New Advances in Seismic Imaging for Exploration and Solid Earth Geophysics IV Posters
Presiding: S Rondenay, Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology; G Schuster, University of Utah
S41B-01 0830h
Broadband surface wave dispersion measurements across North America from ambient seismic noise
Ambient seismic noise contains a significant component of Rayleigh wave energy that appears to be excited by oceanic microseisms and atmospheric forcing. These signals constitute a wavefield in which the phase is randomized by a multiplicity of sources and by scattering. Cross- correlations of the ambient wavefield between pairs of receivers can be used to extract the Rayleigh wave part of Green's function and, therefore, provide a new source of surface wave information that is particularly useful in the context of arrays of broadband seismometers such as PASSCAL experiments, the emerging USArray, or other national deployments. The method produces numerous inter-station measurements that are not sampled by traditional observational methods based on earthquake waves. The method to extract surface wave dispersion measurements from ambient noise was first applied systematically to records from USArray Transportable Array stations in California. These observations were used to construct high-resolution short-period (7-18 s) surface wave dispersion maps and to image the principal crustal geological units. In addition, it has been previously shown that reliable dispersion meausurements can be obtained at intermediate to long periods (20 - 100 sec). In particular, cross-correlations of several months of ambient seismic noise observed at several station-pairs across North America result in coherent broadband waveforms with dispersion characteristics similar to Rayleigh-wave tomography maps constructed using earthquakes waves. We extend this work by computing cross-correlations and obtaining intermediate and long period surface wave group velocity measurements for paths connecting about one hundred permanent broadband stations in North America many of which constitute the Advanced National Seismic System (ANSS). We discuss the promise and limitations of the method to improve information about the crust and upper mantle across North America.
S41B-02 0830h
3-D Wave Equation Based Pre-Stack Depth Migration and Directional Illumination Analysis
3-D pre-stack depth migration has proven to be a critical process in identifying the potential targets for oil and gas exploration. A heterogeneous medium with strong velocity contrast and complex subsurface structures presents a great challenge to seismic imaging. Wave equation based migrations are accurate but usually very expensive, prohibiting their uses in production applications. Recent advances in efficient 3-D wave field extrapolators and the continued increase in computing performance are making application of these techniques to full 3-D pre-stack wave equation depth imaging more affordable. In particular, screen propagators in midpoint-offset coordinates using local velocity perturbations provide fast and accurate wave field continuation. In the case of complex structures with large velocity contrast, such as salt-related structures where the velocity of salt body can be 2-3 times higher than the surrounding sediments, imaging shadow zones may be present in the sub-salt and around the salt boundaries. In such situation, illumination study is a useful estimate to investigate whether these imaging shadows are caused by the acquisition geometry under a specific geological model or caused by the inaccuracy of migration algorithm. However, traditional ray-based illumination methods are based on high-frequency asymptotic approximation, causing fundamental limitations in applications to complicated sub-salt regions. Wave equation illumination method is a more powerful and more accurate tool to analyze the illumination strength of a specific target event and amplitude variations. Such information can also be used for optimizing the acquisition geometry design. A number of benchmark synthetic and field data examples will be presented.
S41B-03 0830h
Elastic Scattering from a Quarter Space; Numerical Tests
We are investigating elastic and acoustic wave scattering from an elastic quarter space within water. Understanding rough surface scattering is critical for marine seafloor acoustics. The simplest scattering is from a two-dimensional edge. Within an elastic quarter space we observe multiple downward-scattered and transmitted phases including transmitted P and S waves, P and S head waves, P and S diffracted waves, and interface waves. The forward and backward scattered waves in the water can originate from the direct water wave as well as converted wave phases within the quarter space. At some locations in the water, there are as many as four distinct arrivals. In many seafloor survey systems, some of these diffracted waves could be interpreted as coming from more distant scattering points on the seafloor or from sub-seafloor volume scattering points, hence leading to incorrect interpretations of the observed data.
S41B-04 0830h
Improved Teleseismic Green's Functions and Crustal Multiples
Over the past two decades, "receiver functions" have proven to be a useful tool to investigate crustal structure. As they represent a first-order approximation to the S-wave component of the teleseismic-P Green's function, receiver functions provide valuable information on physical properties related to shear modulus. However, the use of the P-component seismogram as a proxy for the source precludes the recovery of information on discontinuous structure involving constrasts in compressional modulus. By deconvolving improved estimates of complex source time functions generated by earthquakes, one may move beyond the conventional receiver function paradigm to a more accurate approximation to the Green's function. Using a method described by Baig et al. [2005, submitted to JGR], we present estimates of P-component teleseismic-P Green's functions at several stations of the Canadian National Seismic Network that clearly show the Ppmp crustal multiple. The identification and characterization of this phase in studies of the lithosphere will afford better constraints on lithology and represent a significant narrowing of the gap between active- and passive-source seismic imaging.
