|WikiProject Geology||(Rated Start-class, Mid-importance)|
OK - more to do. Need to separate the cold diapiric stuff from the hot diatremic stuff. Later - :-) Vsmith 12:53, 6 December 2005 (UTC)
Yeah, especially as diatremes are not diapirs. Rolinator 00:50, 22 December 2005 (UTC)
- True - separation is planned - I'm kind slow though :-). I see you have made a diatreme page - good, but it too needs a bit of a re-write. Ah well, someday ... Cheers, Vsmith 02:52, 22 December 2005 (UTC)
Dicussion on causes of diapirism
The original paper by Mrazec described diapirs from Roumania where salt cores have 'pierced across' and breached an overlying anticline. The folds are assymetric, compressive in style suggesting some 'tectonic energy' driving the intrusion. These pierced folds are very different to Gulf Coast salt domes.
Further, I don't believe that 'bouyancy' is essential for intrusion. The driving force for any intrusion is the weight of the adjacent overburden/cover which 'sinks' into the source layer, driven by gravity. The one factor that allows this release of gravitational energy is the MOBILITY of the intruding material. For example, sea water will rise up a hole/crack in an ice floe, forming an 'intrusion', yet the ice is less dense than the water otherwise the ice would sink. Even mercury will rise up a hole in a styrafoam block floating on it.
Sure, salt is less dense than typical host rocks/overburden, but its most important property, for intrusion, is its ability to flow(especially if wet and hot)in response to very low stresses. It is a time-dependant fluid, or 'rheid', rather like pitch/bitumen (and glass, marble etc). A tombstone (see old marble examples) or window (see glass in old cathedrals) of rock salt will droop and flow like treacle, given enough time
The lavalamp is a poor example of a 'diapir'.
A mobile intruding material that is LESS dense than the overburden simply has the potential to EXtrude (the cover sinks completely). If more dense, the intrusion will penetrate the cover to some point, then cease.
This is especially the case in extensive terrains where intrusion can be 'permitted' by tensional zones in the cover; even 'pull aparts' in compressive zones.
Forget about 'low density'; mobility is the key to intrusion.
Perhaps we should forget about 'diapirs' and just call them 'intrusions (salt, granite, shale, mud, breccia etc)'?
Cheers tmount 00:41, 27 February 2007 (UTC)
- This is over a year since the comment was posted, but I'd like to clarify that the density difference IS driving this. What is described above uses examples of holes in the upper material. Because of buoyancy and isostasy, no less-dense material will hover above a more-dense material - a volume proportional to the ratio in densities will be submerged. But for rocks, we are looking at layers - there are no holes cut in them, and if this density instability didn't exist, the salt would be very immoble because it would only be able to move upwards at a distance of the density ratio - so much less than the thickness of the salt layer. The importance is that, with a layer of salt covered by a denser layer, the density contrast allows Rayleigh-Taylor instabilities to form and cause domal structures of salt to intrude the overlying rocks
- The mobility is also important for setting the short time-scales required for intrusion, but the fact that salt domes form is set by the density imbalance.
- Second, and less related, glass flowing over ~1000 year time-scales is a myth. It's viscosity requires orders of magnitude more time. The "thinning" is related to the Medieval process of thinning glass for windows on a rotating table - the outer edges of the glass became thicker, and were put at the bottom of the windows, probably for stability.
Appreciate comment on glass viscosity and withdraw reference to'cathedral windows' although input from a physicist on time-dependent fluids/viscosity or rheids would be useful. Salt is certainly a rheid, especially when wet/hot.
