Mobil Oil Corporation

Thomas E Shirley

Robert A Meek

Stanley J Laster

Andrew F Linville, Jr

Mark E Willis

Rambabu P Ranganayaki

Paul G Griffith

Stan Y Lee

Method for removing coherent noise from a series of seismic traces. Each seismic trace included in the series is Fast Fourier Transformed into the frequency domain. For each frequency of the resultant f-x representation, a series of N complex numbers, one corresponding to each trace, is selected. The selected series of complex numbers corresponding to each frequency are processed to produce autoregressive model coefficients from which a polynomial rooting provides complex roots related to wavenumbers and decay constants. Amplitudes and phases for complex non-stationary spatial sinusoids associated with the estimated wavenumbers are then determined. For all frequencies f, the collection of determined wavenumbers, decay constants, amplitudes and phases constitute an f-k representation of the seismic traces. The produced f-k representation is examined to identify components corresponding to seismic data and the f-k components which correspond to seismic data are summed to reconstruct frequency components of an f-x representation of the acquired seismic traces with coherent noise removed.

Method for removing coherent noise from seismic data

An improved method for filtering noise bands from seismic data is described. The data is analyzed to determine the frequency component of the undesired noise, and a precise autocorrelation function is generated and used to design a filter to remove the undesired frequency component from the data. The filter can be band pass, band reject, notch, low pass, or high pass, and multiple bands of noise can be removed from the seismic data in a single filtering operation.

Method of digital filtering

A seismic record is enhanced through trace restoration and extrapolation. A set of traces surrounding a missing trace or adjacent an edge of the seismic record are transformed into the frequency domain using a fast Fourier transform. Each frequency of the resultant x-f representation a series of complex numbers, one corresponding to each trace, is selected. Autoregressive model coefficients are estimated for this complex series using a recursive least squares method. A complex number is estimated for trace restoration and extrapolation from the series of complex numbers and autoregressive coefficients. After all frequency components are thus processed restored or extrapolated traces are obtained by an inverse Fourier transform.

Method for restoring and extrapolating seismic traces

Method of interpolating additional seismic traces between existing seismic traces obtained during the exploration of a subsurface formation. Seismic traces are recorded upon a datum surface in accordance with standard seismic data acquistion methods. Forward and backward wave equation datuming of the data set at the same velocity estimates new measurements for originally acquired data deemed unsatisfactory by propagating measured seismic traces to a fictitious datum plane using wave theory to create a second set of seismic traces and datuming the second set of seismic traces, again using wave theory, back to the original datum plane to create a third set of seismic traces. The third set of seismic traces will include new seismic traces to be used in place of the originally acquired unsatisfactory data.

Method for seismic trace interpolation using a forward and backward

Method for restoring seismic data to missing or severely noise contaminated traces included in a seismic section. The seismic section is transformed into a data localizing space and muted to retain only the localized data components of the seismic section. The muted transform space is inverse-transformed into the x-t domain to produce first estimates of the seismic data to be restored. The first estimates of the restored data are substituted for the corresponding traces in the original seismic section. Fully restored traces corresponding to the missing or severly noise contaminated traces are determined from the single estimate of the restored traces.

Method of restoring seismic data

Method of restoring null traces in seismic records. The seismic record is transformed from the x-t domain to the f-k domain by a series of fast Fourier transforms. The f-k transform of the seismic record is filtered to retain only coherent events and the filtered f-k transform is inverse-transformed back to the x-t domain. A first determination of the restored trace corresponding to the null trace is selected and substituted for the null trace. The first determination of the restored trace is used to determine a fully restored trace.

Seismic trace restoration using F-K filtering

Method and apparatus for removing sinusoidal noise from seismic traces comprised of a seismic data component and a sinusoidal noise component. Seismic traces acquired by conventional seismic exploration of a subsurface formation are utilized to produce a frequency domain amplitude spectrum of the traces. The amplitude spectrum is analyzed to determine a line frequency equivalent to the sinusoidal noise frequency to be removed. A filter order N (number of filter weights) is selected and the exact autocorrelation function of the selected sinusoidal frequency is determined for lags 0, 1, . . . , N-1. An autocorrelation matrix is produced utilizing the determined autocorrelation values and predictive deconvolution techniques are applied to the autocorrelation matrix to produce a convolutional filter which will remove sinusoidal noise at the determined frequency while leaving other frequencies minimally attenuated. The sinusoidal noise is then removed from the seismic data by convolving the acquired seismic traces with the filter.

Method and apparatus for removing sinusoidal noise from seismic data

An analog-to-digital converter for use in seismic data processing applications is disclosed which features variants on delta modulation and delta-sigma modulation methods. A series of single bits is output by an analog-to-digital converter based on comparison of a predicted analog signal with the actual analog signal. A low pass filter and subsampling operation is applied to the output digital bits to provide digital words corresponding to the analog signal amplitude. Plural integrator stages are shown which allow reduction of quantization error.

Growing community of inventors

Richardson, TX, United States of America

Thomas E Shirley