Data-driven Layer-stripping Strategy for 3-D Joint Refraction and Reflection Travel-time Tomography

77th Earth Science for Energy and Environment (EAGE) Conference & Exhibition 2015, 1-4 June 2015, Madrid, Spain.-- 5 pages, 3 figuresAs alternative to the most common tomography methods that use either first arrivals or reflection travel times, the joint refraction and reflection method combines both types of phases to retrieve the velocity field and the geometry of reflecting boundaries. This combination provides a denser coverage of the study area. Moreover, since refractions only depend on velocity, they reduce the negative effect of velocity¿depth trade-off intrinsic to reflection travel times. However, this method has issues if several reflectors are inverted at once, which arise from the said dependency between velocity and depth parameters, and the difficulty in including sharp contrasts in velocity models that is inherent to travel-time tomography. Here we present TOMO3D, a code for 3-D joint refraction and reflection travel-time inversion, together with a data-driven layer-stripping strategy that strongly mitigates these issues. The code inherits state-of-the-art forward and inversion solvers from TOMO2D (Korenaga et al., 2000), and the former has been parallelised using a combination of multi-processing (MP) and message passing interface (MPI) extensionsThe first author was funded by Repsol through the CODOS collaboration project with CSIC, and this work was done at the Grup de Recerca de la Generalitat de Catalunya 2009SGR146: Barcelona Center for Subsurface Imaging of the Institut de Ciències del Mar (CSIC). Part of this work was also funded by the MICINN project POSEIDON (CMT2010-21569)Peer Reviewe