GNGTS 2017 - 36° Convegno Nazionale

682 GNGTS 2017 S essione 3.2 Fig. 3 - a) Comparison between CCP used to generate the synthetic data (red bars) and CCP estimated by ten different LM inversions in which a double dispersion model is assumed. b) As in a) but for the PSO algorithm. In a) and b) the ten LM and PSO tests are represented by different colours. c) L2 norm data misfit for each LM and PSO inversion test. d) And e) eigenvectors in model space and singular values of the Jacobian matrix for a double dispersion model, respectively. method. The analysis of the residual function maps and of the eigenvectors in model space, shows that the FWI-IP, is a severely ill-conditioned problem mainly characterised by elongated valleys in the objective function and less affected by the presence of multiple minima. This work demonstrated that the LM method is better suited than the PSO to invert IP data. We are currently working on the next step of our research, that is the application of the LM and PSO algorithms to invert field IP data. References Aleardi M., and Tognarelli A. (2016). The limits of narrow and wide-angle AVA inversions for high Vp/Vs ratios: An application to elastic seabed characterization. Journal of Applied Geophysics, 131, 54-68. Aster R., Borchers B., and Thurber C. (2005). Parameter Estimation and Inverse Problems. Elsevier Academic. Major J., and Silic J. (1981) Restrictions on the use of Cole-Cole dispersion models in complex resistivity interpretations. Geophysics 46, 916–931. Martinez J., Muñiz Z., and Tompkins M. (2012). On the topography of the cost functional in linear and nonlinear inverse problems. Geophysics, 7 , W1-W15 Ollson, P. (2015). Optimization of time domain induced polarization data acquisition and spectral information content. Lund University. Pelton W., Ward S., Hallof P., Sill W. and Nelson P. (1978). Mineral discrimination and removal of inductive coupling with multifrequency IP. Geophysics, 43 , 588-609. Sajeva A., Aleardi M., Galuzzi B., Stucchi E., Spadavecchia E., and Mazzotti A. (2017). Comparing the performances of four stochastic optimisation methods using analytic objective functions, 1D elastic full-waveform inversion, and residual static computation. Geophysical Prospecting. In print. Doi: 10.1111/1365-2478.12532