3-D Operator for Structure-Independent Seismic Noise Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for noise estimation in 3-D seismic data are biased by geological structures and fail to accurately distinguish between noise and valid structure, leading to unreliable data reliability assessment in seismic exploration.
Innovation Solution
A 3-D operator is developed to suppress seismic structure effects on noise identification, using a 3x3 mask derived from performance requirements, applied pixel-by-pixel to 3-D post-stacked data, and extended from 2-D to 3-D to provide a structure-independent noise estimation, facilitating the identification of noisy regions through color or gray-scale displays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If correlation theory is applied to estimate noise in 3-D seismic data, then noise estimation is provided, but the method is biased by geological structures such as faults where traces are not correlated
Solution Approach 1:
The patent extracts and removes the influence of geological structures from the noise estimation process. By applying a mask that suppresses structure-dependent components, the method isolates the random noise component that is independent of geological structures, thereby eliminating the bias introduced by faults and other structural features
Solution Approach 2:
The patent applies different processing characteristics to different regions of the seismic data. The mask is designed to have different weights in different directions (horizontal vs. vertical), applying stronger suppression in the horizontal direction where structural bias is more pronounced, while maintaining better noise estimation in the vertical direction where correlation is more reliable
2Ease of manufacture
If simple two-tap difference filters are used to suppress image structure, then noise estimation is achieved, but edge information is left in the filtered images requiring special post-processing
Solution Approach 1:
The patent extends the noise estimation from 2-D to 3-D by incorporating the vertical dimension into the mask processing. The 3-D mask operates on volume data rather than just surface images, allowing noise estimation to consider correlations in all three spatial dimensions simultaneously, which eliminates the need for post-processing corrections
3Measurement precision
If 2-D Laplacian filters are used to suppress image structure, then noise variance estimation is improved, but anisotropic effects remain that affect 3-D data
Solution Approach 1:
The patent generalizes the 2-D Laplacian filter approach to 3-D by creating a volumetric mask that operates on 3-D seismic data. The mask coefficients are extended from 2-D spatial relationships to 3-D spatial relationships, maintaining the structure-suppression capability while eliminating anisotropic effects through proper weighting in all three dimensions
Solution Approach 2:
The patent creates a universal noise estimation method that works for both 2-D and 3-D seismic data. The mask design is formulated in a way that can be applied to data of any dimensionality, making the method versatile and adaptable to different data types and processing requirements
Data Source
AI summary
A system and method identify and display random noise in three dimensional (3-D) seismic data utilizing a 3-D operator to reduce the effects of seismic structure on noise identification. The 3-D operator is derived using statements of required performance in 3-D. The 3-D operator is applied on a pixel-by-pixel basis to each of the pixels in the 3-D post-stacked data to display images in a 3-D display or to output an estimate of noise that is substantially independent of the image structure. The resulting display is generated in colors to indicate noise amplitude to facilitate location of noisy regions in the original display.