Seismic Inversion and Its Applications to Exploration and Global Earth Structures II
Presiding: P D Anno, ConocoPhillips; P S Routh, Boise State University
S44A-01 INVITED 15:30h
The Inverse Scattering Series and Seismic Exploration
The inverse-scattering series is the only direct multi-dimensional inversion procedure that permits all processing objectives associated with inversion to be accomplished without knowing, or determining, either the slowly or rapidly varying part of the medium through which the waves propagate. The latter property explains why these methods are a direct response to the challenge of removing multiples and imaging beneath a complex ill-defined overburden, e.g., salt, basalt, and karsted sediments. The removal of free surface multiples, the attenuation of internal multiples, and the depth imaging and inversion of primaries are achieved by four distinct subseries, and each operates without knowledge or determination of the velocity model. We will exemplify these task-specific subseries concepts, and algorithms, with synthetic and field data. Reference: Weglein, A. B., Araujo, F. A., Carvalho, P. M., Stolt, R. H., Matson, K. H., Coates, R., Foster, D. J., Shaw, S. A. and Zhang, H. "Topical Review: Inverse-scattering Series and Seismic Exploration." Inverse Problems 19, (2003):R27-R83
S44A-02 INVITED 15:45h
Wave Field Attenuation and the Task-Specific Inverse Scattering Series
Task-specific portions of the inverse scattering series (ISS) are currently under examination as to their potential for accomplishing various key seismic processing objectives in the absence of accurate knowledge of the medium. Study of the effect of Q on terms in the forward scattering series suggests that the same mechanisms that accomplish reflector location in the ISS (the imaging subseries) should act to compensate for the suppression of high frequencies of the signal that is associated with visco-acoustic propagation. Here we develop more specifically the basics of the non-linear data operations that act to generate a Q compensation operator without prior knowledge or specific determination of Q. This study at its least ambitious aims to describe the effect on ISS algorithms due to the presence of Q; at its most ambitious it aims to develop a practical means to re-provision suppressed signal resolution in cases where little is known a priori about the medium properties. Reference: Innanen, K. A., and Weglein, A. B., Construction of Absorptive/Dispersive Wave Fields with the Forward Scattering Series, 2003: Journal of Seismic Exploration, 12: 259--282.
S44A-03 INVITED 16:00h
Seismic Inversion for Hydrocarbon Exploration, Crustal Structure and Whole Earth Modeling
Seismic inversion employing travel time, amplitudes and full waveform attributes is a challenging task. In order to maintain accuracy and computational speed, it is important to make judicial choice of optimization methods, onstraints, a priori information, forward modeling operator and data. We have applied seismic inversion to a variety of applications including (1) full waveform inversion for estimating rock properties within a hydrocarbon reservoir, (2) joint inversion of seismic travel time and gravity data for delineating plate boundaries and crustal structure offshore Taiwan, (3) crustal velocity structure from waveform inversion of shear-couple PL waves, and (4) anisotropic structure of the core-mantle transition zone. Although all these applications make use of the same forward modeling algorithms and optimization routines, they are carefully designed to address the disparate scale and data types. For example, a semi-analytic approach for differential seismogram calculation is used for full waveform inversion in the first application while arc-parameter basis functions are used to represent the subsurface in the joint inversion of seismic and gravity data. Forward caculation is tailored to rapidly compute the response of the crustal structure for the shear coupled PL waves. To model anisotropic D" layer, we pre-compute the respose of the crust and the mantle layers (without D") for use in each iteration of the waveform inversion. In addition to deriving best fit models we also estimate parameter uncertainty and tradeoff between different parametes using Bayesian statistics. Such results are interpreted in terms of the physics of the problem being investigated.
http://www.ig.utexas.edu
S44A-04 INVITED 16:15h
Kinematics of Reverse Time Imaging
Prestack depth migration produces image gathers, which present the kinematic information implicit in seismic data in the form of redundant images of the subsurface. Most effective velocity estimation techniques rely on analysis of image gathers. This talk will describe a computation of image gathers using the full ("two-way") wave equation, known as reverse time imaging, in which the redundant parameter in the image is (subsurface) offset. Similar computations are routinely accomplished using depth extrapolation ("one-way" wave equation). Amongst other advantages, the two-way computation avoids the aperture limitation inherent in the one-way approach. Analysis also shows that two-way reverse time image gathers are free of kinematic artifacts (energetic events not corresponding to actual reflectors), at least near zero (subsurface) offset. Such kinematic artifacts often contaminate other forms of prestack imaging. This new construction of image gathers may provide a basis for reliable velocity analysis, hence imaging, in regions of severe refraction. such as gas chimneys and partial melts.
S44A-05 16:30h
Inversion of the Teleseismic P Coda for Lithospheric Structure: Examples From Ontario and California
The coda of the teleseismic P wave has become one of the most powerful tools for unravelling fine-scale receiver-side structure, using both single stations and sparse or dense arrays of seismometers. Determining structural information from the coda is an inverse problem that may be treated using either linear or nonlinear methods, depending on what ad hoc assumptions are made about the nature of the coda waves and the structures that generate them. We will review some of the principal methods used in coda imaging and inversion, and examine two methods in greater detail: a non-linear search algorithm applied to single-station data in the presence of anisotropy and dip, and a linearized tomographic inversion of scattered-wave energy in the coda. Examples of applying these methods to detect thinly laminated mantle anisotropy beneath southern Ontario and features correlated with seismicity in California will be given.
S44A-06 16:45h
Source Estimation from Pre-Stack Seismic Data using Nonlinear Bounded Minimization Technique
Source signature estimation is an important problem in seismic data processing and inversion. Inaccurate estimation of the source function leads to incorrect estimates of Earth parameters. In exploration seismics, one common approach is to estimate the source independently of the Earth parameters. These processing-oriented techniques usually make assumptions about wavelet phase and/or the statistics of reflectivity. In this paper we utilize differential (non-parallel) moveout over offset in prestack gathers to estimate the source function (Minkoff et. al, 1997). Our choice of prior information favors oscillatory wavelets and spiky reflectivity traces when input data constraints are weak. The inversion operates on input data having no moveout correction. Therefore wavelet stretch related to moveout correction, which typically degrades bandwidth and resolution, is not an issue. We propose a nonlinear inversion method that minimizes a mixed-norm objective function (L1 norm of source and L2 norm of AVA parameters) subject to fitting the L2 norm of the data. A full-Newton interior point method accommodates our bounds on reflectivity and the mixed norms on model parameters. The system of equations arising from this minimization procedure is solved using a conjugate gradient algorithm. Field data and synthetic data examples illustrate that the inversion methodology effectively deconvolves the source contribution, and at the same time estimates reflectivity with high resolution. We jointly recover mixed-phase wavelets and AVA parameters to within a scale factor. Moreover, the inverted intercept and gradient traces estimated along with wavelets exhibit much greater time resolution than, say, stacked data. We attribute this resolution in part to the inherent designature of the inversion.
http://cgiss.boisestate.edu/~routh/research.html