- I don't know what the viscosity is, but I'm sure you could look it up. It should have an Arrhenius-type relationship with respect to temperature, and I'm-not-sure-what with respect to water content. Awickert (talk) 01:15, 23 February 2009 (UTC)
But it is my understanding that real-time measures of viscosity are meaningless in geological time frames . . . 'rheidity', or 'time-dependent viscosity' is the key, with flow of pitch being an example. Pure halite has 'no (measurable) ultimate shear strength', meaning it will flow, in time, in response to (any?) stress . . presumably a fly landing on it will eventually sink into the 'solid'. Even quicker for hot and wet salt. Expert input required here, and I am trying to recall research I did 38 years ago! Tjmount (talk) 20:27, 24 February 2009 (UTC)
- It depends on the stress-strain relationship. If it behaves as a Newtonian fluid, then the viscosity will hold across all time-scales, because there is a linear relationship between stress and strain. However, if the rheology is more complicated, then variability in the strain over time can be observed. I'd like to find a curve for salt, but I haven't looked too hard. Awickert (talk) 20:49, 24 February 2009 (UTC)
- Thanks, a useful paper,and for its reference to hot and wet salt. In the Flinders model I am also suggesting that, at any given scale, fine breccia particles ('fluidised' by 10 to 15% salt) acted as a fluid matrix (by granular flow, like grain) to entained xenoclasts. ie Fractal geometry. How do you measure the viscosity of that !? Tjmount (talk) 01:43, 25 February 2009 (UTC)
Agree with most of your comments re layer instabilities etc, and the role of bouyancy in forming the classic US Gulf coast domes . . . and lava lamps.
However, my work was done in the Flinders Ranges of South Australia where the intruded host material was demonstrably unyielding and BRITTLE at salt emplacement (no Rayleigh-Taylor waves here), as evidenced by dyke wedging, roof grabens (extension normal to S1, bounding faults aligned parallel to S1=profile planes of anticlines), very sharp contacts with shape matching ('continental drift' style) across intrusions, stoped host blocks (to kilometres long = 'titaniclasts') etc. Instead of your 'holes' (above), I suggest the complex of intruding salt dykes were emplaced up S1/S2 structural planes, parallel to S1, normal to S3 principal stress directions, which at the regional scale was the direction of tectonic compression. Locally, the orientation of dyke planes varied with local stress field directions, giving complexity to the salt dyke swarms. In particular, smaller salt breccia dykes invade the two conjugate joint plane sets on the folds, as well as the usual axial and profile plane joints/faults. So where the overburden is thick, S2 is great and S3 becomes the direction of minimum stress . . . and even EXTENSION, as in the cores of massive Flinders Ranges regional anticlines.
- OK - I see - so basically, in this area, structural control was dominant, and salt intruded in dykes, generally along planes normal to both /sigma_2 and /sigma_3. Am I correct? And when you say extension, I would generally assume the least compressive stress direction to have some extension occur - do you mean that /sigma3 was negative (and therefore a tensional stress)? Awickert (talk) 01:15, 23 February 2009 (UTC)
Yes, I think you have got it . . . stuctural control, dykes into planes normal to S3=least or even tensional stress with extension as the dykes 'wedged' in as the fold grew under regional compression.Tjmount (talk) 20:34, 24 February 2009 (UTC)
These are 'passive / permitted intrusions'
In a sence, the heavy fractured/blocky brittle overburden has foundered or sunk into the fluid salt layer, with the salt being further energised by lateral teconic/compressive forces.
I am trying to think of an example, but perhaps a glass of water filled mostly with crushed ice where the glass walls restrain lateral movement (ie 'tectonic forces') . . . the mobile (but, note, denser!)water rises and fills the voids between the clasts to a level way above 'isostatic' expectations.
- I don't think I agree with this analogy, but I think I know what you're trying to say. In a glass, or any container, ice will behave isostatically, with about 10% above the surface, regardless of confinement. However, the important thing with your area would be both the density and the low viscosity (high mobility) of the salt. The salt wants to go up, and when you sqeeze the sides of the "glass", it shortens in the horizontal directions and therefore lengthens in the vertical (being isovolumetric) and oozes in around the cracks in-between the "ice". Awickert (talk) 01:15, 23 February 2009 (UTC)
Yes, I need help on this one and the glass model may not be the best. The point is, for a fractured brittle overburden, a mobile (but relatively dense, as for water up cracks in less dense/brittle ice breccia) material can move up a long way. Potentially to the surface where it 'extrudes' if it is less dense. How far vertically would a wet/hot (250deg C?)and mobile halite/anhydrite mush, with entrained dolostone/siliciclastic/dolerite clasts, intrude in a very thick (total to 40km) overburden of lithified rift sediments; as in the Flinders Ranges. First, on isostatic principles, then adding compressive energy? Think 'a salt -fluidised mush, based on a very poorly sorted 'fractal' breccia with clasts across a wide size range (dust to several km) such that the fines act as a mobile matrix (eg granular flow, like wheat) to larger clasts. Pore space could be as low as 10 to 15 % with the voids filled with 'fluid/plastic' salt which allows particle rotation and flow of the breccia. 'Salt rheidised/fluidised' in my terms. After emplacement, the mass cools, salt is leached out, and the intrusion becomes solid (with no obvious salt masses) in time, as in the Flinders Ranges. Zeolite facies cements (calcite, talc, silica, pure low-T adularia, magnesioriebeckite, dusty heamatite from dedolomitisation reactions during intrusion) etc lock up the breccias. Tjmount (talk) 21:09, 24 February 2009 (UTC)
All this is very like Mrazec's Type 'diapirs' where salt has been injected across the cores of compressive folds.
