the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Strike-slip faulting in extending upper plates: insight from the Aegean
Abstract. During gravitational collapse of orogenic systems or in hot extending back-arc systems, normal faulting is often associated with strike slip faulting whose origin remains enigmatic. The formation of major strike slip fault zones during subduction upper plate extension driven by slab-roll back can be related to slab tearing at depth. In the Aegean, where back-arc extension driven by southwest-ward migration of the Hellenic trench (slab rollback) has occurred since at least 30 Ma, the co-existence of normal faulting and a multiple strike-slip fault zones is observed since the onset of the westward extrusion of Anatolia, but before the onset of slab tearing that occurs in the Pliocene. Here we show how strike slip faults and normal faults can coexist in a hot deforming continental lithosphere. Our 3D numerical models with two deformation stages (initial pure extension followed by combined shortening and extension) can explain the Aegean tectonics. Several rifts form during the purely extensional stage that, during the second deformation stage, are either fully reactivated as strike-slip faults, or remain active but rimmed by dextral and sinistral strike-slip faults. This suggests that the extension driven by slab rollback and shortening driven by westward extrusion of Anatolia interact in space and time in the Aegean domain to create a complex tectonic pattern with coeval active normal faulting (e.g. Corinth and Evvia rifts) and dextral strike-slip faulting (e.g. the North Anatolian and Myrthes-Ikaria faults). These results show that strike slip faults in extending domain can be a sign of shortening at high angle to the extension direction.
- Preprint
(4866 KB) - Metadata XML
-
Supplement
(92 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-569', Haralambos Kranis, 29 Mar 2024
General comments
The paper by Faucher et al. is an interesting approach to the complexly deforming Aegean region. It is scientifically significant and falls well within the scope of SE. The model setup and boundary conditions are valid and the model resolution is sufficient enough to display the crustal structural features. Nonetheless, the intrinsic limitations of the method lead to certain inconsistencies between the model outcome and the actual structural configuration, as described below:
The outcome of the final model (Fig 3D) correctly predicts the formation of E-W striking normal faults (roughly corresponding to the active rifts of Corinth and Evvia), alongside with ENE-WSW to NE-SW dextral strike-slip faults, (which is the case in the central and north Aegean, i.e. splays/branches of the North Anatolian Fault), conjugate to WNW-ESE to NW-SW sinistral s.s. faults. A fault zone of such strike and kinematics does occur in reality, the Katouna Fault System, NW of the gulf of Corinth (Perouse et al., 2017) (note that it is incorrectly shown in Fig 1C as a normal fault).
The conjugate arrangement of NE-SW dextral NW-SE sinistral s.s. faults is not (exactly) the case in the broader Aegean region, however. As shown in the paper by Sakellariou and Tsambouraki-Kraounaki (2018) (which is properly acknowledged in the m/s and used as a basis for Fig 1C), the upper-plate deformation in the South Aegean is characterized by NW-SE sinistral s.s. faults.
Overall, the paper by Faucher et al is a step forward, towards the understanding the complex deformation of the broader Aegean region, albeit with some limitations, posed by the model setup and boundary conditions. The symmetry displayed in the outcome of the model possibly arises from the equality between the amounts of compression and extension applied to it and the strict orthogonality between the compression and extension axes, as trench retreat is faster than the Anatolian extrusion and these two end-member vectors are not strictly perpendicular to each other. Moreover, the effect of the Anatolia extrusion in the northern part (i.e. north of the Maliakos-Amvrakikos latitude) is much less pronounced; probably the Vc amount imposed along the x-direction could follow a gradient, increasing towards the “south” edge of the model.
Specific comments
See attached pdf with notes and comments on Fig 1.
The “Pelagonian” fault is currently active, as evidenced by recent (2022-today) seismic activity; based on focal mechanism solutions, it is indeed a left-lateral fault boundary.
Technical corrections - typing errors
l.41: “the Oligocene”, “the Miocene”
l.150 “NW-striking strike-slip faults”
References
Pérouse, E., Sébrier, M., Braucher, R. et al. Transition from collision to subduction in Western Greece: the Katouna–Stamna active fault system and regional kinematics. Int J Earth Sci (Geol Rundsch) 106, 967–989 (2017). https://doi.org/10.1007/s00531-016-1345-9.
Sakellariou, D. and Tsampouraki-Kraounaki, K.: Plio-Quaternary extension and strike-slip tectonics in the Aegean, in: Transform Plate Boundaries and Fracture Zones, edited by: Duarte, J., Elsevier, 2018
- RC2: 'Comment on egusphere-2024-569', John Naliboff, 10 Apr 2024
-
CC1: 'Comment on egusphere-2024-569', Frank Zwaan, 30 Apr 2024
Interesting models!
