Seismic Data Noise Attenuation via Component Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional seismic data processing methods, particularly two-way wave equation techniques like reverse time migration, are computationally intensive and costly, especially when dealing with complex geological settings and large datasets, and they struggle to accurately image steeply dipping reflectors and reduce noise in simultaneous source reverse-time migration (SS-RTM) images.
Innovation Solution
The method involves generating multiple initial subsurface images using unique random encoding functions, decomposing them into components, identifying and averaging similar and dissimilar components, and applying noise attenuation techniques such as curvelet transforms to generate enhanced images with reduced interference noise, thereby improving computational efficiency and image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If two-way wave equation techniques (reverse time migration) are used to improve subsurface imaging accuracy, then imaging quality and accuracy are improved, but computational cost and processing time increase significantly
Solution Approach 1:
The patent segments the computational process by separating signal components from noise components through multiple encoding functions and iterations. By decomposing the imaging process into distinct signal extraction and noise attenuation stages, the method achieves accurate subsurface imaging while reducing overall computational burden through targeted processing of specific signal characteristics.
Solution Approach 2:
The patent performs preliminary noise attenuation and signal separation through multiple encoding functions before final image reconstruction. By pre-processing the seismic data to remove interference patterns and isolate signal components in advance, the method reduces the computational complexity of the final imaging step while maintaining high accuracy.
2Reliability
If conventional two-way wave equation techniques are applied to complex geological settings with many shot records, then imaging completeness is improved, but computational demand increases dramatically
Solution Approach 1:
The patent applies partial action by using multiple encoding functions with different random seeds to process subsets of the data. Instead of processing all shot records simultaneously with full computational resources, the method uses multiple passes with varying encoding schemes to gradually build the complete image, reducing peak computational demand while maintaining imaging completeness.
Solution Approach 2:
The patent changes parameters by varying the random encoding functions across multiple iterations and using different curvelet transform thresholds. By adjusting these parameters in successive passes, the method achieves comprehensive imaging of complex geological structures while distributing computational load to avoid excessive peak demands.
3Measurement precision
If noise attenuation techniques are applied to SS-RTM images, then signal-to-noise ratio is improved, but processing complexity increases
Solution Approach 1:
The patent extracts signal components from noise by applying curvelet transforms and identifying coherent signal patterns separate from incoherent noise. By isolating and removing noise components through mathematical transformation and thresholding, the method improves signal-to-noise ratio while keeping processing complexity manageable through efficient transform algorithms.
Solution Approach 2:
The patent uses curvelet transforms as an intermediary mechanism to bridge the original seismic data and the final enhanced image. This intermediate transformation domain allows for selective noise attenuation while preserving signal characteristics, reducing the direct complexity of noise filtering in the spatial domain.
4Object-affected harmful factors
If multiple encoding functions are used to separate signal and noise components, then noise reduction effectiveness is improved, but computational overhead increases
Solution Approach 1:
The patent applies periodic action by using multiple encoding functions with different random seeds in successive iterations. Each encoding pass periodically processes the data with a different random pattern, progressively improving noise reduction effectiveness while distributing computational overhead across multiple smaller, manageable steps rather than one large computation.
Data Source
AI summary
The invention includes a method for reducing noise in migration of seismic data, particularly advantageous for imaging by simultaneous encoded source reverse-time migration (SS-RTM). One example embodiment includes the steps of obtaining a plurality of initial subsurface images; decomposing each of the initial subsurface images into components; identifying a set of components comprising one of (i) components having at least one substantially similar characteristic across the plurality of initial subsurface images, and (ii) components having substantially dissimilar characteristics across the plurality of initial subsurface images; and generating an enhanced subsurface image using the identified set of components. For SS-RTM, each of the initial subsurface images is generated by migrating several sources simultaneously using a unique random set of encoding functions. Another embodiment of the invention uses SS-RTM for velocity model building.