2026.06.24(星期三)10:00 报告会 Prof.  Andrzej Górszczyk Institute of Geophysics of the Polish Academy of Sciences in Warsaw

2026-06-23


 

北京地球物理学会地球物理前沿论坛logo(左).jpg   北京大学地球物理学术报告会logo(右).png


Integrated Crustal-Scale Seismic Imaging in the Eastern Nankai Trough: Advanced Full-Waveform Inversion and

Multi-Dataset Workflow

Prof.  Andrzej Górszczyk

Institute of Geophysics of the Polish Academy of Sciences in Warsaw

2026.06.24(星期三)10:00,理科二号楼 2821


摘要:

High-resolution imaging of crustal-scale structures using seismic data remains inherently challenging due to strong nonlinearity of FWI, cycle-skipping effects, and limited constraints imposed by acquisition geometry. These difficulties are particularly pronounced for long-offset OBS data and 2D towed-streamer MCS surveys, where sparse sampling and velocity-depth ambiguity hinder robust model reconstruction.

We present an integrated multiscale imaging workflow applied to seismic data from the eastern Nankai Trough (Tokai area). The approach combines first-arrival traveltime tomography, Laplace-Fourier and time-domain FWI, and PSDM of MCS data. To mitigate cycle skipping and reduce inversion nonlinearity, the starting velocity model is iteratively refined using FAT assisted by FWI quality control. Progressive frequency continuation, relaxation of time damping, and offset-based layer stripping are employed to hierarchically inject information into the inversion. Additionally, we investigate the graph-space optimal transport misfit function, which improves convexity and enables convergence even in the presence of large initial kinematic errors.The resulting FWI velocity model provides a high-resolution image of the entire crust and serves as a reliable background model for PSD M o f the MCS data. Further refinement through slope tomography minimizes residual move-outs and enhances reflector focusing. The integration of wide-angle OBS data and short-spread MCS data significantly improves imaging across scales, linking large-scale velocity structures with short-wavelength reflectivity patterns.

Our results demonstrate that advanced inversion strategies combined with multi-dataset integration substantially enhance crustal imaging in complex subduction environments and provide new perspectives for the design of future OBS-based crustal-scale seismic experiments.


报告人简介:

Andrzej Górszczyk, is an associate professor working at the Institute of Geophysics of the Polish Academy of Sciences in Warsaw, the Department of Seismic Research. He specializes in advanced seismic data processing, high-resolution seismic imaging, and velocity model building – areas that are central to understanding complex structures of the Earth’s crust and lithosphere.His research has made significant contributions to seismic imaging methodology, particularly through the development and application of full-waveform inversion and other cutting-edge seismic techniques to reconstruct detailed regional-scale models of crustal and subduction-related structures. These methods have improved our ability to visualize and interpret geological features that are otherwise difficult to resolve, advancing both basic geoscience knowledge and practical applications such as seismic hazard assessment and natural resource exploration.

He leads and participates in multiple nationally and internationally funded research projects. These include ongoing grants from the National Science Centre of Poland that focus on optimizing seismic acquisition and imaging techniques for enhanced crustal-scale understanding and on elucidating the influence of volcanic processes on lithospheric structure using multi-parameter seismic imaging.He also collaborates with leading research institutions worldwide. His projects involve partnerships with scientists from Japan (e.g., JAMSTEC) and across Europe, reflecting his active role in global efforts to advance seismic imaging science.