The conjugate strike-slip structures remind me of those we found in our analogue models (Zwaan et al. 2019, and references therein), and those in the recent analogue modelling Liu et al (2024) study. It could be interesting to compare these works with these new numerical results. In the analogue models, deformation is driven by a basal boundary condition (hence a decoupling layer can make the system switch from conjugate strike-slip faults to normal faults), whereas here it seems to be "side wall" boundary conditions in these numerical models, which could prevent such decoupling effects?
NB: it may also be interesting to check the Guillaume et al. (2022) analogue modelling paper, which makes things even more interesting by adding structural inheritance to the mix.
- Zwaan et al. (2019): https://doi.org/10.5194/se-10-1063-2019
- Guillaume et al (2022): https://doi.org/10.5194/se-13-1393-2022
- Liu et al. (2024): https://doi.org/10.1029/2023TC008127
Citation: https://doi.org/10.5194/egusphere-2024-569-CC1 -
CC2: 'Comment on egusphere-2024-569', jun liu, 30 Apr 2024
Nice models!
Followed Franks' comments, in our analogue experiments basal boundary condition (Liu et al., 2024), we observe systematic change in fault pattern from compartments of conjugate sets of oblique-slip normal fault to one-sided sets of oblique normal faults to horst-and-graben structures with decreasing strain rates (~extension velocity). As we all knnow, the strength of viscous layer is a function of strain rate έ and viscosity η (σ1-σ3 = ηέ ), so in our models, fast models are recogonised as with strong ductile layer and low models are recogonised as with weak ductile layer. We attribute strain rate dependent of faulting structures to the increasing vertical coupling between the basal rubber sheet and the sand layer, controlled by the rate-dependent strength of the viscous layer.
you can also check similar numerical modelling paper by Pang et al., (2018) (http://doi.org/10.1002/2017jb014011). they also have "side wall" boundary conditions with shortening and extensnion. their models also show from conjugate strike-slip faults to horst-and-graben structures by adding weak lower crust.
It would be great to visualize these models in a more quantitative way to link data in the field examples, shch as slip partitioning.
Citation: https://doi.org/10.5194/egusphere-2024-569-CC2 -
CC3: 'Comment on egusphere-2024-569', Matthias Rosenau, 13 May 2024
Dear authors, all,
Congratulations! I came across this preprint on my search for 3D numerical models allowing us to compare with our analogue models with similar kinematic boundary conditions, i.e. horizontal lateral shortening and longitudinal extension. I see a nice discussion going on and just want to add my perspective.
During my literature search, I found few models with a comparable specific boundary condition although, as stated above, there have been multiple numerical rifting papers now in 3D and at crustal scale. However, I see those looking at different aspects of the general rifting problem and not so much on multiphase rifting. We have also run some numerical models with the Aspect code with similar kinematic boundary as presented here and our results look surprisingly similar (at least qualitatively) given that we had very likely very different parameters (e.g. mesh resolution, domain size, rheologies, etc.) as we aimed at simulating our analogue experiments. In conclude for myself that the general patterns seem to be stable first-order features although details matter definitely at second order. One may want to see a larger number of models for a more systematic parameter study but this seems beyond the aim of the current paper the models are used to test a specific regional tectonic hypothesis.
I second the reviewer #2's point that the numerical methods and assumptions should be explained in more detail especially because ASPECT has become a tool widely used in the community and we should share as much information as possible. I am not familiar with the sharing of data in the numerical community but I think sharing the input files and/or some sort of set of files and codes allowing to reproduce the results should be provided. From the model numbering in figure 2 it seems a couple of additional models (2, 3, 5, 6) have been run but not presented here. They may help however in discussing the sensitivity of the models to specific parameter choices and may therefore serve as benchmarks. If this is the case I would suggest adding them in an appendix.
Overall, I think the paper tackles an important issue and can be, after some mainly technical modifications, an important contribution to triggering the study of 3D aspects of tectonic deformation and complementing existing numerical and analogue models.
Citation: https://doi.org/10.5194/egusphere-2024-569-CC3 -
RC3: 'Comment on egusphere-2024-569', Guillaume Duclaux, 18 May 2024
Review of "Strike-slip faulting in extending upper plates: insight from the Aegean", by Agathe Faucher, Frédéric Gueydan and Jeroen Van Hunen.
This short communication investigates the development of strike-slip structures at the surface associated with coeval extension and shortening using 3D thermomechanical models in the context of the Aegean. 3D modelling work like this one are important to better understand strain partitioning in orogens and active tectonic systems like back-arcs. The manuscript is illustrated with 3 figures and fairly written. Some references about 3D models with comparable BCs are missing (e.g. Le Pourhiet, L., et al. (2018). Continental break-up of the South China Sea stalled by far-field compression. Nature Geoscience, 11(8), 605-609.), and recent analogue papers have also been suggested in the open comments.
