Ghosting and marine signature deconvolution: A prerequisite for detailed seismic interpretation

Geophysics ◽  
1983 ◽  
Vol 48 (11) ◽  
pp. 1468-1485 ◽  
Author(s):  
Dushan B. Jovanovich ◽  
Roger D. Sumner ◽  
Sharon L. Akins‐Easterlin
Keyword(s):  
Gulf Coast ◽  
Seismic Line ◽  
Near Surface ◽  
Seismic Wavelet ◽  
Phase Errors ◽  
Air Gun ◽  
Statistical Hypotheses ◽  
Seismic Sections ◽  
Sonic Logs ◽  
Zero Phase

Detailed lithologic interpretation of seismic sections and/or pseudo‐sonic logs generated from seismic data requires that the seismic trace can be modeled as a reflection series convolved with a zero‐phase broadband wavelet. Ghosting and marine signature deconvolution processing is a prerequisite for assuring that the seismic wavelet on a marine CDP section will be zero phase. A deterministic approach to deconvolution is centered around the concept of abandoning the purely statistical method of wavelet estimation and actually measuring the seismic wavelet. A proper signature recording for marine data is, therefore, a crucial component of deterministic deconvolution. Another important element in the deterministic deconvolution sequence is the application of a deghosting filter to remove near‐surface reflections. Proper application of a deghosting filter significantly improves the correlation between log synthetics and the seismic trace. It has been found that statistical deconvolution schemes, because of the number of statistical hypotheses required to produce a deconvolution filter, produce residual wavelets that are highly variable in character and whose average phases cover the entire phase spectrum, modulo 2π. Examples of a Gulf Coast marine line which was shot with Aquapulse™, air gun, and Maxipulse™ sources by the RV Hollis Hedberg are presented to demonstrate the differences between statistical and deterministic deconvolution processing sequences. It will be shown, using sonic logs from wells adjacent to the seismic line, that the deterministic deconvolution sections for all three sources are close to zero phase while the statistical deconvolution sections have residual average phase errors between 180 and 270 degrees. The deterministic deconvolution sections have a high degree of correlation among themselves and to the wells adjacent to the line, while the statistical deconvolution sections correlate poorly to each other and to the wells. Synthetic seismograms and their impedance logs, and the seismic sections and their corresponding pseudo‐sonic logs, are used to demonstrate how deconvolution influences lithologic interpretation. ™Western Geophysics Co.

Empirical observations relating near‐surface magnetic anomalies to high‐frequency seismic data and Landsat data in eastern Sheridan County, Montana

Geophysics ◽  
1991 ◽  
Vol 56 (10) ◽  
pp. 1553-1570 ◽  
Author(s):  
John A. Andrew ◽  
Duncan M. Edwards ◽  
Robert J. Graf ◽  
Richard J. Wold
Keyword(s):  
Seismic Data ◽  
Magnetic Anomalies ◽  
Oil Fields ◽  
Data Types ◽  
Seismic Line ◽  
Near Surface ◽  
The North ◽  
Basement Faults ◽  
Magnetic Basement ◽  
Seismic Sections

Our empirical synergistic correlations of aeromagnetic and seismic data and a Landsat lineament interpretation revealed lineations on the magnetic map that have expression on seismic sections. We observed a conjugate set of northwest‐southeast and northeast‐southwest trending magnetic lineaments (zones which offset and truncate near‐surface magnetic anomalies). We believe these OZs (offset zones) represent lateral faults in a wrench‐fault system. Lateral offsets appear to be 100s of meters to a few kilometers (fractions of a mile to a few miles). We observed a direct correlation between OZs and vertical faults in seismic data. Faults on seismic sections extend from near the surface to near the seismic basement. The faults are most pronounced in the Upper Cretaceous reflectors and seem to disappear with depth. Fault throws are inconsistent (reversing throw across faults). OZs trend northeast‐southwest in the north half of the study area and both northeast‐southwest and northwest‐southeast in the south half. The OZ direction of northeast‐southwest in the north half of the survey is confirmed with seismic data. The northwest‐southeast seismic line contains numerous faults and the northeast‐southwest seismic line contains few faults. Most northeast‐southwest faults do not appear to reach seismic basement and are not seen in an interpretation of the magnetic basement. In two cases, northwest‐southeast OZs and correlative Landsat lineaments coincide with mapped magnetic basement faults. These magnetic basement faults can be seen in seismic data too. Faults trending northwest‐southeast may represent Precambrian faults reactivated during the Laramide Orogeny. Movement along these faults possibly generated the northeast‐southwest faults. Most oil fields have an associated near‐surface magnetic anomaly. Other near‐surface magnetic anomalies occur over obvious, untested (in 1985), seismic character or amplitude anomalies in seismic events which correlate with producing intervals in the oil fields. This synergistic correlation is the most important single observation from our study. Different data types and interpretation techniques identified the same geologic trends and prospective geographical areas. This fundamentally important information is often lost in bickering over which filter or processing technique to use or in arguments over which data type is “more important” than others. Further, if the synergistic correlation of data types were not done, the importance of the anomalous features in each individual data type may not have been recognized.

