Uploaded January 2024 | Updated September 2026, 1 week ago
Watch the latest installment of the SIAM Geosciences Webinar Series on waveform inversion, led by Jorn Zimmerling from the University of Uppsala. This series is organized by the SIAM Activity Group on Geosciences: siam.org/membership/activity-groups/detail/geosciences.
Zimmerling's webinar delves into the challenges of full waveform inversion, a technique for estimating properties of complex mediums with variable coefficients in wave equations. While widely used, existing inversion methods face limitations, particularly in nonlinear least squares data fit optimization due to the non-convex nature of the objective function. The complexities arise from nonlinear and intricate mappings, limited data capturing, and band-limited signals lacking low-frequency content. The talk explores an innovative approach utilizing reduced order models (ROMs), specifically data-driven ROMs derived directly from measurements without knowledge of the inaccessible medium's wave field. These ROMs, computed from probing signals, offer a unique perspective, capturing essential features of wave propagation and demonstrating surprisingly effective approximation properties that enhance waveform inversion.
0:00 Introduction
1:25 Webinar
56:55 Q&A
#webinarseries #geophysics #seismic #earthscience #geosciences #appliedmathematics
Watch the latest installment of the SIAM Geosciences Webinar Series on waveform inversion, led by Jorn Zimmerling from the University of Uppsala. This series is organized by the SIAM Activity Group on Geosciences: siam.org/membership/activity-groups/detail/geosciences.
Zimmerling's webinar delves into the challenges of full waveform inversion, a technique for estimating properties of complex mediums with variable coefficients in wave equations. While widely used, existing inversion methods face limitations, particularly in nonlinear least squares data fit optimization due to the non-convex nature of the objective function. The complexities arise from nonlinear and intricate mappings, limited data capturing, and band-limited signals lacking low-frequency content. The talk explores an innovative approach utilizing reduced order models (ROMs), specifically data-driven ROMs derived directly from measurements without knowledge of the inaccessible medium's wave field. These ROMs, computed from probing signals, offer a unique perspective, capturing essential features of wave propagation and demonstrating surprisingly effective approximation properties that enhance waveform inversion.
0:00 Introduction
1:25 Webinar
56:55 Q&A
#webinarseries #geophysics #seismic #earthscience #geosciences #appliedmathematics
