GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 1.3 GNGTS 2023 Lithospheric modeling in Iran and the Arabian Peninsula from tomographic data: first results G. Maurizio 1 , M. Capponi 2 , D. Sampietro 2 , C. Braitenberg 1 1 Dept. of Mathematics and Geoscience, Trieste University, Trieste, Italy 2 Geomatics Research & Development srl, Lomazzo, Italy Introduction Use of seismic tomography as constraint modeling the density of the lithosphere is a valid process to reduce ambiguities in interpretation and increase the validity of defining the physical properties of the lithosphere. Through different empirical laws it is possible to define an acceptable starting density model, with its uncertainties, directly from a Vs or Vp velocity model. The area selected for the work is a vast territory encompassing Iran and the Arabian Peninsula. It is an area of high scientific interest and subject to numerous geophysical studies, both for scientific aspects, for example, because of its location in a collision zone between the African and Eurasian plates, linked among other causes also to the rift of the Red Sea and the Gulf of Aden, and for more practical aspects such as, mining, which is already taking place in some parts of the area. With this presentation, we want to show the first results of the density model realized through a Bayesian inversion applied to an optimized starting model of the aforementioned area of the Middle East. The starting 3D model for the inversion was derived from local tomography, using a simplified version of the Brocher's relation for velocity-density conversion, and recalculating new coefficients based on observations of the gravity field. Processing and results We adopt the velocity cube, of the tomography model of Kaviani et al. (2020). We define the area of interest in a polygon that goes from 10°N to 41°N in latitude and from 36°E to 64°E in longitude, and to a depth of 105 km, leaving unchanged the original resolutions of the starting geometry, with a spatial resolution of 0.25° and vertical resolution varying from 1 km for the first 60 km depth, up to the final 7 km at 105 km depth. At this point, a least-squares collocation procedure (Moritz, 1978) was carried out, to interpolate the starting model with the global tomography model of Simmons et al. (2015). This was performed because of the non-total coverage over the studied area. Once the seismic velocity cube was estimated, the model was divided into five parts: water, sediment, crust, and mantle, and a separate crustal layer in the Red Sea. Specifically, the Moho depth was obtained using the vertical velocity gradient method as presented in Tadiello and Braitenberg (2021), except for the southeastern zone along the Red Sea suture, which had strong