3 4 0B. physics COEXISTING CONTRACTION-EXTENSION CONSISTENT WITH BUOYANCY OF THE CRUST AND UPPER MANTLE IN NORTH-CENTRAL ITALY.

Size: px

Start display at page:

Download "3 4 0B. physics COEXISTING CONTRACTION-EXTENSION CONSISTENT WITH BUOYANCY OF THE CRUST AND UPPER MANTLE IN NORTH-CENTRAL ITALY."

Andrew Hoover
6 years ago
Views:

_the united.ntos abdus salam1111111 II11111IlIuu ui educational. CCl~.

1 _the united.ntos abdus salam II11111IlIuu ui educational. CCl~ ii11111 and internationial XA a rn centre physics COEXISTING CONTRACTION-EXTENSION CONSISTENT WITH BUOYANCY OF THE CRUST AND UPPER MANTLE IN NORTH-CENTRAL ITALY Abdelkrim Aoudia Alik T. Ismail-Zadeh Giordano Chimera Antonelia Pontevivo and Guliano F. Panza 3 4 0B

2 Available at: it/-pub off IC/2002/ 102 United Nations Educational Scientific and Cultural Organization and International Atomic Energy Agency THE ABDUS SALAM NTERNATIONAL CENTRE FOR THEORETICAL PHYSICS COEXISTING CONTRACTION-EXTENSION CONSISTENT WITH BUOYANCY OF TILE CRUST AND UPPER MANTLE IN NORTH-CENTRAL ITALY Abdelkrim Aoudial Dipartimento di Scienze della Terra, UniversitA degli Studi di Trieste, via E. Weiss 1, Trieste, Italy and The Abdus Salam International Centre for Theoretical Physics, SAND Group, Trieste, Italy, Alik T. Ismail-Zadeh Geophysikalisches Institut, Universitlit Karlsruhie, Hertzstr. 16, Karlsruhe, Germany, International Institute of Earthquake Prediction Theory and Mathematical Geophysics, Russian Academy of Sciences, Warshavskoye shosse 79-2, Moscow, Russian Federation and The Abdus Salamn International Centre for Theoretical Physics, SAND Group, Trieste, Italy, Giordano Chimera, Antonella Pontevivo Dipartimento di Scienze della Terra, UniversitS degli Studi di Trieste, via E. Weiss 1, Trieste, Italy and Giuliano F. Panza Dipartimento di Scienze della Terra, UniversitA degli Studi di Trieste, via E. Weiss 1, Trieste, Italy and The Abdus Salam International Centre for Theoretical Physics, SAND Group, Trieste, Italy. MIRAMARE - TRIESTE September Author for all correspondence. -mnail: aoudia~dst.units.it

3 Abstract The juxtaposed contraction and extension observed in the crust of the Italian Apennines and elsewhere has, for a long time, attracted the attention of geoscientists and is a long-standing enigmatic feature. Several models, invoking mainly external forces, have been put forward to explain the close association of these two end-member deformation mechanisms clearly observed by geohyica 12 and geologica investigations. These models appeal to interactions along plate margins or at the base of the ithosphere such as back-arc extension 7 or shear tractions from mantle flow5, or to subduction processes such as slab roll back, retreat or pull 5 8 and detachment 9 "' 0. We present here a revisited crust and upper mantle model that supports delamination processes beneath North-Central Italy and provides a new background for the genesis and age of the recent magmatism in Tuscany. Although external forces must have been important in the building up of the Apennines, we show that internal buoyancy forces solely can explain the coexisting regional contraction and extension.