The paper very briefly introduces the tectonic context of the Aegean system since 30 Ma, then focuses on a series of 3D numerical models’ results presentation and discussion. Four numerical models, simple in design and relatively convincing in their output are presented. The model domain is relatively limited spatially with a model box of 100x100km horizontally and 50 km vertically. The first three models are generic and explore strain partitioning in different settings, and the fourth one is hybrid and supposedly developed in the context of the Aegean extension with Anatolia extrusion.
I am no expert in the Aegean system evolution, so I can't really comment on section 1.
I present below some key points for which I have concerns followed with a list of minor comments and detailed critics about the figures. I sincerely hope these will help the authors with improving their contribution.
I have quite a few critics about the modelling work presented here. Numerical models visualisation and analysis could very much be improved. I get back onto that later when discussing the figures individually. For example, colour scales are changing from figures to figures which solely display the strain rate second invariant (I assume this is the second invariant and not the magnitude of the strain rate tensor - not sure if this is correct?), the line drawing interpretations sometimes seem arbitrary. It would be neat to show other scalars such as cumulative strain, some isotherms. As they are designed now it feels like we miss a lot of information from these simulations.
A table summarizing all models setup is necessary. In Figure 2 the three presented models are named Model 1, Model 4, and Model 7. In the text (line 95-96) they are referred to as model M_e, M_c, and M_ec respectively, and the author indicate those are the "end-member solutions". I suppose other velocity settings have been tested. Is that correct and could you briefly comment on the other models’ behavior? This table should at least be presented in supplementary material. For consistency I recommend you use only M_c, M_e, and M_ec in the text and figures.
Other critical information regarding the models is missing. For example: What is the top surface boundary condition? What are the top and bottom temperatures (or top temperature and basal heat flow) of the model? A graph showing at least the initial viscosity profile and the geotherm would be much appreciated. The initial 920˚C temperature at 30 km depth in the model seems rather high and can't be buffered through latent heat during partial melting as it is not considered in the model.
According to Fig 1, in the context of the Aegean system the Anatolia extrusion velocity is ~40% smaller than the trench retreat. That is very different than in Model M_ec_t (presented in Fig 3 and the last pages of the manuscript) or with the text in lines 127-128. Have you explored similar V_e/V_c ratios? How would changing the relative velocities along the orthogonal walls impact the model? You suggest (line 115-116) that "shortening and extension should be of roughly the same magnitude to trigger strike-slip", so does the Aegean qualifies? Or is the model applicable for this particular test case?
For all this reasons I would recommend the authors profoundly rework the figures and the manuscript as in its current form this communication is not worth publication in Solid Earth.
Minor comments:
+ line 26: STEP faulting: not sure why STEP is spelled in uppercase.
+ line 30: Amorgos: not shown in fig 1C. Please add the fault system name to the map.
+ line 36: Please ensure you use the same name for this domain (i.e. the Aegean microplate) throughout the manuscript. It is defined as Aegean/Anatolian microplate line 52. Also, the spelling of microplate changes from micro-plate to microplate.
+ line 41: I would suggest a small edit here: "metamorphic core complexes (e.g. Rhodopes in the Oligocene and Cyclades in the Miocene)".
+ line 46: Here you mention plutons. Plutons of what rock type? Gabbro? Salt? Granite?
+ line 62: Would inherited structures (and which ones) play a major role in the evolution of the Aegean system? This would be an interesting discussion point.
+ line 68: At what depth is the flat Moho in the Aegean today? This would be an interesting feature to add to the map in Fig 1C (e.g. showing crust isopaches).
+ line 76: please define the top BC here. Moreover, what BC is applied to the walls parallel to the inflow of outflow for models M_e and M_c?
+ line 80: You should refer to the concentration of radiogenic elements or the radiogenic heat production and not the "rate of radiogenic elements".
+ line 82: "(" missing before "see Kronbichler et al., 2012[...]".
+ line 82-83: Although ASPECT is more and more used in the tectonic community it would be nice to write down which equations are solved somewhere... at least in supplementary. I assume the system of equations are for highly viscous incompressible fluid motion. Is that so? Please provide a little more details about the physics used in the code.
+ line 83: I would recommend adding yield or frictional in "[...] with Drucker-Prager yield plasticity." I can't see either from the text or Table S1 whether there is any weakening imposed for the frictional rheology. Is there any weakening function applied to the cohesion and/or friction angle?
