Seismic Wavefield Deghosting Using Time-Domain Covariance Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for deghosting seismic wavefields in geophysical surveys face challenges due to the lack of regularly and densely sampled data, which prohibits effective deghosting at many angles of incidence, especially in cross-line directions for towed-streamer acquisitions, and often introduce errors through interpolation.

Innovation Solution

A method for deghosting seismic wavefields in the time domain that calculates the incidence vector and angle of incidence using covariance matrices from vectorial measurements, allowing for correction and combination with scalar measurements without requiring transformation into the Fourier domain, enabling deghosting with spatially irregularly sampled data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If data is interpolated to achieve regular and dense sampling for Fourier domain transformation, then deghosting can be performed at many angles of incidence, but interpolation errors are introduced into the measured data set

Engineering Contradiction:
Improvedeghosting accuracyVSAvoiddata errors
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent extracts the essential information needed for deghosting (angle of incidence) directly from the irregularly sampled data through covariance matrix analysis, rather than attempting to reconstruct the full regular grid through interpolation. This extracts only the necessary deghosting parameters while avoiding the introduction of interpolation errors throughout the entire data set.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the approach from working in the Fourier domain (which requires regular sampling) to working directly in the time domain with irregularly sampled data. By calculating the angle of incidence directly from time-domain covariance matrices, the method transforms the problem into one that can be solved with the actual sampled data without requiring parameter changes that would introduce errors.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If Fourier domain transformation is used for deghosting, then effective deghosting can be achieved, but regularly and densely sampled measurements are required which are difficult to obtain in cross-line direction

Engineering Contradiction:
Improvedeghosting effectivenessVSAvoidapplicability to irregular sampling
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

Instead of transforming irregular data into the Fourier domain through interpolation, the patent inverts the conventional approach by performing deghosting calculations directly in the time domain using the irregularly sampled data as-is. The angle of incidence is calculated directly from time-domain statistics rather than from Fourier-transformed regular grid data.

Inventive Principle:
Principle #13The other way round (Inversion)

Solution Approach 2:

The patent substitutes the mechanical requirement of regular grid sampling and Fourier transformation with a statistical approach using covariance matrices calculated directly from the irregularly sampled time-domain data. This replaces the rigid structural requirement with a flexible statistical method that adapts to any sampling pattern.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach effectively deghosts seismic wavefields in the time domain, reducing errors associated with interpolation and enabling accurate subsurface imaging even with sparse data sampling, applicable to marine and land-based seismic surveys.

Implementation Method 1

The reflection from the sub-surface may, however continue upwards to the surface of the water, where it may again be reflected by the boundary between the water and the air above the water. Because the water-air boundary is a near perfect reflector, the seismic wavefield reflected from the water-air boundary may have a reflection coefficient of minus one and its propagation direction may change

Methodology Applied
Scientific EffectSeismic wave reflection: Reflection

Implementation Method 2

The pressure sensor may, for example, be a hydrophone that records scalar pressure measurements of a seismic wavefield

Methodology Applied
Scientific EffectAcoustic pressure detection: Sound

Implementation Method 3

Scientists and engineers conduct 'surveys' utilizing, among other things, seismic and other wave exploration techniques to find oil and gas reservoirs within the Earth. These seismic exploration techniques often include controlling the emission of seismic energy into the Earth with a seismic source of energy

Methodology Applied
Scientific EffectSeismic wave propagation: Sound

Data Source

PatentEP2810099B1Method and apparatus for processing seismic data
Publication Date: 2023.05.24 NUTEC SCI
  • EP2810099B1 patent drawingFigure 1
  • EP2810099B1 patent drawingFigure 2
  • EP2810099B1 patent drawingFigure 3

AI summary

Methods, apparatuses, and systems are disclosed for processing seismic data. In some embodiments, a set of vectorial measurements and a set of corresponding scalar measurements of a seismic wavefield may be obtained at a seismic receiver (103). An angle of incidence of the seismic wavefield at a first instance of time may be determined by calculating an incidence vector of the seismic wavefield at the seismic receiver at the first instance of time (214), with the incidence vector derived from a measure of correlation of at least one of the vectorial measurements (212). A component of a vectorial measurement may be corrected with the determined angle of incidence of the seismic wavefield at the first instance of time (216), and the corrected component may be combined with a scalar measurement that corresponds to the first instance of time (218).