An Improved shear velocity model beneath the Iranian plateau using adjoint noise tomography

We perform an adjoint waveform tomography using Rayleigh-wave empirical Green‘s functions (EGFs) at periods 10-50 s to improve a pre-existing 3-D velocity model of the crust and uppermost mantle beneath the Iranian plateau. The starting model was derived from a conventional surface-wave dispersion tomography based on high-frequency ray-theory assumption to invert of a quasi-3-D shear-wave velocity model. We use the EGFs from the same study that were derived from cross correlation of continuous seismic noise. Adjoint tomography refines the initial model by iteratively minimizing the frequency-dependent travel-time misfits between synthetic and observed EGFs measured in different period bands. Our new model covers the known tectonic units such as the Central Iranian Block, Zagros fold-and-thrust belt, Sanandaj-Sirjan metamorphic zone and Urumieh-Dokhtar magmatic arc.

Overall, the adjoint tomography provides images with better lateral resolution and depth sensitivity and more realistic absolute velocity values due to the inclusion of finite-frequency waveforms. The use the numerical spectral-element solver in adjoint tomography provides accurate structural sensitivity kernels, which helps to obtain more robust images rather than those generated by ray-theory tomography. The final model adjusts the shapes of velocity anomalies at crustal depth more specifically for the eastern Zagros. The final model significantly improves the initial model at the upper mantle depths and provides a higher resolution for the shape of velocity anomalies.