconvolution and corridor stacking). It has been widely used for estimate an acoustic impedance pseudo-log in travel time similar to the one obtain from sonic and density logs (Blanco and Canadas, 1992). However, the method deals with the simplest problem which relies on the one dimensional propagation model considering plane compressional waves travelling at notmal incidence. This technique has some limitations: 3D geometrical spreading is not taken into account, VSP in deviated wells and offset VSP cannot be inverted. In this paper, we propose a new approach to overcome these limitations: source and pseudo-logs estimation from VSP inversion in cylindrical coordinates.
-
-
-
AVO Calibration using borehole data
Authors P. Armstrong, B. Chmela and S. LeanyAVO analysis of surface seismic data has become a widely used technique in recent years. Its success has varied greatly and the use of log data to model anticipated AVO anomalies bas not always succeeded in explaining the AVO response observed on CMP gathers. The reasons for this are many and include reflectivity mismatches between the surface seismic and log data, tuning effects, geometric effects, processing-related issues and anisotropy. Without an independent measure of the seismic AVO response and the ability to include both anisotropy and tuning effects in the forward modelling, a satisfactory explanation of the origins of the surface seismic AVO response is very difficult to achieve. This paper addresses both of these requirements, investigating the use of a Walkaway VSP to measure both AVO and anisotropy and employing anisotropic plane-wave modelling which retains the full detail of the log information in the reservoir. These techniques have great potential for improving the interpretation and calibration of AVO responses at well locations.
-
-
-
An example of walkaway processing based on spectral matrix filtering
Authors J. L. Mari, F. Huguet and J. LaurentTwo walkaway seismic profiles were recorded in the vicinity of a vertical well. The first profile, oriented north-south, was composed of 103 offset VSPs. The maximum source-offset was 800 m. The average distance between two source positions was 10 m. The second profile, oriented east-west, was composed of 60 offset VSPs with a maximum offset of 500 m. The source was a vibroseismic source. The receivers in the vertical well were 15 permanent vertical sensors, clamped in the 540-680 m depth-interval. The distance between two permanent sensors was 10 m.
-
-
-
Wavefield decomposition on offset VSP data using parametric separation
Authors P. Gavin, J. Marc and A. EssabarSeparation of Upgoing and Downgoing waves is an essential and important process in study of Vertical Seismic Profile. Commonly used two dimensional transform techniques like FK or t-p, model implicitly input data as the superposition of a large number of plane waves. The idea of modelling array waveform data as the small number of direction-dependant wavefied has been developed by Seeman and al [1] to separate Upgoing from Downgoing waves. In 1990, Esmersoy [2] proposed a multicomponent separation (including optimization for non linear parameters) by introducing parametric approach to model Downgoing P waves and SV wavefield. In this way, we use this techniques and we propose to include a measure of the quality of the wave separation at each depth and each frequencies by studying the condition number.
-
-
-
Reducing smiles in a wavefield propagator VSP migration
Authors M. Williams and G. M. JacksonThe migration of well seismic data (VSPs and Walkaways) is typically dominated by large ambiguity arcs or smiles. These are much more dominant than on surface data migrations because of the narrow zones of illumination and variable fold of coverage within these zones. Normal practice is to limit the aperture of the migration around an expected dip (e.g. v.d. Poel and Cassels, 1989). This is easily done in a Kirchhoff or GRT type migration (integral solutions) in which the summation of data samples is limited to selected traces. The calculation of the Green functions is typically ray-based. When the overburden geology becomes complex, Green's functions become multivalued and ray methods become less stable. This makes diffraction stack (integral) migration methods less attractive and wavefield propagators a more viable option. Unfortunately the flexible implementation of aperture limitation to reduce smiles is more difficult. The propagation direction of energy at the depth migration bin is not explicitly contained in the formulation. Here we present a modified version of the split-step Fourier shot gather migration of Lee et. al. (1991) which allows the selection of particular propagation directions at the level of interest. Forward modelling using rays can be used to deduce the necessary propagation directions for the selection of reflectors of the expected dip.
