Random‐Objective Waveform Inversion of 3D‐9C Shallow‐Seismic Field Data. Issue 9 (3rd September 2021)

• Record Type: Journal Article
• Authors: Pan, Yudi; Gao, Lingli; Bohlen, Thomas
• Is Part Of: Journal of geophysical research. Volume 126:Issue 9(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-09-03
• Subjects: Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1029/2021JB022036
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: Robustness and uncertainty estimation are two challenging topics in full‐waveform inversion (FWI). To overcome these challenges, we present the methodology of random‐objective waveform inversion (ROWI), which adopts a multi‐objective framework and a preconditioned stochastic gradient descent optimization algorithm. The use of one shot per iteration avoids using redundant data and reduces the computational cost. The Pareto solutions represent a group of most likely solutions and their differences quantifies the model uncertainty associated with the trade‐off between conflicting objective functions. Due to the high dimensionality in the data and model spaces, it is prohibitively expensive to check the Pareto optimality of all solutions explicitly. Thus, we decompose the original multi‐objective function into shot‐related subproblems and use the Pareto solutions of the subproblems for trade‐off analysis. We apply ROWI to a field multi‐component shallow‐seismic data set acquired in Rheinstetten, Germany. The 3D near‐surface model is successfully reconstructed by ROWI and the main target, a refilled trench, is delineated. We compare the results estimated by ROWI and a conventional least squares FWI to prove the high efficiency of ROWI. We run six ROWI tests on the field data with different solution paths to prove the robustness of ROWI against the random solution path. The validity of the reconstructed model is verified by multiple 2D ground‐penetrating radar profiles. We estimate 246 Pareto solutions of multi‐objective subproblems for trade‐off analysis. Another four ROWI tests starting from different poor initial models are performed, whose results prove the relatively high robustness of ROWI against the initial model. Plain Language Summary: Seismic waveform contains abundant information about the physical properties of the Earth. The random‐objective waveform inversion (ROWI) method provides an efficient way to estimate the Earth model by randomly probing a subset of seismic data and fitting one of its waveform characteristics. We can further evaluate the trade‐off in the reconstructed subsurface model by comparing the differences among the admissible results that explain the data equally well. Here, we promote the ROWI method to 3D media and apply it to a 3D 9‐component shallow‐seismic field data set. The ROWI result nicely reveals the existence of a refilled ancient trench, and the validity of it is verified by multiple ground‐penetrating radar profiles. Several ROWI tests with different setups show that ROWI is robust against its stochastic nature and poor starting models. Trade‐off analysis shows that the boundaries of the refilled trench and another trench‐like structure are less reliable compared with the other parts in the reconstructed model. Overall, the field example proves the validity and efficiency of ROWI in reconstructing the Earth's shallow subsurface. Key Points: We present the methodology of random‐objective waveform inversion (ROWI) and apply it to a 3D‐9C shallow‐seismic field data set We prove the high robustness of ROWI against random solution paths and poor initial models The Pareto solutions of multi‐objective subproblems are analyzed in the model space for trade‐off analysis