4 Twenty years ago an anomaly within the upper mantle in North Central Italy was delineated on the basis of the analysis of seismic surface wave dispersion measurements". Since then, a large amount of data has been available. Deep seismic sounding profiles 3 """, from the Tyrrhenian to the Adriatic coast along a stripe perpendicular to the Apennine chain, samples the crust, while upper mantle structures were fairly well resolved by regional P-wave tomographyl '1. To date, an integrated description of the structure and physical properties of the uppermost mantle and its transition to the crust is missing. To fill this gap we investigate the lithosphere-asthenosphere structure along a stripe from the Tyrrhenian to the Adriatic coast combining new tomnographic inversions of surface wave with other pertinent available data. We consider a large number of new local and regional group velocity measurements sampling the Adriatic margin (Fig. a), the Umbria- Marche Apennines (Fig. b), new and published' 5 phase velocity measurements sampling Italy and surroundings, and deep P-wave seismic soundings which crossing the whole Peninsula, from the Tyrrhenian to the Adriatic coasts, go through the Umnbria-Marche area 3 ". We obtain locally averaged dispersion curves and the corresponding standard errors (Fig. lc) in discrete points'1 6 of the area under investigation and then proceed with a non-linear inversion' 7, where the unknown independent parameters are S-wave velocities and thickness of layers. The resolving power of our surface waves data set is suitable for studies of the in-depth changes of the physical properties of the uppermost 250 km of the Earth interior. Furthermore, the well-tested inversion method we use does not require the assumption of any a priori starting model. Thus the data, as well as the methodology, provide a well understood foundation for our work. We do not choose the solutions of the inverse problem corresponding to the minimum root mean square (rms), since they are the most affected by the presence of possible systematic errors 16. From the set of solutions, we accept the solution (Fig. d) corresponding to the rms closest to the average value, which was computed from all solutions, and hence reducing the projection of possible systematic 2

5 errors into the structural model. f alternatively a median of all solutions in a set is used as representative, the resulting picture is not significantly different from Fig. d. The non-linear inversion of the Umnbria-Marche local dispersion measurements, spanning the period range from 0.8s to 4s, provides the ranges of S-wave velocity (Vs) and thickness for each stratigraphic unit, identified in the pertinent part of the deep seismic soundings, that give us P-wave velocity (Vp) and density. We thus estimate the Vp/Vs ratio for each stratigraphic unit. Fixing the upper crust parameters we resolve the deeper structure from the Tyrrhenian to the Adriatic coast, inverting the longer period dispersion measurements, reaching periods above 100s. In this inversion we assume that the mantle is made by Poissonian material and the density is determined from empirical average relations with seismic wave velocities' 8 According to the resolution length of the data used, the crust exhibits clear layering and lateral variation in thickness (Fig. 2a): it is about 25 km thick below the Tuscany Magmatic Complex (TMC), about 30 km below the extensional thick sedimentary basins and about 35 km below the Umbria-Marche geological Domain (UMD). Similarly a variation of lithospheric thickness is observed: the lithospheric mantle (lid) is thin (about 30km) below the TMC, while it is about 70 km thick below UMD (Fig. 2a). A lithospheric root, more than 120 km wide, between the TMC and UMD, reaches a depth of at least 130 km. All along the profile a mantle wedge with maximum thickness of about 30 km below the TMC, is in-between the crust and the lid (Fig. 2a). The wedge exhibits very low velocities and hence decouples the underlying lid from the crust. Our model provides a new background for the still debated genesis of the TMC. Extensive petrological and geochemical investigations' revealed a large variety of rock types. The rocks are rich in incompatible elements and their crustal-like isotopic signatures are consistent with a genesis in a sub-crustal anomalous mantle that may correspond to our identified low-velocity mantle wedge that was probably metasomatized by addition of subduction-related upper crustal material. Such metasomatic modification could have taken place during the Alpine eastward subduction as revealed by the similar compositions of the TMC lamproites with those identified in the western 3

