CONTINUOUS CALIBRATION OF MARINE SEISMIC SOURCES*

The success of signature deconvolution in optimizing both signal-to-noise ratio and time resolution in the seismic section depends critically upon obtaining an accurate estimate of the far-field source signature. Various deterministic estimation schemes have been proposed in recent years, most of which involve direct monitoring of source output within the water layer. As an alternative to elaborate and error-prone source monitoring schemes during data acquisition, a simple modification to any source array permits subsequent estimation of far-field signatures directly from reflected signal. The new method requires the inclusion within any chosen source array of a simple point source, the “reference” source. Initial experiments employed a water gun as the reference source, characterized by a concise implosive signature with peak-to-peak amplitude of approximately 2 bar·m within the seismic sprectrum. In operation the reference source is fired shortly before the main array (typically 2 s during initial trials) and the usual record length is extended by a similar amount. Each recorded trace then comprises two results: the subsurface response to the reference source signal followed by the response of the same subsurface to the main array. The disparities in source amplitudes and NMO differentials ensure that interference effects are negligible in the main recording. Time- or frequency-domain methods can be employed to extract the main array signature from the dual dataset or to invert this to some preferred wavelet simultaneously. As an additional benefit the reference source yields excellent high-resolution profiles of the shallow geology.     

The effects of source and receiver motion on seismic data1

The effects of source and receiver motion on seismic data are considered using extensions of the standard convolutional model. In particular, receiver motion introduces a time-variant spatial shift into data, while source motion converts the effect of the source signature from a single-channel convolution in time to a multichannel convolution in time and space. These results are consistent with classical Doppler theory and suggest that Doppler shifting can introduce distortions into seismic data even at relatively slow acquisition speeds. It is shown that, while both source and receiver motion are known to be important for marine vibroseis acquisition, receiver motion alone can produce significant artifacts in marine 3D data. Fortunately, the convolutional nature of the distortions renders them amenable to correction using simple deconvolution techniques. Specifically, the effects of receiver motion can be neutralized by applying an appropriate reverse time-variant spatial shift, while those due to source motion can be addressed by introducing time-variant spatial shifts both before and after standard, deterministic, signature deconvolution or correlation.     

How to Remedy Non-optimal Seismic Data by Seismic Processing

—Seismic data processing mostly takes into account the statistics inherent in the data to improve the data quality. Since some years the deterministic approach for processing shows many advantages. This approach takes into account e.g., the source signature, with the knowledge of its amplitude and phase behavior. The transformation of the signal into an optimized form is called wavelet processing. By this step an optimal input for deconvolution can be produced, which needs a minimum- delay signal to function well. The interpreter needs a signal which gives the optimum resolution, which is accomplished by the zero-phase transformation of the input signal. The combination of different input sources such as Vibroseis and Dynamite requires a phase adoption. All these procedures can be implemented via Two-Sided-Recursive (TSR-) filters. Spectral balancing can be accomplished very effectively in time domain after a minimum delay transform of the input signals. The DEKORP data suffer from a low signal/noise ratio, so that special methods for the suppression of coherent noise trains were developed. This can be done by subtractive coherency filtering. Multiple seismic reflections also can be suppressed by this method very effectively. All processing procedures developed during recent years are now fully integrated in commercial software operated by the processing center in Clausthal.     

An Approximation to Sparse-Spike Reflectivity Using the Gold Deconvolution Method

