Erroneous Phase Velocities from Frequency-Wavenumber Spectral Sections

Summary Two-dimensional cross-sections of finite frequency-wavenumber spectra can easily be misinterpreted, since leakage of energy occurs along lines of constant wavenumber. In particular, the signal phase velocity determined from measurements on cross-sections normal to the frequency axis can be incorrect. We discuss an algorithm which corrects this situation. Examples using real and synthetic data are given.     

An algebraic formulation of geophysical inverse problems

Many geophysical inverse problems derive from governing partial differential equations with unknown coefficients. Alternatively, inverse problems often arise from integral equations associated with a Green's function solution to a governing differential equation. In their discrete form such equations reduce to systems of polynomial equations, known as algebraic equations. Using techniques from computational algebra one can address questions of the existence of solutions to such equations as well as the uniqueness of the solutions. The techniques are enumerative and exhaustive, requiring a finite number of computer operations. For example, calculating a bound to the total number of solutions reduces to computing the dimension of a linear vector space. The solution set itself may be constructed through the solution of an eigenvalue problem. The techniques are applied to a set of synthetic magnetotelluric values generated by conductivity variations within a layer. We find that the estimation of the conductivity and the electric field in the subsurface, based upon single-frequency magnetotelluric field values, is equivalent to a linear inverse problem. The techniques are also illustrated by an application to a magnetotelluric data set gathered at Battle Mountain, Nevada. Surface observations of the electric ( E _y ) and magnetic ( H _x ) fields are used to construct a model of subsurface electrical structure. Using techniques for algebraic equations it is shown that solutions exist, and that the set of solutions is finite. The total number of solutions is bounded above at 134 217 728. A numerical solution of the algebraic equations generates a conductivity structure in accordance with the current geological model for the area.     

Three-dimensional massively parallel electromagnetic inversion—I. Theory

An iterative solution to the non-linear 3-D electromagnetic inverse problem is obtained by successive linearized model updates using the method of conjugate gradients. Full wave equation modelling for controlled sources is employed to compute model sensitivities and predicted data in the frequency domain with an efficient 3-D finite-difference algorithm. Necessity dictates that the inverse be underdetermined, since realistic reconstructions require the solution for tens of thousands of parameters. In addition, large-scale 3-D forward modelling is required and this can easily involve the solution of over several million electric field unknowns per solve. A massively parallel computing platform has therefore been utilized to obtain reasonable execution times, and results are given for the 1840-node Intel Paragon. The solution is demonstrated with a synthetic example with added Gaussian noise, where the data were produced from an integral equation forward-modelling code, and is different from the finite difference code embedded in the inversion algorithm     

A fast converging and anti-aliasing algorithm for Green's functions in terms of spherical or cylindrical harmonics

Geophysical observables are generally related to earth structure and source parameters in a complicated non-linear way. Consequently, a large number of forward modelling processes are commonly necessary to obtain a satisfactory estimate of such parameters from observed data. The most time-consuming part of the forward modelling is the computation of the Green's functions of the different earth models to be tested. In this study, we present a fast converging algorithm: the differential transform method for the computation of Green's functions in terms of spherical or cylindrical harmonics. In this method, a deconvolutable high-pass filter is used to enhance the numerical significance of the far-field spectrum of Green's functions. Compared with existing fast converging algorithms such as the Kummer's transformation and the disc factor method, the differential transform method is more efficient except for the extremely near-source region. The new method can be used to suppress numerical phases (non-physical seismic signals) associated with the aliasing effect that may arise in synthetic seismograms when the latter are computed from a windowed wavenumber (or slowness) spectrum. The numerical efficiency of the new method is demonstrated via two representative tests.     

Solution of the H-polarization induction problem by a wavenumber integral equation approach

Summary. The H -polarization induction problem is solved in terms of an integral equation, which in the horizontal direction is transformed into the wavenumber domain. By this transformation the usual complicated integral expressions for the Green's tensor elements are removed. By extracting asymptotic features from the system of linear equations, we reduce the number of equations considerably independent of whether the horizontal variation in the conductivity is continuous or discontinuous. Likewise we reformulate the problem so that arbitrary conductivity contrasts may be studied. The method is finally tested by comparing with analytic solutions, and good agreement is achieved. Furthermore the numerical results indicate that a small amount of wavenumbers is required.     

