This is my point of view, as a former student of geology.

Supercontinents are divided into pieces by the apparition of rifting ( first phase of Wilson cycle). We can assume that the mechanisms which control the rifting are directly responsible for supercontinents decay :

1) Thinning of the upper crust which leads to the ascension of magma from the asthenophere (hotter so less dense), and finally provokes the apparition of a rift ;

2) Local enhancement of earth material fusion, which leads to the production of magma, and finally apparition of a new rift.

We can imagine that two systems are implied : first, gravitational modifications triggering the thinning of the crust by stretching the supercontinent, and then (or at the same time), reorganization or acceleration of hot material circulation in the mantle, leading to a massive ascension of magma, leading to a rifting phase. When looking at atmospheric datas, supercontinent spliting periods (or supposed so), generally correspond to a huge increase of carbon dioxyde accumulation in the atmosphere, which is correlated to an high volcanic activity. Therefore, we can assume that the modification of the internal earth activity is the most important actor in supercontinent spliting.

I hope it helped your reflexion. Regards.

A good synopsis on the subject of supercontinent cycles, including mantle processes related to formation and breakup was published by Nance and colleagues (2013), see here:

https://www.researchgate.net/publication/235834618_The_Supercontinent_Cycle_A_Retrospective_Essay?ev=prf_pub

Article The Supercontinent Cycle: A Retrospective Essay

I cast some doubt about the possibility that progressive accumulation of heat may be responsible of supercontinent destruction:in this case we could not observe large and old cratonic areas. Moreover around a supercontinent the older an heavyer oceanic crust sourrounds the continental crust and this should stabilize the supercontinent...

I suspect that gravitational effects in the Earth-Sun-Moon system may be involved in the activation of the buoyancy of the superplumes that start to destroy supercontinents infact the major LIPs are contemporary to the largest coastal placers, testifying a role for important mareal (= gravitational) effects.

What is wrong with the simple assumption that allowing a dozen or so plates to independently move in a "random walk" that there will be periods when they are close together as a group - and periods when they are dispersed?

Where does the assertion that oceanic crust is older than the continental crust it surrounds come from? Generally not the case....

To Keith: The assumption that the plate motion is random is probably wrong. I am more confident that the walk of the plates is "cahotic" . in deterministic sense, Chaos is often charaterized by "strange atrractors" and the very occurrence of supercontinents is by itself the prove that such attractors do characterize plate motions.

About the second observation I agree with you that when supercontinents start to divide some NEW margin is bounded by very young oceanic crust, But this is not my point. When continents collide in a unique mass if they are in the middle of a supercontinent they are not bounded by oceans, if they are on asupercontinent border, their oceanward portion has the age of the margin that is certainly older thas the spreading axis which drove the continent towards the supercontinent... .

The occurrence of the super - continent does not prove anything - as left to their own devices - in a purely random sequence of movements there will be times when the plates are closer to each other - and other times when they are dispersed. Statistically the "entropy" state of all being stable and equidistant will be the minimum energy condition.

I am afraid I don't follow your argument about the age of the oceanic crust - please tell us if you don't believe in subduction - as I don't wish to waste my time getting involved in an argument with NCGT.....

The distinction between “Random” and “Cahotic” behaviour is fundamental in our field as well as that of “Phase Space” and should be well understood. For those who are curiuous and like to spend some time to see things from a different point of view, the reading of this paper in Gcubed may be helpful and is not a waste of time:

https://www.researchgate.net/publication/235903990_Isotopic_evidence_of_chaotic_imprint_in_upper_mantle_heterogeneity

Article Isotopic evidence of chaotic imprint in upper mantle heterogeneity

Your original question presumes the correctness of Plate Tectonics, Subduction,

random, or driven motion of continents on a planet of relatively static radius

and surface area, and the validity of convection currents driving crustal motions

either directly or indirectly.

Since rock, and materials like concrete are relatively strong in compression,

but relatively weak in tension, rock is more likely to rupture in a tension zone,

whether that is on the top of the crust, or the bottom of the crust.

Study continents for the location of where radial tension fractures occur, and where lateral compression wave forms occur, or simply observe that rivers

are wider at the mouth than inland, and folded and faulted mountain ranges

tend to be in the interior of continents, and that the largest continents are

higher above the ocean level, and the smallest of continental fragments are submerged ( Under new Zealand ).

