3.3.7. Uplift of the Calabria-Apennine orogenic belt

The Neogene evolution of the Central Mediterranean was dominated by progressive roll-back of the subducting Alpine Tethys slab, which induced the opening of the Liguro-Provencal (30-16.5 Ma) and Tyrrhenian (12 Ma to present-day) back-arc basins. During the last 10 My, the subducting slab was progressively deformed and disrupted resulting in the opening of slab windows. These are tomographically imaged beneath the Sicily Channel and the Apennines (Fig. 49). This process led to a reduction of the active subduction zone to a width of less than 300 km and to the formation of the Calabrian arc, and probably had a bearing on the thermal regime, volcanism and slab kinematics of the latter. Although the Calabrian subduction zone has been studied by geological, seismological, and geochemical methods, several aspects are still unclear. These include the shallow geometry of the subduction zone, the cause and the mechanism of the Calabrian uplift, the temporal and spatial relationships between orogenic and anorogenic volcanism, and the question of whether subduction and accretion processes are still active.

Fig. 49. a) NE-SE striking tomographic cross-section from the Gulf of Lyon (left) across Calabria into the Ionian Sea (right) and b) three-dimensional image of the upper mantle beneath Italy and the Tyrrhenian Sea (after tomographic model PM0.5 of Piromallo and Morelli, 2003). The cross-section in a) shows in blue the 1000 km long Calabrian slab that penetrates the 410 km discontinuity and flattens out above the 660-km discontinuity. The 3D model in b) shows the reduced width of the Calabrian slab. The green isosurface encloses the volume characterized by velocity anomalies larger than +0.8% relative to average mantle velocities. The layers at 250 and 650 km depth are shown in coloured transparency; blue: regions of higher than average velocity (cold material); red: areas of lower than average velocity (hot material) (after Faccenna et al., 2005).

Fig. 49. a) NE-SE striking tomographic cross-section from the Gulf of Lyon (left) across Calabria into the Ionian Sea (right) and b) three-dimensional image of the upper mantle beneath Italy and the Tyrrhenian Sea (after tomographic model PM0.5 of Piromallo and Morelli, 2003). The cross-section in a) shows in blue the 1000 km long Calabrian slab that penetrates the 410 km discontinuity and flattens out above the 660-km discontinuity. The 3D model in b) shows the reduced width of the Calabrian slab. The green isosurface encloses the volume characterized by velocity anomalies larger than +0.8% relative to average mantle velocities. The layers at 250 and 650 km depth are shown in coloured transparency; blue: regions of higher than average velocity (cold material); red: areas of lower than average velocity (hot material) (after Faccenna et al., 2005).

TOPO-EUROPE aims at assessing a number of key issues by integrating different approaches, including seismology, geochemistry, reflection seismics, radiometric dating, geomorphology, fluid dynamics, and thermomechanic modelling:

Fig. 50. GPS vectors for permanent GPS stations in Italy in the EurAsia fixed reference frame (D'Agostino and Selvaggi, 2004). Differential motions between Sicily (reflecting the motion of Nubia) and the rest of the Italian Peninsula are accommodated in northeast Sicily, notably at the Messina site of large earthquakes (e.g. 23.12.1908).

Fig. 50. GPS vectors for permanent GPS stations in Italy in the EurAsia fixed reference frame (D'Agostino and Selvaggi, 2004). Differential motions between Sicily (reflecting the motion of Nubia) and the rest of the Italian Peninsula are accommodated in northeast Sicily, notably at the Messina site of large earthquakes (e.g. 23.12.1908).

i) Active status of the Calabria subduction process: Paleomagnetic data show that evolution of the Calabrian orogenic arc ended during the mid-Pleistocene (Gattacceca and Speranza, 2002; Mattei et al., 2004). Geodetic data (Fig. 50) show a convergence rate of only a few millimetres per year for this region (Hollenstein et al., 2003; D'Agostino and Selvaggi, 2004). In the tectonic framework of the Calabrian arc, the present deep seismicity along the Wadati-Benioff plane may possibly be related to progressing slab break-off rather than to active subduction. To shed light on this process, a re-analysis of the deep and shallow stress regimes of the subducting slab is required. Combined with the analyses of the large amount of available reflection- seismic profiles, this can shed light on the tectonic style and structural geometries of the underplating/accretion process.

ii) Shallow geometry of the subducting slab: At depths shallower than about 50-70 km, the geometry of the subducting slab is poorly known. Reflection-seismic data show a 15° dipping reflector beneath the Calabrian orogenic wedge that is interpreted as the Moho and that can be traced over a distance of about 60 km (Cernobori et al., 1996). In contrast, at deeper levels, the geometry of the slab is constrained by the Wadati-Benioff plane (Selvaggi and Chiarabba, 1995). Tomographic analyses will shed light on the shallow geometry of the subducting slab.

Fig. 51. SW-NE oriented topographic swath profile across the Central Apennines. Elevation points from the 40 km-wide swaths are projected into the profile and maximum, minimum and mean elevations are calculated for 2 km intervals. Thick dashed line is obtained by fitting a 4th order polynomial fit to the mean elevation. Highest peak is around 3000 meter. Two wavelengths of topography are evident. The long-wavelength topography may be attributed to dynamical mantle support whereas the shorter wavelength topography may be related to elastic flexure in response to produced by normal faulting (modified after Bartolini et al., 2003).

