Analytic Wavefield Decomposition for Reflectivity Model Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current geophysical imaging techniques, such as Reverse Time Migration (RTM) and Least Squares Reverse Time Migration (LSRTM), face challenges in generating accurate reflectivity models due to noise in seismic data and the complexity of subsurface structures, leading to incomplete recovery of true reflectivity models and convergence issues.
Innovation Solution
The method involves generating analytic source and residual wavefields, decomposing them into down-going and up-going components, calculating a gradient vector and source illumination factor, and using a preconditioned gradient vector to update the reflectivity model, employing wavefield decomposition based on analytic signals to improve convergence and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional RTM or LSRTM techniques are used to generate reflectivity models, then subsurface imaging can be performed, but the accuracy is reduced due to noise in seismic data and complexity of subsurface structures
Solution Approach 1:
The wavefield is segmented into down-going and up-going components through decomposition of the seismic data. This segmentation allows separate processing of different wavefield components, enabling more accurate reconstruction of the reflectivity model by selectively using the down-going source wavefield and up-going residual wavefield components.
Solution Approach 2:
An intermediary iterative optimization process is introduced between the seismic data and the final reflectivity model. This process uses gradient calculation and preconditioning as intermediate steps to refine the reflectivity model incrementally, reducing the impact of noise and structural complexity on the final accuracy.
2Reliability
If conventional imaging techniques are used, then subsurface structures can be visualized, but convergence issues occur due to noise and structural complexity
Solution Approach 1:
The method implements feedback through an iterative optimization process where the reflectivity model is continuously refined. Each iteration uses the previously obtained model to generate new wavefields, calculate gradients, and update the model, creating a feedback loop that progressively improves convergence despite noise and structural complexity.
Solution Approach 2:
The imaging process transitions from a static conventional approach to a dynamic iterative process. The reflectivity model evolves dynamically through multiple iterations, with each iteration adapting the model based on the calculated gradient and preconditioning, enabling the system to converge reliably even in the presence of noise and complex structures.
3Measurement precision
If wavefield decomposition based on analytic signals is employed, then convergence and accuracy are improved, but computational complexity increases
Solution Approach 1:
The computational process is segmented into distinct phases: analytic signal generation, wavefield decomposition into down-going and up-going components, gradient calculation, and preconditioning. This segmentation allows each computational step to be optimized independently and enables parallel processing of different wavefield components, managing the overall computational complexity.
Solution Approach 2:
The wavefield decomposition into down-going and up-going components is performed as a preliminary action before the main iterative optimization process. This preliminary processing organizes the seismic data in a form that facilitates more efficient gradient calculation and model updating, reducing the computational burden during the iterative refinement stages.
4Measurement precision
If iterative optimization processes are used to improve reflectivity model accuracy, then model quality increases, but computational time increases
Solution Approach 1:
The wavefield decomposition and source illumination factor calculation are performed as preliminary actions before the iterative optimization begins. This preliminary processing prepares the data in an optimized form that accelerates convergence during the iterative stages, reducing the total computational time required to achieve high accuracy.
Solution Approach 2:
The method transforms the seismic data into the analytic signal domain, changing the parameter representation from real-valued to complex-valued wavefields. This parameter change enables more efficient gradient calculation and preconditioning operations, improving the rate of convergence and reducing the computational time required for iterative optimization.
Data Source
AI summary
The present disclosure describes methods and systems, including computer-implemented methods, computer program products, and computer systems, for generating a reflectivity model for a subsurface area. One method includes: receiving a set of seismic data associated with the subsurface area; generating analytic source wavefields; generating analytic residual wavefields based on the set of seismic data and an initial reflectivity model; decomposing the analytic source wavefields and the analytic residual wavefields to obtain down-going and up-going components of the analytic source wavefields and the analytic residual wavefields; calculating a gradient vector using the down-going components of the analytic source wavefields and the up-going components of the analytic residual wavefields; calculating a source illumination factor using the down-going components of the analytic source wavefields; calculating a preconditioned gradient vector, based on the gradient vector and the source illumination factor; and generating an updated reflectivity model based on the preconditioned gradient vector.