Sensitivity of the dispersion correction to Q error

Geophysics ◽  
1997 ◽  
Vol 62 (1) ◽  
pp. 288-290 ◽  
Author(s):  
Richard E. Duren ◽  
E. Clark Trantham
Keyword(s):  
Amplitude Spectrum ◽  
Careful Attention ◽  
Dispersion Correction ◽  
Event Time ◽  
Seismic Wavelet ◽  
Reflection Time ◽  
Seismic Image ◽  
Time Zones ◽  
Seismic Sections ◽  
Zero Phase

A controlled‐phase acquisition and processing methodology for our company has been described by Trantham (1994). He pointed out that it is careful attention to wavelet phase that leads to improved well ties and a more geologically accurate seismic image. In addition, we prefer zero‐phase wavelets on our seismic sections. For a given amplitude spectrum they have the simplest shape and the highest peak; further, the peak occurs at the reflection time of the event. This alignment is important since the seismic wavelet generally broadens with increasing depth with a zero‐phase wavelet remaining symmetrical about the event time. Our experience has been that a true zero‐phase section can be tied over the entire length of a synthetic trace without having to slide the synthetic trace to tie different time zones.

Nonstationary phase estimation using regularized local kurtosis maximization

Geophysics ◽  
2009 ◽  
Vol 74 (6) ◽  
pp. A75-A80 ◽  
Author(s):  
Mirko van der Baan ◽  
Sergey Fomel
Keyword(s):  
Statistical Estimation ◽  
Phase Estimation ◽  
Time Varying ◽  
Seismic Wavelet ◽  
Optimization Framework ◽  
Phase Rotation ◽  
Seismic Sections ◽  
The Stability ◽  
Kurtosis Maximization ◽  
Zero Phase

Phase mismatches sometimes occur between final processed seismic sections and zero-phase synthetics based on well logs — despite best efforts for controlled-phase acquisition and processing. Statistical estimation of the phase of a seismic wavelet is feasible using kurtosis maximization by constant-phase rotation, even if the phase is nonstationary. We cast the phase-estimation problem into an optimization framework to improve the stability of an earlier method based on a piecewise-stationarity assumption. After estimation, we achieve space-and-time-varying zero-phasing by phase rotation.

LEAST‐SQUARES INVERSE FILTERING AND WAVELET DECONVOLUTION

Geophysics ◽  
1977 ◽  
Vol 42 (7) ◽  
pp. 1369-1383 ◽  
Author(s):  
A. J. Berkhout
Keyword(s):  
Seismic Data ◽  
Broad Band ◽  
Detailed Comparison ◽  
Inverse Filtering ◽  
Borehole Data ◽  
Seismic Wavelet ◽  
Phase Property ◽  
Seismic Sections ◽  
Phase Spectra ◽  
Zero Phase

Detailed comparison between borehole data and seismic data has taught that, in general, conventional seismic inverse filtering is not effective enough to produce desirable deconvolution results, i.e., seismic sections with broad‐band zero‐phase wavelets. Application of conventional seismic reverse filters has the advantage that very little information is needed from the user. However, as is shown in this paper, the phase spectra of these filters may be seriously in error, even if the seismic wavelet has the minimum‐phase property. In wavelet deconvolution the phase spectrum of the filter is correct, provided a good estimate of the seismic wavelet is available. In this paper, wavelet deconvolution is compared with Wiener filtering. The main conclusions are illustrated by examples.

Air‐gun source instabilities

Geophysics ◽  
1987 ◽  
Vol 52 (9) ◽  
pp. 1229-1251 ◽  
Author(s):  
Bill Dragoset ◽  
Neil Hargreaves ◽  
Ken Larner
Keyword(s):  
Large Amplitude ◽  
Small Amplitude ◽  
Ocean Surface ◽  
Data Interpretation ◽  
Central Portion ◽  
Air Gun ◽  
Individual Signature ◽  
Band Limited ◽  
Zero Phase

