I know that things are not as simple, but for the sake of argument I'll assume a supercontinent cycle of $\sim$400 Myr periodicity is active on Earth.

If that is the case, then my naive understanding was that roughly all of Earth's crust is reformed on this timescale, and information about the past is forgotten due to melting in the subducted zones.

However I stumbled upon a research article that cites "Most exposed continental crust that can be dated back to the Archaean eon (4 billion to 2.5 billion years ago) comprises of..." as summary of another article.

So there is a fraction of crust that survives this process and is several cycles old. Do we know the fraction of material in the crust that survived this process?

  • $\begingroup$ The title asks about crustal material in general but your quote talks about continental crust. Which are you asking about? $\endgroup$
    – Spencer
    Mar 3, 2017 at 0:42
  • 1
    $\begingroup$ Re my naive understanding was that roughly all of Earth's crust is reformed on this timescale ... This is not the case. $\endgroup$ Mar 3, 2017 at 3:20
  • $\begingroup$ @DavidHammen: That's, as you can read, is exactly what this question is about. Feel free to correct any wrong assumptions in an answer. $\endgroup$ Mar 3, 2017 at 11:39

2 Answers 2


There is two parts to this answer, first continental crust tends to persist and oceanic tends to be recycled. This is largely due to density, continental crust is less dense and tends to "float" compared to the mantle and is very difficult to push back into the mantle for remelting. Only a tiny fraction of continental crust ever gets subducted mostly as sediment on oceanic crust, so basically by piggybacking on subducting oceanic crust. This is why continent to continent convergent boundaries form mountains with no subduction, they can only push down by pushing even more material up, in effect packing together like wads of compressing clay.

Second continental crust is reworked by other processes, erosion, faulting, rifting, mountain building, metamorphic processes, etc. These are regional effects and thus more random, so how much they alter can vary quite a bit from continent to continent. Exposed un-reworked continental crust is rare but this is due to the accumulation of events reworking smaller regions and the tendency of sediment covering it up. There is much more unexposed old continental crust buried under the sediment, often referred to as the shield portion of the craton, Shield is the portion of the continental crust that is not sediment and is not tectonically active (even some shield is partially metaphorically altered, but this is where you find unaltered material) There is quite of bit of ancient shield (brown in this picture) although it still makes up a small percentage of the overall continental crust. Even then only a fraction of it has survived all the continental cycles and is completely unaltered since the first continents formed.

This image illustrates how much of the oldest continental crust is still around(purple).

enter image description here

Basically oceanic crust is guaranteed to be reworked eventually, check the age of oceanic crust, while continental crust has a more random factor involved meaning some in the more stable regions of some continents has not been reworked. That is why they are only small scattered areas.

  • $\begingroup$ Ah, so we could say that continental plate boundaries contribute next to nothing to the reworking, right? And also why is oceanic crust denser? Are the minerals formed under aquaeous condition different from those that reside on continents? $\endgroup$ Jul 2, 2017 at 22:49
  • $\begingroup$ Oceanic crust is much closer in composition to the mantle, it basically is cooled mantle, sorta (there is a layer in between). Tt spreading centers separating plates leave a gap magma(mantle material) flow up forming new oceanic plate, Just like the mantle it is denser than continental crust so it sits deeper in the mantle. Water flows down hill filling in lowland. It helps if you remember the planet was once entirely molten, so continental crust started as basically the lighter "scum" floating on a sea of molten rock. $\endgroup$
    – John
    Jul 3, 2017 at 2:42
  • $\begingroup$ All magma basically starts as oceanic (basaltic) it only becomes like continental material (more rhyolitic) if it has to burn through a continent to reach the surface and thus picks up a lot pf molten continental crust (silica) $\endgroup$
    – John
    Jul 3, 2017 at 2:44
  • $\begingroup$ continental boundaries contribute a lot to reworking, especially of oceanic crust, just not as much to continental crust, continental crust is more stable, basically all oceanic crust will get subducted and reworked eventually. The oldest oceanic crust we have is only 380 mya. $\endgroup$
    – John
    Jul 3, 2017 at 2:56
  • $\begingroup$ Please add a description of the first map, so that we can figure out how it contributes to the answer. For the second image, please verify whether it is basement rock or not. The surface rock of the Atlantic Coastal Plain in eastern North America is mostly less than 50 million years old, and not just isolated bits. $\endgroup$
    – Spencer
    Jul 4, 2017 at 12:14

The fraction that has survived this "supercontinent cycle" is very small and can be found in only a handful of places where it outcrops....generally as part of interior cratons in Canada, Greenland, etc. However, even within these Archaen shield rocks, the volume of "original crust" is itself very small, hard to find, and subject to controversy with respect to chemistry.

Original crustal material that dates back to a pre-tectonic (Hadean) regime would have to have been formed from cooling of molten material and would itself contain a very specific chemical signature that is itself, theorized but not known with certainty.

Given how small that volume is, and how difficult it is to access many of these rocks, it's unlikely that anyone will pony up the money for a deep seismological survey of their extent at depth, so, a solid number for volume will not be forthcoming soon.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.