There is a real need to rethink these mechanisms, towards the development of a 'grand unified theory if all intrusions' in which we can all insert our piece of the model.
- To your last comment: I think that the biggest hard thing is just to know the 3D stress field around the intrusion, because the intrusion modifies the regional stress field, and the rheologies of the intrusion (variable, in the case of magmas, as they cool) and the country rock (probably brittle, but needs to be shoved out of the way, I've never heard a great conservation of rock mass about this) need to be dealt with.
- Your field work sounds like pretty cool stuff!
- Awickert (talk) 01:15, 23 February 2009 (UTC)
What my Flinders work showed was that all the traditional Gulf salt dome/lava lamp/Ramberg centrifuge thinking might not be the full story. In South Oz, at least, I became increasing nervous about geoworkers attributing every structural disruption, conglomerate, or facies change to the ubiquitous 'energetic punch up through the adjacent cover of a bouyancy driven low density salt 'diapir'.
Circular reasoning was common . . . 'here is a conglomerate . . . thus a diapir nearby . . . which must have shed conglomerates.
I quickly got a bad reputation by asking for proof that the conglometic clasts 'once were resident in a diapir'
Similarly, that latest trend is to attribute local facies changes in 'mini basins' to 'salt withdrawal and diapir emplacement' but this is not proof, especially where an active block faulted basement underlay the rift.
I could only see gentle, synteconic, passive emplacement of breccias of low salt content/low or higher density . . as above.
'Bouyancy' is something I have never really understood, like 'aerodynamic lift' over a wing. Seems we should get back to physical first principles (eg thermodynamics) and abandon comfortable assumptions.
It seems reasonable that all intrusions (diapirs, batholiths, salt domes, diatremes, igneous stocks, mid-oceanic dykes, lava lamps, mud up between the toes, sauces in puddings, C/Nimbus thunderheads/thermals, mantle plumes,) must all be driven by a single common set of simple principles (GUT of Intrusions). Glaciers and rivers probably follow the same rules, but upside down, air is the host, and more horizontal than vertical!
My challenge is to ask if the potential energy of a system can be lowered in a gravitational field by the migration of matter where mobility of the invading material is the key, subordinate to the effects of relative densities. Tjmount (talk) 21:57, 24 February 2009 (UTC)
- I'll put my comments down here to break up the mess. To your top comments, I don't know enough about the geologic setting or the rheological properties of salt to be able to say anything. To your bottom comments, I believe that there are already good theories for intrusions. They are driven by multiple factors. Gradients in density (buoyancy, i.e., mantle plumes), determine the rate of intrusion and determine how far it will intrude until it reaches a balance in pressure. Large overburden pressure (i.e., stepping on mud) also determine the rate of intrusion (because the rate is pressure-driven). The rheology/viscosity/mobility of the material determines how quickly the material gets to its final state: fast in water, slower in molasses, geologic time-scales in the mantle. To answer your question: if the pressure gradient is low, but the viscosity is very low, then the intrusion can still occur quickly. But the final state will always, as you say, minimize gravitational potential energy.
- I don't think of glaciers and rivers as intrusions, although I suppose you could think of them that way, always being pushed downwards ... I guess the density of air is so low that I don't think about it as much as, say, rock.
