TSK 11 Göttingen 2006 Schwarz
 Evolution and structure of
 the Upper Rhine Graben —
 quantitative insights from nu-
 merical modelling approaches
 Vortrag
 Michael Schwarz1
 Introduction
 The Upper Rhine Graben forms the ma-
 jor segment of the Cenozoic Rift sys-
 tem of Western Europe. Although the
 rift was the target of many seismic
 and geological investigations, the style
 of lithospheric extension below the in-
 ferred faults, the depth to detachment,
 and the amounts of horizontal extension
 and lateral translation are still being de-
 bated. In this study, the date base to
 the Upper Rhine Graben was subjected
 to a finite element approach in order to
 include thermomechanical processes of
 the lithosphere as well as erosion and
 sedimentation. The study concentrated
 on the consequences of extension and
 lateral translational events on the struc-
 ture and evolution in terms of basin
 geometry, sediment layer thicknesses,
 Moho elevation, and shoulder uplift on
 a lithospheric scale. The numerical ap-
 proach was three dimensional in order
 to incorporate the lateral crustal hetero-
 genities in the Upper Rhine area and the
 varying ambient stress field.
 The thermomechanical simulation of a
 real rift requires the knowledge of the
 parameters controlling its structure and
 evolution. Furthermore, field data are
 needed for assessing the modelling re-
 sults. Both preconditions could be met
 by the production of comparative data
 sets as well as by a parameter study
 1 Geologisches Institut, Universität Freiburg,
 Albertstr. 23B, 79104 Freiburg i. Breisgau
 before the modelling of the rift evolu-
 tion. The critical validation of the re-
 search level allowed the extraction of
 the parameters to be determined. The
 results of the parameter study already
 gave some cues on the points of contro-
 versy mentioned above.
 In anticipation of the parameter study,
 a hypothesis on continental rifting pro-
 cesses was formulated. It describes the
 consequences of the potential factors
 controlling the vertical displacements of
 the graben, shoulder, and Moho under
 simplified conditions. Opposite to other
 concepts, the hypothesis also takes into
 account the mechanical behaviour of
 faults as a primary factor. The numeri-
 cal results of the parameter study were
 compared with the forecasts of the hy-
 pothesis in order to identify additional
 processes acting in the more complex
 setting of the Upper Rhine Graben area.
 Moreover, the comparison allowed dis-
 closing functional relationships between
 the vertical displacements and the con-
 trolling parameters. Apart from these
 insights in the continental rifting pro-
 cess the parameter variations rendered
 some important results specific to the
 Upper Rhine rift system. They are in
 detail:
 1. The vertical displacements in the
 rift system are controlled primar-
 ily by the friction, depth, and ge-
 ometry of the border faults in the
 brittle domain. The consequences
 of the temperature and rheology in
 the creep regime are of minor im-
 portance. The same holds for the
 effects of the erosion and sedimen-
 tation.
 2. The boundary faults are sub-listric
 down to maximum depth of 15 to
 16km. The geometry of the faults
 1
Schwarz TSK 11 Göttingen 2006
 remains the same during the rift-
 ing. In the pre-rift setting, they
 flatten from a dip angle of around
 65° to some 40° at greater depth.
 Beneath, the deformation is accom-
 modated by ductile creep without a
 need for discrete shear zones in the
 lower crust and upper mantle.
 3. The apparent frictional coefficients
 mostly lie around 0.3 bis 0.4, but
 at every point on the fault surfaces
 lower than 0.5.
 4. There are no crustal horizons where
 a considerable restoration of the
 isostatic equilibrium takes place.
 5. High viscosities can be excluded at
 any depth in the lower crust. The
 variability of the lower crustal com-
 positions is of no consequence for
 the rift evolution.
 6. The upper crustal creep behaviour
 can be simulated only with much
 higher viscosities as it is predicted
 by the quarzite rheology. The re-
 quirement can be followed by us-
 ing the creep parameters of a felsic
 granulite.
 These results were put into the actual
 modelling of the Rhine Graben evolu-
 tion. Therein, the implementation of
 the thermomechanical processes and the
 balancing on a lithospheric scale allowed
 reconstructing the vertical displaments
 of the graben, Moho, and shoulder over
 time. The comparative data sets are
 matched with a rift evolution in two
 phases. An extension being approxi-
 mately orthogonal to the Rhine Graben
 is replaced by lateral translation which
 leads to a reactivation of the rift in a
 sinistral sense. The horizontal exten-
 sion of 7.5 to 8.5 km and a sinistral dis-
 placement across the entire graben of
 4.5 km at most are necessary in order
 to accommodate the sedimentary thick-
 nesses. Dextral and sinistral displace-
 ments along the border faults take place
 during the period of orthogonal exten-
 sion as well. These displacements are
 located at the fault segments where the
 friction coefficients change laterally due
 to a switch of the rift polarity. There,
 the graben block is extended parallel to
 the graben accompanied by a reduction
 of the principal stress in the same direc-
 tion which causes the lateral displace-
 ments.
 The conduction keeps pace with the
 advection of heat at any time during
 the rift evolution. There is no need
 for a thermal input of subcrustal ori-
 gin for initiation of the rift. No thermal
 anomaly is created by the rifting in the
 Upper Rhine Graben area. The mod-
 elling results confirmed the ideas of the
 graben as an example of a passive rift.
 The numerical outcomes can serve as
 decision guidance for solving conflic-
 tive positions about the geodynamics of
 the graben system. The fit between
 model and reality gives preferences for
 an evolution of the graben in two pe-
 riods with different kinematics. The
 lateral displacements calculated in the
 study lie at the upper end of values
 which are inferred from structural and
 geophysical observations in the Rhine
 Graben region. Other stretching di-
 rections than orthogonal would result
 to additional strike-slip displacements
 above this threshold and, therefore,
 have to be declined. In the period of lat-
 eral translation during the Neogene, the
 regional strike slip setting disintegrates
 into different tectonic regimes along its
 strike. The configuration resembles the
 recent kinematics in the Rhine Graben
 region. Thus, the study approves the
 2
TSK 11 Göttingen 2006 Schwarz
 assumption of regional stress field being
 nearly constant since the beginning of
 the Miocene.
 The modelling results refer only to pe-
 riods of sedimentary record. They
 contain no information about the
 tectonosedimentary evolution in times
 with hiatuses. The hiatuses are located
 predominantly in the period of rift-
 parallel translation. This is at conflict
 with the calculated displacements which
 are already at the top of the values in-
 ferred for the Upper Rhine Graben from
 other observations. However, the results
 show that the graben subsidence, espe-
 cially in a strike slip regime, is highly
 sensitive to the mechanical properties of
 the border faults and their orientation
 to the local stress field. Slight modifi-
 cations of these factors can result in a
 sticking behaviour of these faults over
 large distances which may account for
 the hiatuses.
 Apart from continental rifting, south-
 western Germany was affected by the
 migration of the Alpine peripheral fore-
 bulge into the Rhine Graben region.
 The modelling outcome permits sepa-
 rating the vertical displacements due to
 rifting from those related to the bulging
 processes. Future thermomechanical
 modelling studies require an implemen-
 tation of these processes in order to
 achieve a holistic reconstruction of all
 geodynamic processes which were active
 in the Upper Rhine Graben area during
 the Cenozoic. Therefore, this study is
 regarded only as a first step to under-
 stand the interaction between extension,
 and lateral translation in a spatially and
 temporally varying stress field with a
 pre-existing structural inventory.
 3