+ line 100-101: There seems to be some strain rate concentration in the mantle for the pure extension model at least. Could you comment on the surface spacing of the resulting grabens? And possibly the resulting topography as I assume this the feature used to define the fault trace in Fig 2 and 3.
+ line 104: How high is the resulting topography for model Mc? Why is the fault spacing so different than for Ve model? You mention no tectonic inheritance in the model’s setup, so I'm curious about what is controlling strain localization and shears/faults spacing. It's a shame we can't visualize the relative thickening of the highly radiogenic upper crust and the depleted lower crust layers. I might have made a mistake in my back of the envelope calculation but in the absence of surface processes I am surprised the Moho depth is only 50km at 4 Ma for a full-rate inflow velocity of 2cm/yr... please see comment about figure 2 below.
+ line 109-111: this should be in the discussion section rather than the results.
+ line 112-113: I'm afraid you do not clearly show any proper fault rotation... Because solely the strain rate is used to visualize the deformation it is not clear whether structures rotate, or new structures replace them at a different angle. Maybe an animation would help visualizing the progressive evolution of faults orientation through time. Alternatively, a quantitative analysis of the fine strain through time would be more convincing.
+ line 113: I suppose you mean thickening rather than "shortening".
+ line 134: Why is the time period 15-5 Ma? I don't believe the extrusion of Anatolia is over yet and the Hellenic trench retreat is still taking place too.
+ line 137: "looks more homogeneous and with ongoing activity of the three former rift systems" what do you mean by that? The former rifts activity is not visible in Fig 3B.
+ line 141-142: from the figure I really can't see the side rifts evolving into anything.
+ line 147-148: If the model really applies to the Evvia and Corinth system I suggest you have a look at the work of Pechlivanidou et al. paper (Pechlivanidou, S., et al. (2022). Contrasting geomorphic and stratigraphic responses to normal fault development during single and multi-phase rifting. Frontiers in Earth Science 9: 748276). But AFAICT this region is dominated by multiple generations of normal faults rather than normal faults and strike-slip faults.
Figure 1:
+ You mention in the text (lines 60-61) that according to Brun et al., 2016 extension rates have changed through time. Adding estimates of those rates in Fig 1A, 1B (red arrows) and in the text would be very good. This is also key to validate the models where V_c and V_e chosen for the modelling approach.
+ Please add a bounding box on Fig 1A and 1B inset maps to highlight the region corresponding to the close-up views.
+ the [C] is missing in Figure 1C. Could you please also mark distinctively the trace of the Hellenic trench in the map? This is one of the main features of this figure, yet it doesn't clearly appear. As noted earlier, the Amorgos fault system should also be added to the map.
Figure 2:
+ Arrows for Ve and Vc should be drawn on both x-normal and y-normal faces respectively. As it is displayed here it looks like Ve and/or Vc are solely applied on one face. That questions whether the announced velocities are full-rate or half rate. I understand from the text (lines 73-74) 1 cm/yr is half rate. Is that correct? This is very important for estimating the finite strain and moho depth, and it must be clarified in the figure.
+ The strain rate and the marked fault traces are not superimposed, which is totally understandable. This is because the faults traced represent early structures (I believe the fault traces are based on topographic changes delimiting grabens for Model 1), but it is nevertheless intriguing, and you should explain on which premise these black lines are drawn. For the Extension model it is rather surprising to see active shears crosscutting pre-existing structures.
+ The strain rate color scale's range is different for every snapshot (same comment applies for figure 3). I believe this makes the comparison between setting a little less obvious.
+ For Model 7 (M_c): I assume the x-normal walls have free-slip BCs, so it behaves like a 2D model. If you apply a constant full rate inflow velocity of 2 cm/yr (1 cm/yr on each y-normal face) during 4 Ma, the Moho depth should reach 150 km... here it is marked at 50km depth. Either the full rate inflow velocity is 1cm/yr or the model duration is 2 Ma (or I did something wrong with my calculation).
Figure 3:
+ I understand these are snapshot of a single model with evolving velocity BCs through time. The model runs with pure extension in the x-direction (and free-slip BC along the y-normal walls?) for 7.5 Ma, then orthogonal extension in the x-direction and compression in the y-direction for an extra 7.5 Ma. This is clear from caption, but not so clear from the figure itself. Indeed, for 3B I get when reading that "8Ma, Extension=Compression" that the model has the same amount of extension that compression for 8Ma... Maybe something like 8Ma (7.5 Ma extension + 0.5 extension&compression)? Or this is just me... sorry for being a bit pedantic here.
+ The faults traces in black at the surface are just obscure. I assume that in 3B the faults traces should be rather close to 3A, at least for the central rift. This is definitely not the case.... Why? This really questions the visualization and interpretation of the model results. Same goes for the white shear zones after.