Wiener deconvolution is generally used to improve resolution of the seismic sections, although it has several important assumptions. I propose a new method named Gold deconvolution to obtain Earth’s sparse-spike reflectivity series. The method uses a recursive approach and requires the source waveform to be known, which is termed as Deterministic Gold deconvolution. In the case of the unknown wavelet, it is estimated from seismic data and the process is then termed as Statistical Gold deconvolution. In addition to the minimum phase, Gold deconvolution method also works for zero and mixed phase wavelets even on the noisy seismic data. The proposed method makes no assumption on the phase of the input wavelet, however, it needs the following assumptions to produce satisfactory results: (1) source waveform is known, if not, it should be estimated from seismic data, (2) source wavelet is stationary at least within a specified time gate, (3) input seismic data is zero offset and does not contain multiples, and (4) Earth consists of sparse spike reflectivity series. When applied in small time and space windows, the Gold deconvolution algorithm overcomes nonstationarity of the input wavelet. The algorithm uses several thousands of iterations, and generally a higher number of iterations produces better results. Since the wavelet is extracted from the seismogram itself for the Statistical Gold deconvolution case, the Gold deconvolution algorithm should be applied via constant-length windows both in time and space directions to overcome the nonstationarity of the wavelet in the input seismograms. The method can be extended into a two-dimensional case to obtain time-and-space dependent reflectivity, although I use one-dimensional Gold deconvolution in a trace-by-trace basis. The method is effective in areas where small-scale bright spots exist and it can also be used to locate thin reservoirs. Since the method produces better results for the Deterministic Gold deconvolution case, it can be used for the deterministic deconvolution of the data sets with known source waveforms such as land Vibroseis records and marine CHIRP systems.     

A COMPARISON OF STATISTICAL AND DETERMINISTIC WIENER DECONVOLUTION OF DEEP-TOW SEISMIC DATA*

Wiener ‘spiking’ deconvolution of seismic traces in the absence of a known source wavelet relies upon the use of digital filters, which are optimum in a least-squares error sense only if the wavelet to be deconvolved is minimum phase. In the marine environment in particular this condition is frequently violated, since bubble pulse oscillations result in source signatures which deviate significantly from minimum phase. The degree to which the deconvolution is impaired by such violation is generally difficult to assess, since without a measured source signature there is no optimally deconvolved trace with which the spiked trace may be compared. A recently developed near-bottom seismic profiler used in conjunction with a surface air gun source produces traces which contain the far-field source signature as the first arrival. Knowledge of this characteristic wavelet permits the design of two-sided Wiener spiking and shaping filters which can be used to accurately deconvolve the remainder of the trace. In this paper the performance of such optimum-lag filters is compared with that of the zero-lag (one-sided) operators which can be evaluated from the reflected arrival sequence alone by assuming a minimum phase source wavelet. Results indicate that the use of zero-lag operators on traces containing non-minimum phase wavelets introduces significant quantities of noise energy into the seismic record. Signal to noise ratios may however be preserved or even increased during deconvolution by the use of optimum-lag spiking or shaping filters. A debubbling technique involving matched filtering of the trace with the source wavelet followed by optimum-lag Wiener deconvolution did not give a higher quality result than can be obtained simply by the application of a suitably chosen Wiener shaping filter. However, cross correlation of an optimum-lag spike filtered trace with the known ‘actual output’ of the filter when presented with the source signature is found to enhance signal-to-noise ratio whilst maintaining improved resolution.     

A derivative-based interpretation approach to estimating source parameters of simple 2D magnetic sources from Euler deconvolution, the analytic-signal method and analytical expressions of the anomalies

The major advantage of using either the analytic‐signal or the Euler‐deconvolution technique is that we can determine magnetic‐source locations and depths independently of the ambient earth magnetic parameters. In this study, we propose adopting a joint analysis of the analytic signal and Euler deconvolution to estimate the parameters of 2D magnetic sources. The results can avoid solution bias from an inappropriate magnetic datum level and can determine the horizontal locations, depths, structural types (indices), magnetization contrasts and/or structural dips. We have demonstrated the feasibility of the proposed method on 2D synthetic models, such as magnetic contacts (faults), thin dikes and cylinders. However, the method fails to solve the parameters of magnetic sources if there is severe interference between the anomalies of two adjacent magnetic sources.     

