Non‐linear 2‐D traveltime inversion in complex media: Application to the Southern Apennines thrust belt (Italy)

Abstract

Summary A new traveltime inversion method based on a non linear approach and multi scale process has been applied to a seismic data set acquired with a non conventional acquisition layout in a thrust-belt region. First arrivals and a main reflected phase have been hand picked. A first inversion is realized with only the first arrival traveltimes in order to obtain a 2D velocity image. The latter has been then used as background model for the interpretation of the reflected phase using another non linear multi scale inversion technique. Finally, the whole data set has been jointly inverted. The final velocity images are compared in order to assess the resolution and smearing effect. Moreover, the availability of a VSP survey allows us to independently assess the reliability of our results. Introduction In May 1997 an active seismic experiment was performed by Enterprise Oil Italiana S.p.A. on a test line in the Southern Apennines overthrust-belt (Italy). A high quality data set was acquired with a global offset acquisition configuration (Dell’Aversana et al., 2000) by deploying 160 receivers and 233 sources along a profile 14600 m long. The investigated area is characterized by a high structural complexity (i.e. strong vertical and lateral velocity variations) and presents a rugged topography. In such a context, seismic imaging of the crust by conventional reflection seismic is a difficult task. Aimed at investigating whether complex shape reflectors and velocity structures can be properly imaged by non-linear traveltimes inversion, we used the global offset data set to test three new methods for the traveltime inversion of first-arrivals and reflected phases. First, the information of the first-arrivals have been used to obtain an accurate 2-D P-velocity model by tomography. Then, we use the retrieved image as background velocity model for the traveltimes inversion of near-vertical/wide angle reflections in order to define the shape and the location of the corresponding interface. Finally, both the first- and the secondary arrival times are jointly inverted. The comparison of the three images obtained with the different techniques allows us to discuss the problem of resolution and smearing effect when the target medium is very complicated. A VSP survey performed in a nearby well allows for the verification of the results.