CRUSTAL DEFORMATION IN SUBDUCTION ZONES

Document Sample
CRUSTAL DEFORMATION IN SUBDUCTION ZONES Powered By Docstoc
					GNGTS – Atti del 19° Convegno Nazionale / 14. 05


C. Pauselli (1), C. Federico (1) and J. Braun(2)
(1)
      Dipartimento di Scienze della Terra, Università degli Studi di Perugia
(2)
      Researc h School of Earth Sciences, The Australian National University, Canberra, Australia



               CRUSTAL DEFORMATION IN SUBDUCTION ZONES:
                 INFORMATION FROM NUMERICAL MODELING

Abstract. In this work a fully coupled, thermo -mechanical, 2D, finite-element model has been applied
in order to study the geodynamical evolution of t wo orogens: the Lachlan Fold Belt, LFB (Victoria,
Australia) and the Northern Apennines, NA (Italy). For both t he orogens, their geodynamical evolution
is still a matter of debate and different models have been proposed. In this study, the model
experiments are based on deformation of the crust induced by kinematic basal boundary conditions
simulating the subduction of part of the lithosphere. This does not mean that the simulations done
want to prove that the subduction is the unique answer to the problem, but they underline that the
mantle subduction cannot be rejected. The developed finite element code used in this work is based
on the Dynamical Lagrangian Remeshing (DLR) method (B raun And Sambridge, 1994). In the studied
models, the lithosphere is regarded as a bidimensional non-linear Maxwell viscoelastic body capable
of brittle failure at low pressure and of viscous creep at elevate temperat ure. B ecause it is based on a
Lagrangian method, in which the numerical grid is attached to material particles and advected with the
deformation, the DLR algorithm is ideally suited to track material boundaries and properly represents
free surface. The dynamical mesh refinement is done through the insertion of nodes in regions of large
deviatoric strain. Into the model obt ained with the DLR there is no a priori c onstraint on the fault
geometry or on the rheology, but lithos phere compositionally layered has been assumed and the
rheology of each element is not determined a priori but is a function of the temperature, the state of
stress and t he deformation history. The aim of this work is to study the influenc es of temperature and
the crustal rheological stratification in an orogen driven by subduction. At the same time, the numeric al
simulations have provided insight into the dy namics of t he LFB and NA system. As regard as the LFB,
in particular, it was noted that the well-documented eastward propagation of t he deformation front that
led to the formation of a set of west-dipping imbricate thrusts is consistent with the mechanic al
behavior of a thick brittle layer, decoupled from the underlying oceanic crust by a weak decollement. In
the NA case, the deformation induced by a subduction zone ret reat in which an anomalous
temperature field is considered, has been analyzed. The performed simulations underlined the
presence of subduction slab-retreat that determines the formation of a volume deficit above the
subduction window and a collapse of the previously formed structures was shown.


  DEFORMAZI ONE CROSTALE IN ZONE DI SUBDUZIONE: I NFORMAZI ONI DA MODELLAZIONE
                                   NUMERICA

Riassunto. In questo lavoro, un modello termo-meccanico ad elementi finiti, è stato applicato allo
studio dell'evoluzione geodinamica di due orogeni: la Lac hlan Fold Belt (Victoria, Australia) e
l'Appennino S ettentrionale (Italia). Per entrambi gli orogeni, la loro evoluz ione geodinamica è
argomento di dibattito e numerosi modelli sono stati proposti. In questo studio, le simulazioni effettuate
sono bas ate su un modello di deformazione crostale indotta da condizioni cinematiche simulanti la
subduzione di una part e della litosfera. Questo non significa che le simulazioni vogliono provare che la
subduzione è l'unica risposta al problema, ma che la subduzione non può comunque essere rifiutata.
Il programma usato è DLR (Dynamical Lagrangian Remeshing, Braun and Sambridge, 1994). Tale
programma ha come caratteristica principale quella di poter lavorare anche in presenza di grandi
deformazioni, ridefinendo la griglia di partenza attraverso l'inserimento di nuovi nodi nelle regioni
maggiormente deformate. L'analisi viene effettuata su modelli bidimensionali assumendo che la
litosfera si comporti come un corpo visco -elastico non-lineare (Max well) in cui la viscosità è
dipendente dallo stato di stress e dalla temperatura. In questo modo, la reologia di ogni elemento non
è determinata a priori, ma è funzione della temperat ura, dello stato di stress e della storia deformativa
precedente. Scopo principale di questo lavoro è quello di studiare gli effetti che le variazioni litologiche
e di temperat ura hanno sulle deformazioni crostali finali . Le simulazioni effettuat e hanno anche fornito
utili indicazioni di carattere generale sulla storia geodinamica degli orogeni Australiano ed
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


Appenninico. P er quanto riguarda l'orogene Australiano, in particolare è stato verificato come la
presenza di uno strato relativamente più 'duttile' rispetto al materiale circostante, possa essere
responsabile della propagazione verso est del fronte compressivo caratterizzant e la Lachlan Fold Belt.
Nel caso dell'Appennino, si è studiata la deformazione crostale in uno scenario geodinamico
caratterizzato dall'arretrament o di un piano di subduzione, e dalla presenza di un campo anomalo di
temperatura. Le modellizzazioni effettuate hanno evidenziato che la presenza di un piano in
arretramento determina un "collasso" delle strutture precedentemente formate.



