GNGTS 2015 - Atti del 34° Convegno Nazionale

-10%) and high values (in our case, 10%) on a checkboard pattern superimposed to the final tomographic model. Using the same acquisition geometry in terms of sources (earthquakes) and receivers used for the inversion, it is possible to create a new set of travel times used for a new inversion whose result will point out the resolvability of the system. In our analysis, looking the results of the differences between the checkboard pattern and the computed inverted model, we observe velocity values of the misfit very close to zero in a rather wide zone at the middle of the model. At each tomographic iteration, we calculated the root mean square of time residuals (the difference between computed and picked travel times). In our case, we stopped the procedure after 4 iterations because the solution in terms of variation of seismic events locations seems immediately stable and reliable (absolute values around 0.15 s in term of root mean square of time residuals, corresponding to an error of 2.5% with respect to the picked times on the whole model). In the final tomographic model, only 5 of a total of 180 earthquakes showed a variation in X-Y direction greater than 10 km; moreover, the most part of these events are located at the borders of the working area; while, as we can expect, in the central part, the variation of seismic events location during tomographic iterations is almost nil. Furthermore, the path of variation of these events is more scattered in X-Z respect to X-Y direction, according to theoretical assumptions for which the uncertainty in the vertical component is always bigger than the errors in the horizontal ones. Results and discussion. Fig. 2 displays the final velocity model, which represents the inversion result associated to the minimum root mean square in terms of time residuals; it shows how the P-velocity changes at four different depth: 4, 8, 12 and 16 km. First, we can observe that the greater the depth of investigation, the smaller will be the imaging of the inner Earth solved because the studied zone is not homogeneously covered by the rays. In fact, we can find very few events that occur below 20 km, in according to the seismicity distribution of our area. In particular, in Figs. 2a and 2b, we can observe an anomalous body in the western sector with P velocity values around 7 km/s beneath central Friuli, at a depth between 4 and 8 km, in agreement with a tomographic investigation already performed (Amato et al. , 1990) and that probably suggests the presence of a structure at a depth of 6 km, which could represent a wedge of metamorphic rocks (Chiarabba and Amato, 1994). Other important aspects are represented by the presence of some anomalous bodies in eastern part of the model, corresponding to External Dinarides in western Slovenia, where we find P velocity in a range between 6.5 and 7 km/s at a depth from 4 to 12 km (Figs. 2a, 2b, 2c) which seems to follow the characteristic trend oriented in NW-SE direction of dinaric thrusts. Finally, in Fig. 2d we can observe that the crustal roots of External Dinarides are deeper than those of the southern Alps, in according to the interpretation of recent active geophysical investigations (Brückl et al. , 2007). In the investigation area, we extracted some sections from 3-D final velocity model (Fig. 1a); here we report only two sections: the E-F section in in south-eastern Alps domain and the M-N section, which results to be perpendicular respect to the trend of dinaric faults. Final tomographic results were compared to geological sections performed for the Friuli Venezia Giulia area to verify the existence of a direct connection between the interpretation of geological sections and the trend of velocity values in Earth’s interior. Different geological formations that we can find are summarized in the stratigraphic column (Tab.1) available from a work of Ponton (2010). In particular, the section E-F (Dogna-Forni di Sopra) represents a longitudinal section between Carnia’s sector and JulianAlps, which includes a portion of important crustal thickness of bedrock. Other important structural element in the central part of the section is the But- Chiarsò line, which appears as an inverse fault with a high angle. 70 GNGTS 2015 S essione 1.2