+ Is there any solid block rotation at the surface of the model in C or D? Imaging rotated faults drawn in A away for the center of the model would be very interesting. This type of analysis is really missing here and could bring valuable insights into the Aegean system itself. What happened to the normal faults’ terminations close to the y-orthogonal walls when the conjugate strike-slip faults evolve?
+ As for figure 2 the color scale range is changing between time steps.
Guillaume Duclaux
Nice - 18/05/2024
Citation: https://doi.org/10.5194/egusphere-2024-569-RC3
Status: closed
-
RC1: 'Comment on egusphere-2024-569', Haralambos Kranis, 29 Mar 2024
General comments
The paper by Faucher et al. is an interesting approach to the complexly deforming Aegean region. It is scientifically significant and falls well within the scope of SE. The model setup and boundary conditions are valid and the model resolution is sufficient enough to display the crustal structural features. Nonetheless, the intrinsic limitations of the method lead to certain inconsistencies between the model outcome and the actual structural configuration, as described below:
The outcome of the final model (Fig 3D) correctly predicts the formation of E-W striking normal faults (roughly corresponding to the active rifts of Corinth and Evvia), alongside with ENE-WSW to NE-SW dextral strike-slip faults, (which is the case in the central and north Aegean, i.e. splays/branches of the North Anatolian Fault), conjugate to WNW-ESE to NW-SW sinistral s.s. faults. A fault zone of such strike and kinematics does occur in reality, the Katouna Fault System, NW of the gulf of Corinth (Perouse et al., 2017) (note that it is incorrectly shown in Fig 1C as a normal fault).
The conjugate arrangement of NE-SW dextral NW-SE sinistral s.s. faults is not (exactly) the case in the broader Aegean region, however. As shown in the paper by Sakellariou and Tsambouraki-Kraounaki (2018) (which is properly acknowledged in the m/s and used as a basis for Fig 1C), the upper-plate deformation in the South Aegean is characterized by NW-SE sinistral s.s. faults.
Overall, the paper by Faucher et al is a step forward, towards the understanding the complex deformation of the broader Aegean region, albeit with some limitations, posed by the model setup and boundary conditions. The symmetry displayed in the outcome of the model possibly arises from the equality between the amounts of compression and extension applied to it and the strict orthogonality between the compression and extension axes, as trench retreat is faster than the Anatolian extrusion and these two end-member vectors are not strictly perpendicular to each other. Moreover, the effect of the Anatolia extrusion in the northern part (i.e. north of the Maliakos-Amvrakikos latitude) is much less pronounced; probably the Vc amount imposed along the x-direction could follow a gradient, increasing towards the “south” edge of the model.
Specific comments
See attached pdf with notes and comments on Fig 1.
The “Pelagonian” fault is currently active, as evidenced by recent (2022-today) seismic activity; based on focal mechanism solutions, it is indeed a left-lateral fault boundary.
Technical corrections - typing errors
l.41: “the Oligocene”, “the Miocene”
l.150 “NW-striking strike-slip faults”
References
Pérouse, E., Sébrier, M., Braucher, R. et al. Transition from collision to subduction in Western Greece: the Katouna–Stamna active fault system and regional kinematics. Int J Earth Sci (Geol Rundsch) 106, 967–989 (2017). https://doi.org/10.1007/s00531-016-1345-9.
Sakellariou, D. and Tsampouraki-Kraounaki, K.: Plio-Quaternary extension and strike-slip tectonics in the Aegean, in: Transform Plate Boundaries and Fracture Zones, edited by: Duarte, J., Elsevier, 2018
- RC2: 'Comment on egusphere-2024-569', John Naliboff, 10 Apr 2024
-
CC1: 'Comment on egusphere-2024-569', Frank Zwaan, 30 Apr 2024
Interesting models!
The conjugate strike-slip structures remind me of those we found in our analogue models (Zwaan et al. 2019, and references therein), and those in the recent analogue modelling Liu et al (2024) study. It could be interesting to compare these works with these new numerical results. In the analogue models, deformation is driven by a basal boundary condition (hence a decoupling layer can make the system switch from conjugate strike-slip faults to normal faults), whereas here it seems to be "side wall" boundary conditions in these numerical models, which could prevent such decoupling effects?
NB: it may also be interesting to check the Guillaume et al. (2022) analogue modelling paper, which makes things even more interesting by adding structural inheritance to the mix.