Electric field at the seafloor due to a two-dimensional ionospheric current

The relation between the seafloor electric field and the surface magnetic field is studied. It is assumed that the fields are created by a 2-D ionospheric current distribution resulting in the E-polarization. The layered earth below the sea water is characterized by a surface impedance. The electric field at the seafloor can be expressed either as an inverse Fourier transform integral over the wavenumber or as a spatial convolution integral. In both integrals the surface magnetic field is multiplied by a function that depends on the depth and conductivity of the sea water and on the properties of the basement. The fact that surface magnetic data are usually available on land, not at the sea surface, is also considered. Test computations demonstrate that the numerical inaccuracies involved in the convolution method are negligible. The theoretical equations are applied to calculate the seafloor electric fields due to an ionospheric line current or associated with real magnetic data collected by the IMAGE magnetometer array in northern Europe. Two different sea depths are considered: 100 m (the continental shelf) and 5 km (the deep ocean). It is seen that the dependence of the electric field on the oscillation period is weaker in the 5 km case than for 100 m.     

Seismogram synthesis using normal mode superposition: the locked mode approximation

Summary. A normal mode superposition approach is used to synthesize complete seismic codas for flat layered earth models and the P-SV phases. Only modes which have real eigenwavenumbers are used so that the search for eigenvalues in the complex wavenumber plane is confined to the real axis. In order to synthesize early P -wave arrivals by summing a number of'trapped'modes, an anomalously high velocity cap layer is added to the bottom of the structure so that most of the seismic energy is contained in the upper layers as high-order surface waves. Causality arguments are used to define time windows for which the resulting synthetic seismograms are close approximations to the exact solutions without the cap layer. The traditional Thomson—Haskell matrix approach to computing the normal modes is reformulated so that numerical problems encountered at high frequencies are avoided and numerical results of the locked mode approximation are given.     

A transformational approach to geophysical inverse problems

A set of coordinate transformations is used to linearize a general geophysical inverse problem. Statistical and analytic techniques are employed to estimate the parameters of such linearization transformations. In the transformed space, techniques from linear inverse theory may be utilized. Consequently, important concepts, such as model parameter covariance, model parameter resolution and averaging kernels, may be carried over to non-linear inverse problems. I apply the approach to a set of seismic cross-borehole traveltimes gathered at the Conoco Borehole Test Facility. the seismic survey was conducted within the Fort Riley formation, a limestone with thin interbedded shales. Between the boreholes, the velocity structure of the Fort Riley formation consists of a high-velocity region overlying a section of lower velocity. It is found that model parameter resolution is poorest and spatial averaging lengths are greatest in the underlying low-velocity region.     

Inversion of a normally incident reflection seismogram by the Gopinath–Sondhi integral equation

Summary. The one-dimensional acoustic wave equation has been transformed to two coupled first-order equations whose inverse solution is obtained through application of the Gopinath and Sondhi integral equation. A scattering solution of the Schrödinger wave equation for an explosive source leads us to express the kernel of the Gopinath–Sondhi integral equation in terms of a seismic reflection response. A numerical solution of the integral equation obtained by a trapezoidal rule yields a continuous impedance profile whose derivative has step-like discontinuities. The method is illustrated with computer model studies.     

Efficient global matrix approach to the computation of synthetic seismograms

Summary. A numerically efficient global matrix approach to the solution of the wave equation in horizontally stratified environments is presented. The field in each layer is expressed as a superposition of the field produced by the sources within the layer and an unknown field satisfying the homogeneous wave equations, both expressed as integral representations in the horizontal wavenumber. The boundary conditions to be satisfied at each interface then yield a linear system of equations in the unknown wavefield amplitudes, to be satisfied at each horizontal wavenumber. As an alternative to the traditional propagator matrix approaches, the solution technique presented here yields both improved efficiency and versatility. Its global nature makes it well suited to problems involving many receivers in range as well as depth and to calculations of both stresses and particle velocities. The global solution technique is developed in close analogy to the finite element method, thereby reducing the number of arithmetic operations to a minimum and making the resulting computer code very efficient in terms of computation time. These features are illustrated by a number of numerical examples from both crustal and exploration seismology.     