S41B-05 0830h
Seismic Image Reconstruction by Reducible Transformations
Diffractions of seismic wave cause distortions and misplacements of subsurface structures in the reflection seismograms. A reasonable reconstruction of complex geometry, existing below the Earth's surface in multi-dimension, is possible if there is a reducible transformation, which can be based on wave-theoretical foundations. One such is a Radon transformation used with success in medical imaging. Several authors applied the projection and back-projection ideas inherent in Radon (also known as Tau-P) transformation for seismic imaging. This paper presents a new non-recursive approach implemented at every point in the transform domain with a threshold operator. The approach in this paper is based on a double Radon transformation of the seismic reflection data as a function of midpoint, offset, and time. In the back projection step, the intercept time is modified using midpoint, offset, and velocity. Satisfactory implementation on synthetic data is presented for a vertically inhomogeneous medium with some lateral velocity variation.
S41B-06 0830h
Double Tau-P Transformed Surfaces and Deterministic-Stochastic Muting for Partial Annihilation of Multiple Reflection
This paper presents a new synergic combination of double Tau-P transformation and Deterministic-Stochastic muting for the removal of multiple reflections from the seismic data. The seismic data contains both primary and multiple reflections. The multiple reflections, appearing as coherent events, render the seismic sections uninterpretable by introducing spurious boundaries. This paper attempts to improve the reflection data quality by partial annihilation of multiple reflections. There are several approaches, e.g., Common Midpoint Stacking, Slant Stacking (often 1-D Tau-P transformation), Predictive Deconvolution, and their individual variations used by many authors. Each of these methods has both merits and demerits. Recently inversion based methods were introduced for solving the formidable multiple problems arising due to sub-salt, overhang, and other complex geometries. The present paper takes a different attempt to attenuate multiples from irregular and dipping surfaces. The reflection seismic data are considered as a function of midpoint and offset coordinates and intercept time. A two dimensional Tau-P transformation of the space coordinates reduces the function domain primary and multiple surfaces into special curves. Deterministic-Stochastic muting is applied on the transformed slices. An inverse transform reproduces the seismic data significantly removing multiples. A computer algorithm based on theoretical development is successfully applied on synthetic seismic data.
S41B-07 0830h
Phase Space Migration - a Fast Migration Algorithm for Densely Sampled Data
We present a migration method in the coupled ray-parameter domain that is fast and efficient for seismic data that are densely sampled in the source-receiver configuation space. The method is based on slant stacking over both shot positions and offsets for all the recorded data. If the data acquisition geometry permits, both in-line and cross-line source positions and offsets can be incorporated into a multidimensional phase velocity space which is regular even for randomly positioned input data. By noting the maximum time dips that are present in the shot gathers and constant offset sections, the number of plane waves required can be estimated and this generally results in a data reduction of at least one and possible two orders of magnitude. The required travel time computations for depth imaging are independent for each particular plane wave component and thus can be used for either the source or the receiver plane waves during extrapolation in phase space, reducing considerably the computational burden. Even so, each source and receiver plane wave component must be combined with all other receiver and source components for a complete diffraction summation. Since only vertical delay times are required, many travel time techniques can be employed and the problems with multi-pathing and first arrivals are either reduced or eliminated. Further, the shot plane wave integral can be pruned to concentrate the image on selected targets. In this way the computation time can be further reduced and the technique lends itself naturally to a velocity modelling scheme where for example, horizontal and then steeply dipping events are gradually introduced into the analysis. Of course this imaging scheme can be implemented in parallel using a distributed architecture like a PC cluster to compute various plane wave sections since they are independent of each other. The common ray-parameter image gathers can be used exactly like common angle image gathers for residual migration velocity analysis. The migration method lends itself to imaging in anistropic media since phase space is the natural domain for analysis.
http://www.ig.utexas.edu
S41B-08 0830h
Suppression of Statics and Migration Velocity Errors by Reduced Time Migration of Surface Seismic Data
One of the difficulties in seeing beneath salt is that the migration velocity in the salt and above is not well known. This can lead to reduced quality of migration images beneath the salt. There are two ways to remove the kinematic effects caused by the errors in the overburden velocity model: reduced-time migration and interferometric migration. Interferometric migration is very expensive because it extrapolates data from the surface to the reference layer, followed by migration of the extrapolated data to below the reference interface. In comparison, reduced-time migration is as inexpensive as standard Kirchhoff migration because it does not extrapolate the data but rather shifts the data with a time difference between the calculated and natural arrival times of a reference reflection, τsgref and ~ τsgref, where s and g denote the source and receiver locations on the surface and τsgref is calculated by raytracing through the estimated overburden velocity model. This has the effect of removing the error in computing traveltimes in an uncertain overburden model and the statics associated with the shot and receiver. Since the calculated reference reflection times depend on both the overburden velocity model and the depth of the reference layer, a rough estimation of the reference layer depth is usually needed in RTM migration. The reduce-time migration can be implemented in three steps: 1) Pick the natural arrival times ~ τsgref; 2) shift the data by ~τsgref; and 3) migrate the shifted data and apply the calculated static shift τsgref during migration. Test results on synthetic data show the effectiveness of RTM in removing both source/receiver static errors and the timing errors due to an uncertain overburden model above the reference layer. Results with the marine field data show RTM can mitigate the timing error effects associated with the shallow gassy muds.
