GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 3.3 ______ ___ GNGTS 2023 We therefore selected two areas of the western Mediterranean strongly affected by salt domes: the Sardo-Provençal Basin, west of Sardinia (WS) and the Balearic Basin, south of the Balearic Islands (SB) (Fig. 1a).  Methods Since the two analyzed basins are characterized by the presence of salt diapirs deforming the base of PQ, the isochronous map shows an alignment of anomalies along the profiles associated with the strong variations in time-depth of the interpreted horizon. Such anomalies are not shown in the regions between the profiles, where, however, we can safely assume a regular and continuous distribution of diapirs. It is clear that a regional map including areas not covered by 2D seismic profiles is not able to reproduce a realistic state of the basin. For this reason, we decided to explore a way to represent the most likely salt distribution by using different parameters of the contouring algorithm and extrapolating the diapirs distribution: in this way, we do not create a real map, but a realistic one that can give an idea of the halokinetic effect. Applied algorithms Point data obtained from picking the interpreted horizons were used to calculate a structural map using the “convergent interpolation algorithm”: this is an iterative extrapolation algorithm that increases the resolution with each iteration and consists of three steps of parameter selection:  1. Refining - changes the grid resolution; 2. Snap - regrids the data; 3. Smoothing - minimizes the curvature of the grid. The extrapolation distance is initially set to twice the initial grid interval. This algorithm preserves general trends in areas with little data, while respecting details in areas where data have been interpreted. Some of the salt structures can be seen on the seafloor: we used the bathymetric information through a surface trend because we expected that the deeper salt distribution would have a similar trend. The actual trend calculated from the surface considers two variables (a and b) using the least square method. The algorithm calculates the residual, performs the interpolation of the residual and back-transforms the result. The resulting surface has a similar basic shape as the trend surface, but is adapted to the input data. The surface used as trend is smooth and defined in the whole zone (Fig. 1b and 1c).