First Break Picking for Seismic Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current seismic data processing methods face challenges in accurately picking first breaks in seismic waves, especially in synthetic datasets, which are crucial for full wavefield travel-time inversion, due to noise and irregularities in seismic survey geometry.
Innovation Solution
A method is developed to determine first breaks in synthetic seismic datasets by identifying amplitude maxima and minima, forming masks based on energy ratios, and predicting first breaks using these masks to enhance accuracy and reduce interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional first break picking methods are used on synthetic seismic datasets, then the process is simpler, but the precision of first break picking deteriorates due to noise and irregularities in seismic survey geometry
Solution Approach 1:
The patent segments the first break picking process into multiple distinct steps: identifying amplitude maxima and minima, computing energy ratios, creating masks to exclude invalid traces, and selecting first breaks from remaining candidates. This segmentation allows each step to address specific aspects of the problem, improving overall precision while managing complexity through systematic decomposition of the picking process.
Solution Approach 2:
The patent introduces intermediate computational elements including energy ratio traces, masks, and candidate selections as mediators between the raw synthetic seismic data and the final first break picks. These intermediaries filter and refine the data progressively, enabling accurate first break identification despite noise and geometric irregularities by acting as intermediate processing layers.
2Measurement precision
If multiple processing steps are implemented to improve first break picking accuracy, then the precision improves, but the processing time and computational complexity increase
Solution Approach 1:
The patent performs preliminary actions by pre-identifying amplitude maxima and minima and computing energy ratios before the actual first break selection. Masks are created in advance to pre-exclude traces with geometric irregularities or noise. These preliminary computations organize the data structure and eliminate invalid candidates beforehand, reducing the computational burden during the final picking stage and optimizing the overall processing efficiency.
3Reliability
If synthetic seismic datasets are used for full wavefield travel-time inversion, then the inversion accuracy can be improved, but the impact of noise and geometric irregularities on first break picking increases
Solution Approach 1:
The patent converts the potentially harmful effects of noise and geometric irregularities into beneficial filtering opportunities. By computing energy ratios and comparing them against thresholds, the method identifies and masks traces affected by noise or irregularities. This transformation turns the presence of problematic data into a basis for systematic exclusion, allowing reliable first break picks to be extracted from the synthetic dataset despite the presence of harmful factors.
Data Source
AI summary
A method for picking first breaks on a synthetic seismic dataset for a full wavefield travel-time inversion is provided. This method includes obtaining a synthetic seismic dataset and for each synthetic trace, determining a plurality of synthetic trace amplitude maxima and minima, determining a modified energy ratio trace, and forming a trace mask based at least in part, on an amplitude of the synthetic trace and on an amplitude of the modified energy ratio trace. The method further includes determining for each synthetic trace, a positive estimated first break based on the plurality of synthetic trace amplitude maxima and the trace mask and a negative estimated first bread based on the plurality of synthetic trace amplitude minima and the trace mask. The method concludes with determining a predicted first break for each synthetic trace from the positive estimate first break and the negative estimate first break.