6 Alps. Furthermore, the about 30 km thick lid we observe below TMC, that could be a relic of the Alpine lithosphere, supports the Alpine subduction. The sub-crustal earthquakes2 seem to cluster in the shallower part of the thick Adriatic lid and in the eastern part of the Jithospheric root, consistently with a slab-like geometry, while the part of the lithospheric root and thin lid to the west seems to be almost free of seismic activity (Fig. 2a). This is seen in many P-wave tomographic images indicating the presence of a high-velocity body of significantly larger volume than that delimited by the earthquakes foci' 0. There are two possible explanations for the existence of a large lithospheric volume with laterally varying mechanical properties under North-Central Italy: (a) the presence of lateral variations in the stiffness of the lithosphere, consistent with the heat flow distribution, and/or (b) the coexistence of the Adriatic west-dipping remnant slab with a possible relic of the old Alpine east-dipping slab consistent with geolgica5,6 ptogia19 and seismic reflection data"" The crust and upper mantle structure beneath North-Central Italy provides clear evidence of a low velocity uppermost mantle with a wedge-like geometry. The high velocity upper mantle underlying this low velocity layer is inferred to be lithospheric mantle material delaminated from the overlying crust. We argue that the eastward migrating extension-contraction reported in the Italian upper crust may correspond to an eastward migrating axis of delamination in the uppermost mantle. The present-day configuration of the crust and upper mantle (sketched in Fig. 2a), that resulted from complex evolutionary stages of thickening and thinning of the lithosphere in interaction with the underlying asthenosphere, may contribute internal forces to the regional geodynarnics. To estimate the contribution of buoyancy to the present-day deformation of the region, we have attempted to fit simultaneously the observed stress data and the earthquake distribution with models of the mantle flow induced by the lithospheric buoyancy. Variations in tectonic (deviatoric) stress are mainly caused by changes of gravitational potential energy 2 (due to 4

7 lateral density heterogeneity of the lithosphere) and by lateral variations of the effective viscosity of the lithosphere. In the uppermost 50 km of the model, we use density values estimated from a highresolution gravimetry survey 23made along our study profile (Fig 2b). For the deeper structures we convert our S-wave velocity model into density taking into consideration temperature efcs The concordance between the gravity anomaly directly computed from our density model (no data fitting) and the observed Bouguer anomaly (Fig. 2b) is a measure of the robustness of our assumptions linked to the density estimates. To avoid boundary effects in the region under study, the modelled cross-section has been extended by 200 km in the horizontal and vertical directions. We prescribe free slip conditions at the boundaries of the extended model. In our model we incorporate surface topography over the cross-section with stress free conditions at this surface. The flow field and tectonic stress are computed employing the equation of momentum conservation and the Galerkin approach 26. Being a least-known physical parameter, the viscosity that exerts influence upon the stress state in the lithosphere resulting in its strengthening or weakening is the only variable parameter in our numerical models; density and geometry of the crust-mantle structure were fixed. We choose a 27 plausible range of viscosity which is in agreement with a recent upper mantle viscosity profile. As shown in Fig. 3, the lateral variation in the viscosity of the lithosphere corresponds to the rather sharp heat flow difference (> 100 mw/in 2 ) observed in the area 28. We consider three models of viscosity profiles: model a is shown in Fig. 3; in models b and c the viscosity of the mantle wedge is one order of magnitude lower than that in model a. In model c, the effective viscosity of the crust beneath the Umbria-Marche region is three times higher than that in model a. The modelled flow field responds directly to the density and viscosity distributions (Fig. 3). The motion of the crust and mantle is caused by the negative buoyancy of the lid and the positive buoyancy of the mantle wedge. The downward motion due to the denser lid sinking in the mantle can be responsible for the subsidence of the Adriatic realm, and the upwelling predicted at the 5