The signature of an air‐gun array can change over a period of time or even from one shot to the next. If the signature variations are large, then deterministic deconvolution, with an operator designed from a single signature or from an average signature, could produce errors significant enough to affect data interpretation. Possible sources of air‐gun instability include changes in gun positions, firing times, and pressures, gun failures, and scattering from the fluctuating rough ocean surface. If an air‐gun array were perfectly stable, after application of signature deconvolution the residual signatures for a sequence of shots would be identically shaped, broadband, zero‐phase wavelets. In practice, air‐gun instabilities lead to two major defects in band‐ limited residual signatures: the central portion of the wavelet can become asymmetrical, and unsuppressed energy can occur in the residual bubble region. Processing experiments done with synthesized air‐gun array signatures show that of all types of air‐gun instabilities likely to occur, only gun dropouts cause signature variations severe enough to affect data interpretation. Gun dropouts produce unsuppressed residual bubble energy that can show up as phantom events on a stacked section or that can obscure small‐amplitude events following large‐amplitude events. Neither gun dropouts nor any other kind of air‐gun instability has a significant effect on resolution within the seismic band. Since gun dropouts do not happen on a shot‐to‐shot basis and other instabilities are unimportant, there is no practical benefit to be gained by deriving and applying individual signature deconvolution operators for each shot. The influence of gun dropouts can be minimized through other actions taken in acquisition and processing.

Complex seismic trace analysis of thin beds

Geophysics ◽  
1984 ◽  
Vol 49 (4) ◽  
pp. 344-352 ◽  
Author(s):  
James D. Robertson ◽  
Henry H. Nogami
Keyword(s):  
Instantaneous Frequency ◽  
Trace Analysis ◽  
Frequency Tuning ◽  
Seismic Energy ◽  
High Amplitude ◽  
Opposite Polarity ◽  
Peak Period ◽  
Seismic Wavelet ◽  
Porous Sandstone ◽  
Seismic Sections

Displays of complex trace attributes can help to define thin beds in seismic sections. If the wavelet in a section is zero phase, low impedance strata whose thicknesses are of the order of half the peak‐to‐peak period of the dominant seismic energy show up as anomalously high‐amplitude zones on instantaneous amplitude sections. These anomalies result from the well‐known amplitude tuning effect which occurs when reflection coefficients of opposite polarity a half period apart are convolved with a seismic wavelet. As the layers thin to a quarter period of the dominant seismic energy, thinning is revealed by an anomalous increase in instantaneous frequency. This behavior results from the less well‐known but equally important phenomenon of frequency tuning by beds which thin laterally. Instantaneous frequency reaches an anomalously high value when bed thickness is about a quarter period and remains high as the bed continues to thin. In this paper, complex trace analysis is applied to a synthetic model of a wedge and to a set of broadband field data acquired to delineate thin lenses of porous sandstone. The two case studies illustrate that sets of attribute displays can be used to verify the presence and dimensions of thin beds when definition of the beds is not obvious on conventional seismic sections.

Near-surface scattering phenomena and implications for tunnel detection

2015 ◽  
Vol 3 (1) ◽  
pp. SF43-SF54 ◽  
Author(s):  
Shelby L. Peterie ◽  
Richard D. Miller
Keyword(s):  
Synthetic Data ◽  
P Wave ◽  
S Wave ◽  
Seismic Line ◽  
Data Set ◽  
Near Surface ◽  
Tunnel Detection ◽  
Impedance Contrast ◽  
Diffraction Imaging ◽  
Imaging Algorithms

Tunnel locations are accurately interpreted from diffraction sections of focused mode converted P- to S-wave diffractions from a perpendicular tunnel and P-wave diffractions from a nonperpendicular (oblique) tunnel. Near-surface tunnels are ideal candidates for diffraction imaging due to their small size relative to the seismic wavelength and large acoustic impedance contrast at the tunnel interface. Diffraction imaging algorithms generally assume that the velocities of the primary wave and the diffracted wave are approximately equal, and that the diffraction apex is recorded directly above the scatterpoint. Scattering phenomena from shallow tunnels with kinematic properties that violate these assumptions were observed in one field data set and one synthetic data set. We developed the traveltime equations for mode-converted and oblique diffractions and demonstrated a diffraction imaging algorithm designed for the roll-along style of acquisition. Potential processing and interpretation pitfalls specific to these diffraction types were identified. Based on our observations, recommendations were made to recognize and image mode-converted and oblique diffractions and accurately interpret tunnel depth, horizontal location, and azimuth with respect to the seismic line.

