Geophysical Inversion via Sparse Domain Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The resource-intensive processes of computing the forward model and adjusting estimates in geophysical inversion methods require significant computational time and resources, necessitating techniques to reduce these demands.
Innovation Solution
Computing the forward model on a sparse subset of the parameter domain and interpolating results to cover the entire domain, while sacrificing some model accuracy for improved processing speed and reduced resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the forward model is computed on the entire parameter domain with full model accuracy, then model precision is improved, but computational time and resource requirements increase significantly
Solution Approach 1:
The parameter domain is segmented into a sparse subset and the remaining portions. The forward model is computed only on the sparse subset of the parameter domain, and the results are interpolated to cover the entire domain. This segmentation approach reduces computational time and resource requirements while maintaining acceptable model accuracy through the use of interpolation techniques.
Solution Approach 2:
Instead of computing the forward model on the entire parameter domain, the method applies partial action by computing it only on a sparse subset. This partial computation is sufficient when combined with interpolation, achieving a balance between computational efficiency and model accuracy without requiring full-domain computation.
2Measurement precision
If the forward model is computed on the entire parameter domain with full model accuracy, then model precision is improved, but resource requirements increase significantly
Solution Approach 1:
The parameter domain is segmented into a sparse subset and the remaining portions. The forward model is computed only on the sparse subset of the parameter domain, and the results are interpolated to cover the entire domain. This segmentation approach reduces computational time and resource requirements while maintaining acceptable model accuracy through the use of interpolation techniques.
Solution Approach 2:
The method creates a simplified copy of the full model computation by computing the forward model only on a sparse subset of the parameter domain. This partial computation serves as a representative sample that, when interpolated, approximates the results of a full computation with significantly reduced resource requirements.
3Productivity
If the forward model is computed on a sparse subset of the parameter domain with interpolation, then computational efficiency is improved, but model accuracy deteriorates
Solution Approach 1:
Instead of computing the forward model on the entire parameter domain, the method applies partial action by computing it only on a sparse subset. This partial computation is sufficient when combined with interpolation, achieving a balance between computational efficiency and model accuracy without requiring full-domain computation.
Solution Approach 2:
Interpolation acts as an intermediary process that bridges the gap between the sparse subset computation and the full parameter domain requirements. The interpolation technique generates intermediate values that approximate the full model results, enabling the system to achieve acceptable accuracy with reduced computational effort.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods of geophysical modeling and inversion are disclosed. A sparse domain is defined for a geophysical model, over which a sparse model result is computed. A full model result is then resolved by interpolation over the sparse domain. The full model result may be used as the forward modeling result in a geophysical inversion process. Reconstruction error, or model error, or both may be used to adjust the sparse domain, the model, or the geophysical basis of the model.