.png){kind=icon}
.png){kind=icon}
Return
Multitrace Seismic Deconvolution via Structural Orthogonal Matching Pursuit
.png){kind=icon}
PRE
AI
DOI：10.1109/LGRS.2025.3623635.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Deconvolution, a key technique for enhancing the seismic data resolution, represents an actively studied topic within seismic exploration. However, its inherently ill-posed nature renders seismic deconvolution a challenging task. Conventional sparse inversion approaches reconstruct subsurface reflection coefficients but often neglect the spatial interdependence between adjacent seismic traces. This omission can lead to deconvolved results exhibiting poor spatial continuity. To address this limitation, this letter introduces a novel multitrace seismic deconvolution framework incorporating spatial continuity derived from reflection structure characteristics. Our proposed methodology comprises three sequential steps. First, reflection structure characteristics are extracted from poststack seismic records. Second, these characteristics are integrated into the conventional orthogonal matching pursuit (OMP) algorithm to guide the localization of sparse reflectivity, effectively imposing spatial constraints. Finally, rather than outputting the full-band reflectivity series, a high-resolution seismic trace is generated by convolving a high-frequency wavelet with the reconstructed reflectivity. This approach mitigates solution nonuniqueness inherent in the inversion and enhances result reliability. The proposed method effectively preserves subtle signals while providing sharper delineation of geological features and ensuring robust spatial continuity, even within complex structural settings. Validation using both synthetic and field datasets demonstrates that our technique produces high-resolution deconvolution results characterized by the superior spatial coherence.
Keywords:
Orthogonal matching pursuit (OMP)
reflection structure
seismic data
seismic deconvolution
Journal
I
IF:
4.4
Papers:
572
Citations:
0