- Zwaan et al. (2019): https://doi.org/10.5194/se-10-1063-2019
- Guillaume et al (2022): https://doi.org/10.5194/se-13-1393-2022
- Liu et al. (2024): https://doi.org/10.1029/2023TC008127
Citation: https://doi.org/10.5194/egusphere-2024-569-CC1 -
CC2: 'Comment on egusphere-2024-569', jun liu, 30 Apr 2024
Nice models!
Followed Franks' comments, in our analogue experiments basal boundary condition (Liu et al., 2024), we observe systematic change in fault pattern from compartments of conjugate sets of oblique-slip normal fault to one-sided sets of oblique normal faults to horst-and-graben structures with decreasing strain rates (~extension velocity). As we all knnow, the strength of viscous layer is a function of strain rate έ and viscosity η (σ1-σ3 = ηέ ), so in our models, fast models are recogonised as with strong ductile layer and low models are recogonised as with weak ductile layer. We attribute strain rate dependent of faulting structures to the increasing vertical coupling between the basal rubber sheet and the sand layer, controlled by the rate-dependent strength of the viscous layer.
you can also check similar numerical modelling paper by Pang et al., (2018) (http://doi.org/10.1002/2017jb014011). they also have "side wall" boundary conditions with shortening and extensnion. their models also show from conjugate strike-slip faults to horst-and-graben structures by adding weak lower crust.
It would be great to visualize these models in a more quantitative way to link data in the field examples, shch as slip partitioning.
Citation: https://doi.org/10.5194/egusphere-2024-569-CC2 -
CC3: 'Comment on egusphere-2024-569', Matthias Rosenau, 13 May 2024
Dear authors, all,
Congratulations! I came across this preprint on my search for 3D numerical models allowing us to compare with our analogue models with similar kinematic boundary conditions, i.e. horizontal lateral shortening and longitudinal extension. I see a nice discussion going on and just want to add my perspective.
During my literature search, I found few models with a comparable specific boundary condition although, as stated above, there have been multiple numerical rifting papers now in 3D and at crustal scale. However, I see those looking at different aspects of the general rifting problem and not so much on multiphase rifting. We have also run some numerical models with the Aspect code with similar kinematic boundary as presented here and our results look surprisingly similar (at least qualitatively) given that we had very likely very different parameters (e.g. mesh resolution, domain size, rheologies, etc.) as we aimed at simulating our analogue experiments. In conclude for myself that the general patterns seem to be stable first-order features although details matter definitely at second order. One may want to see a larger number of models for a more systematic parameter study but this seems beyond the aim of the current paper the models are used to test a specific regional tectonic hypothesis.
I second the reviewer #2's point that the numerical methods and assumptions should be explained in more detail especially because ASPECT has become a tool widely used in the community and we should share as much information as possible. I am not familiar with the sharing of data in the numerical community but I think sharing the input files and/or some sort of set of files and codes allowing to reproduce the results should be provided. From the model numbering in figure 2 it seems a couple of additional models (2, 3, 5, 6) have been run but not presented here. They may help however in discussing the sensitivity of the models to specific parameter choices and may therefore serve as benchmarks. If this is the case I would suggest adding them in an appendix.
Overall, I think the paper tackles an important issue and can be, after some mainly technical modifications, an important contribution to triggering the study of 3D aspects of tectonic deformation and complementing existing numerical and analogue models.
Citation: https://doi.org/10.5194/egusphere-2024-569-CC3 -
RC3: 'Comment on egusphere-2024-569', Guillaume Duclaux, 18 May 2024
Review of "Strike-slip faulting in extending upper plates: insight from the Aegean", by Agathe Faucher, Frédéric Gueydan and Jeroen Van Hunen.
This short communication investigates the development of strike-slip structures at the surface associated with coeval extension and shortening using 3D thermomechanical models in the context of the Aegean. 3D modelling work like this one are important to better understand strain partitioning in orogens and active tectonic systems like back-arcs. The manuscript is illustrated with 3 figures and fairly written. Some references about 3D models with comparable BCs are missing (e.g. Le Pourhiet, L., et al. (2018). Continental break-up of the South China Sea stalled by far-field compression. Nature Geoscience, 11(8), 605-609.), and recent analogue papers have also been suggested in the open comments.
The paper very briefly introduces the tectonic context of the Aegean system since 30 Ma, then focuses on a series of 3D numerical models’ results presentation and discussion. Four numerical models, simple in design and relatively convincing in their output are presented. The model domain is relatively limited spatially with a model box of 100x100km horizontally and 50 km vertically. The first three models are generic and explore strain partitioning in different settings, and the fourth one is hybrid and supposedly developed in the context of the Aegean extension with Anatolia extrusion.
I am no expert in the Aegean system evolution, so I can't really comment on section 1.
