The interplay between global tectonic processes anWord文件下载.docx

1、The purpose of this work is to analyse the interaction betweenglobal tectonics and local scale seismicity in a selected region ofthe Mediterranean, the central Apennines.Several results have beenpublished on the modelling of tectonic processesin the Apennine regionand the study of their inuence on c

2、rustal motionsor stress andstrain accumulation using different approaches（e.g. Malinverno &Ryan 1986; Bassi & Sabadini 1994; Giunchi et al. 1994; Faccennaet al. 1996; Carminati et al. 1999, 2001; Negredo et al. 1999a,b）However, none of these has focused on the relation between tectoniceand the seism

3、ic cycle within a selected seismogenic zone. In thisworkwe study the inuence of global scale tectonics on the seismiccycle of the Colorito fault, responsible for the UmbriaMarche1997 sequence.The Apennines developed in a region affected by the collisionbetween Africa and Eurasia and by the subductio

4、n of the Adriaticlithosphere. The tectonic evolution of the chain started in theOligocene, when a collisional phase took place due to the convergencebetween the SardiniaCorsica block and the Adriatic domain（Boccalettiet al.1983）. Extension, initially in a westeast and evolvingin a northwestsoutheast

5、 direction （Patacca et al. 1990）, startedin the Late Oligocene with the 1 INTRODUCTIONopening of the AlgeroProvencalBasin, west of CorsicaSardinia （Rehault et al. 1984）, it continuedwith the opening of the Northern Tyrrhenian Basin in the MiddleLateMiocene （Lavecchia 1988; Sartori 1989） and migrated

6、 southeastwardswith time （e.g. Spadini et al. 1995）. This eastward migration through time of the extensional tectonics in the TyrrhenianBasin and of the external compressional front is a peculiarity ofthe evolution of the Apennine chain （Ricci Lucchi 1986） and hasbeen interpreted as being due to dif

7、ferent geodynamic mechanisms.Malinverno & Ryan （1986） proposed an interpretation based on amechanism of trench retreat or roll-back of the subducting Adriaticlithosphere that would cause the opening of a backarc basin. An alternative interpretation has been proposed （e.g. Wezel 1982）, whichconsiders

8、 the asthenospheric upwelling related to the rifting processin the Tyrrhenian Sea, as a forcing geodynamic mechanism. Furthermore, Cavinato & De Celles （1999）proposed that the bimodal stateof stress in the Apennines is maintained by cornerow in the mantlewedge beneath the crest of the range. Subduct

9、ion in central Italy issuggested by the presence of seismicity down to 90 km （Selvaggi &Amato 1992） and by geochemical analyses of the northern Apennine arc magmatism （Serri et al. 1993）. However, Spakman （1990）,on the basis of tomographic results and the reduced seismicity at intermediate levels, s

10、uggested that the slab may be totally or partiallydetached, while Amato et al. （1993） interpreted the presence of ahigh-velocity body down to 200250 km as evidence of a continuousslab. It is thus clear that the tectonic setting of northern Apenninesis very complex and that anal explanation of the ro

11、le of the severalongoing processes has not been reached yet.A nite-element technique is used to model the stress patternalong a vertical cross-section in central Italy perpendicular to theApennines: this approach improves our understanding of the activetectonic processes in the region under study. I

12、n order to analysethe inuence on the stress eld from various tectonic mechanisms,several different models have been performed, each one characterized by different driving forces. We investigated the effects of thenegatively buoyant subducted Adriatic lithosphere, the effects ofthe extensional SWNE o

13、riented forces expected from the counterclockwise motion of the Adriatic Plate and the effects of small-scaleconvection at the bottom of the Tyrrhenian lithosphere driven by thechange from a passive to active rifting mode （Huismans et al.2001）.The stress distribution induced in the seismogenic part

14、of the crustby each tectonic mechanism is compared with the stress pattern ofthe region deduced from earthquake focal mechanisms, boreholebreakout analysis, in situ stress measurements and young geological deformation features （Frepoli & Amato 1997; Montone et al.1999）. These data are complemented b

15、y recent GPS baseline ratesobtained from new GPS receivers installed along the transect understudy.In Fig. 1 （WSM2000, Mueller et al. 2000） the trace of the modelled section is indicated by the letters A-A. Moving along this direction, from west to east, stress indicators change from extensional（red

16、 symbols） in Tuscany and in the internal sectors of the Apenninebelt, to compressional （blue symbols） in the Adriatic Basin. Similarstress maps have been compiled by Reba et al.（1992） and Montoneet al. （1999）Stress changes, obtained by modelling the seismic cycle alonga shallow low-angle normal faul

17、t, with strike perpendicular to theprole under study, have been superimposed on the stresseld generated by the different tectonic processes. The same approach alreadyused for the simulation of the Mexican subduction zone （Gardi et al.2000） has been followed in the present work, although the modellin

18、gof the seismic cycle is more realistic in the present study. In fact,the locked and free alternating phases on the fault, reproducing thestickslip process, are simulated in a different way with respectto the application carried out by Gardi et al. （2000）. In that studythe locking of the fault was o

19、btained by imposing a zero x- andy-displacement to the nodes of the grid dening the fault, whilethe earthquake was simulated by imposing a xed displacement tothe same nodes in order to reproduce the dislocation. In the presentanalysis, locking is reproduced by impeding the relative motion ofthe two

20、sides of the fault but leaving the fault free to move in thexy-plane. To simulate the occurrence of an earthquake, the two sidesof the fault are left free to move one with respect to the other, along africtionless interface, driven solely by the stress accumulated duringthe loading period of the act

21、ive tectonics. The approach developedin the present study is more realistic than that in Gardi et al.（2000）and represents a substantial improvement with respect to previousmodelling. It is worth emphasizing that this new method facilitatesrelating the slip obtained during a series of subsequent eart

22、hquakesto the earthquake recurrence time and active tectonics.2 MODEL DESCRIPTIONIn order to simulate the sinking of the Adriatic Plate underneaththe Apennines and the seismic cycle on a shallow low-angle normal fault, the MARC package （MSC.Marc 2001） based on a niteelement technique has been used.

23、The simulations were performedunder a quasi-static approximation and the models are purely mechanical, since the coupling between the momentum and the temperature equations is not taken into account. Two models are considered,each one representing a 2-D vertical cross-section （Fig. 2） perpendicular

24、to the strike of the Apennines and of the UmbriaMarcheseismogenic zone, as shown in Fig. 1 by the trace A-A. The 2-Ddomain is subdivided into quadrilateral plane-strain elements, creating an irregular mesh. The element density is increased aroundthe discontinuities in order to ensure a good approxim

25、ation of thesolution. The modelled section is 400 km wide and 420 km deep.The simplied geometry of the plates and the thickness of the different layers have been dened following, for the crustal levels, theresults of the seismic line CROP-03 （e.g. Pialli et al. 1998）, andfor greater depths, the results of tomographic studies （Piromallo &Morelli 1997; Chimera et al. 2003）. Since the depth reached by theplunging lithosphere is still a matter of debate, two hypotheses havebeen tested: a depth of 90 km as inferred from subcrustal intr