3D Surface Multiple Prediction for Irregular Marine Seismic Geometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for removing surface multiples in marine seismic surveying are prone to errors due to regularization and deregularization processes, particularly when dealing with irregular acquisition geometries and non-coincident source and receiver locations.
Innovation Solution
A three-dimensional surface multiple prediction algorithm that minimizes regularization processes by operating on pairs of subsurface lines and predicting multiples at correct locations, offsets, and azimuths without assuming nominal geometry, allowing for accurate prediction of surface multiples in marine seismic data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If regularization and deregularization processes are used to handle irregular acquisition geometries, then the method can process non-coincident source and receiver locations, but errors are introduced in surface multiple removal
Solution Approach 1:
The patent segments the surface multiple prediction into independent contributions from different subsurface reflectors. Each reflector's multiple contribution is predicted separately using its own convolution of traces, avoiding the need to regularize the entire dataset. This segmentation allows accurate handling of irregular geometries without introducing errors through global regularization processes.
Solution Approach 2:
The patent moves from traditional 1D or 2D trace-based processing to a 3D surface multiple prediction approach. By considering spatial coordinates (x, y, z) and computing multiples at correct 3D locations, offsets, and azimuths, the method accurately handles irregular acquisition geometries without requiring regularization to a nominal geometry.
2Ease of manufacture
If traditional surface multiple prediction methods are used, then processing can be simplified, but accuracy decreases for irregular acquisition geometries
Solution Approach 1:
The patent performs preliminary organization of seismic traces into subsets associated with specific subsurface reflectors before conducting convolutions. This preliminary action allows the method to maintain simplicity while improving accuracy, as each convolution operates on pre-organized data with known geometric relationships, eliminating the need for complex regularization while preserving prediction accuracy for irregular geometries.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces errors in surface multiple removal by directly addressing irregularities in acquisition geometry, providing accurate predictions of surface multiples at their correct locations, enhancing the efficiency and accuracy of seismic data processing.
Implementation Method 1
convolving the pairs of recorded traces to generate a plurality of convolutions
Data Source
AI summary
A method and apparatus for predicting a plurality of surface multiples for a plurality of target traces in a record of seismic data. In one embodiment, the method includes creating a file containing information regarding a plurality of pairs of recorded traces. Each pair of recorded traces is substantially closest to a desired shot-side trace and a desired receiver-side trace. The method further includes convolving the pairs of recorded traces to generate a plurality of convolutions and stacking the convolutions for each target trace.