I present below some key points for which I have concerns followed with a list of minor comments and detailed critics about the figures. I sincerely hope these will help the authors with improving their contribution.
I have quite a few critics about the modelling work presented here. Numerical models visualisation and analysis could very much be improved. I get back onto that later when discussing the figures individually. For example, colour scales are changing from figures to figures which solely display the strain rate second invariant (I assume this is the second invariant and not the magnitude of the strain rate tensor - not sure if this is correct?), the line drawing interpretations sometimes seem arbitrary. It would be neat to show other scalars such as cumulative strain, some isotherms. As they are designed now it feels like we miss a lot of information from these simulations.
A table summarizing all models setup is necessary. In Figure 2 the three presented models are named Model 1, Model 4, and Model 7. In the text (line 95-96) they are referred to as model M_e, M_c, and M_ec respectively, and the author indicate those are the "end-member solutions". I suppose other velocity settings have been tested. Is that correct and could you briefly comment on the other models’ behavior? This table should at least be presented in supplementary material. For consistency I recommend you use only M_c, M_e, and M_ec in the text and figures.
Other critical information regarding the models is missing. For example: What is the top surface boundary condition? What are the top and bottom temperatures (or top temperature and basal heat flow) of the model? A graph showing at least the initial viscosity profile and the geotherm would be much appreciated. The initial 920˚C temperature at 30 km depth in the model seems rather high and can't be buffered through latent heat during partial melting as it is not considered in the model.
According to Fig 1, in the context of the Aegean system the Anatolia extrusion velocity is ~40% smaller than the trench retreat. That is very different than in Model M_ec_t (presented in Fig 3 and the last pages of the manuscript) or with the text in lines 127-128. Have you explored similar V_e/V_c ratios? How would changing the relative velocities along the orthogonal walls impact the model? You suggest (line 115-116) that "shortening and extension should be of roughly the same magnitude to trigger strike-slip", so does the Aegean qualifies? Or is the model applicable for this particular test case?
For all this reasons I would recommend the authors profoundly rework the figures and the manuscript as in its current form this communication is not worth publication in Solid Earth.
Minor comments:
+ line 26: STEP faulting: not sure why STEP is spelled in uppercase.
+ line 30: Amorgos: not shown in fig 1C. Please add the fault system name to the map.
+ line 36: Please ensure you use the same name for this domain (i.e. the Aegean microplate) throughout the manuscript. It is defined as Aegean/Anatolian microplate line 52. Also, the spelling of microplate changes from micro-plate to microplate.
+ line 41: I would suggest a small edit here: "metamorphic core complexes (e.g. Rhodopes in the Oligocene and Cyclades in the Miocene)".
+ line 46: Here you mention plutons. Plutons of what rock type? Gabbro? Salt? Granite?
+ line 62: Would inherited structures (and which ones) play a major role in the evolution of the Aegean system? This would be an interesting discussion point.
+ line 68: At what depth is the flat Moho in the Aegean today? This would be an interesting feature to add to the map in Fig 1C (e.g. showing crust isopaches).
+ line 76: please define the top BC here. Moreover, what BC is applied to the walls parallel to the inflow of outflow for models M_e and M_c?
+ line 80: You should refer to the concentration of radiogenic elements or the radiogenic heat production and not the "rate of radiogenic elements".
+ line 82: "(" missing before "see Kronbichler et al., 2012[...]".
+ line 82-83: Although ASPECT is more and more used in the tectonic community it would be nice to write down which equations are solved somewhere... at least in supplementary. I assume the system of equations are for highly viscous incompressible fluid motion. Is that so? Please provide a little more details about the physics used in the code.
+ line 83: I would recommend adding yield or frictional in "[...] with Drucker-Prager yield plasticity." I can't see either from the text or Table S1 whether there is any weakening imposed for the frictional rheology. Is there any weakening function applied to the cohesion and/or friction angle?
+ line 100-101: There seems to be some strain rate concentration in the mantle for the pure extension model at least. Could you comment on the surface spacing of the resulting grabens? And possibly the resulting topography as I assume this the feature used to define the fault trace in Fig 2 and 3.
+ line 104: How high is the resulting topography for model Mc? Why is the fault spacing so different than for Ve model? You mention no tectonic inheritance in the model’s setup, so I'm curious about what is controlling strain localization and shears/faults spacing. It's a shame we can't visualize the relative thickening of the highly radiogenic upper crust and the depleted lower crust layers. I might have made a mistake in my back of the envelope calculation but in the absence of surface processes I am surprised the Moho depth is only 50km at 4 Ma for a full-rate inflow velocity of 2cm/yr... please see comment about figure 2 below.
+ line 109-111: this should be in the discussion section rather than the results.