      INTRODUCTION

      The geodynamical history of an orogen could be only completely described
using all the geological and geophysical data acquired over the years. Thus, it will be
useful to integrate field geology (structural, geological, stratigraphic analysis), that
gives information about the shallow structure, with geophysical investigation
(gravimetry, seismic soundings), that offers a tool to understand the deeper geometry
of a structure. Along with the geological a nd geophysical investigations, quantitative
analysis of the geological structures has been developed in the last years. In
particular, with recent advances in computing power, numerical simulations have
considerably improved the understanding of the geomec hanical and geodynamical
behavior of a particular region. Among the numerical techniques used in geology, the
Finite Element Method (FEM) is the most widely used to solve the static and
dynamical problems of a geological structure with complex geometry and linear
and/or non-linear rheology.
      In recent years, numerical simulations have represented a novel way of testing
geodynamical models. FEM modeling has been used to constrain the geodynamical
evolution of an orogen or to emphasize factors controlling the belt evolution. For
example, the evolution of an orogen in a convergent regime has been analyzed
looking at the thermo-mechanical consequences of crustal shortening (e.g., Batt &
Braun, 1999), at the behavior of the continental lithosphere during plate convergence
(Pysklywec et al., 2000) or at the cause for syn-orogenic extension for a variety of
rheological models (see e.g., Willett, 1999). Several works have been dedicated to
subduction zone roll-back to shed some light on the physical parameters controlling
back-arc extension (e.g., Whittaker et al., 1992; Hassani et al., 1997) or to evaluate
the effect of a retreating subduction zone on deformation (Waschbusch & Beaumont,
1996). In this work we want to focus the attention on two factors within already
proposed scenario of the geodynamical evolution of an orogen: the temperature and
the crustal rheological stratification. In order to analyze the importance of these two
different parameters we have used a coupled thermo-mechanical finite element
model of crustal deformation driven by mantle subduction. The developed finite
element code used in this work is based on the Dynamical Lagrangian Remeshing
(DLR) method (Braun and Sambridge, 1994). We have modified the DLR code in
order to correctly introduce the rheological stratification, thermal anomaly and to
simulate a slab retreat. The modified code DLR has been applied to two different
belts: the Lachlan Fold Belt (here and after LFB) (Eastern Australia) and the Northern
Apennines (here and after NA) (Central Italy). These two orogens are extremely
different as regards as the geographical position or the geological setting, but for
both the orogens their geodynamical evolution is still a matter of debate and different
models have been proposed. We have chosen to study these two belts for two
reasons: the first is that among the geodynamical models proposed, for both the
orogens, one involves the presence of a subduction plane. The second reason is that
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


they represent a key cases for the controlling factors that we want to focalize: the
temperature and the crustal rheological stratification in an orogen driven by
subduction. In particular, the LFB has allowed us to study the influence of the
rheological stratification on the final deformation. In fact, it has bee n well accepted
(Glen, 1992) that the mid-Palaeozoic deformation on the LFB was controlled by sub-
horizontal shearing along a basal decollement, that is a weak layer within the crust.
The NA has been chosen to look at the influences of the temperature in a n orogen
driven by subduction. In fact, it has been shown that a key role on present
deformation seems to be attributed to the thermal conditions present during its
evolution (e.g., Mongelli et al., 1989; Della Vedova et al., 1991; Doglioni et al., 1998).
At the same time, the numerical simulations have provided insight into the dynamics
of the LFB and NA system.
      In what follows we present just a brief description of the geological setting of the
two orogens, then the numerical modeling and, at the end of the paper, results and
conclusions.


      GEOLOGICAL PROBLEM

     The Northern Apennines (NA) is situated in the Central Italy and it is the result
of the convergence between the previously formed Alpine orogen and the Adria
promontory (Boccaletti et al., 1980) (Fig. 1).