Identification of calibration errors in helicopter electromagnetic (HEM) data through transform to the altitude-corrected phase-amplitude domain

Geophysics ◽  
2006 ◽  
Vol 71 (2) ◽  
pp. G27-G34 ◽  
Author(s):  
Yusen Ley-Cooper ◽  
James Macnae ◽  
Terry Robb ◽  
Julian Vrbancich
Keyword(s):  
Measurement Errors ◽  
Thin Sheet ◽  
Inductive Limits ◽  
Near Surface ◽  
Response Parameter ◽  
Phase Errors ◽  
Dimensionless Amplitude ◽  
Phase Amplitude ◽  
Different Response ◽  
System Geometry

We investigate the properties of EM signals in several different response-parameter domains to identify calibration errors in helicopter electromagnetic (HEM) data. In particular, we define a dimensionless response parameter α, derived from frequency-domain data, that is numerically identical to the historic wire-loop response parameter, and is closely related to the thin-sheet and half-space response parameters. The arctangent of α is the phase ϕ of the secondary field. We further define a dimensionless amplitude response parameter β, calculated as the ratio between inductive limits estimated from the data and from system geometry. The inductive limit calculated from geometry provides an initial altitude correction to the data amplitudes. Additional data corrections further correct phase effects and altimeter variations. The amplitude and phase errors in calibration become independent differences between the data and the fitted model in the ϕβ domain. This investigation was undertaken in the response-parameter domain rather than in the data domain, allowing the analysis to be independent of absolute values of conductivity and system frequencies. Statistical analysis in the ϕβ domain of recently acquired data suggests that amplitude calibration errors in HEM data may cause fitted models to have systematic depth errors of 1 to 2 m for near-surface conductors; variable altitude measurement errors are about 1.5 m over seawater.

Near‐surface faulting effects on seismic reflection times

Geophysics ◽  
1983 ◽  
Vol 48 (8) ◽  
pp. 1140-1142 ◽  
Author(s):  
W. Honeyman
Keyword(s):  
Seismic Reflection ◽  
Small Time ◽  
Time Curve ◽  
Near Surface ◽  
Common Depth Point ◽  
Offset Distance ◽  
Large Spread ◽  
Offset Time ◽  
Seismic Sections ◽  
Zero Offset

The depth conversion of seismic reflection records has been the subject of many papers, particularly where faults or other geologic features are present. The common‐depth‐point (CDP) stacked seismic sections with large spread lengths of the order of 2 km have resulted in different interpretation problems. Al‐Chalabi (1979) considered the effect on stacking velocities of subsurface inhomogeneities where different rays in the CDP gather do not penetrate the same type of earth column. He showed that small time delays of 10 msec produce steps in the hyperbolic offset distance‐time curve of the CDP gather and produce stacking velocity variations of the order of ten percent. Levin (1973) considered a time delay in only one ray of the CDP gather and its effect on both stacking velocity and the zero‐offset time [Formula: see text]. This paper models the effect of near‐surface faults on the zero‐offset time [Formula: see text] of deeper layers as determined by the CDP method. This is particularly important since the zero‐offset time is plotted on the processed final record.

Use of seismic stratigraphy for Minnelusa exploration, northeastern Wyoming

Geophysics ◽  
1984 ◽  
Vol 49 (6) ◽  
pp. 715-721 ◽  
Author(s):  
Reverend Francis D. Raffalovich ◽  
Terrell B. Daw
Keyword(s):  
Seismic Data ◽  
Seismic Stratigraphy ◽  
Synthetic Seismograms ◽  
Powder River Basin ◽  
Paper Briefly ◽  
Seismic Attribute ◽  
Seismic Line ◽  
Processing Techniques ◽  
Sonic Logs ◽  
Section Form

While Minnelusa sands have yielded significant reserves in Wyoming’s Powder River Basin, geologic complexities have made these sands an elusive target. This paper briefly describes the development of a technique which was used successfully in the exploration of Minnelusa sands. This tehnique can be applied to many stratigraphic exploration programs. Sonic logs, which are key logs in defining Minnelusa sands, in the C-H field were used to construct synthetic seismograms. These synthetics were then organized in cross‐section form to define whether a change in Minnelusa sands would yield an identifiable change on the synthetics. The “idealized” seismic response did show an obvious lateral change from upper sand to no upper sand conditions, and a pilot seismic line was shot using a Vibroseis® source. This line, which was shot through the C-H field, successfully showed the updip limits of the upper Minnelusa sands. A subsequent seismic program was acquired and other leads and prospects were identified, including prospects that were drilled and successfully completed in the Rozet area. However, a number of other wells conformed to Murphy’s law. In addition to standard processing techniques, high‐resolution processing and seismic attribute processing was done on some of the seismic data, yielding differing degrees of success. By closely coordinating geologic and geophysical principles, a useful stratigraphic‐seismic methodology was developed which has application to a wide variety of exploration problems. ™Trade and service mark of Conoco Inc.

Sign in / Sign up

Export Citation Format

Share Document