Imagine dear reader a bleak winter morning in the fair City of Edinburgh in or about 1924 (in reality it was probably Durham) with rain driving from the west almost horizontally through the grey streets to sweep out over the Firth of Forth driven by a wind that could persuade even a Scotsman to wear trousers. Arthur began his daily ritual of a good clensing bowl of porridge little suspecting that this was to be a day of epiphany; not in the strictly eclesiastical sense though no less divine for there in his seething convective bowl of porridge he was startled to behold the Atlanic Ocean as clumps of yet uncooked oats rafted across the surface of the boiling water. And so (as Arthur's uncle Sherlock was wont to say) was formed the League of Porridge Stirrers based on a concept so powerful that even today it has become a Mantra to be chanted by all should they ever hope to get a grant to study any aspect of the crustal behaviour of the earth. If you have forgotten how it goes sing along with me. If you cant carry a tune dont worry there is only one note. Yeh the mantle of the earth is a seething convective quivering mass that doth rend continents and scatter them asunder by its convective power.
Arthur's model can be summarised by the following illustration (pl1.gif)..
see big copy
Dont get me wrong, it was a great idea and given what was known about the earth on that bleak Scottish morning it deserved great credit but today we need to consider it with an eye to its real value or lack thereof and Im sure that Arthur as a good Scotsman (OK I confess, I dont know for sure if Arthur was an indigenous Scot but it fits the artistic theme) would agree that true value is what we should seek.
Consider then that convection as a means of driving the lithospheric plates hither and yon about the face of the earth has some fatal flaws. Here is a good one. OK one of these convection cells just happens to start up - for whatever reason - under an as yet unsundered continental plate and just rips it apart as illustrated below (pl2.gif).
see big copy
You will agree that this is a notable manifestation of instability yet after its birth in such a paroxism of instability the rift in the plate immediately settles down to become one of the most incredibly stable geologic features on the face of the earth which may continue for periods in excess of 100 million years. This is no less remarkable because the rift is also one of the most vigorous of geological features.
Next consider that if two cells start up under a continental plate and create two parallel rifts then unless a subduction zone also starts up within the continental plate between the rifts or at the edge of one of the rifts then the centres of the rifts must move apart (pl3.gif).
see big copy
This means that if the energy supply to each ridge is to be maintained then the convection cells will have to go along with the centres of the rifts and move apart in exact lock step with the rifts. This appears to be what we have in the case of the spreading ridges on each side of Africa in the Atlantic and Indian Oceans. So you tell me this, which is in control, the spreading ridge or the underlying convection cell? Im sure that Arthur would agree that it is the ridge that is in control and it is dragging this meek and mild convection cell across the face of the earth like a naughty child being pulled by the ear. Is this the same mighty convection cell that only a little while ago did the Arnold Schwartzenegger thing and ripped the plate apart. Come on, you have to agree that there is a glaring inconsistency here.
So we appear to have a rule here that all proposed plate tectonic mechanisms must obey which I will call Retic's RULE NUMBER 1 and this is:
NUMBER 1: THE ENERGY SUPPLY TO THE SPREADING RIDGE MUST FOLLOW THE RIDGE WHEREVER IT MAY CHOOSE TO GO EVEN IF IT COMMITS SUICIDE IN A SUBDUCTION ZONE.
In other words, the spreading ridge is in complete control of its own energy supply.