Fig. 1 - Simplified geological map of the Northern Apennines. 1a) P ost -orogenic sediments; 1b) P ost-
orogenic magmatic rocks; 2) L.s. Ligurids; 3a) Non-metamorphic Tuscan carbonate sequence; 3b)
Non-met amorphic Tuscan siliciclastic sequence (Macigno); 4) Met amorphic Tuscan carbonate and
siliciclastic sequenc e; 5a) Umbria -Marche carbonate sequence; 5b) Umbria-Marche siliciclastic
sequence (Marnoso Arenacea); 5c) Marche foredeep sediments (modified after D’Offizi et al., 1994).
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


     The Lachlan Fold Belt (LFB) is part of the Tasman Orogen (Eastern Australia)
which is a fragment of an originally much longer system formed along the Pacific
margin of Gondwanaland during Paleozoic time (Fig. 2).




Fig. 2 - (a) Geological map of Australia showing the position of the Laclhlan orogen within the
Tasmanides and major Late Proterozoic and early Paleozoic element in cratonic Australia (aft er Foster
et al., 1999); (b) Map of the Lachlan Fold Belt showing structural subzones (numbered), structural and
aeromagnetic trends (thin lines), fault traces (heavy lines), the Wagga Metamorphic Belt (shaded
region) (after Gray and Foster, 1997).

      The differences between these two belts are not only geographical but regard
different factors: the age of deformation is largely Ceno zoic for the NA, whereas is
Paleozoic for the LFB; the NA was developed for the interaction between the Africa
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


and Eurasia plate whereas the LFB is part of the Tasman Orogen (Eastern Australia)
which is a fragment of an originally much longer system formed along the Pacific
margin of Gondwanaland. But, for both the orogens, among the numerous
geodynamical models proposed, there is one that hypothesizes the existence of a
subduction plane.
      As regards the LFB, different authors have explained the marked differences
across the belt (in lithology, magmatic rock type, metamorphic grade and structural
trends, Fig. 2), notably the timing of deformation in the various provinces, proposing
a tectonic model that involves the presence of three subduction zones (Gray and
Foster, 1997). The various deformation fronts are related to 'accretionary style'
deformation at the leading edges of overriding plates, in an inferred southwest
Pacific-type subduction setting from the Late Ordovician to the early Caboniferous,
along the former Gondwana margin. (Foster et al., 1999). In the western LFB, fault
zones within the chevron folded turbidite sequence reflect an imbrication process
within an oceanic setting that is most likely to be related to subduction.
      This interpretation is supported by (a) the widespread occurrence of
intermediate pressure metamorphism associated with a relatively low geothermal
gradient, and (b) the presence of relict intermediate to high-pressure metamorphic
assemblages within associated fault-bounded mafic volcanic rocks. It is well
accepted (Glen, 1992) that the mid-Palaeozoic deformation in the Western LFB was
controlled by sub-horizontal shearing along a basal decollement. Recent 40Ar-39Ar
dating of movement along the major faults in the area clearly indicates that
deformation propagated from the boundaries of the orogen towards its interior
between 460 Ma (latest Ordovician-earliest Silurian) and 390 Ma (Early to Middle
Devonian).
      The presence of a plane of subduction for the NA is still a great matter o f
debate. Possible evidence of an Apenninic subduction has been documented by
structural, geophysical and volcanic data suggesting that the subduction is still active.
The results obtained from tomographic studies (Amato et al., 1993; Spakman et al.,
1993; Selvaggi and Chiarabba 1995; Piromallo and Morelli, 1997) have provided
images of a subducting plate in the upper mantle of Italy. The data show the
presence of an almost vertical cold body underneath the Northern Apenninic arcs
that could be interpreted as a slab. In the upper 200 km, below the Northern
Apenninic arc, the slab is considered continuous by Amato et al. (1998) and
references therein, on the basis of the results of studies on seismic tomography,
deep seismic anisotropy, and the state of stress. The east-ward directed roll-back of
the Adriatic slab is considered to be responsible for the eastward migration of both
extensional and compressional fronts. Doglioni (1991), in the framework of a
delamination-subduction model of the Adriatic-Ionic continental lithosphere, proposed
that the motor of the process is the eastward ''relative'' asthenospheric mantle flow.
In that scenario, the geodynamic evolution of the Apennines is controlled by the
westward dipping and passive roll back and eastward retreat of the Adriatic
lithosphere. Almost simultaneously, an active asthenospheric rise is intruded on the
rear of the retreating and rolling back lithosphere.
      Starting from the already described geological settings, simulations will be
performed in order to investigate, in the example of the NA, the behavior of an
orogen driven by a slab retreat an by an anomalous hot pulse whereas, in the case of
the LFB, the behavior of an orogen driven by subduction with the presence of a weak
decollement in the crust. We want to underline that, although the experimental setup
bears some resemblance with the tectonic setting of the NA and of the LFB, the aim
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


of the simulations is not to exactly ''reproduce'' all characteristics of the two orogens.
Moreover, in order to emphasize the influences of rheological stratification and
temperature, we have simplified the initial geodynamical scenario both for the LFB
and the NA.