- Awickert (talk) 02:02, 25 February 2009 (UTC)
- That is an excellent summary of the 'good theories of intrusions' and your input is much appreciated . . . I have had my quota here I suspect and have some catch up reading to do after 38 years . . .
Importantly, the Wiki aim must be to provide good reliable information to a range of users, including those not so interested in acaddemic discussions.
So . . .
can we now EDIT the Wiki 'diapir' definition page to give an extended definition of intrusions (using modern concepts, based on your words, above) and show where 'diapirs' fit into the suite (and refer to Mrazec).
Begin with an introduction on 'intrusion theory'(link in Wiki?) and the 'minimization of gravitational potential energy' . . . then explain the difference between (i) the usual Gulf-type mobile host salt domes/lava lamps and the continuum to(ii)the 'Flinders' brittle host model.
Finish with a comment on the importance of mobility and density (bouyancy) in the intruding medium, noting that an intrusion can be more dense or less dense than the host.
Introduce the idea of structurally-contolled passive/permitted (Flinders) intrusions to contrast with energetic/bouyant salt domes.
Add a photo example of a brittle-host intrusion, as well as the lava lamp.
Perhaps there are other experts (MIT?) who can contribute to a concise, useful, and high quality Wiki entry?
Here is the link/reference to the 1975 work on diapirism in the Flinders Ranges that prompted many of the comments, above: http://web4.library.adelaide.edu.au/theses/09PH/09phm928.pdf Tjmount (talk) 23:04, 25 February 2009 (UTC)
- Hi to both of you. I keep meaning to contribute to this discussion, but every time I get here there seems to be a new comment. I don't know that much about igneous intrusion mechanisms but I have worked extensively on salt tectonics in both extensional and contractional regimes. The behaviour of salt varies hugely depending on the purity of the salt, its water content, the temperature and the strain rate. The same body of salt may completely detach faulting above and below in one phase of extension but simply fault straight through during a later, more rapid, phase of extension. I've often been told that salt can't possibly fault because it's a newtonian fluid, but there are plenty of examples of faulted salt out there. At high strain rates, it will behave elastically with fracturing and brittle faulting just like any other rock. There are some amazing examples of salt thrust imbricates in deepwater Angola  for instance.
- As regards active versus passive salt diapirism there's plenty of evidence for both mechanisms. The whole Triassic sequence in the Central North Sea is a record of continuous movement of the underlying late Permian Zechstein salt with clear evidence (caprock development) of near emergence of the highs in the classic passive 'downbuilding' model. In the same area there is evidence of a much later phase of salt structure rise and piercement associated with two phases of inversion, active diapirism at least at the start.
- I'll do my best to help out as a new version starts to appear. Mikenorton (talk) 22:13, 26 February 2009 (UTC)
Great to have your input on this . . . thinking it through last night, what has really troubled me over the last 38 years is how the classic model for a diapir has become the 'US Gulf bouyant salt dome with bendy host' while the original Mrazec definition (1908, 1912) relates to breached (ie Greek 'pierced across') anticlinal cores under strong lateral compression and with a brittle host/overburden.
What we all need is a simple generic definition of 'an intrusion', then see how the Roumanian 'Type Location' diapirs fit into the classification . . . then ask if Gulf domes are really diapirs at all.
- Edits sound great. Unfortunately, I'm the one of the three of us who has never worked on salt tectonics, so while I'd be happy explaining the physics, I don't feel qualified to start talking about them: the only ones I'm familiar with are the Gulf ones, which seem to be the "classics" that are explained here.
From what I know, Diapir is more of the R-T instability one (feel free to correct me), so maybe the above info would be better on Salt tectonics.Awickert (talk) 06:12, 27 February 2009 (UTC)
- [Update] Clearly I have reading issues. So the definition of diapir has changed, huh? That makes me wonder which definition should be here; maybe Mikenorton could help.
