GNGTS 2017 - 36° Convegno Nazionale

GNGTS 2017 S essione 3.1 599 Fig. 3 - (a) and (b) display respectively the evolutions of the data misfits and the model misfits in two-grid genetic algorithm FWI, corresponding to the true model shown in Fig. 1a. The black lines are related to the smallest data misfits, while the red in (a) to the mean data misfits and in (b) to the mean model misfits. Cyan lines show when frequency marching occurs. (c) and (d) show observed and predicted seismograms, respectively as results produced by two-grid genetic algorithm FWI and FWI with the local optimization method, corresponding to the true model shown in Fig. 1a. The black traces are the observed data while the red are the predicted. The seismograms are trace- wise normalized. Conclusion. The near-surface Vs models predicted by two-grid genetic algorithm FWI fairly reproduce the long wavelength structures of the models and thus they are suitable starting models for FWI with a local optimization method. The latter further improves the prediction of the model structures and the actual shear wave velocities. However, the role of the starting model is crucial: small errors in the starting models could be very much amplified in the final predictions, notwithstanding the further improvements of the data misfits. Acknowledgments We would like to acknowledge the authors of the IFOS code (by KIT), with which we carried out our tests of FWI with local optimization methods. References Aleardi M. and Mazzotti A.; 2017: 1D elastic full-waveform inversion and uncertainty estimation by means of a hybrid genetic algorithm – Gibbs sampler approach. Geophysical Prospecting, 65, 64-85, doi: 10.1111/1365- 2478.12397. Bunks C., Saleck F. M., Zaleski S. and Chavent G.; 1995: Multiscale seismic waveform inversion. Geophysics, 60(5), 1457-1473, doi: 10.1190/1.1443880. Groos L., Schäfer M., Forbriger T. and Bohlen T.; 2014: The role of attenuation in 2D full-waveform inversion of shallow-seismic body and Rayleigh waves. Geophysics, 79(6), R247–R261, doi: 10.1190/GEO2013-0462.1. Köhn D., De Nil D., Kurzmann A., Przebindowska A. and Bohlen T.; 2012: On the influence of model parametrization in elastic full waveform tomography. Geophys. J. Int., 191, 325–345, doi: 10.1111/j.1365-246X.2012.05633.x. Mazzotti A., Bienati N., Stucchi E., Tognarelli A., Aleardi M. and Sajeva A.; 2017: Two-grid genetic algorithm full- waveform inversion. The Leading Edge, 35(12), 1068-1075. doi: 10.1190/tle35121068.1. Sajeva A., Aleardi M., Stucchi E., Bienati N. and Mazzotti A.; 2016: Estimation of acoustic macro models using a genetic full-waveform inversion: Applications to the Marmousi model. Geophysics, 81(4), R173-R184, doi: 10.1190/geo2015-0198.1. Thorbecke J. W. and Draganov D.; 2011: Finite-difference modeling experiments for seismic interferometry. Geophysics, 76(6), H1-H18, doi: 10.1190/geo2010-0039.1. Xing Z. and Mazzotti A.; 2016: Rayleigh Waves Modelling Complexities in the Perspective of Full Waveform Inversion of Surface Waves - Synthetic Examples. Near Surface Geoscience 2016, Barcelona, Spain, Extended Abstracts, doi: 10.3997/2214-4609.201601909. Xing Z. and Mazzotti A.; 2017a: Two-grid full waveform Rayleigh wave inversion by means of genetic algorithm with frequency marching. 79 th EAGE Conference and Exhibition 2017, Paris, France, Extended Abstracts, doi: 10.3997/2214-4609.201701412. Xing Z. and Mazzotti A.; 2017b: Surface Wave FWI on Complex Models - The Robustness of the Inversion to Assumptions and Forward ModelingApproximations. Near Surface Geoscience 2017, Malmö, Sweden, Extended Abstracts, doi: 10.3997/2214-4609.201702019.