Fig. 3 - Numerical modeling: a) model LFB1. Initial and basal kinematic boundary conditions are
reported; b) model APP1 showing initial and basal kinematic boundary conditions. S, the subduction
point, dynamically determined separation point in the crustal flow (see text for details).



      NUMERICAL MODELING

      The continental crust has been numerically studied through a fully coupled
thermo-mechanical finite element numerical code written and developed by Braun
and Sambridge (1994). The two-dimensional version of the quasi-static force balance
equation has been solved using the Dynamical Lagrangian Remeshing method
proposed by Braun and Sambridge (1994). In the models the lithosphere is regarded
as a non-linear Maxwell visco-elastic body capable of brittle failure at low pressure
and of viscous creep at elevated temperature. In our cases three rheological datasets
have been selected to build three ''synthetic materials'': a normal strength material
(NORM), quartz-dominated rheology based on the quartzite rheology of Paterson
and Luan (1990), a medium strength material (MED) based on the Adirondak
granulite rheology of Wilks and Carter (1990), a soft material (SOFT) in which the
strength is artificially reduced by a factor of 10 with respect to the MED material, and
a strong material (STRONG) based on the Anheim dunite rheology of Chopra and
Paterson (1981). The reader should not regard these materials as representing rocks
that may be present in the various parts of the crust/mantle system under
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


consideration, but rather as proxies for the strength heterogeneities that exist within
the Earth's lithosphere.
       All models have a rectangular geometry with a thickness (D) of 50 km and a
length (L) of 900 km and in Fig. 3a the experimental setup performed for the LFB is
reported whereas the Fig. 3b represents the one for the NA.
       In order to minimize the effects of the boundary conditions on the final results,
we choose a L/D ratio equal to 18. In all numerical experiments the deformation style
results from shortening driven by subduction. Shortening is imposed on the system
by applying a velocity distribution along its base. The ''S'' point in Fig. 3a-b (similar to
the ''S'' point in the model of Willett et al., 1993) represents the separation point in the
flow. To simulate subduction, the velocity is imposed in such way as cause
convergence under a block that is considered fixed. The imposed conditions imply
that the mantle subducts whereas the crust undergo thrusting. The top of the model
is allowed to move in any direction (Fig. 3a-b). The velocity distribution as well as the
other mechanical and thermal boundary conditions are shown on Fig. 3a-b: note in
Fig. 3 the region where an anomalous thermal pulse is applied. The convergence
velocity (v 0 ) is chosen to be 10 mm/a and is imposed to the right-hand side boundary
of the model as well as along the base of the model to the right of a so-called
''subduction window'' of width (w). Within the subduction window, the components of
the velocity are imposed to be:

                                         u  v0 cos 
                                         v  v0 sen 

for the simulations of the LFB and

                                       u  v0 cos   v r
                                       v  v0 sen 

for the simulations of the NA, where v r is the assumed velocity of the slab-retreat.
       To the left of the subduction window, both velocity components are set to zero.
The subduction window is a first-order (cinematic) representation of the assumed
subduction process. The width (w) and the ''dip'' () characterizing the subduction
window determine the flux of material being consumed by the subduction.
       The fixed left-hand side of the model represents a ''stable'' continental block
made of a 30 km thick crust (MED material) and a 20 km thick mantle (STRONG
material) that in the case of the LFB represents the stable Delamerian orogen. The
lower part of the lithospheric mantle is not included in simulations. The right-hand
side represents a piece of lithosphere that is being subducted beneath t he continent.
The boundary conditions are set in such way that is possible to ensure that only the
lower part of the subducting plate, made of STRONG material, is forced to be
subducted beneath the continent.
       The 20 km-thick upper layer of the LFB represents a thick sedimentary layer
(and possibly part of the oceanic crust) that is mechanically decoupled from the
underlying mantle. In these models a decollement made of SOFT material has been
introduced in the sedimentary layer assumed to be made of MED and NORM
material (Fig. 3a) in order to verify the influences on the final deformation of a
stratified crust. On the contrary, because the aim of the simulations performed for the
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