- Mikenorton, do you have a good copyright-free diagram of the stress-strain constitutive relationship for salt? I feel like something like that would really help salt tectonics and explain (by referencing different parts of the diagram) where (a) tectonic stresses would be most important, (b) where R-T-instability-style intrusion would be most important, and (c) what the brittle failure point of salt is. If we could add temperature (is it a simple Arrhenius relationship?) or water content, that would be great too. We could use that to outline different dominant mechanisms for salt intrusion and deformation. Awickert (talk) 21:43, 27 February 2009 (UTC)
- Unfortunately I don't have a diagram like that. There's a paper here  by Janos Urai and Chris Spiers that has a lot of information about deformation mechanisms in salt, although it concentrates on 'solution -precipitation' a form of pressure solution, which does produce a linear viscous (newtonian) type of behaviour, in contrast to dislocation creep, which is non-linear. Another one here  from the same group also discusses flow laws and mechanism for 'dry' & 'wet' salt. I need to add a section on salt rheology in the salt tectonics article, I think. I can't find much on the brittle/elastic behaviour of rock salt during natural deformation (as opposed to mining induced) though. Mikenorton (talk) 22:48, 27 February 2009 (UTC)
As a small team of three, we may be approaching the limits of our expertise, reverting to personal comfort zones, with the potential for interesting but endless discussion?
We are wandering.
I began by suggesting that Marazec's original 'diapir' term may have been hijacked by Gulf Coast users (while not allowing for other types of intrusion) and admitted to needing a lot of help on the physics. AW's claimed strength is 'physics rather than salt tectonics', while MN is focused on salt rheology.
Lets' get back to defining the problem here and devise a plan of attack . . . as noted, above, the aim is to provide the average Wiki user with a fair and useful definition of 'diapirs'. So, is our work plan to: (i) to go back to Mrazec's original definition, (ii) review subsequent use, (iii) consult with current experts (eg Google . . . John K Warren, Mark G. Rowan, Vendeville & Jackson), (iv) draft a definition (and one that allows for uncertainty in scope), (v)obtain agreement, (vi) publish & maintain in Wiki?
- Definition. Thanks for the focus. It looks like the Mrazec original definition is being kept today. In terms of the dictionary definitions I've been able to find: Google's list is all right, Britannica's is really good, as long as you're OK with closing their "please pay" box. It says that diapir is the general term, as Tjmount tried to say and I was too dense to get, and can include salt domes, dikes, and anything in-between, depending on the tectonic environment. So I think we should re-structure the article to say that. I'll start. Awickert (talk) 22:45, 1 March 2009 (UTC)
- Great, that sounds like real progress. Agree that Britannica definition is very good, but I would modify the section that attributes the rise of mobile material (implied = 'more dense') to 'lateral forces'. Again, I contend that (some) vertical rise is possible for a more dense but mobile source material into a lighter cover, especially into a brittle fractured overburden under tension.
And see my earlier (above)suggested layout for the Wiki note, but add 'the effect of regional compression'. Also need some nice diagrams, as well as the cursed lava lamp.
Later continuation of discussion on causes of diapirism
As a worker in salt, extensional, and inversion tectonics, I thought I'd add my two cents..haha...
Do read papers by Vendeville and Jackson 1992 - The rise and fall of diapirs during extension...they are quite helpful and easy to read... This is another one that will talk about Trusheim (who first tried to understand this concept)www.beg.utexas.edu/presentations/2002_presentations/vendeville_gcags02.pdf
For all of the physics people in this group...Weijermars et al., 1993 - Rheological and tectonic modeling of salt provinces...that paper has loads of those diagrams you mentioned before...
The defination of a diapir is not as simple as it may seem, because it refers to both a process (diapirism) and a geologic structure (salt diapir- a salt mass that has flowed in a ductile manner to discordantly pierce or intrude its overburden; which again is really still refering to a process). Perhaps you may just want to call the article salt structures...then you can include diapirism, along with other structures like rollers, anticlines, salt glaciers, and allocthonous salt sheets, etc.
