The effect of an a priori known 3-D crustal model in teleseismic tomography of upper-mantle structure is investigated. We developed a 3-D crustal P -wave velocity model for the greater Alpine region, encompassing the central and western Alps and the northern Apennines, to estimate the crustal contribution to teleseismic traveltimes. The model is constructed by comparative use of published information from active and passive seismic surveys. The model components are chosen to represent the present large-scale Alpine crustal structure and for their significant effect on the propagation of seismic wavefields. They are first-order structures such as the crust-mantle boundary, sedimentary basins and the high-velocity Ivrea body. Teleseismic traveltime residuals are calculated for a realistic distribution of azimuths and distances by coupling a finite-difference technique to the IASP91 traveltime tables. Residuals are produced for a synthetic upper-mantle model featuring two slab structures and the 3-D crustal model on top of it. The crustal model produces traveltime residuals in the range between -0.7 and 1.5 s that vary strongly as a function of backazimuth and epicentral distance. We find that the non-linear inversion of the synthetic residuals without correcting for the 3-D crustal structure erroneously maps the crustal anomalies into the upper mantle. Correction of the residuals for crustal structure before inversion properly recovers the synthetic slab structures placed in the upper mantle. We conclude that with the increasing amount of high-quality seismic traveltime data, correction for near-surface structure is essential for increasing resolution in tomographic images of upper-mantle structure.
574ZUTimes Cited:17Cited References Count:45