NA is to explore the influence of the roll-back, the rheological models of the NA have
been simplified to avoid the influence of other factors on the final results (Fig. 3b).
       No dataset exists that can inform us on the thermal characteristics of the
lithosphere beneath the LFB in the early Palaeozoic. An arbitrary value of the basal
temperature (750C at 50 km) that is consistent with the moderately low geothermal
gradient observed nowadays in borehole data from the area has been chosen (Offler
et al., 1998).
       The roll-back of the west-directed slab simulations for the NA, implies a
substitution of the lithosphere by the underlying mantle and, as a result an
astenospheric rise is present at the rear of the retreating and rolling back lithosphere.
In order to simulate the increase in heat flow connected with this process, models are
performed taking into account the influence of an anomalous temperature field (the
thermal anomaly in Fig. 3b). In this way, starting with a fixed temperature equal all
along the bottom of the model, at each time step of the simulation to the left of the
''S'' point, a temperature increment was imposed. To evaluate this temperature
anomaly, it was assumed that the heat flux density on the surface at the beginning of
the Apennine orogenesis (about 30 Ma, Oligocene time), was equal to 50-55
mW/ m 2 , a current value measured on stable platforms such as the Iblea and Apula
(Atlas of Geothermal resources in Europe, 1994). In order to obtain an initial thermal
field (T=0) characterized by a surface heat flow equal to 50-55 mW/ m 2 , we have
chosen obtain a geothermal gradient of about 20 C/km that give an imposed
temperature at the base of the model equal to 1000 C. The present geothermal
gradient for the hotter part of NA, where the astenospheric dome is supposed to be,
was estimated to be, on the average, greater than 40 C/km (see e.g., Mongelli et al.,
1989; Pauselli and Federico, 2001). Consequently, after 30 Ma (i.e. at the end of the
simulation), the basal temperature of the model to the left of the ''S'' point will reach
approximately double its initial value. For sake of simplicity, a linear increase in
temperature was assumed in the model.


      RESULTS

      LFB

      Here will be reported some results about the models performed for the Lachlan
Fold Belt. More details could be found in Pauselli & Braun (2001).
      The results of the models will be plotted by means of the contours of the second
invariant of the deviatoric part of the strain rate tensor, that gives information on the
instantaneous (or current) deformation field. The Fig. 4 represe nts the distribution of
the strain rate for the LFB1 model, at four different time steps during the simulation.
GNGTS – Atti del 19° Convegno Nazionale / 14. 05




Fig. 4 - Result LFB1 experiment: contour plots of the second invariant of the deviatoric pert of the
strain rat e tensor: different symbols indicate where diffuse strain distribution (folding) localized shear
(fault), sedimentation and erosion happen during the simulation.

      Looking at the shallow structures, the overall pattern of deformation within the
orogen shows a progression from a diffused strain distribution, which could be
associated with the development of open folds (panels a), to a situation where quasi -
rigid blocks are rotated along discrete crustal-scale faults (panels b-d). With time,
there is also a propagation of the deformation field from the S point towards east.
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


This is caused by the horizontal stress differences resulting from the surface
topographic gradient and is facilitated by dextral shear along the low-viscosity
decollement located at about 10 km at depth. Note that the vertical deflection of the
oceanic lithosphere under the weight of the thickened crustal column has been
neglected . This simplification leads to underestimating the amount of shortening and
thickening required to induce the deformed field able to construct the orogen. We
must note that a pair of conjugate shear zones rooting into the decollement appears
in panel b, in the pro-side, east of the S point. In this work the terms of pro-side and
retro-side are used following the convention proposed by Willett et al., 1993: thus the
pro-side is the zone that is east of the ''S'' point while the retro-side is the zone west
of the ''S'' point. With increasing strain, one of the two dipping shear planes is
abandoned whereas the other continues to accumulate deforma tion as is shown in
panel d. This behavior of the model can be interpreted as the evolution of the system
from a situation, where deformation of the crustal layer is accommodated by shearing
along its base together with the development of large scale open folds, to a situation
where localized shearing (faulting) propagates upwards from the basal decollement.
In the early stages of deformation, the regions that are comprised between each pair
of dipping shear zones experience uplift and surface erosion. The regions directly
adjacent to each of these structures experience sediment accumulation. With the
progressive broadening of the orogen and the initiation of new structures, areas that
experienced accumulation become themselves subject to uplift and erosion. This
leads to reworking and transport of previously deposited sediments towards the right
boundary of the orogen. It is interesting to note that even if the main deformation field
is progressively moving toward east, a movement toward west on the subducti on
window is also present.
       During the simulation a shearing E-dipping structure is also located through all
the lithosphere west of the ''S'' point, in the retro -side, that in our model is assumed
to be the Delamerian Orogen (Fig. 3a).


