GNGTS 2015 - Atti del 34° Convegno Nazionale

GNGTS 2015 S essione 3.3 143 white arrow in Fig. 1B). Nevertheless, the results are consistent in all the domains, and allow to correctly mark the sedimentary sequence, its lateral boundaries, and even the salt dome intrusion. Such area is indeed a complex target for all the existing picking procedures, due to the high velocity contrast between the salt and the intruded sediments, as well as the abrupt lateral changes in the horizon continuity. Fig. 1 – Application of the picking procedure to a reflection seismic profile. The figure shows: A) the cosine phase section, and B) the resulting picked reflections, superimposed on the amplitude profile, after applying on the horizons energy thresholds equal to 50% of the peak reflection strength in each energy package. Positive amplitudes are marked in green, negative amplitudes in red. The profile images a sedimentary basin (S), the seismic basement (B) and a salt dome (D). The white arrow indicates an example of distinct parallel reflections automatically grouped within the same energy package, preventing the weaker one from being displayed. Refer to the text for further details. Road inspection (GPR) . In this section, we analyze a common offset (CO) GPR data set acquired over an airport runway. The recording system was a Malå Geoscience ProEx GPR, connected to 500 MHz ground-coupled bistatic shielded antennas. An exemplary profile, acquired using a 0.185 ns sampling interval and an average 0.19 m trace spacing, is shown in Fig. 2A. The applied processing sequence consists of DC removal, drift correction, background removal, and band-pass filtering. Amplitude recovery algorithms were not used, and so the signal shows a steady decrease of amplitude with depth (Fig. 2A). Nevertheless, as previously discussed, the cosine phase allows to correctly track laterally continuous events (Fig. 2B), even without any applied gain function. More chaotic results can be noticed in the deeper part of Fig. 2B, where the signal-to-noise ratio is lower, preventing the identification of clear coherent structures. However, most of these artificial horizons are automatically disregarded during the described phase-based horizon grouping process, whose results are displayed in Fig. 2C. In fact,