Fig. 51. SW-NE oriented topographic swath profile across the Central Apennines. Elevation points from the 40 km-wide swaths are projected into the profile and maximum, minimum and mean elevations are calculated for 2 km intervals. Thick dashed line is obtained by fitting a 4th order polynomial fit to the mean elevation. Highest peak is around 3000 meter. Two wavelengths of topography are evident. The long-wavelength topography may be attributed to dynamical mantle support whereas the shorter wavelength topography may be related to elastic flexure in response to produced by normal faulting (modified after Bartolini et al., 2003).

iii) Origin of the Calabrian and Apennines uplift: Marked uplift of the Calabrian-Apennines commenced probably during mid-Pleistocene. The wavelength of this uplift (Fig. 51) suggests that it is controlled by sub-lithospheric processes, analogous to those proposed for the remainder of the Apennines (Bordoni and Valensise, 1998). Several 2D models have been advanced to explain this process, such as friction decrease along the subduction surface (Giunchi et al., 1996) or slab break-off (Westaway, 1993). To solve this problem it is necessary to: i) analyze and date the oldest and less manifest evidence for uplift; ii) verify the continuity and integrity of the subducting slab at shallow levels; iii) develop models for 3-D simulations of the above-mentioned processes.

iv) Spatial and temporal evolution of mantle sources of volcanism in the southern Tyrrhenian Sea: In the southern Tyrrhenian Sea, orogenic and anorogenic (OIB) volcanic rocks were emplaced contemporaneously. Although these volcanic rocks have been variably interpreted, there is general agreement that the OIB volcanism is related to mantle flow around the edges of the subducted slab or through slab windows (Doglioni et al., 2001; Trua et al., 2003; Faccenna et al., 2004; Faccenna et al., 2005). To further the understanding of this issue, it is necessary to assess the geochemical and isotopic characterization of the pre-orogenic mantle and to define the contribution of subducted sediments, as well as the spatial and temporal relationship between orogenic and anorogenic magmatism in some key areas where this is still poorly constrained.

Fig. 52. a) Five stages in the evolution of the B08 subduction system analogue model that is contained in a rectangular plexiglas tank (34 cm high, 58 cm long, 14-30 cm wide) and illuminated from the side. A silicone plate is used to simulate the long-term viscous behaviour of the subducting lithospheric slab whilst different glucose syrups are used to simulate the upper and lower mantle. Viscosity ratios for the slab/upper mantle and the lower/upper mantle are 350 and 30, respectively. Note that during stages IV and V the slab flattens out at the upper/lower mantle boundary. b) Diagram comparing the geological timing of trench migration in the Central Mediterranean during the last 40 My (crosses) with the results of different analogue models (A03, B08, B07, C01). The curve for model B08 (green diamonds in panel b, and shown in panel a), which involved a non-homogeneous mantle, fits best with natural observations and differs significantly from the best-fitting exponential curve of model A03, which simulates gravity-driven subduction into a homogeneous mantle.

Fig. 52. a) Five stages in the evolution of the B08 subduction system analogue model that is contained in a rectangular plexiglas tank (34 cm high, 58 cm long, 14-30 cm wide) and illuminated from the side. A silicone plate is used to simulate the long-term viscous behaviour of the subducting lithospheric slab whilst different glucose syrups are used to simulate the upper and lower mantle. Viscosity ratios for the slab/upper mantle and the lower/upper mantle are 350 and 30, respectively. Note that during stages IV and V the slab flattens out at the upper/lower mantle boundary. b) Diagram comparing the geological timing of trench migration in the Central Mediterranean during the last 40 My (crosses) with the results of different analogue models (A03, B08, B07, C01). The curve for model B08 (green diamonds in panel b, and shown in panel a), which involved a non-homogeneous mantle, fits best with natural observations and differs significantly from the best-fitting exponential curve of model A03, which simulates gravity-driven subduction into a homogeneous mantle.

v) Reconstruction of the mantle flows induced by the subduction process: The simulation of narrow subducting slabs (<300 km) is complex because 2D return flow models of mantle flow are unable to simulate this. Narrow slabs are characterized by intense mantle flow around their edges (Dvorkin et al., 1993; Funiciello et al., 2004). This process is also invoked to explain some characteristics of the southern Tyrrhenian and Mt Etna magmatic provinces (Gvirtzman and Nur, 1999; Doglioni et al., 2001; Marani and Trua, 2002). The SKS anisotropy of the mantle beneath Calabria suggest toroidal mantle flow around the edges of the subducting slab (Civello and Margheriti, 2004). Mantle material upwelling from below the slab and flowing towards the orogenic arc may have generated local decompression (Kincaid and Griffiths, 2003), temperature increase (Davaille and Lees, 2004), modification of mantle composition (Marani and Trua, 2002), slab erosion, and acceleration of slab roll-back (Dvorkin et al., 1993). These processes will be analyzed by seismological methods (particularly by considering mantle anisotropy), by geochemical and volcanologic methods applied in areas where both orogenic and anorogenic volcanic rocks occur, and by modelling methods aimed at quantitatively determining these processes in the central Mediterranean (Fig. 52).