Here is another conundrum for the porridge stirrers. Hawaii is a very compelling example of a mantle plume over which the Pacific Plate grinds inexorably so that if the plate is being carried by a flow of the mantle how does the mantle contrive to avoid shearing off the top of the plume or just simply stirring it up right down to the core. It might appear that the only way in which the plume could remain in tact would be to go with the flow and head for Japan along with the plate and the mantle. To me it seems inescapable that the mantle beneath Hawaii is notably stable in order to support the plume pipe for extremely long periods despite the overlying turmoil of the plate passing by. There have been some attempts to get round this by having the convection confined to the top most layer of the mantle, or even to the 100 km thick partial molten zone or layer. This certainly wont work if you expect to create a convection cell that extends from the East Pacific rise to Japan and have it be no more than 100 km in vertical extent. So what I call the ball bearing concept came along, this being that the partially molten zone is convecting locally in the form of a multitude of cells. This sounds reasonable because a wide flat bottomed porridge pot would almost certainly do this. A good point is that this would serve to - as it were - act as a sheet of ball bearings over which the plate could ride but where is the drive towards Japan in all this? So at this point the porridge stirrers invoke a couple of ideas to help the plate on its way. These are illustrated in the following illustration.(pl4.gif)
see big copy
At the young end of the plate the stirrers invoke sliding down the spreading ridge as a means of persuading the plate to head off on a journey of 8000 km to Japan. It seems it must be downhill all the way from the Galapagos Islands to the Japan Trench. Well perhaps it is, although of course the plate which this force is pushing towards Japan is spherically curved so that the concept of downhill is a bit hard to grasp when the push at the east Pacific rise would be a vector pointing into deep space at the Japan Trench (to be fair that is a cheap shot). However, let us not forget the dear old Atlantic because if the push as the plate slides off the mid atlantic ridge is responsible for the South American Plate heading west, then the concept that the Andean Mountains are somehow downhill from the sub ocean ridge is more than my humble intellect can grasp.
I suspect that the porridge stirrers know very well that they are in trouble with the down slope push of the ridge so at the old end of the plate they invoke some assistance from the subduction zone in the form of a supposed pull. Now this one really offends me. We are asked to believe that the plate as it goes down the throat of the subduction zone spontaneously increases in density to the point where it exceeds the density of the zone into which it is plunging and becomes negatively buoyant and begins to sink thereby pulling the rest of the plate in with it. If you believe this then you believe that the handle on your grandmothers old fashioned clothes wringer was turned by the clothes passing throught the wringer rather than by your granny's muscular arm. There seems to be good evidence that the plate does increase in density but this implies that energy is expended to compact it and that the plate must experience resistance to entering the subduction zone from this compacting force. If the plate spontaneously compacts itself why does it wait till it gets to the subduction zone to do this. Come on! The plate's natural inclination is to float on the mantle because of its density being lower than that of the mantle. You can experience this piece of physics personally. First find a very deep swimming pool take a deep breath and swim downwards. This is hard work because your natural positive buoyancy tries to force you back to the surface - thank goodness. However if you are determined and get deep enough your body becomes compacted to the point where you become neutrally or even negatively bouyant and you no longer have to work to keep going down. After this you can just stop working and if a rope extended from you to surface you could even pull something down with you. However, before you got to the depth of negative buoyancy you really had to work hard against the forces that were trying to get you back to the surface. So where does the plate get the energy to overcome the forces which are trying to force it back to the surface. Well the already compacted plate is perhaps helping a bit by pulling but there appears to be a deficit here otherwise how can the whole thing start when there is no already compacted plate to provide any pull. Also we should never forget the Atlantic where there is no subduction zone to provide a pull so how does that South American plate move westwards particularly when it is colliding with the East Pacific plate coming the other way. We might also ask why are there no subduction zones in the Altlantic?
It seems that we need a few more of Retic's Rules so here is
NUMBER 2: ANY PLATE MECHANISM HAS TO EXPLAIN HAWAII.
NUMBER 3: SUBDUCTION ZONES ARE NOT SELF SUSTAINING.
NUMBER 4: DOWNSLOPE PUSH BY RIDGES MAKES NO SENSE IN THE SOUTH ALANTIC/SOUTH AMERICAN PLATE.
The idea that the Atlantic spreading zone is fed by hot spots of which Iceland is put forward as the prime example, needs some explanation by its proponents. Consider that if the spreading on either side of Africa implies that either one or both of those spreading centres are moving over the mantle then the ridge in the north Atlantic must also be moving - presumably westward. For Iceland to remain on the ridge it too must be heading westward so that if Hawaii is stationary in the mantle the distance between Hawaii and Iceland must be changing. So here we have two hot spots one of which seems to be fixed in the mantle while the other is mobile in that same mantle and more remarkably, able to regulate its motion through the mantle so as to always remain directly centred on the mid Atlantic ridge. This Icelandic hot spot is not only mobile it is remarkably intelligent as well.