非稳态地震记录时变子波估计

地震子波估计是地震资料处理和解释中的一个关键问题,子波估计的可靠性会直接影响反褶积和反演的准确度.现有的子波估计方法分为确定型和统计型两种类型,本文通过结合这两类方法,利用确定型的谱分析法和统计型的偏度最大化方法,分别提取时变子波的振幅和相位信息,得到估计的时变子波.这种方法不需要对子波进行任何时不变或相位等的假设,具有对时变相位的估计能力.进而利用估计时变子波进行非稳态反褶积,提高地震记录的保真度,为精细储层预测和描述提供高质量的剖面.理论模型试算验证了方法的可行性,通过实际地震资料的处理应用,表明该方法能有效地提取出子波时变信息.     

基于最小反演拟合差的重磁场源深度计算方法

以等效源及位场物性反演为基础,本文提出一种新的求取重磁场源深度的方法.该方法将一层等效源以一定的间隔从浅部向深部移动,并将等效源作为初始模型进行反演,当反演拟合差最小时,停止反演,此时的等效源底深即为所求场源的中心深度.由于仅需要反演一层等效源,比传统的物性反演计算时间大大减少,并且不需要进行深度加权约束.理论模型数据处理结果表明该方法能够获得较准确的场源深度:以长宽比为7.5的薄板模型为例,深度计算误差约为1个点距(25 m);以长宽比为0.5~1.5的厚板模型为例,深度计算误差小于1个点距(25m).将该方法应用于实测航磁梯度数据,计算的磁源中心深度在200~250m之间,钻井资料显示该异常由埋藏深度在200~300m的闪长岩引起,计算结果与钻井资料较吻合.     

Euler deconvolution of the analytic signal and its application to magnetic interpretation

Euler deconvolution and the analytic signal are both used for semi‐automatic interpretation of magnetic data. They are used mostly to delineate contacts and obtain rapid source depth estimates. For Euler deconvolution, the quality of the depth estimation depends mainly on the choice of the proper structural index, which is a function of the geometry of the causative bodies. Euler deconvolution applies only to functions that are homogeneous. This is the case for the magnetic field due to contacts, thin dikes and poles. Fortunately, many complex geological structures can be approximated by these simple geometries. In practice, the Euler equation is also solved for a background regional field. For the analytic signal, the model used is generally a contact, although other models, such as a thin dike, can be considered. It can be shown that if a function is homogeneous, its analytic signal is also homogeneous. Deconvolution of the analytic signal is then equivalent to Euler deconvolution of the magnetic field with a background field. However, computation of the analytic signal effectively removes the background field from the data. Consequently, it is possible to solve for both the source location and structural index. Once these parameters are determined, the local dip and the susceptibility contrast can be determined from relationships between the analytic signal and the orthogonal gradients of the magnetic field. The major advantage of this technique is that it allows the automatic identification of the type of source. Implementation of this approach is demonstrated for recent high‐resolution survey data from an Archean granite‐greenstone terrane in northern Ontario, Canada.     

Vibroseis deconvolution: A comparison of pre and post correlation vibroseis deconvolution data in real noisy data

Vibroseis is a source used commonly for inland seismic exploration. This non-destructive source is often used in urban areas with strong environmental noise. The main goal of seismic data processing is to increase the signal/noise ratio where a determinant step is deconvolution. Vibroseis seismic data do not meet the basic minimum-phase assumption for the application of spiking and predictive deconvolution, therefore various techniques, such as phase shift, are applied to the data, to be able to successfully perform deconvolution of vibroseis data.This work analyzes the application of deconvolution techniques before and after cross-correlation on a real data set acquired for high resolution prospection of deep aquifers. In particular, we compare pre-correlation spiking and predictive deconvolution with Wiener filtering and with post-correlation time variant spectral whitening deconvolution. The main result is that at small offsets, post cross-correlation spectral whitening deconvolution and pre-correlation spiking deconvolution yield comparable results, while for large offsets the best result is obtained by applying a pre-cross-correlation predictive deconvolution.     

