# Research activities

The research activities winth the Geophysical Research Group from MINES ParisTech is organised along three main axis:

- Inversion Velocity Analysis: the objective is to derive an initial model for subsequent Full Waveform Inversion. It consists of analysing focusing panels (e.g. Common Image Gathers). For the correct model, different images of the subsurface should be consistent. The recent advances are
- the derivation and introduction of a direct inverse (instead of the adjoint of the Born modelling operator) for much more macro-model gradients (Chauris and Cocher, 2017)
- the extension beyond reflected waves, with the incorporation of multiples (Cocher and Chauris, 2017) and transmitted waves (Lameloise and Chauris, 2016)

- Stochastic tomography: in collaboration with the Geostatistics team, we have developed a research activity dedicated to the stochastic geophysics in order to reliably estimate uncertainties associated to the velocity model and to the seismic event locations (Gesret et al., 2015). We recently focused on first arrival traveltime tomography with development of:
- efficient algorithms for both the forward problem (Noble et al., 2014) and the inverse problem (Boterro et al., 2016; Luu et al., 2018)
- parsimonious parameterizations of the velocity model (Boterro et al., 2016; Belhadj et al., 2018)

- Geomechanics: our main contributions are the development and the study of mechanical models of frictional faults and frictional fault networks. We use these models to understand seismicity and aseismic deformations in diferent contexts (natural tectonic activity or induced by fluid operations at depth). Recent advances concern:
- the understanding of earthquake nucleation on a heterogeneous Dieterich-Ruina rate-and-state frictional fault (Dublanchet, 2018)
- the understanding of seismic activity induced by a fluid injection into a heterogeneous rate-and-state fault (Almakari et al., 2018)

The three axis are complementary. In a more long-term prespective, the stochastic approach is the central part to evaluate reliable uncertainties