Exact synthetic seismograms for an inhomogeneity in a layered elastic half-space

Summary. The propagation of a pulsed elastic wave in the following geometry is considered. An elastic half-space has a surface layer of a different material and the layer furthermore contains a bounded 3-D inhomogeneity. The exciting source is an explosion, modelled as an isotropic pressure point source with Gaussian behaviour in time.
The time-harmonic problem is solved using the null field approach (the T matrix method), and a frequency integral then gives the time-domain response. The main tools of the null field approach are integral representations containing the free space Green's dyadic, expansions in plane and spherical vector wave functions, and transformations between plane and spherical vector wave functions. It should be noted that the null field approach gives the solution to the full elastodynamic equations with, in principle, an arbitrarily high accuracy. Thus no ray approximations or the like are used. The main numerical limitation is that only low and intermediate frequencies, in the sense that the diameter of the inhomogeneity can only be a few wavelengths, can be considered.
The numerical examples show synthetic seismograms consisting of data from 15 observation points at increasing distances from the source. The normal component of the velocity field is computed and the anomalous field due to the inhomogeneity is sometimes shown separately. The shape of the inhomogeneity, the location and depth of the source, and the material parameters are all varied to illustrate the relative importance of the various parameters. Several specific wave types can be identified in the seismograms: Rayleigh waves, direct and reflected P -waves, and head waves.     

Calculation of the elasto-dynamic Green's function in layered media by means of a modified propagator matrix method

Summary. The calculation of the two-dimensional elasto-dynamic Green's function for a stratified medium is investigated. The solution is represented in the form of an inverse Fourier integral which is to be integrated along a properly chosen path in the complex wavenumber plane. The integrand is computed using a modified propagator matrix method.
This method is based on a mixed formulation using the propagator matrix and the matrix of minors of the propagator matrix (compound matrix). The major advantages of this approach are the elimination of the numerical loss of precision problems associated with the Thomson-Haskell formulation, without losing the attractive tractability and compactness of the propagator matrix method. This modified method is first mathematically derived, and theoretical seismograms are then presented for two examples.     

The theory of finite frequency body wave synthetic seismograms in inhomogeneous elastic media

b
A new method is presented by means of which one can compute finite frequency synthetic seismograms for media whose velocity and density are continuous functions of two or three spatial variables. Basically, the method is a generalization of the familiar phase integral method, to which it reduces in a stratified medium. For a given source location the travel-time and distance functions needed to compute synthetics are obtained by numerically tracing rays through the model. This information is then used to evaluate a double integral over frequency and take-off angle at the source. The solution obtained reduces to the geometrical optics solution wherever that is valid but it also works in shadows and at caustics without knowing explicitly where these may be located. The method can be used as a spectral method, in which the integral over take-off angle is evaluated first, or as a slowness method, in which the frequency integral is evaluated first.     

Green's functions for modelling the EM response of a target near a dipping contact

Summary. Although the integral equation method has shown promise for realistic modelling of electromagnetic exploration techniques, it has so far been limited to targets in horizontally layered hosts. This note shows how the Green's functions appropriate for horizontal layers can be used for targets near a dipping contact.     

On the relative scattering of P- and S-waves

Summary. Using a single scattering approximation, we derive equations for the scattering attenuation coefficients of P- and S -body waves. We discuss our results in the light of some recent energy renormalization approaches to seismic wave scattering. Practical methods for calculating the scattering attenuation coefficients for various earth models are emphasized. The conversions of P - to S -waves and S- to P -waves are included in the theory. The earth models are assumed to be randomly inhomogeneous, with their properties known only through their average wavenumber power spectra. We approximate the power spectra with piecewise constant functions, each segment of which contributes to the net, frequency-dependent, scattering attenuation coefficient. The smallest and largest wavenumbers of a segment can be plotted along with the wavevectors of the incident and scattered waves on a wavenumber diagram. This diagram gives a geometric interpretation for the frequency behaviour associated with each spectral segment, including a 'transition' peak that is due entirely to the wavenumber limits of the segment. For regions of the earth where the inhomogeneity spectra are concentrated in a band of wavenumbers, it should be possible to observed such a peak in the apparent attenuation of seismic waves. We give both the frequency and distance limits on the accuracy of the theoretical results.     