Receiver function method in reflection seismology

The receiver function method was originally developed to analyse earthquake data recorded by multicomponent (3C) sensors and consists in deconvolving the horizontal component by the vertical component. 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 receiver function signal is the local PS wave's impulse response generated at impedance contrasts below the 3C receiver.We propose to adapt this technique to the wide‐angle multi‐component reflection acquisition geometry. We focus on the simplest case of land data reflection acquisition. Our adapted version of the receiver function approach consists in a multi‐step procedure that first removes the P wavefield recorded on the horizontal component and next removes the source signature. The separation step is performed in the τ?p domain while the source designature can be achieved in either the τ?p or the t?x domain. Our technique does not require any a priori knowledge of the subsurface. The resulting receiver function is a pure PS‐wave reflectivity response, which can be used for amplitude versus slowness or offset analysis. Stack of the receiver function leads to a high‐quality S wave image.     

Prestack nonstationary deconvolution based on variable-step sampling in the radial trace domain

The conventional nonstationary convolutional model assumes that the seismic signal is recorded at normal incidence. Raw shot gathers are far from this assumption because of the effects of offsets. Because of such problems, we propose a novel prestack nonstationary deconvolution approach. We introduce the radial trace （RT） transform to the nonstationary deconvolution, we estimate the nonstationary deconvolution factor with hyperbolic smoothing based on variable-step sampling （VSS） in the RT domain, and we obtain the high-resolution prestack nonstationary deconvolution data. The RT transform maps the shot record from the offset and traveltime coordinates to those of apparent velocity and traveltime. The ray paths of the traces in the RT better satisfy the assumptions of the convolutional model. The proposed method combines the advantages of stationary deconvolution and inverse Q filtering, without prior information for Q. The nonstationary deconvolution in the RT domain is more suitable than that in the space-time （XT） domain for prestack data because it is the generalized extension of normal incidence. Tests with synthetic and real data demonstrate that the proposed method is more effective in compensating for large-offset and deep data.     

An Improved Analytic Signal Technique for the Depth and Structural Index from 2D Magnetic Anomaly Data

This paper presents a new inversion method for the interpretation of 2D magnetic anomaly data, which uses the combination of the analytic signal and its total gradient to estimate the depth and the nature (structural index) of an isolated magnetic source. However, our proposed method is sensitive to noise. In order to lower the effect of noise, we apply upward continuation technique to smooth the anomaly. Tests on synthetic noise-free and noise corrupted magnetic data show that the new method can successfully estimate the depth and the nature of the causative source. The practical application of the technique is applied to measured magnetic anomaly data from Jurh area, northeast China, and the inversion results are in agreement with the inversion results from Euler deconvolution of the analytic signal.     

Can Euler deconvolution outline three-dimensional magnetic sources?

Severe limitations of the standard Euler deconvolution to outline source shapes have been pointed out. However, Euler deconvolution has been widely employed on field data to outline interfaces, as faults and thrust zones. We investigate the limitations of the 3D Euler deconvolution–derived estimates of source dip and volume with the use of reduced-to-the-pole synthetic and field anomalies. The synthetic anomalies are generated by two types of source bodies: (1) uniformly magnetized prisms, presenting either smooth or rough interfaces, and (2) bodies presenting smooth delimiting interfaces but strong internal variation of magnetization intensity. The dip of the first type of body might be estimated from the Euler deconvolution solution cluster if the ratio between the depth to the top and vertical extent is relatively high (>1/4). For the second type of body, besides dip, the source volume can be approximately delimited from the solution cluster envelope, regardless of the referred ratio. We apply Euler deconvolution to two field anomalies which are caused by a curved-shape thrust zone and by a banded iron formation. These anomalies are chosen because they share characteristics with the two types of synthetic bodies. For the thrust zone, the obtained Euler deconvolution solutions show spatial distribution allowing to estimate a source dip that is consistent with the surface geology data, even if the above-mentioned ratio is much less than 1/4. Thus, there are other factors, such as a heterogeneous magnetization, which might be controlling the vertical spreading of the Euler deconvolution solutions in the thrust zone. On the other hand, for the iron-ore formation, the solution cluster spreads out occupying a volume, in accordance with the results obtained with the synthetic sources having internal variation of magnetization intensity. As conclusion, although Euler deconvolution–derived solutions cannot offer accurate estimates of source shapes, they might provide a sufficient degree of reliability in the initial estimates of the source dip and volume, which may be useful in a later phase of more accurate modelling.     