      NA

      The obtained results are plotted contouring the second invariant of the
deviatoric part of the strain rate tensor also in this case.
      The APP1 model (Fig. 5) shows a first, crustal-scale structure (panel a)
localized within the subduction window. With the retrograde slab migration, the
development of a volume deficit is observed. In fact, from at the beginning of the
simulation, a basin is formed above the subduction window (panels a and b). The
material deficit increases as the model evolves, causing the basin to grow i n size.
Extension is observed in the pro-side, as the model responds to the material deficit
created by the subduction zone retreat. The particular feature of this geological
setting is that no high range of shortening and thickening occurs. The plot of the
model surface morphology at different steps of the simulation (Fig. 6) shows an
increasing of the thickening on the pro-side followed by a ''collapse'' of the structures
on the retro-side.
      It is worth pointing out that, looking at the geometry of the subduction plane
(panels a to d), the performed model represents a case in which a very quick
increase in the angle of subduction occurred. The rapid verticality of the subduction
plane may look unrealistic, but it is interesting to note that a similar morpho logy, i.e. a
volume deficit in the subduction window is obtained for an analogous model
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


proposed by Waschbusch and Beaumont (1996), in which the angle of subduction is
constant and equal to 30° during all the simulation.
       Unfortunately, the complexity of the applied boundary conditions induces in out
case a high mesh deformation that stops the simulation from continuing for a period
longer than about 7 Ma (Fig. 5, panel d). For that reason it is impossible to compare
the obtained deformation and morphology with those of the NA. However the already
described particular features of the model emphasized the influence of the slab-
retreat on the NA dynamics.
       The effect of the probable alternances of relative cold and hot phases (Doglioni
et al., 1998), during the geodynamical evolution of the NA, have been studied in the
APP2 (Fig. 7).




                                                              50 km




Fig. 5 - Results of the APP1 experiment: Countour plots of the second invariant of the deviatoric part
of the strain rate tensor.
GNGTS – Atti del 19° Convegno Nazionale / 14. 05




Fig. 6 - Model morphology for the different steeps of the APP1 experiment.

      Here, after a period in which the geothermal gradient is kept very low (less than
15 C/km), the geothermal gradient has been increased. The increase of the
temperature has been set in such way that at the last step the temperature field is
comparable with the present thermal state of the NA (Fig. 3 of the work of Pauselli &
Federico, 2001).
      Looking at the panel a in Fig. 7, we discover that the formation of the basin
above the subduction window, shown in the APP1 model, is not here observed. On
the contrary, the diffuse, brittle behavior of the material due to the low initial value of
the geothermal field, tends to generate a material excess and an orogenic belt is
formed (Fig. 7, panel a).
      It is interesting to note that, looking both at the contouring of the second
invariant of the deviatoric part of the strain rate tensor (Fig. 7) and at the model
morphology (Fig. 8), the previously formed orogen (panel a in Fig. 7 and curve a in
Fig. 8) is destroyed with the increase in temperature. In particular, one of the two
dipping shear planes is abandoned whereas the other continues to accumulate
deformation (panel d). In this case the material deficit, due to the subduction slab-
retreat and responsible of the basin formation has caused the collapse of the orogen
and areas, that previously experienced uplift and erosion become, themselves
subject to accumulation.
      Looking at the curve d in Fig. 8, it is also interesting to note that the large
increase of the temperature on the rear of the retreating and rolling back lithosphere,
generates an increase of the model morphology on the retro-side. This increase is
not observed in the previous simulation where only low temperatures are reached.


      FINAL REMARKS

      The performed numerical simulations have focalized how the presence of a
crustal rheological stratification and different thermal conditions during the orogeny
influence the final deformation. At the same time, the numerical simulations have
provided some insights into fundamental processes that control the evolution of the
LFB and NA, in what follows we present the main conclusions.
GNGTS – Atti del 19° Convegno Nazionale / 14. 05




                                                                50 km




Fig. 7 - Results of the APP2 experiment: contour plots of the invariant of the deviatoric part of the
strain rate tensor.

       The results from LFB1 model, show that the well-documented eastward
        propagation of the deformation front, that led to the formation of a set of west-
        dipping imbricate thrusts, is consistent with the mechanical behavior of a thick
        brittle layer decoupled from the underlying oceanic crust by a weak
        decollement. The mechanical model also supports the idea that shortening of
        the turbidite sequence was first accommodated by large scale folding and,
        later in the orogenic cycle, by upward propagation of shearing from the basal
        decollement. Movement along these late crustal-scale faults led to the
        exhumation of a strong, sub-horizontal mica fabric that may have developed
        within the decollement (Foster et al., 1999). Micas and whole rock 40 Ar- 39 Ar
        dating also suggest a progression in sediment source within the western LFB
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


          (Foster et al., 1999). Ages from detrital micas found in the western part of the
          orogen indicate that they were last deformed during the late Proterozoic-
          Cambrian Delamerian-Ross Orogeny. Ages from detrital micas in the eastern
          part of the orogen are characterized by a deformation age (Silurian and
          Devonian), that post-date the initiation of fault movement in the western
          region. This suggests a progressive reworking of sediments from west to east
          which is clearly supported by the results of the numerical model.




