The second main type of faults found in extensional regimes, listric faults can be defined as curved normal faults in which the fault surface in concave upwards; its dip decreases with depth. These faults also occur in extension zones where there is a main detachment fracture following a curved path rather than a planar path. Hanging wall blocks may either rotate and slide along the fault plane (eg slumps), or they may pull away from the main fault, slipping instead only along the low dipping part of the fault. Roll-over anticlines will often form between bedding planes and the main fault plane as a result of the flexing between the two.

Hey Jonas, I've been asking the same question to myself many times but I haven't come across a single answer. Maybe there isn't such. I hope some experts on passive margins will show up here. I'm not sure if a "perfectly curved shape" exists in nature or rather only in Move-driven cross-sections. And even if it exists here or there, is it a true shape of a rupture plane or a result of its passive deformation due to the ongoing extension? As you certainly know, in thin-skinned (supra-salt) extensional domains, a relatively constant (~Coulomb) cut off in the upper brittle part is of the section is observed and it is only within the ductile detachment where the fault abruptly flattens. A kind of domino pattern is created... But this kind of extension is gravity driven and may not be a good analogue to crustal-scale processes. However, when I see a normal fault drawn with a landslide-like curved trajectory from the surface down to the brittle/ductile transition, I become suspicious...

Hi Jonas. As we've sometimes discussed, there is also the question of the local stress orientations in a crustal profile. Could rotation of principal stresses away from horizontal give a "correctly" oriented stress relative to a near-flat detachment? Stresses are reoriented to some degree around elastic faults. Crustal scale detachments however, flatten in the ductile realm. Lukasz's general observations are also interesting. To hear somebody actually state that there is a relatively constant cut off in the brittle domain is a bit of a rarity. Probably because this issue has been ignored.

There are suggestions that fluid pressure or other weaknesses at depth may play a role? Clearly, if you have an extremely weak layer somewhere at depth, with relatively strong material above it (and below it), then it is likely that the stuff above it will fracture down to the weak layer, but extension will then localise in that weak (flat) layer over very long distances. The hanging wall material will deform as necessary (and according to the rules of Mohr Coulomb behaviour) above it, as it extends.

My first point incidentally, gets back to the (apparently raging) argument between "low angle initial fault -ers" and "rotate to low angle -ers". The first option appears at first sight to be mechanically impossible in the upper crust. Unless as I suggested earlier, there is a big rotation of principal stress orientations away from horizontal for some reason.

I greatly appreciate David's comment. Let me dig a bit. How the s1 deviation would work for conjugate systems of normal faults? Although in most extensional provinces there is a preferred polarity of normal faults that results from crustal-scale asymmetry of rifting, symmetric cases are also known. And in the latter case the faults also flatten to a common detachment despite of opposite dips. The s1 would rotate locally in opposite directions, for each fault separately? What about a passive upwarping of a deep segment of normal fault due to a ductile flow in the detachment caused by unroofing? Don't you you think that listric geometries are more common in mature (high beta) extensional provinces, while for incipient extension planar geometries dominate? So, could planar evolve into listric with increasing beta?

Hi Lukasz

Your example is very good. And of course, you are correct. Symmetric cases confound the idea of rotated stress directions (at initiation, at least). So if they do work at all, then not for all cases.

Passive upwarping of deeper segments has already been proposed as a natural consequence of "flexural cantilever" type, extension models. And developing that line of argument a little leads almost automatically to an evolution from planar to listric faults (I would think at least).

There is quite some literature on all this. I'm not expert enough in any of it to know whether the central question asked by Jonas has ever been properly addressed. Roger Buck, Luc Lavier and several others have written extensively on the topic. Perhaps time to dig a little?

Dear all,

thanks for your contributions! Our discussion seems to cover much of what has been argued around listric normal faults, with some almost inevitable excursions to low-angle normal faults (primary or rotated?) or the question why systems of normal faults sometimes exhibit a consistent asymmetry (tilting of the entire system?). I share the concerns about interpretational bias but still think that there are some well-documented examples of truly listric geometries. My question was originally aimed at cases where a weak stratigraphic horizon provides a detachment level. I like the idea that with increasing displacement faults may evolve from an abrupt transition between two planar segments - ramp and detachment - to a more gently curved listric shape. This change could be driven by a tendency to minimize internal deformation of the hanging wall and would involve the (repeated?) shortcutting of unfavourably kinked geometries. I guess that this process should be governed by local stress reorientation, but couldn´t figure out how exactly that could work.

Hi Payman,

I didn´t mean to challenge Lukasz´s point about regional stress orientation, but was thinking about a more local phenomenon. Anderson´s theory is a useful simplification but not a strict rule as evidenced by many EQ focal mechanisms that have none of the principal stresses horizontal or vertical. Variable stress trajectories are probably inevitable in inhomogeneous materials.

I am not sure whether your intruiging argument about 2θ gradually going to zero in the ductile realm with increasing temperature really holds. Wouldn´t we expect the envelope of decreasing shear strength to asymptotically approach the abscissa, implying that even a very small Mohr circle would still touch it at 2θ similar to 90°?

Thrust faults going vertical at depth were popular before plate tectonics, but I am skeptical if any well documented example exists. The general absence of subvertical contraction structures (if I am prejudiced the right way there) would strengthen my belief that flat stretches of faults are governed by subhorizontal zones of prescribed mechanical weakness rather than gradually changing properties of otherwise homogeneous materials.