Elimination of the water-layer response from multi-component source and receiver marine electromagnetic data

This paper presents the theory to eliminate from the recorded multi‐component source, multi‐component receiver marine electromagnetic measurements the effect of the physical source radiation pattern and the scattering response of the water‐layer. The multi‐component sources are assumed to be orthogonally aligned above the receivers at the seabottom. Other than the position of the sources, no source characteristics are required. The integral equation method, which for short is denoted by Lorentz water‐layer elimination, follows from Lorentz' reciprocity theorem. It requires information only of the electromagnetic parameters at the receiver level to decompose the electromagnetic measurements into upgoing and downgoing constituents. Lorentz water‐layer elimination replaces the water layer with a homogeneous half‐space with properties equal to those of the sea‐bed. The source is redatumed to the receiver depth. When the subsurface is arbitrary anisotropic but horizontally layered, the Lorentz water‐layer elimination scheme greatly simplifies and can be implemented as deterministic multi‐component source, multi‐component receiver multidimensional deconvolution of common source gathers. The Lorentz deconvolved data can be further decomposed into scattering responses that would be recorded from idealized transverse electric and transverse magnetic mode sources and receivers. This combined electromagnetic field decomposition on the source and receiver side gives data equivalent to data from a hypothetical survey with the water‐layer absent, with idealized single component transverse electric and transverse magnetic mode sources and idealized single component transverse electric and transverse magnetic mode receivers. When the subsurface is isotropic or transverse isotropic and horizontally layered, the Lorentz deconvolution decouples into pure transverse electric and transverse magnetic mode data processing problems, where a scalar field formulation of the multidimensional Lorentz deconvolution is sufficient. In this case single‐component source data are sufficient to eliminate the water‐layer effect. We demonstrate the Lorentz deconvolution by using numerically modeled data over a simple isotropic layered model illustrating controlled‐source electromagnetic hydrocarbon exploration. In shallow water there is a decrease in controlled‐source electromagnetic sensitivity to thin resistors at depth. The Lorentz deconvolution scheme is designed to overcome this effect by eliminating the water‐layer scattering, including the field's interaction with air.     

Optimum pilot sweep

The successful application of high-resolution seismic methods requires evaluating each element in the seismic system and ensuring that each part of the system contributes optimally to the success of the method. Unfortunately, unlike data processing, seismic signal generation is not carefully optimized. The purpose of our study was to optimize the source signal in order to better coordinate field operations with subsequent data processing to achieve their common objective. We developed an iterative method for a rational frequency distribution of the energy of a seismic source. The method allows the optimum amplitude spectrum of a source signal to be calculated, thus providing the best data quality at the end of the processing. We assume that the source signal is affected by a total transfer function, by the reflectivity function of a target interval, and by ambient noise, whose characteristics, if not known, can be estimated or measured in practice. The transfer function includes data processing other than the correlation stage and the final trace-optimizing filter. The variance of a reflectivity estimate is considered to be a measure of the data quality and improvement of the characteristic corresponds to a decrease in the variance. For this reason, a constrained Wiener deconvolution filter is used as the final trace-optimizing filter. It not only minimizes the variance of a reflectivity estimate but also ensures a specific signal-to-noise ratio. The method is made feasible by following the Vibroseis technique, primarily because of the versatility of the technique in controlling the signal spectrum. With the optimum amplitude spectrum obtained, the corresponding optimum pilot sweep can be readily calculated. Examples using synthetic data are presented to illustrate the method.     