-
-
-
A Neural network picker for VSP data
Authors H. Dai and C. MacBethArtificial neural networks (ANN) are presented as a way of automatically picking shear-wave arrival onset times in VSP data. Picking is achieved by utilizing the relative vector modulus of two horizontal component recordings as the neural network input. A discriminant function, F(t) , determined by the output of the trained ANN, is then employed to pick the arrival onset. The results demonstrate this neural network architecture is successful after training by a few selected datasets.
-
-
-
Zero moveout (ZMO) stacking
Authors P. Bunchen and A. KraganskyThis paper describes a simple stacking procedure based on summing the traces of a common shot or common mid-point gather, without the usual relative time shifts. We refer to this procedure as zero moveout or ZMO stacking. The stacking procedure may include an offset dependent weighting function, which is determined by the dimensionality of the data. It is shown that for the correct choice of weighting function, the resultaat stacks are exact zero-offset one dimensional seismograms for horizontally layered media. We explore the practical implementation of ZMO stacks towards source estimation. The principle can also be used for inversion.
-
-
-
Multi-dip radon stack as an alternative to conventional stacking
Authors M. M. N. Kabir and D. J. VerschuurConventional velocity stack is performed after a hyperbolic moveout correction and straight summation of the data (CMP gathers). This process is equivalent to summing the data along hyperbolic paths. Although the Radon transform (RT) has never been designed to perform such operation as stacking, it is very interesting to note the basic definition of the forward generalized Radon transform (GRT) which is obtained by summing the original data along a path. If this path happens to be a hyperbolic one, then in principle, the forward GRT domain would indeed be equivalent to a velocity stack domain. And picking suitably in this transform domain should enable us to obtain an equivalent of a velocity stack. This is the idea that lays the foundation of the proposed method.
-
-
-
From well logs to surface seimics using an equivalent medium
More LessMeasurements carried out at different stales may not agree in their observations. An equivalent medium approach can link measurements carried out at different scales. I apply high frequency (Dix) and low frequency (Schoenberg&Muir) averages to a well log and compare the results to a conventional velocity analysis of surface seismic data.
-
-
-
Velocity analysis incorporating 3D pre-stack time migration
Authors R. G. Williams and N. J. CooperConventional industry processing of seismic data includes an approximate NMO and DMO prior to velocity analysis. The NMO and DMO imaging steps are approximate only in the sense that an approximate (simple) velocity field is used; the algorithms themselves are accurate. After velocity analysis, NMO is re-applied with the more detailed velocity field derived from the analyses. In the constant velocity case, DMO is almost independent of velocity and the iterative procedure described above plus post stack migration yields an accurate image.
-
-
-
Velocity analysis as shape parameter estimation of wideband events
Authors U. Spagnolini and F. RoccaGeophysical processing requires the evaluation of the kinematical parameters of reflected wavelets (wideband events). Reflection tomography requires the estimation of traveltime and ray parameters; velocity analysis estimates the curvature (Shape Parameter: SP) from reflected or migrated wavefronts. In any case, SP estimation entails die definition of a reliable coherence measurement of wideband events along a pre-assigned time-space window. Better algorithms allow a simplification of the SP picking and reduces the spurious peaks due to incoherent and interfering noise. This paper introduces and compares the joint use of matched filter and Karhunen-Loeve (K-L) decomposition for high resolution SP estimation of wideband events.
-
-
-
The application of stacking and migration in refraction seismology
Authors E. Bruckl and F. KohlbeckFor decades the conceptions of stacking und migration have been welt established processes in reflection seismology. These processes utilize the information contained in the total wavefield. However, in contrast to this evolution, most modern developments in refraction seismology restrict themselves to the evaluation of traveltimes. Although the way of the implementation of the complete refracted wavefield is indicated by the wavefron- or ray-methods, applications still remain the exception. Based on these classical approaches and some more recent research work (Hill 1987, Bruckl 1987, Bruckl 1991) a method for processing and interpretation of refracted wavefields inciuding stacking und migration has been developed. These new wavefield techniques offer to the user a range of facilities which comprise optimization of the interpretation model, quality control, improved accuracy, and resolution.