...or Simply the result of the rooting of a fault on a décollement layer, as shown by Brun and Mauduit, Tectonophysics, 2008

I think that listric faulting is related to rotational landslide. More compact layers go down due to bedrock soft rocks (esp. claystones). This is driven by gravity.

Greetings! What about this paper?

Kreszek et al., 2007: "Mechanics of fault and expulsion rollover systems developed on passive margins detached on salt: insights from analogue modelling and optical strain monitoring". From: JOLLEY, S. J., BARR, D., WALSH, J. J. & KNIPE, R. J. (eds) Structurally Complex Reservoirs. Geological Society, London, Special Publications, 292, 103–121. DOI: 10.1144/SP292.6

Enjoy!

Dear Jonas Kley: Since in your query refers to “seismic interpretation”, without clarifying whether the domains is time or depth, let me alert you to the distortion that exists in the geometry of subsurface faults in general, and in listric imaging in this particular case (flattening downward) if you are working on time-domain.

I would like to make an alert that this incongruity may shoot in part from misconceptions, misinterpretation inherent not only on the fact that you are in two-way-travel time instead depth, but also by seismic migration artifacts, structural complexity and SNR quality of the seismic data amongst others. Consequently, many interpretations of fault geometry are possible, and of course, how this geometry develops.

It is difficult to speak whether faults are planar or listric on time (TWT, two-way time) seismic domain. Fault geometries imaging on time-sections will appear listric even if their true depth geometry is planar, due to the increase in velocity with depth (Fig. 1). Furthermore, the apparent dips of faults in time can be misleading, making it difficult to assess geological validity of structural interpretation. As advise Lyon et al. (2005), a comparison between time and pseudo-depth seismic sections show us that many of the listric faults are planar in geometry (Fig. 2). Once I talk about a “pseudo-depth”, it means that: in the presence of a lack of significant lateral velocity variation (velocity measure obtained from compressional sonic log / check-shot / VSP from wells, and / or processing - stacking -velocities from seismic), in such a way that it would be possible to apply a constant-laterally-gradient velocity-function to time migrated seismic section (2D) or volume (3D) to produce a pseudo-depth data set, whichever PSTM, or PSDM scaled to time or poststack-time migration.

Therefore, as a Geophysicist interpreter we must examined the impact of the choices in depth conversion on structural validation: the velocity model choices in seismic depth conversion have an impact on the final depth image and hence the structural geometry of interpretations, e.g., the imaging of the footwall should not expected impacted by “shadow-zones” or incorrect migration velocities.

Please, see this paper: “Improving fault geometry interpretation through ‘pseudo-depth’ conversion of seismic data in the Penola Trough, Otway Basin”, by Lyon, P.J., P.J. Boult, A. Mitchell and R.R. Hillis. PESA Eastern Australasian Basins Symposium II. Adelaide, 19–22 September, 2004.

Best regards,

Mario E. Sigismondi

Hi Darko and Mario,

thanks for your interesting comments! Krezsek et al.`s is a nice paper I hadn´t seen before. Seeing planar faults as curved on time-domain seimics is a valid concern as you demonstrate. However, I still think that there are truly listric faults, in particular if they flatten into a subhorizontal detachment (thin-skinned extension). I agree with Jan that they are very similar to rotational landslides (which in turn demonstrate that listric geometries can develop without synkinematic sedimentation).

Jonas

Hi Jonas,

Thanks for raising this issue! In order to develop a (sub)horizontal flattened listric fault segment, even in a thin-skinned succession with a ductile/"rafting" detachment/footwall layer, involvement of lithospheric-scale extensional processes has often been the case. These flattened sections are extremely difficult to document, either in seismic or in the outcrop scale. Another thing, we need to look closely the eventual footwall deformations as well, prior outlining listric or finite length faults.

Darko

The formation of listric normal faults is an interesting topic in structural geology. The following work investigates the role of fluid overpressures on the formation of many observed listric faults.

Dear Kley,

I think the main confusion arising from the assumption that Coulomb-Mohr criterion can be applied all along the detachment fault. In fact, it is not possible, since the rheology of the rock from the surface to the deeper parts of the earth is changing from brittle to ductile. Therefore, Coulomb-Mohr criterion only can be applied for the shallower (brittle) parts of the fault while it gradually changes to von Mises plastic (ductile) yield criterion at depth. Also, one should consider the increase in the pore-water pressure at depth, like in the case of Beer Can experiment of Hubbert & Rubey (1959). A combination of these mechanisms can easily explain the geometries of detachment faults. So, stress configurations at the shallower parts of the fault control the deformation but at depth, there is no more differential stress.

The listric geometry is in fact inevitable from the surface to lower crust since material properties are changing as the ambient temperature is increasing and local stress configuration is changing due to sedimentation in the half-graben part of the fault. The fault geometry approximates the geometry of the composite failure envelope (parabola) that combines all the failure criteria from Tensile to Mohr, Coulomb, and to von Mises plastic yield.

Really inclusive discussion.

Nuretdin KAYMAKCI

Hi Nuretdin,

thanks for your comment (and sorry for a really late reaction). I am not sure your point answers my question. The brittle-to-ductile argument only holds for faults that reach as deep as 10 km or so, or shallower ones which flatten into plastic salt. More importantly, a transition from the Columb-Mohr to the von Mises failure criterion would only predict a decrease of the fault´s dip angle from around 60° to 45° (with the Mohr circle touching the horizontal criterion at 2 theta=90°) , but not to a subhorizontal detachment.