DEBUBBLING: A GENERALIZED LINEAR INVERSE APPROACH*

On seismograms recorded at sea bubble pulse oscillations can present a serious problem to an interpreter. We propose a new approach, based on generalized linear inverse theory, to the solution of the debubbling problem. Under the usual assumption that a seismogram can be modelled as the convolution of the earth's impulse response and a source wavelet we show that estimation of either the wavelet or the impulse response can be formulated as a generalized linear inverse problem. This parametric approach involves solution of a system of equations by minimizing the error vector (ΔX = X^obs– X^cal) in a least squares sense. One of the most significant results is that the method enables us to control the accuracy of the solution so that it is consistent with the observational errors and/or known noise levels. The complete debubbling procedure can be described in four steps: (1) apply minimum entropy deconvolution to the observed data to obtain a deconvolved spike trace, a first approximation to the earth's response function; (2) use this trace and the observed data as input for the generalized linear inverse procedure to compute an estimated basic bubble pulse wavelet; (3) use the results of steps 1 and 2 to construct the compound source signature consisting of the primary pulse plus appropriate bubble oscillations; and (4) use the compound source signature and the observed data as input for the generalized linear inverse method to determine the estimated earth impulse response—a debubbled, deconvolved seismogram. We illustrate the applicability of the new approach with a set of synthetic seismic traces and with a set of field seismograms. A disadvantage of the procedure is that it is computationally expensive. Thus it may be more appropriate to apply the technique in cases where standard analysis techniques do not give acceptable results. In such cases the inherent advantages of the method may be exploited to provide better quality seismograms.     

CONTROLE OPTIQUE DES TRAITEMENTS NUMÉRIQUES DÉCONVOLUTION OPTIQUE DES TRACES SISMIQUES *

A checking method of digital multiple elimination and of deconvolution processing using computers and based on optical autocorrelation is first presented. Comparison between autocorrelograms before and after a single or several processing steps allows to estimate, on one hand, the strength of the deconvolution obtained, known by the study of the central parts which is in fact the signal autocorrelation, on the other hand, the multiple elimination given by the study of side parts of the autocorrelogram. Further, an optical deconvolution procedure, is presented. For this, it is supposed that the signal is known and optically reproduced in the same way as the one of a trace. This is achieved by sphero-cylindrical optics allowing trace to trace processing. Deconvolution is carried out in the spectral domain by inserting a filter in the Fourier plane of the optical unit, the transmission law of which expresses the Fourier transform of the antisignal. This filter device introduces a holographic technique called Fourier holography, in such a way phases as well as amplitudes are preserved. Several results are presented from a synthetic section and also from field sections.     

预条件共轭梯度反褶积的改进及其应用

预条件共轭梯度反褶积方法是结合盲反褶积的实现,运用基于Krylov子空间上优化的预条件共轭梯度法,完成反射系数的反演.用该方法处理地震资料时可提高资料频率,展宽有效频率宽度.但由于地震数据对不同频带的信噪比有差异,若直接运用该反褶积处理常伴随分辨率提高的同时出现信噪比显著降低的现象.对于此,本文采取如下方法的改进措施：①在时间域上,当地震数据的振幅较大时,对应的反褶积数据的振幅取值与原地震数据的振幅相等;②在频率域上,当地震数据的频谱幅值大于一定阀值时,对应的反褶积数据的频谱取做原地震数据的频谱.由本文所给的数值算例可以看出,此两项改进方法可取得较好的实用效果.     

Estimation of the seismic wavelet through homomorphic deconvolution and well log data: application on well-to-seismic tie procedure

Wavelet estimation and well-tie procedures are important tasks in seismic processing and interpretation. Deconvolutional statistical methods to estimate the proper wavelet, in general, are based on the assumptions of the classical convolutional model, which implies a random process reflectivity and a minimum-phase wavelet. The homomorphic deconvolution, however, does not take these premises into account. In this work, we propose an approach to estimate the seismic wavelet using the advantages of the homomorphic deconvolution and the deterministic estimation of the wavelet, which uses both seismic and well log data. The feasibility of this approach is verified on well-to-seismic tie from a real data set from Viking Graben Field, North Sea, Norway. The results show that the wavelet estimated through this methodology produced a higher quality well tie when compared to methods of estimation of the wavelet that consider the classical assumptions of the convolutional model.