-
-
-
Determination of shallow velocity-depth model from seismic refraction data by coherency inversion
Authors E. Landa, S. Keydar and A. KravtsovThe reflection survey is the dominant method of seismic prospecting. However, reflection waves clone are not always effective for studying the upper part of the subsurface. The strong surface waves and non-ray effects do not give stable and reliable results in the upper part of the reflection seismic section. Unlike reflections, refraction waves can be observed outside the zone of interference with surface and reflection waves. Seismic refraction waves have different applications in seismic prospecting.
-
-
-
Evaluation of sea bottom features from seismic calibration data - A case history
Authors D. Monk, R. Flud and M. PlumleeDuring 1992 a seismic survey was conducted in the West Delta area of the gulf of Mexico. The water depth in the area of the survey varied from approximately 20 feet to over 200 feet, and for this reason, seismic acquisition using a bottom cable technique was selected. Additionally the use of hydrophones and geophone was chosen, because of the potential to minimize ghost and reverberation effects associated with the variable water layer. It was known prior to the commencement of the survey, that in some areas, the "hard" water bottom was overlain by soft mud flows with remarkable physical properties including velocity in some cases of less than 1000 feet/sec. (ref. Meeder et.al. (1988)). It was anticipated that the mud flows may present significant problems in both acquisition and processing of the seismic data, and an attempt was made prior to the seismic survey, to map the sea floor features as part of a hazard survey.
-
-
-
Ray-mapped focusing - A migration velocity analysis for Kirchoff pre-stack depth imaging
Authors J. P. Jeannot and I. BerrangerFocusing analysis was first described for time migration by Yilmaz and Chambers (1984) and adapted to depth migration by Jeannot and al. (1986). Although the focusing equations were established for a constant velocity medium and a flat, horizontal reflector, focusing analysis has been used extensively for velocity model-building, even in the presence of severe lateral velocity variations. One of the main flaws of conventional focusing analysis is its sensitivity to dip. Audebert (1993) proposed residual zero-offset migration of the focused energy so that the same CRP will be imaged as downward continuation proceeds. We show how this concept can be implemented in Kirchhoff migration.
-
-
-
Kirchoff migration/inversion in thinly layered media
Authors M. T. Widmaier, S. A. Shapiro and P. HubralAmplitude-preserving prestack Kirchhoff migration (Schleicher et al., 1993) is a well discussed tool to image reflector positions and to analyse offset-dependent reflectivity. Additionaly to the known weighting function of Kirchhoff-summation, we introduce a new correction to take into account the influence of thinly layered media on wave propagation.
-
-
-
Anti-aliased Kirchoff migration
Authors D. E. Lumley, J. E. Claerbout and D. BevcDiscrete space-time aliasing of Kirchhoff migration images arises when the migration operator summation trajectory is too steep for a given seismic trace spacing and frequency content. This operator aliasing is independent of data aliasing, and can seriously degrade the focusing quality of subsurface reflectivity images, especially in the presente of large reflector dip and sparse or irregular 3-D acquisition geometries. We have developed a method based on local time- and dip-variant temporal filtering to suppress Kirchhoff migration operator aliasing. In a numerically efficient implementation, we apply local N-point anti-aliasing triangle filters as three-point filter operations after a step each of causal and anticausal temporal trace integration. Our approach is very conservative in its memory requirements, as demonstrated by a 3-D implementation on a massively parallel Connection Machine CM-5. We compare our new anti-aliased Kirchhoff migration to a conventional aperture-weighted Kirchhoff migration, in applications to both 2-D and 3-D marine seismic data. The results indicate that our anti-aliasing method enhances the resolution of steep salt-sediment interfaces and faults, and suppresses false reflections caused by conventional Kirchhoff migration operator-aliasing artifacts.
-