GNGTS 2017 - 36° Convegno Nazionale

GNGTS 2017 S essione 3.1 543 REPROCESSING OF OLD REGIONAL SEISMIC LINES FOR NEW GEOTHERMAL TARGETS: THE CASE OF CROP-03 G. Cassiani 1 , M. Giustiniani 2 , R. Nicolich 2 , P. Primiero 3 , C. Strobbia 4 , U. Tinivella 2 1 Dipartimento di Geoscienze, Università degli Studi di Padova, Italy 2 Istituto nazionale di Oceanografia e di Geofisica Sperimentale - OGS, Sgonico (TS), Italy 3 Seismometrix srl, Angera, Italy 4 Realtimeseismic, Pau, France Introduction. The seismic data of the CROP-03 regional line were reprocessed using a series of alternative approaches for the perturbation corrections and noise attenuation, for the data conditioning and for the imaging (i.e, Giustiniani et al. , 2015). The reprocessing of the seismic lines is part of an ongoing effort on different regional lines in Europe, including the French ECORS and German DEKORP. Despite the relatively low number of channels with respect to modern 2D onshore seismic programs, these old data have a remarkable quality and a high raw signal-to-noise ratio, due to the very large source effort. Large explosive charges or very long sweeps with high vertical stacking were used to acquire data with long offset, 6 kilometres, to image the deep crustal structures. These data today can be used to locate and image deep geothermal targets and better understand the regional geology: thanks to the improvement of the processing and imaging algorithms approaches and software, the reprocessing is often able to improve legacy images. The data processing with pre-stack time and depth migration also allows obtaining better velocity models, for the regional and local seismological applications, including the calibration of the models for the monitoring of the seismicity. The processing of the seismic data, after the reconstruction of the geometry, a crucial challenge with some of the ECORS data, follows a specific workflow designed to handle noisy land data with a crooked line geometry and considerable perturbations. In this work, we present the reprocessing CROP-03 which was recorded across the northern Apennine Arc, with nearly E–W direction, from Tyrrhenian to the Adriatic Seas. The profile allowed exploring the crust to understand the origin of the main heath flow anomalies, looking also for hidden resources and new exploitable geothermal reservoirs. Method. As far as the CROP 3 which is discussed in this paper, the data processing started with an accurate near-surface characterization, based on the exploitation of the information in the refracted arrivals with a non-linear diving-wave tomography. The quality of the data and the accuracy of the first break picking is excellent, due to the large source effort with explosive sources. An automatic first break picking algorithm has been tested to assess the possibility of rapidly picking the full offset range. The non-linear diving wave tomography has been performed with a multiscale gridding approach, to obtain the best compromise between the investigation depth, the stability of the deep velocity section and the resolution in the first hundreds of meters. A smoothed version of the refraction tomography velocity section was used as starting model for the acoustic Full Waveform Inversion, to further refine and improve the near-surface velocity model. The investigation depth exceeds 1 km in some areas due to the long offset, and shows an excellent correlation with the geological sections, and with the reflection images. The velocity sections show large lateral velocity variations, with areas characterized by large shallow velocities and basins with lower velocity sediments. The velocity model, besides being the key for the near-surface perturbation corrections, is used in the velocity-modelbuilding for the pre-stack depth migration. The velocity model can also be used to calibrate microseismic localization algorithms. The perturbation corrections were computed using both a wave-equation datuming (WED) process, published by Giustiniani et al. (2015) and a static correction to a floating datum in depth. The WED consists of a downward and upward continuation of the wavefield between two arbitrary chosen surfaces, to image the subsurface from a different reference surface and