3.2. Geodynamics.

 

Although earthquakes are widely exploited on the basis of the short time scales associated with the rupture of the fault and wave propagation, studies on the slow built up of stress due to tectonic loading are still largely missing. In order to emphasize the impact of the GOCE mission on the recovery of the density structure at the upper mantle level, which is crucial in a modern approach to seismic hazard studies, we simulate the accumulation of deviatoric stress in central Italy.

A 2D viscoelastic dynamic model allows to quantify the stress distribution due to the geodynamics processes likely active in central Italy: near north-south convergence between Africa and Eurasia. upwelling of hot asthenosphere underneath Tuscany. and underthrusting/subduction of the Adriatic lithosphere under the Apennines.

 

 

Model results are compared with the earthquake distribution and seismotectonic observations. These findings are thus relevant for understanding the process of tectonic loading of the seismogenetic regions and earthquake nucleation.

The geometry of the central, shallow part of the model is constrained by earthquake distribution, tomographic images and seismic data. Seismic tomography and petrological data do not allow, on the other hand, to constrain with sufficient accuracy the density anomalies at lithospheric and upper mantle level. A substantial improvement in the knowledge of the density structure necessary to constrain seismotectonic models is expected from high resolution gravity missions (GOCE).

We have performed a sensitivity analysis varying the density contrast at sublithospheric level, that are likely to exist in the area. The first two displays the model predicted distribution of the equivalent von Mises (deviatoric) stress that correlates with the distribution of earthquakes. The earthquakes falling in a 40 km wide band centered on the modelled section have been projected (circles).

We have verified that varying the density contrast in the subducted slab, Fig. 3.2.1, or in the asthenosphere, Fig. 3.2.2, produces major modification in the pattern of deviatoric stress Fig 3.2.3 and Increasing the density contrast, induces large deviatoric stresses at greater depth in the subducting plate as in Fig 3.2.4. As a matter of fact, deep earthquakes can only be reproduced by deep seated density anomalies.

 

Fig. 3.2.3 and 3.2.4 shows the effects of density variations within the slab at crustal level, where the most damaging sesimicity is concentrated. Again we notice the sensitivity of the extensional stress at the surface (blues) to variations in the density of the subducted slab. An increase in the density, reduces the extension at the surface and increases compression in proximity of the mantle wedge.

These results clearly show the impact of upper mantle density structures on the distribution of deviatoric stress and on the style of extension or compression at convergent margins.

These results indicate the major impact of the lateral variations in the density structure of the lithosphere and upper mantle in controlling the stress pattern in seismogenetic region. The GOCE mission will allow to constrain the density structure at the global level for all the seismogenic regions of the earth, thus contributing in a substantial way in studies of seismic hazard. It is in fact now well appreciated that statistical approaches based on historical records of seimicity cannot be sufficient to mitigate seismic risk, since the slow time scales of earthquakes due to tectonic loading can be of several thousands of years, thus beyond the possibility of historical seismic records. A new appraisal in the field of seismic hazard can be gained by means of the modelling of the slow built up of stress due to tectonic loading and by means of the comparison of the predicted deformation pattern with GPS surveying.

From the examples shown in these figures, it is clear that a detailed modelling of stress built up necessitates the knowledge of the density anomalies in the lithosphere and upper mantle, that seismic tomography and petrology alone cannot provide. Only from the inversion of the gravity data from GOCE, we can obtain a worldwide pattern of the density structures in the uppermost portion of the planet, that has a major impact, together with the relation motions of the plates and rheology of the crust, in controlling the earthquake nucleation in seismogenetic regions.

In figures 3.2.5 - 3.2.7 we show the gravity anomaly in mgals along the same profile of previous figure. Figure 3.2.5 shows the total gravity anomaly over the subduction structure. Figure 3.2.6 corresponds to a density anomaly of the subducted slab of 40 kg/m3, and fig 3.2.7 correspond to a density anomaly of 100 kg/m3. Both figure 3.5.6 and 3.5.7 clearly shows that the wavelength and intensity of the gravity signal from deep seated density anomalies is larger than the error expected by GOCE (see the figures: 2.8, 2.10 and 2.11), and consequently GOCE will substantially improve our knowledge of the density structure with depth.

The same sensitivity analysis has been performed in a three dimensional modelling, to study the effects of density variations in the subducted plate on vertical deformation, that impact the interpretation of sea level rise due to tectonic activity in the central Mediterranean, in particular in the Adriatic sea. Increasing the density contrast in the subducted Adriatic plate to from 0 kg/m3 (Figure 3.2.8) to 40 kg/m3 (Figure 3.2.9) , we notice a substantial widening of the regions subject to subsidence, responsible for a sea level rise of the same amount along the coastlines.

 

 

It is remarkable that a density increase in the subducted slab is responsible for a global increase of the subsidence of the coastlines, and consequently for a sea level rise. Using a scaling argument in terms of sensitivity of sea level rise as a function os density anomalies, a density variation of 10 kg/m3, responsible for a gravity anomaly of about 3 mgals at the surface, wavelengths of 100-150 km, is responsible, along the western coast of the Italian peninsula, for a sea level rise of 0.2 mm/yr, comparable to the sea level change due to viscosity variation in the Glacial Isostatic Adjustment in the far field.

The density variations at the level of the upper mantle, exactly of the size expected to be retrieved by GOCE by simultaneous inversion of gravity and tomographic data, can be responsible for sea level variations of the same amount of those due to Glacial Isostatic Adjustment in the far field with respect to Pleistocenic ice sheets, of 0.1-0.2 mm/yr.

These results clearly show that the knowledge of the upper mantle structure expected from the GOCE mission, will be crucial to constrain, at the global scale, the stress pattern that impact our understanding of earthquake distribution, and to build physical models of sea level changes that impact our understanding of ongoing global changes.