 |
| Titel |
Slab geometry depending on absolute velocities and possibilities of cyclicity, Insights from numerical modelling. |
| VerfasserIn |
Gibert Gaelle, Hassani Riad, Emmanuel TRIC, Monfret Tony |
| Konferenz |
EGU General Assembly 2010
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 12 (2010) |
| Datensatznummer |
250043480
|
|
|
|
|
|
| Zusammenfassung |
| Recent analogical models of Heuret et al (2007) and Guillaume et al (2009) showed that the
kinematical and mechanical role of the overriding plate is crucial for the slab geometry. We
used Hassani et al (1997) 2D finite element numerical code to model subduction
with parameters close to analogue models conditions. Our results are coherent with
considered previous analogue models. As shown by Heuret et al (2007), after the
subducting plate reaches the 660 km discontinuity modeled by a rigid foundation, we
obtained two different styles of subduction depending on the overriding plate velocity
vop:
if vop >0, the slab lays forward on the 660km discontinuity (style 1)
if vop -¤0, the slab lays backward on the discontinuity (style 2).
We also light up two different processes: the subduction evolves in a steady-state regime when
vsp -¤0 or 2vop + vop -¤0 (where vsp stands for the subducting plate velocity) and in non
steady-state regime when one of these two conditions is not fulfilled. This result is coherent
with Guillaume et al (2009) non steady-state models at vop=0. We checked these conditions
by conducting several simulations varying subduction velocities (relative velocity), absolute
plates velocities and plate viscosities. When the process is not steady-state, the slab is
periodically folding on the 660 km discontinuity leading to episodes of slab flattening
(minimal dip of 13Ë ) followed by slab steepening phases (maximal value of 65Ë
). The folding period is essentially controlled by slab viscosity and subduction
velocity.
We then compared our models with the natural case of the Chilean flat slab. Fukao et al
(2001) show that the Andean slab does not penetrate the 660 km discontinuity which make
possible comparison between our models and nature. For the Andean case, actual
velocities are vop- 4.3 cm.an-1 and vsp- 2.9 cm.an-1 for a relative velocity close to
vs = vop + vsp- 7.2 cm.an-1, which places us in the case of style 1 with cycles. If we model
those velocities with a viscosity of η=2Ã1024 Pa.s, we obtain cycle duration of ca.
26 Ma with a period of ca. 3.6 Ma of low dip, with a minimal value of ca. 15Ë .
Haschke et al (2002) work on Andean flat slab region magmatic rocks suggest that
evolutionary cycles occurred during the Andean orogeny. Episodes of slab steepening and
shallowing with a duration of 30-40 Ma might have occurred, explaining magmatic
gaps of ca 5-10. Simplifications imposed by numerical modeling could explain the
difference between our values and Andean values. Still, our values are in the range
of Haschke et al (2002) estimation of the Andean case. Our models are coherent
with Kay et al (2002) results on South American overriding plate: when the Nazca
plate slab was steep, the South American margin underwent extension, and that
when shallowing occurred, compression occurred in the margin. Therefore, style 1
with cycle may be considered as a possible answer for this natural speacial case. |
|
|
|
|
|
|