Automatic differentiation in geophysical inverse problems

Automatic differentiation (AD) is the technique whereby output variables of a computer code evaluating any complicated function (e.g. the solution to a differential equation) can be differentiated with respect to the input variables. Often AD tools take the form of source to source translators and produce computer code without the need for deriving and hand coding of explicit mathematical formulae by the user. The power of AD lies in the fact that it combines the generality of finite difference techniques and the accuracy and efficiency of analytical derivatives, while at the same time eliminating 'human' coding errors. It also provides the possibility of accurate, efficient derivative calculation from complex 'forward' codes where no analytical derivatives are possible and finite difference techniques are too cumbersome. AD is already having a major impact in areas such as optimization, meteorology and oceanography. Similarly it has considerable potential for use in non-linear inverse problems in geophysics where linearization is desirable, or for sensitivity analysis of large numerical simulation codes, for example, wave propagation and geodynamic modelling. At present, however, AD tools appear to be little used in the geosciences. Here we report on experiments using a state of the art AD tool to perform source to source code translation in a range of geoscience problems. These include calculating derivatives for Gibbs free energy minimization, seismic receiver function inversion, and seismic ray tracing. Issues of accuracy and efficiency are discussed.     

On azimuthal eigenwavenumbers associated with Laplace's tidal equations

Laplace's tidal equations for the case of an ocean of constant depth bounded by meridians were considered by two authors at a specific frequency as an eigenvalue problem in the azimuthal wavenumber. A finite spectrum of eigenwavenumbers was found. That eigenvalue problem is re-examined by means of asymptotic techniques and numerical integration of the governing equation of the problem. At low frequencies a formula connecting the frequency and the number of eigensolutions is established. It is shown that at a given frequency the spectrum of eigenwavenumbers is wider than that reported, but (for this type of solution) the meridional boundary conditions are satisfied approximately only for the case of very low frequencies.     

Generalized Born inversion of seismic reflection data

Summary . Born inverse methods give accurate and stable results when the source wavelet is impulsive. However, in many practical applications (reflection seismology) an impulsive source cannot be realized and the inversion needs to be generalized to include an arbitrary source function. In this paper, we present a Born solution to the seismic inverse problem which can accommodate an arbitrary source function and give accurate and stable results. It is shown that the form of the generalized inversion algorithm reduces to a Wiener shaping ***filter, which is solved efficiently using a Levinson recursion algorithm. Numerical examples of synthetic and real field data illustrate the validity of our method.     

Computation and visualisation of the accuracy of old maps using differential distortion analysis

The accuracy of old maps can hold interesting historical information, and is therefore studied using distortion analysis methods. These methods start from a set of ground control points that are identified both on the old map and on a modern reference map or globe, and conclude with techniques that compute and visualise distortion. Such techniques have advanced over the years, but leave room for improvement, as the current ones result in approximate values and a coarse spatial resolution. We propose a more elegant and more accurate way to compute distortion of old maps by translating the technique of differential distortion analysis, used in map projection theory, to the setting where an old map and a reference map are directly compared. This enables the application of various useful distortion metrics to the study of old maps, such as the area scale factor, the maximum angular distortion and the Tissot indicatrices. As such a technique is always embedded in a full distortion analysis method we start by putting forward an optimal analysis method for a general-purpose study, which then serves as the foundation for the development of our technique. Thereto, we discuss the structure of distortion analysis methods and the various options available for every step of the process, including the different settings in which the old map can be compared to its modern counterpart, the techniques that can be used to interpolate between both, and the techniques available to compute and visualise the distortion. We conclude by applying our general-purpose method, including the differential distortion analysis technique, to an example map also used in other literature.     

CLASSIFICATION:OLDTIMERS AND NEWCOMERS

Classification and regression techniques are among the most used tools by chemometricians.Withclassification,the two classic methods are discriminant analysis and SIMCA.In this paper we discuss theconnection between these two methods and introduce two new ones of the same family:DASCO(discriminantanalysis with shrunken covariances)and RDA(regularized discriminant analysis).We demonstrate on bothsimulated and real data sets that their performance is superior to the old favorites.This is especially truein small-sample/high-dimension settings typical in chemistry.