Seismic Data Noise Attenuation via 4D Volume Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise filtering methods in seismic data acquisition are insufficient in attenuating noise, leading to reduced accuracy in subsurface information obtained from seismic data.
Innovation Solution
A method involving the formation of a four-dimensional volume using seismic data with dimensions of trace number, time, shot number, and cable number, and applying a random noise attenuation filter to this volume to filter out uncorrelated noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional noise filtering methods are applied to seismic data, then some noise is attenuated, but the level of attenuation is insufficient and noise remains in the filtered record
Solution Approach 1:
The patent transforms the noise attenuation problem from a one-dimensional signal processing task to a four-dimensional volume processing task. By organizing seismic data into a four-dimensional volume with dimensions of trace number, time, shot number, and cable number, the system can apply filtering operations that exploit correlations across multiple dimensions. This dimensional expansion enables the random noise attenuation filter to distinguish between random noise (which lacks correlations across dimensions) and genuine geological signals (which exhibit correlations), thereby achieving superior noise attenuation while preserving subsurface information accuracy.
2Object-affected harmful factors
If noise filtering is applied to reduce noise in seismic data, then noise attenuation increases, but the complexity of data processing increases
Solution Approach 1:
The patent segments the seismic data into a four-dimensional volume structure with distinct dimensions (trace number, time, shot number, cable number). This segmentation allows the random noise attenuation filter to operate independently on each dimension while exploiting correlations across dimensions. By breaking down the complex noise attenuation problem into manageable dimensional components, the system achieves effective noise reduction through a systematic approach that processes data in organized segments rather than as a monolithic complex operation.
3Measurement precision
If a random noise attenuation filter is applied to a four-dimensional volume, then uncorrelated noise is removed, but the computational requirements increase
Solution Approach 1:
The random noise attenuation filter leverages the inherent correlations present in the four-dimensional seismic data volume to achieve noise attenuation. Instead of requiring external reference signals or complex processing algorithms, the system uses the data's own structural correlations across trace number, time, shot number, and cable number dimensions to distinguish between random noise and genuine signals. This self-service approach allows the filter to automatically identify and remove uncorrelated noise patterns while preserving correlated geological information, reducing the need for additional computational resources.
Data Source
AI summary
Noise may be filtered or attenuated from seismic data by building a four-dimensional volume using the acquired seismic data and then applying a random noise attenuation filter to the four-dimensional volume. The dimensions of the four-dimensional volume may include a trace number dimension, a time dimension, a shot number dimension, and a cable number dimension. The random noise attenuation filter may filter portions of the acquired seismic data if the seismic data is not correlated with respect to other seismic data in the four dimensional volume.