8 western part of the profile contributes to the tectonic uplift in Tuscany, where the horst-graben structure is observed 3 5. Therefore the predicted flow field is in agreement with the regional geological observations and at the same time it explains the delamination of the crust from the underlying mantle lithosphere and how the mantle wedge was emplaced. Young magmatismn at the surface and high heat flow values, in agreement with the computed upward flow field, suggest that this wedge may represent a partially molten mantle. The ages of the TMC rocks ranging from 8-7 to 0.2 Ma 19, with a tendency to decrease from west to east, is in agreement with the computed eastward flow of the low-velocity mantle wedge. We propose that this low-velocity mantle wedge feeds TMC. Fig. 4 shows the state of tectonic stress in the lithosphere resulting from our modelling. By assigning different effective viscosities to the crust and to the mantle wedge, we find that the maximum horizontal compressional stress is associated with the lid below the Umbria-Marche Apennines, in the depth range of 50 to 80 km. This finding is in good agreement with the compressional mechanisms of well-constrained fault-plane solutions 21. The sub-crustal seismicity distribution (see Fig. 2), correlates well with the region of high shear stress in the models. A larger viscosity contrast between the lid and mantle wedge does not affect strongly the stress field (see Fig. 4a-4b) while the higher effective viscosity of the crust, the larger shear stress is (Fig. 4c). In the crust the models predict compressional regime at the eastern part of the profile, as already reported in the literature ',' and tension ust below the Umbria -Marche Apennines, where the 1997 normal faulting earthquake sequence 29took place. Our models predict also tension in-between local compression below the thick Quaternary extensional sedimentary basins (Fig. 4). On the left-hand side of the modelled profile the crust exhibits tension where the TMC outcrops and where the underlying mantle wedge and the asthenospheric low velocity layer are uprisen (Fig. 4). This localised uplift-subsidence type of motion is in agreement with a radial type of exeso'i proximity of Quaternary vertical intrusions 20. 6

9 We have shown here that the buoyancy forces that result from the density distribution in the lithosphere govern the present day deformation within North-Central Italy and can explain regional coexisting contraction and extension. In particular, the peculiar geometry of the crust-upper mantle system controls the distribution of stresses, while the related kinematics is mainly controlled by the lateral variations in density. In this area, the reported unchanged time-space stress field, inferred from the magnetic anisotropy of Plio-Pleistocene sedimnents 30, and the distribution of the thick extensional Quaternary sedimentary basins, in places where our model predicts high magnitude tension (Fig 4), makes us think that buoyancy is the prevailing mechanism in this slow-rate deforming area since, at least, Pleistocene times. 7

10 REFERENCES 1. Frepoli, A. & Amato, A. Contemporaneous extension and compression in the northern Apennines from earthquake fault-plane solutions. Geophys. J. Int. 129, (1997) 2. Montone, P., Amato, A. & Pondrelli, S. Active stress map of Italy. J. Geophys. Res. 104, (1999) 3. Pialli, G_, Barchi, M. & Minelli, G. (eds) Results of the CROP03 deep seismic reflection profile (Mem. Soc. Geol. It., 52, 1998). 4. Lavecchia, G., Brozzetti, F., Barchi, M., Menichetti, M., & Keller, J.V.A. Seismotectonic zoning in east-central Italy deduced from an analysis of the Neogene to present deformations and related stress fields. Bull. Geol. Soc. Am. 106, (1994). 5. Doglioni, C., Mongelli, F. & Pialli, G. Boudinage of the Alpine belt in the Apenninic back-arc. Mem. Soc. Geol. It. 52, (1998). 6. Doglioni, C., Harabaglia, P., Merlini, S., Mongelli, F_, Peccerrilo, A. & Piromallo, C. Orogens and slabs vs. their direction of subduction. Earth-Science Reviews, 45, (1999). 7. Facenna, C., Becker, T.W., Lucente, F.P., Jolivet, L. & Rossetti, F. History of subduction and back-arc extension in the Central Mediterranean. Geophys. J. Int. 145, (2001). 8. Negredo, A.M., Barba, S., Carminati, E., Sabadini, R. & Giunchi, C. Contribution of numeric dynamic modeling to the understanding of the seismotectonic regime of the northern Apennines. Tectonophysics 315, (1999). 9. Carminati, E., Giunchi, C. & Sabadini, R. Numerical modeling of the dynamics of the northern Apennines. Mem. Soc. Geol. t. 52, (1998). 8

