Iterative Source Signature Inversion for Seismic Data Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Seismic surveys face challenges in accurately distinguishing and separating seismic energy sources due to equipment drift and phase differences, leading to source-generated data contamination, which can only be detected after significant survey completion, making re-running surveys costly and inefficient.
Innovation Solution
A method for source separation in seismic data using coded seismic sources, where composite seismic data is inverted with an initial estimated source signature, errors are computed, and the signature is revised iteratively to minimize contamination, ultimately using the optimal source signature for accurate separation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple seismic vibrators operate simultaneously with phase differences to improve survey efficiency, then productivity increases, but source-generated data contamination occurs due to equipment drift and phase errors
Solution Approach 1:
The patent implements an iterative feedback mechanism where the inversion process computes errors reflecting source-generated contamination, then revises the source signature estimates to minimize this contamination. This closed-loop feedback continues until convergence, automatically correcting phase errors and drift without requiring perfect initial phase relationships between vibrators.
Solution Approach 2:
The patent performs preliminary determination of optimal source signatures through iterative inversion before final data processing. By pre-characterizing the actual source signatures delivered to the ground (accounting for equipment drift and phase errors), the system prepares corrected reference data that enables accurate source separation in subsequent processing, preventing contamination issues rather than merely detecting them.
2Ease of operation
If conventional inversion with fixed source signatures is used, then processing is simple, but source separation accuracy deteriorates due to equipment drift and phase errors
Solution Approach 1:
The patent transforms the static source signature determination into a dynamic, adaptive process. Instead of using fixed, predetermined source signatures, the system iteratively refines source signature estimates based on actual recorded data and computed errors. This dynamic adaptation allows the processing to automatically compensate for equipment drift and phase errors, significantly improving source separation accuracy while maintaining computational feasibility through automated iteration.
3Reliability
If surveys are re-run to correct phase errors, then data quality improves, but loss of time and increased cost occur
Solution Approach 1:
The patent replaces the mechanical approach of physically re-running surveys with a computational solution. Instead of repeating field operations to correct phase errors, the system uses iterative mathematical inversion and error computation to determine optimal source signatures from existing data. This substitution of field re-surveying with post-processing computation eliminates time loss and additional costs while achieving the same data quality improvement.
Data Source
AI summary
The invention relates to processing seismic data that includes signals from at least two sources and typically three or four sources where source separation is necessary for geophysical analysis. Specifically, the present invention is an analytical technique that quickly creates a more accurate source signature delivered by analysis of the source generated data contamination present in the separated data. The technique is to invert a segment of the data using a seed source signature and compute an error that reflects the generated data contamination observed in the separated source data. The source signature is iteratively revised as the segment is continually inverted with the goal of finding the optimal source signature that provides the lowest computed error. The source signature that provides the lowest error is, or is very close to, the true source signature and is then used in the separation process for the entire composite data set. This will provide much more information for geophysical interpretation.