Heres alittle bit about diapirism...if you already didn't know... Extension can create "reactive diapirs"...they are reacting to the extension by upwelling along the fault (this is because of gravity...less overburden because extension thins .....but the salt mass does not "pierce the surface"...once diapirs the reach the surface, they are called "passive diapirs"...they will then continue to grow (downbuilding) until the source layer is depleted...It will be an "active diapir" for the hot second between reactive and passive when it breachs the surface...(see [] for a neat animation)
It's also important to mention, like the previous author who mentioned the North Sea, that salt tectonics, depends upon a number of parameters, all of which can vary depending on the region ( and even within the region). For example, salt structures in the Gulf of Mexico may look similar to salt structures elsewhere, but may be created by completely different tectonic processes...detached vs. basement-involved...but thats a conversation for later if interested..Or it can be a combination of both (eastern canada, morracco, brazil, North Sea, etc)...thats when it gets interesting...haha
I will try to write more with references in here. You can then decide which you would like to include in your article...
sorry I just read this...---->>>""In the process, segments of the existing strata can be disconnected and pushed upwards"") NOOOO!!!!!!). This is not what happens during diapirism. This implies that salt can do the pushing...try nailing a rubber "dowel" into a board. Salt is weak...it can only "displace" thick overburden when there is a regional tectonic trigger...I think that you should talk about diapirism in subheading of this article. It is very important, and also fairly simple to understand.
- Thanks - sounds like you know what you're doing - it would be great to see you contribute to the article. The issue with the papers is that the figures aren't free, so they can't be in here - bummah, eh?
- As to the "only tectonic forcing", as far as I know, salt domes can form completely passively as well, just due to a density instability. That's why the rubber/board analogy wouldn't work: over the proper time-scales, it's like a less-dense, less-viscous fluid beneath a more-dense, more-viscous layer. In this case, it would be pushing rock out of the way: even though it flows much more rapidly, if there is nowhere to go, it must go up and displace material above it. Awickert (talk) 00:19, 22 May 2009 (UTC)
- Salt domes can form in a number of tectonic settings and on both passive (i.e. GoM) and active margins (Med Sea). The regional trigger, may not be a basement involved tectonic event such as crustal extension or shortening, but can be caused by differential loading realated to thermal subsidence or like in the case of portions of the GoM, the prograding lobes of the Mississippi delta. If the rate of salt rise exceeds the rate of deposition, a diapir, or "salt dome" can form...if the rate of deposition exceeds the rate of rise, the salt will remain buried and will not move. Now, in the case of the latter...the only way to get it to move when it is that deeply buried, would be a regional tectonic event such as extension or shortening. At that time, the salt will begin to move only if it has somewhere to go. It cannot just rise through a thick (kms) overburden without some type of faulting. Now all of this is assuming that we are talking about pure rocksalt...Salt interbedded with clastic rocks will behave more brittlely, and therefore be less likely to flow in any tectonic setting...
- --->"over the proper time-scales, it's like a less-dense, less-viscous fluid beneath a more-dense, more-viscous layer."...It is interesting that you mention this as I was discussing this with collegues not to long ago...This idea is based on Trusheims idea that all rocks behave like viscous fluids. This may be the case at depth (i.e. around the Moho), but near the surface, the crust behaves brittlely, meaning that rocks break, rather than flow like an extrememly viscous fluid.
- Thanks for the comments - yes, this is what I mean, that passive salt structures move upwards. They will, however, move upwards even if there is a high rate of deposition (though they will not pierce the surface), so I only half agree with you there. I don't think anyone calls dense material deposited over light material a "tectonic trigger", so that was my issue with your above comment; it seems like we do generally agree after your newer comments - so good. And I agree with all of your tectonic comments.
- Yeah, the time-scales thing is a gross oversimplification: the rock above the salt will fracture. What I was trying to get at is the density instability that allows the domes to form. But the large-scale pattern of Rayleigh-Taylor-like salt domes fits this oversimplification well. The issue with your analogy is that it is a long-term density instability that drives the formation of passive-margin salt domes: nailing a piece of salt into a rock would likewise fail, but put a ton of salt under a ton of rock, and with enough time, salt domes will form. This is why the viscous fluid analogy often does work: on a coarse scale, the patterns of faulting can be smoothed into what looks like fluid-like behaviour.
I just want to point out that the moon Enceladus belongs to Saturn, not Jupiter. I don't know whether the author meant "Saturn's moon Enceladus" or "Jupiter's moon Europa". Probably the latter, judging from the Europa article. —Preceding unsigned comment added by Poolio (talk • contribs) 15:37, 13 October 2009 (UTC)