11 10. Wortel, M.J.R. & Spakman, W. Subduction and slab detachment in the Mediterranean- Carpathian region. Science 290, (2000) Panza, G.F., Mueller, S., Calcagnile, G. & Knopoff, L. Delineation of the North Central Italian upper mantle anomaly. Nature 296, (1982). 12. Bally, A.W., Burbi, L., Cooper, C. & Ghelardoni, R. Balanced sections and seismic reflection profiles across the central Apennines. Mem. Soc. Geol. It. 35, (1986). 13. Finetti, I.R., Boccaletti, M., Bonini, M., Del Ben, A., Geletti, R_, Pipan, M. & Sani F. Crustal section based on CROP seismic data across the North Tyrrhenian-Northern Apennines-Adriatic Sea. Tectonophysics 343, (2001). 14. Lucente, F.P., Chiarabba, C., Cimini, G.B. & Giardini, D. Tornographic constraints on the geodynamnic evolution of the Italian region. J. Geophys. Res. 104, (1999). 15. Du, Z, Michellini, A. & Panza, G.F. EuriD: a regionalized 3-D) seismological model of Europe. Phys. Earth Planet. Inter. 105, (1998). 16. Panza, G.F. In: The solution of the inverse problem in geophysical interpretation. (Ed Cassinis R., Plenum Publ. Corp., 39-77, 198 1). 17. Valyus, V.P., Keilis-Borok, V.1 & Levshin, A.L. Determination of the velocity profile of the upper mantle in Europe. Proc. Acad. Sc. USSR, 185 (1969). 18. Ludwig, W.J., Nafe, J.E. & Drake, C.L. In: New concepts of sea floor evolution. (Ed Maxwell, A.E., Wiley-Interscience, New York, 53-84, 1970). 19. Peccerrilo, A., Poli, G. & Donati, C. The Plio-Quaternary magmatism of southern Tuscany and northern Latium: compositional characteristics, genesis and geodynarnic significance. Ofioliti 26, (2001). 9

12 20. Venturelli, G., Thorpe, R.S., Dal Piaz, G.V., Del Moro, A. & Potts, P.J. Petrogenesis of calcalkaline, shoshonitic and associate Oligocene volcanic rocks from the northwestern Alps, Italy. Con trib. Mineral. Petro. 86, (1984). 21. Selvaggi, G. & Amnato, A. Subcrustal earthquakes in the northern Apennines (Italy): evidence for a still active subduction? Geophys. Res. Lett. 19, (1992). 22. Jones, C.H., Unruh, J.R. & Sonder, L.J. The role of gravitational potential energy in active deformation in the southwestern United States. Nature 381, (1996). 23. Marson, IL, Cernobori, L., Nicolich, R., Stoka, M., Liotta, D_, Palmieri, F. & Velicogna,. CROP03 profile: a geophysical analysis of data and results. Mem. Soc. Geol. It. 52, (1998). 24. Goes, S., Govers, R. & Vacher, P. Shallow mantle temperatures under Europe from P and S wave tomnography. J. Geophys. Res. 105, (2000) 25. Naimark, B.N., Ismail-Zadeh, A.T. & Jacoby, W.R. Numerical approach to problems of gravitational instability of geostructures with advected material boundaries. Geophys. J. Int. 134, (1998). 26. Ismail-Zadeh, A.T., Panza, G.E. & Naimark, B.N Stress in the descending relic slab beneath the Vrancea region, Romania, Pure Appi. Geophys. 157, (2000). 27. Forte, A. & Mitrovica, J.X. Deep-mantle high-viscosity flow and thermochemical structure inferred from seismic and geodynamic data. Nature 410, (2001). 28. Della Vedova, B., Bellani, S., Pellis, G. & Squarci, P. In: Anatomy of an Orogen: the Appennines and Adjacent Mediterranean Basins (eds Vai, G.B. & Martini, I.P.) (Kluwer Academic Publisher, 2001) 10