Fig. 8 - Model morphology for the different steps of the APP2 experiment.

           The presence of a weak layer in the LFB1 model, has also shown that with
            the ongoing of the propagation of strain rate on the pro-side, a movement on
            retro-side is also present. The presence of this retro-shear could be related
            to the only one mayor fault dipping toward east: the Woorndoo-Moyston
            fault. The performed simulation emphasizes that the fault represents a back-
            thrusting born at an early stage and reactivated during the eastern migration
            of the thrust. It is possible to postulate, however, that the timing of the well -
            documented late movement (420 Ma) on the westernmost fault of the
            orogen (the Woorndoo-Moyston Fault) may correspond to the time at which
            the eastern margin of the orogen was advected over the subduction zone.
           The APP1 model underlines that the presence of roll-back, that determines
            the formation of a volume deficit above the subduction window, induces the
            formation of a basin. The material deficit increases as the model evolves,
            causing the basin to grow in size. Extension is observed in the pro-side, as
            the model responds to the material deficit created by the subduction zone
            retreat, and at the same time, towards the retro -side, a collapse of the
            previously formed structures is determined.
           The influences of the probable alternances of relative cold and hot events
            during the geodynamical evolution of the NA (Doglioni et al., 1998), has
            been tested in the APP2 model. The obtained results have shown that the
            material deficit due to the subduction slab-retreat observed in APP1 model,
            determines in this simulation the collapse of a previously formed orogen. It
            has been noted that the accretionary prism of the Apennine was formed in
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


          sequence at the front of the pre-existing back-thrust belt of the Alpine belt
          (Doglioni et. al., 1998). In this framework the Alpine belt was progressively
          deformed by the back-arc extension of the Apennine subduction and the
          final deformation in the Apennine shows trace of the interaction of the two
          belts. The our models underline that the mechanism found in APP2 model
          could be responsible for the collapse of the previously formed Alpine
          orogen.
          In addition, an increase in model morphology, connected to the temperature
          increase, is observed on the retro -side of the model. This last result is in
          agreement with the regional uplift that has interested Tuscany (Bonadonna
          et al., 1975) and that seems to be connected with the presence of a hot
          body beneath the western sector of the NA.


                                            REFERENCES

Amato A., Alessandrini B., Cimini G., Frepoli A. and Selvaggi G. (1993) - Active and remnant
       subducted slabs beneath Italy: evidence from seismic tomography and seismicity, Annali di
       geofisica, 36, 201-214.
Amato A. and R. Azzara and C. Chiarabba and G.B. Cimini and M. Cocco and M. Di Bona and L.
       Margheriti and S. Mazza and F. Mele and G. S elvaggi and A. B asili and E. Boschi and F.
       Courboulex and A. Deschamps and S. Gaffet a nd G. Bittarelli and L. Chiaraluce and G. Piccinini
       and M. Ripepe (1998) - The 1997 Umbria-Marche, Italy earthquake sequence: a first look at the
       main shocks and afters hocks, Geophys. Res. Lett., 25, 2861-2864.
Atlas of Geot hermal Resources in Europe (1994) - Surfac e heat-flow density map (Italy), Commission
       of the European Communities. Directorate General XII, Science, Research and Development.
Batt G.E. and B raun J. (1999) - The tectonic evolution of the Southern Alps, New Zealand: insights
       from fully thermally coupled dy namical modeling, Geophys. J. Int., 136, 403 -420.
Boccaletti M., Coli M., Dec andia F.A., Giannini E. and Lazzarotto A. (1980) - E voluzione
       dell'A ppennino Settentrionale secondo un nuovo modello strutturale, Mem. Soc. Geol. It., 21,
       359-373.
Bonadonna F.P., Squarci P. and Taffi L. (1975) - Sollevamento dei sedimenti neogenici (Toscana e
       Lazio), Rapport o interno CNR-IIRG, Pisa.
Braun J. and M. Sambridge (1994) - Dynamical Lagrangian Remeshing (DLR): a new algorithm for
       solving large strain deformation problems and its application to fault-propagation folding, Earth
       Planet. Sci. Lett., 124, 211-220.
Chopra P.N. and Pat erson M.S. (1981) - The experimental deformation of dunite, Tectonophysics, 78,
       453-473.
Della Vedova B., Marson I., Panza G.F. & Suhadolc P. (1991) - Upper mantle properties of the
       Tuscan-Tyrrhenian area: a framework for its recent evolution, Tectonophysics, 195, 311-318.
D'Offizi S., Minelli G. and G. Pialli (1994) - Foredeeps and thrust system in the Northern Apennines,
       Boll. Geof. Teor. Appl., 36, 141-144.
Doglioni C. (1991) - A proposal of kinematic modelling for W -dipping s ubductions-Possible
       applications to the Tyrrhenian-A pennines system, Terranova, 3, 423 -434.
Doglioni C., Mongelli F. and Pialli G. (1998) - B oudinage of the Alpine belt in the Apenninic back-arc.
       Mem.Soc.Geol., 52, 457-468.
Foster D.A., Gray D. R. and B ucher M. (1999) Chronology of deformation within the turbidite -
       dominated, Lachlan orogen: implication for the tectonic evolution of eastern Australia and
       Gondwana, Tectonics, 18 (3), 452-485.
Glen R.A. (1992) - Thrust, extensional and strike-slip tectonics in an evolving Palaeozoic orogen of
       southeastern Australia, Tectonophysics, 214, 341 -380.
Gray D.R. and D.A. Foster (1997) - Orogenic concepts application and definition: Lachlan Fold belt,
       Eastern Australia, American Journal of Science, 297, 859 -891.
Gray D.R. and D.A. Foster (1998) - Character and Kinematics of faults within the turbidite-dominated
       Lachlan Orogen: implications for t ectonic evolution of eastern Australia, J. Struct. Geol., 20,
       1691-1720.
Gray D.R., Foster D.A. and Bucher B. (1997) - Recognition and definition of orogenic events in the
       Lachlan Fold belt, Australian Journal of Earth Science, 44, 489 -501.
GNGTS – Atti del 19° Convegno Nazionale / 14. 05