These are patterns of the results of the globally applied forces that change

the shape and features of continents and of the ocean floors, and change the elevations of both (differentially) based on the vertical and horizontal motion

of plastic and liquid rock inside the earth.

We must all remember that we believe what our mind perceives.

If we believe that the Earth remains a fairly constant radius, then we will

perceive what we observe as being consistent with our belief.

If we believe that all planets and stars grow larger at various rates, then

our observations and perceptions will be consistent with our beliefs.

The most difficult thing we will ever do is throw out our old beliefs, start

over and make new observations without being influenced by our old beliefs,

create new patterns from our observations, and form new theories.

A challenge, list ten attributes of continental crust, and list ten attributes

of oceanic crust, and compare them to multiple Tectonic Theories.

Dear all,

To come back to the original question, it seems that supercontinents cannot survive because their size exceeds a critical value above which subcontinental mantle temperatures are too high and lead to continental rifting. Mantle heat flow beneath continents are anomalously low (they act as if they were insulating - although thermal conductivity does not differ much from oceans), and this insulating effect implies higher and higher temperatures when continental size increases. This is still one hot topic, as you can see in the recent papers by Lenardic, Moresi and others (G3, october 2011; GRL june 2013) or Coltice and others (JGR, april 2010; EPSL, october 2012)

The excess temperature combined with lower average density of continents

could be part of why the centers of continents seem to remain

abnormally elevated.

In calculating the equilibrium pressure depth I came up with a below sea

level depth of 27.792 km for both continental and oceanic materials.

The oceanic side with the water converted to equivalent basalt height and

density is 25.334 km at an equivalent density of 3.007, while the

continental side has an equivalent height of 28.662 km at a density

of 2.6579.

Ocean : 76.179338

Land : 76. 180730.........................close

Cross checking: ( 0.7071 x 25.334 ) + ( 0.2904 x 28.662 )

= 17.9136714 + 8.3234448 = 26.237 1162 km ( up)

27.792 - 26.237 = 1.555 km below sea level is the equivalent

elevation of a planet of uniform density material and no

water cover expressed as a uniform density rock.

That uniform ( near surface density would be:

( 0.7170 x 3.007 ) + ( 0.2904 x 2.6579 ) = 2.904 748 610

which is close to my density equation constant = 2.904 696 055

The average near surface density would be 2904.7 kg / cubic meter

if averaged over the entire surface of the planet.

Then the cross check for equivalent continental crustal thickness:

27.792 + 0.870 = 28.662 ( ( e +1) / e ) = 39.206 km thick

27.792 + 0 = 27.792 ( ( e +1) / e ) = 38.016 km thick

The average is 38.661 km thick continents.

in prior calculations I had settled on a value of 38.825 km

thickness for continents, so 38.661 is fairly close to 38.825.

So, does this mean the Thermal insulative value of continents

raises the continents an addition 0.164 km above what

it would be based on by differential density alone?

I suppose we could look at the thermal co-efficient of expansion

of various rocks to see what the expansion ( loss of density )

is for various rocks, including the magma below the continents.

Interesting problem. This could enhance radial flow of material

out from under continents thus adding additional tensile forces to the

bottom of continents causing bottom up rifting.

The material flow out from under the continents supplies material toward

the mid oceanic spreading ridges. It just takes a few 10's of million of years to

get there.

Heat fluxes and tensile stress in the mantle are much more linked than is commonly thought: thermal diffusivity of olivine and orthopyroxene are strongly anysotropic at upper mantle conditions (Peterman and co worker 1996) thence, since both olivine and orthopyroxene are aligned by flow deformation processes, the upper mantle is expected to be thermally anisotropic. I presume that if a significant volume of the mantle beneath a supercontinent is structured heat fluxes and the consequent continent elevation, may be much higher than the 0,164 km reported in the calculations above. In northern Victoria land, for the switch of the local thermal gradient from 0,5 to 3 °C/km I computed a total uplift of of 1,6 km related to the change of olivine molar volume. (https://www.researchgate.net/publication/222372798_Cenozoic_thermal_evolution_of_lithospheric_mantle_in_northern_Victoria_Land_(Antarctica)_Evidences_from_mantle_xenoliths?ev=srch_pub)

The thermal effect, coupled with mantle structures at colliding continental margins and gravitational instabilities due to tidal effects on the crust , in my opinion play a role that is more significant than simple mantle convection in disruption of supercontinents .

Article Cenozoic thermal evolution of lithospheric mantle in norther...