S41B-09 0830h
An Artificial Neural Network Method for Formation Permeability and Skin Coefficient With Wireline Formation Test Data
Wireline formation testing tools have undergone generations of evolutionary changes and have continued to advance in formation evaluation techniques by mimicking a miniature well test. Compared with drill stem tester (DST), wireline formation testers have better acquisition ability, more detailed and reliable vertical pressure profiles, shorter test time, lower cost and higher efficiency. With the development of new types of wireline formation testers, the flow rate can be controlled and more information may be available. It is desirable to develop an efficient data interpretation model to determine formation mobility and skin coefficients more accurately. In this paper we have developed a back propagation neural network (BPNN) model to simulate the complex relation between pressure response of wireline formation testers and reservoir parameters. The neural network model has been trained by a training set. Of the training average relative error in permeability is 6.49%. The training results in dimensionless skin coefficients are also fairly accurate. These data shows that the neural network model developed in this paper provides accurate mapping relation between pressure responses from wireline formation testers and reservoir parameters, such as permeability and skin coefficient. To check extrapolated ability of the neural network model, 178 sets of data are computed with random reservoir parameters. The predicting results have demonstrated that the predicting average relative error in permeability is 7.75% and the errors in predicting results of skin coefficients are less than 22%. The data interpretation examples demonstrate that the neural network model takes full advantage of data in early to intermediate time to reduce test time. It can determine reservoir permeability and dimensionless skin coefficient simultaneously without complex mathematics and get accurate results.
S41B-10 0830h
Effects of Near-Source Scatterers on Radiation From Seismic Sources
The objective of this research is to study the effects of scattering from near-source heterogeneities on seismic wave radiation. High-contrast heterogeneities strongly scatter both P and S waves and later the source radiation patterns. We investigate these effects using a three-dimensional finite-difference modeling code that includes variable grid, free surface with topography, and seismic attenuation. A perfectly matched layer is incorporated into the code for improving absorption at the boundaries. Two source models are considered: an explosion and a double couple. An explosion source is placed near a finite-length tunnel. The calculations show both P to P and P to S scatterings. The tunnel acts as a secondary source with strong radiation of S waves. Love waves are observed at the surface. Then a double-couple source is placed near high impedance contrast heterogeneity. Complicated scatterings of both P and S waves are observed. In both cases the distance from source to scatterer plays an important role in scattering intensity. Moment tensor inversion of seismograms for an explosion source near a tunnel gives a source model that includes both diagonal and off-diagonal elements. For a double-couple source near a solid scatterer, the source distortion is observable, but less prominent than that of an explosion.
S41B-11 0830h
Application of Time Reversed Acoustics for Seismic Source Characterization
Traditionally an earthquake is located and the source mechanism is determined by using P and S phases. This uses only a limited portion of the information contained in a seismogram. A large part of the information carried by the waveform is not used. In this study we investigate the applicability of the Time Reversed Acoustics (TRA) technique, and thus the whole waveform of the recorded signal, for earthquake locations and source characterization. The basic concept involved in TRA is the fundamental symmetry of time reversal invariance. Injecting the recorded signal, with time running backwards, can focus the wave field to the source. TRA has emerged as an important technique in acoustics with applications to medicine, underwater sound, and many other disciplines. Numerical simulations show that the TRA technique can successfully locate a seismic source inside a layered earth model and can also recover the source time function. Finite difference modeling results show that TRA can determine the fault dip, rupture direction, and rupture length. The method is especially advantageous when data are available only from a sparse station network. Full seismograms contain source information from both waves radiated along the source-station ray path and from waves that radiated in all other directions but scattered toward the receivers. Application of the TRA technique to seismic source characterization requires the Green's function, which can be obtained in two ways. If the earth structure is known then the Green's function can be calculated numerically. To improve the efficiency, the method of constructing a medium response library is developed. This improves computation time significantly. The second approach uses small events (e.g., aftershocks) as an empirical Green's function. The performance of the TRA technique is demonstrated with data from real earthquakes.