Hassani R., Jongsman D. and Chery J. (199 7) - Study of plat e deformation and stress in subduction
       processes using two-dimensional numerical methods. J.Geophys.Res., 102, 17951 -17965.
Mongelli F., Zito G., Ciaranfi N. & Pieri P. (1989) - Interpretation of heat flow density on the Apennine
       Chain, Italy, Tectonophysics, 164, 267-280.
Pauselli C. and Braun J.(2001): Tectonic evolution of the Lachlan fold belt: constraints from numeric al
       model. In prep.
Pauselli C. and Federico C.(2001) - The brittle/dusctile transition along the Crop03 seismic profile:
       relationship with the geological feat ures. Atti del Convegno "evoluzione geologica e
       Geodinamica dell'Appenino" - Boll.Soc.Geol. It., Vol.Spec.- in press.
Offler R., Miller J.Mc., Gray D. R., Foster D.A. and Bale R. (1998) - Crystallinity and b0 spacing K-white
       micas in a Palaeozoic accretionary complex, eastern A ustralia: metamorphism, paleo -
       geotherms and structural style of an underplated sequence, Journal of Geology, 106 495-510.
Paterson M.S. and Luan F.C. (1990) - Quartzite rheology under geologic al conditions, in Deformation
       mechanism, rheology and tectonics, Edited by Knipe R. J. and Rutter E.H. Geol. Soc. Spec.
       Pub. London, 42, 299-307.
Piromallo C. and Morelli A. (1997) - Imaging the Mediterranean upper mantle by P-wave travel time
       tomography, Ann. Geofis., 40, 963-979.
Pysklywec R., Beaumont C. and Fullsack P. (2000) - Modeling t he behavior of the continental mantle
       lithosphere during plate convergence. Geology, 28, 655-658.
Selvaggi G. and Chiarabba C. (1995) - Seismicity and P-wave velocity image of the southern
       Tyrrhenian subduction zone, Geophys. J. Int., 121, 818-826.
Spakman W., van der Lee S. and R. van der Hilst (1993) - Travel-time tomography of the European-
       Mediterranean mantle down to 1400 km, Phys.Earth Planet.Inter., 79, 3-74.
Waschbusch P. and C. Beaumont (1996) - Effect of a ret reacting subduction zone on deformation in
       simple regions of plate convergenc e, J. Geophys. Res., 101(B 12), 28133 -28148.
Whittaker A., Bott M.H.P. and Waghorn G.D. (1992) - Stresses and Plate boundary forces associated
       with subduction plat e margins. J.Geophys.Res., 97, 11933-11944.
Willett S.D. (1999) - Rheological dependence of extension in wedge models of c onvergent orogens.
       Tectonophysics, 305, 419-435.
Willett S.D., Beaumont C. and Fullsack P. (1993) - Mechanical model for the tectonics of doubly
       vergent compressional orogens, Geology, 21, 371 -374.
Wilks K. and N.L Carter (1990) - Rheology of some continental lower crustal rocks, Tectonophysics,
       182, 57-77.