Use of Global-Local FWI of Seismic Land Data Acquired in a Geothermal Exploration Area

In this work we present a workflow to estimate a high-resolution acoustic velocity model of a seismic land dataset acquired in a geothermal exploration area. The data are characterized by poor S/N ratio and lack of reflections in the shallow part, above 2s. The model is obtained by exploiting the transmitted waves in a two-steps full-waveform inversion (FWI) approach. In the first, we perform a global FWI based on the genetic-algorithm (GA) optimization method. Making use of a large search range centered on a gradient-like velocity field, GA are used to estimate a low frequency velocity model. In the second step, the obtained model is used as the starting point for a gradient- based FWI. In the GA inversion we consider frequency lower than 10 Hz, while we extend the frequency up to 20 Hz in the local inversion, adopting a frequency marching scheme. The final velocity model shows an improved resolution in the shallow part of the subsurface. Its features, characterized by many details, can be employed directly as a geological interpretation tool. The reliability of the final velocity model is assessed by the data misfit achieved and by the comparison with the available VSP