Seismic orthogonal decomposition via PCA for fault detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current seismic data processing techniques, such as filtering and smoothing, struggle to effectively detect faults and fractures in subsurface formations due to noise and acquisition footprints, often suppressing small seismic features that are crucial for understanding latent structures in reflection seismology.
Innovation Solution
The method employs principal component analysis (PCA) for orthogonal decomposition of seismic data, separating noise and acquisition footprints from structural information, allowing for the detection of faults and fractures by generating orthogonal components that are orthogonal to each other, thereby preserving high-frequency features associated with latent structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If filtering and smoothing techniques are applied to seismic data, then noise is reduced, but small seismic features associated with faults and fractures are suppressed
Solution Approach 1:
The patent segments the seismic data into different frequency components through spectral decomposition, separating high-frequency signals (containing fault and fracture information) from low-frequency noise. This allows selective processing of different frequency bands to preserve small seismic features while reducing noise.
Solution Approach 2:
The patent extracts high-frequency components from the seismic data that contain information about faults and fractures. By isolating and analyzing these specific frequency components separately from the dominant low-frequency noise, the method preserves small seismic features that would otherwise be suppressed by conventional filtering.
2Measurement precision
If conventional processing techniques are used, then processing is simple, but detection accuracy of faults and fractures is poor
Solution Approach 1:
The patent applies dynamic time warping (DTW) to align seismic traces dynamically based on their actual temporal characteristics rather than using fixed alignment windows. This adaptive approach improves detection accuracy by accounting for variations in wavelet shapes and timing while managing complexity through efficient DTW algorithms.
Solution Approach 2:
The patent changes the parameter space by transforming seismic data from the time domain to the frequency domain through spectral decomposition. This parameter transformation allows for better separation of signal components and improved detection of faults and fractures by analyzing specific frequency bands where these features manifest.
3Loss of information
If spectral decomposition is applied, then high-frequency features are preserved, but processing time increases
Solution Approach 1:
The patent applies partial spectral decomposition by focusing computational resources on specific frequency bands that are most likely to contain fault and fracture information, rather than processing the entire frequency spectrum. This selective approach preserves high-frequency features while reducing overall processing time by avoiding unnecessary computation in frequency bands that do not contain target information.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method can include providing seismic data for a subsurface region that includes a reflector; decomposing at least a portion of the seismic data using principal component analysis (PCA) to generate at least one principal component; and outputting output data based at least in part on the at least one principal component. Various other apparatuses, systems, methods, etc., are also disclosed.