13 29. Amato, A. et al. The Colfiorito, Umbria-Marche earthquake sequence in central Italy (Sept-Nov, 1997): a first look to mainshocks and aftershocks. Geophys. Res. Lett. 25, (1998). 30. Sagnotti et al. Magnetic anisotropy of P1lio-Pleistocene sediments from the Adriatic margin of the northern Apennines (Italy): implications for the time-space evolution of the stress field. Tectonophysics 311, (1999). Acknowledgments The work benefited from discussions with B. Mueller and T. Yanovskaya. Comments by G. Houseman substantially improved an initial version of the manuscript. This research was funded by Italian MIUR, CNR, and STC. A.T.I-Z was supported by the Alexander von Humboldt-Stiftung.

14 FIGURE CAPTIONS Fig. 1. Synoptic view of all dispersion profiles considered (a, b) with an example (c) of the locally averaged observed (black dispersion measurements (vertical bars represent measurement errors), compared with the group and phase velocity values (red computed for the accepted S-wave solution (d) (solid broken red line) among other possible solutions (solid broken black lines). The grey area in (d) delimits the portion of the solution's space, sampled by the simultaneous non-linear inversion of phase () and group (A) velocity of Rayleigh waves. The study profile (blue dotted line in (b)), from the Tyrrhenian to the Adriatic coastlines, crosses the Umbria-Marche area. Fig. 2. (a) Schematic S-wave velocity cross-section of the crust and upper mantle in North-Central Italy and the hypocentres (red *) of the earthquakes (vertical bars indicate the depth error) recorded in the period within a stripe 150 km wide along the study profile (blue dotted line in Fig. 1 b). The section has been obtained juxtaposing eleven selected (see text and Fig. 1) solutions of the inverse problem. In each of these solutions, a bold black segment indicates the average Moho depth. Horizontal arrows directed outside and inside indicate extension and shortening observed in the region, respectively. The topmost line presents the surface topography. (b) Observed 23 Bouguer gravity anomaly (blue) vs. gravity (red) predicted from our density model. Fig. 3. Effective viscosity used and flow field predicted by the model a (bottom panel). A reduction of the effective viscosity of the crust and mantle lithosphere at the left of the model correlates with a high heat flow observed in the region 28 (upper panel). The effective viscosity of the modelled crust (model a) is chosen to be four times less than that of the lid. This decrease in the effective viscosity may be explained by inherited strong regional deformations and a presence of fluids. Fig. 4. Tectonic shear stress and compressional axes (ticks) predicted by models (a), (b) and (c) along the studied profile (tensional axes are perpendicular to the compressional ones). The horizontal ticks indicate thrusting and vertical ticks indicate normal faulting. The viscosity profile for model a is shown in Fig

15 ci~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~ E i.~~~~~~~~~~~~~~~~~... r4 j Pt , WE% ~~~~~~~~~~~~~~~~~---- Figure 1 E 2E14 Obs ~ ~ ~ ~ ~ ~ ~ ~~....

16 20D~~~ 0 50 Distance km, IO 2. O OO , 4' 5.0 S-wave veocity(kns Figure 2 "4

17 A $ m... o 50 ~~~~~~~~~~~~1n 200 Dhstarce (kt) o as Figure 3

18 Fxernsicrn Cotmct:on TMC C Figure 4 16

TRANSITION CRUSTAL ZONE VS MANTLE RELAXATION: POSTSEISMIC DEFORMATIONS FOR SHALLOW NORMAL FAULTING EARTHQUAKES

IC/99/80 United Nations Educational Scientific and Cultural Organization and International Atomic Energy Agency THE ABDUS SALAM 1NTKRNATIONAL CENTRE FOR THEORETICAL PHYSICS TRANSITION CRUSTAL ZONE VS MANTLE