Seismic Multiple Prediction via Geometry Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing seismic data processing methods face difficulties in accurately predicting and removing surface multiples due to geometry-related time and amplitude effects, leading to errors in seismic data sets, especially in marine surveys where water-bottom multiples are a significant issue.
Innovation Solution
A method that corrects geometry-related time and amplitude effects in seismic data by identifying target and desired traces, convolving corrected recorded traces, and stacking convolutions to improve multiple prediction accuracy, which involves compensating for effects like surface-consistent statics, structural dips, and anisotropy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If surface-related multiple attenuation is used to remove surface multiples from seismic data, then multiple attenuation capability is improved, but prediction accuracy deteriorates due to geometry-related time and amplitude errors
Solution Approach 1:
The patent applies parameter changes by correcting geometry-related time and amplitude effects in the seismic data before performing multiple attenuation. This involves adjusting time parameters (travel time corrections) and amplitude parameters (impulse response corrections) to account for variations in seismic wave propagation through different geological layers, thereby improving prediction accuracy while maintaining multiple attenuation capability
Solution Approach 2:
The patent replaces direct mechanical wave propagation modeling with a signal processing approach that uses recorded seismic traces to infer and correct geometry-related effects. By substituting complex physical modeling with data-driven correction methods, the system achieves more accurate multiple predictions without requiring detailed geological models
2Adaptability or versatility
If seismic data is interpolated and extrapolated to improve sampling, then data completeness is improved, but prediction error increases
Solution Approach 1:
The patent implements feedback by using the corrected and interpolated seismic data to generate multiple predictions, then comparing these predictions against the actual recorded data. This feedback loop allows the system to identify and correct interpolation errors, adjusting the correction parameters to minimize prediction error while maintaining data completeness
Solution Approach 2:
The patent applies preliminary action by performing time and amplitude corrections on the seismic data before interpolation and extrapolation. This preliminary correction ensures that the data used for interpolation is already optimized, reducing the propagation of errors through subsequent processing steps and improving final prediction accuracy
Data Source
AI summary
A method for predicting a plurality of surface multiples for a plurality of target traces in a record of seismic data acquired in a survey area. The method includes selecting a target trace and identifying two or more desired traces for multiple prediction based on the target trace. After identifying the desired traces, the method identifies one or more recorded traces for each desired trace. Each identified recorded trace is described as being substantially close to one of the desired traces. The method then includes correcting the identified recorded traces for one or more geometry-related effects associated with the survey area and convolving the corrected recorded traces to generate a plurality of convolutions. After convolving the corrected recorded traces, the method then stacks the convolutions.