+ line 112-113: I'm afraid you do not clearly show any proper fault rotation... Because solely the strain rate is used to visualize the deformation it is not clear whether structures rotate, or new structures replace them at a different angle. Maybe an animation would help visualizing the progressive evolution of faults orientation through time. Alternatively, a quantitative analysis of the fine strain through time would be more convincing.
+ line 113: I suppose you mean thickening rather than "shortening".
+ line 134: Why is the time period 15-5 Ma? I don't believe the extrusion of Anatolia is over yet and the Hellenic trench retreat is still taking place too.
+ line 137: "looks more homogeneous and with ongoing activity of the three former rift systems" what do you mean by that? The former rifts activity is not visible in Fig 3B.
+ line 141-142: from the figure I really can't see the side rifts evolving into anything.
+ line 147-148: If the model really applies to the Evvia and Corinth system I suggest you have a look at the work of Pechlivanidou et al. paper (Pechlivanidou, S., et al. (2022). Contrasting geomorphic and stratigraphic responses to normal fault development during single and multi-phase rifting. Frontiers in Earth Science 9: 748276). But AFAICT this region is dominated by multiple generations of normal faults rather than normal faults and strike-slip faults.
Figure 1:
+ You mention in the text (lines 60-61) that according to Brun et al., 2016 extension rates have changed through time. Adding estimates of those rates in Fig 1A, 1B (red arrows) and in the text would be very good. This is also key to validate the models where V_c and V_e chosen for the modelling approach.
+ Please add a bounding box on Fig 1A and 1B inset maps to highlight the region corresponding to the close-up views.
+ the [C] is missing in Figure 1C. Could you please also mark distinctively the trace of the Hellenic trench in the map? This is one of the main features of this figure, yet it doesn't clearly appear. As noted earlier, the Amorgos fault system should also be added to the map.
Figure 2:
+ Arrows for Ve and Vc should be drawn on both x-normal and y-normal faces respectively. As it is displayed here it looks like Ve and/or Vc are solely applied on one face. That questions whether the announced velocities are full-rate or half rate. I understand from the text (lines 73-74) 1 cm/yr is half rate. Is that correct? This is very important for estimating the finite strain and moho depth, and it must be clarified in the figure.
+ The strain rate and the marked fault traces are not superimposed, which is totally understandable. This is because the faults traced represent early structures (I believe the fault traces are based on topographic changes delimiting grabens for Model 1), but it is nevertheless intriguing, and you should explain on which premise these black lines are drawn. For the Extension model it is rather surprising to see active shears crosscutting pre-existing structures.
+ The strain rate color scale's range is different for every snapshot (same comment applies for figure 3). I believe this makes the comparison between setting a little less obvious.
+ For Model 7 (M_c): I assume the x-normal walls have free-slip BCs, so it behaves like a 2D model. If you apply a constant full rate inflow velocity of 2 cm/yr (1 cm/yr on each y-normal face) during 4 Ma, the Moho depth should reach 150 km... here it is marked at 50km depth. Either the full rate inflow velocity is 1cm/yr or the model duration is 2 Ma (or I did something wrong with my calculation).
Figure 3:
+ I understand these are snapshot of a single model with evolving velocity BCs through time. The model runs with pure extension in the x-direction (and free-slip BC along the y-normal walls?) for 7.5 Ma, then orthogonal extension in the x-direction and compression in the y-direction for an extra 7.5 Ma. This is clear from caption, but not so clear from the figure itself. Indeed, for 3B I get when reading that "8Ma, Extension=Compression" that the model has the same amount of extension that compression for 8Ma... Maybe something like 8Ma (7.5 Ma extension + 0.5 extension&compression)? Or this is just me... sorry for being a bit pedantic here.
+ The faults traces in black at the surface are just obscure. I assume that in 3B the faults traces should be rather close to 3A, at least for the central rift. This is definitely not the case.... Why? This really questions the visualization and interpretation of the model results. Same goes for the white shear zones after.
+ Is there any solid block rotation at the surface of the model in C or D? Imaging rotated faults drawn in A away for the center of the model would be very interesting. This type of analysis is really missing here and could bring valuable insights into the Aegean system itself. What happened to the normal faults’ terminations close to the y-orthogonal walls when the conjugate strike-slip faults evolve?
+ As for figure 2 the color scale range is changing between time steps.
Guillaume Duclaux
Nice - 18/05/2024
Citation: https://doi.org/10.5194/egusphere-2024-569-RC3
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
387 | 131 | 30 | 548 | 55 | 13 | 13 |
- HTML: 387
- PDF: 131
- XML: 30
- Total: 548
- Supplement: 55
- BibTeX: 13
- EndNote: 13
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1