Marine Seismic Data Denoising via Tau-p Domain Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for denoising marine seismic data, particularly the vertical component (Z) in ocean bottom cable acquisitions, often fail to preserve the signal and effectively isolate noise due to the overlap of signal and coherent noise areas, especially when using traditional mathematical transformations.
Innovation Solution
A method involving transforming seismic data from the time-space domain to the tau-p domain, calculating the down-going energy, and using sparse Radon transforms to isolate noise regions based on energy ratios between pressure and vertical geophone data, allowing for noise subtraction in the time-space domain without scaling the Z component.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mathematical transformations are used to denoise marine seismic data, then noise can be isolated, but the signal is degraded or lost due to overlap of signal and coherent noise areas
Solution Approach 1:
The patent transforms the data from the conventional time-space domain to the tau-p domain (time-slowness domain), changing the parameter space in which noise and signal are represented. This domain transformation allows for better separation of noise and signal components, enabling effective noise isolation without degrading the signal content.
2Measurement precision
If the Z component is scaled to match P component energy, then noise can be attenuated, but the original signal amplitude information is altered
Solution Approach 1:
Instead of scaling the Z component to match P component energy (which alters signal amplitude), the patent extracts only the noise portion from the Z component in the tau-p domain. By isolating and removing just the noise component while preserving the original Z component's signal amplitude, this approach achieves effective noise attenuation without modifying the original signal characteristics.
3Measurement precision
If standard denoising techniques are applied to the Z component, then noise is reduced, but the technique assumes noise is random or distinguishable which is not always true for coherent noise
Solution Approach 1:
The patent addresses the limitation of standard denoising techniques by transforming the data into the tau-p domain, adding a new dimension (slowness) to the analysis. This dimensional transformation enables the separation of coherent noise from signal based on their different trajectories in the time-slowness domain, making the denoising technique applicable to coherent noise types that cannot be handled by conventional methods assuming random or easily distinguishable noise.
Data Source
AI summary
Computing device, computer instructions and method for denoising marine seismic data recorded with first and second seismic sensors. The method includes receiving first seismic data recorded with the first sensor in a time-space domain; receiving second seismic data recorded with the second sensor in the time-space domain, calculating with a processor models of the first and second seismic data in a transform domain that is different from the time-space domain; calculating in the transform domain an energy related to a down-going energy based on the models of the first and second seismic data; determining a noise in the transform domain corresponding to the second seismic data based on the calculated energy; reverse transforming the noise from the transform domain into the time-space domain; and denoising the second seismic data by subtracting the noise in the time-space domain from the second seismic data.