OK let us say Iceland is not moving then the whole of Europe must be moving and on a globe this movement will be south south east. Correspondingly the north American plate must be exactly matching the movement of Europe but in the opposite direction. Now surely the resistance to the motion of Europe and N America over the mantle can not be so remarkably equal that they achieve a miraculous equality of motion and leave Iceland dead in the middle. Is this reasonable? Personally I think it would be miraculous, so the mid Atlantic ridge must be moving over the mantle along with Iceland. Consider this, if material is moving towards the ridge crest from whatever direction, is it not also reasonable that when that material gets to the ridge it may decide to move along the ridge just as easily as emerging at the crest of the ridge. If a northerly flow developed under one section of ridge and a southerly flow under another section the collision of these two flows would leave no option but eruption to surface. This would be a zone of convergence of mass and heat which would be a hot spot for all purposes and it would stay with the ridge because the ridge would be its parent.
So we now need Retic's rule
NUMBER 5: ICELAND MUST BE MOVING OVER THE MANTLE SO HOW CAN IT BE A DEEP SEATED HOT SPOT?
So its a question more than a rule, so sue me.
Now here is another thing to ponder. Pacific spreading rates are approximately 4 times greater than those in the Atlantic and you might think that this should be one of the most important features that any plate tectonic mechanism should explain. It appears that we can infer that the energy supply to the spreading in the Pacific must be approximately 4 times greater than that in the Atlantic but why? If the porridge stirrers have an explanation for this I have not seen it.
Rule NUMBER 6: ANY PLATE DRIVE MECHANISM MUST EXPLAIN THE DIFFERENCE IN SPREADING RATES BETWEEN THE ATLANTIC AND THE PACIFIC.
OK one more. Back arc spreading ie. spreading in the small ocean in the back arc basin behind a major island arc, seems to be happening and the Sea of Japan is a good example but if its driven by convection how does the puny cell below the basin overcome the apparently massive compression which is happening at the edge of the basin in the subduction zone of the arc? (pl5.gif).
see big copy
Just as in the case of the Andes and the South American plate we have a plate being driven against the flow of the incoming oceanic plate and this seems to be a bit too much to ask of the old porridge pot.
It seems to me that the porridge stirrers league have ignored the wise words of William of Occam (or Ockham) who was a good old fashioned Yorkshireman who could not abide crap. He said "Entia non sunt multiplicanda praeter necessitatum" or "Entities are not to be multiplied beyond necessity" or "Any true scientist will search for the explanation which involves the minimum of complication before going on to introduce unnecessary complications". The League's insistance that convection is essential, leads them into all manner of twists and turns that get further and further away from the edicts of the good William.
Now I hear you saying that its all too easy to be a critic and pull down great builders like Arthur. I agree and so Im going to provide you with a model for plate motion which overcomes the problems we were just discussing and a few more beside.
Lets get right to it. The Retician model is as follows. THE PLATES ARE DRIVEN BY AN OSCILLATORY ELEVATION OF THE OCEANIC CRUST OF THE EARTH GENERATED BY AN ALTERNATING CHANGE OF STATE AND CHANGE OF VOLUME IN THE PARTIALLY MOLTEN ZONE MODERATED BY A SEASONAL OSCILLATION OF WATER AND ATMOSPHERIC LOAD BETWEEN THE CONTINENTAL AND OCEANIC SURFACE OF THE EARTH.
So is there a seasonal oscillation of load between the continents and the oceans? Indeed there is. But I hear you complain that this mass would be so insignificant that it would not have any measureable effect on the mass that is the earth. Not so. The atmospheric mass that moves into and out of Siberia each year causes the whole earth to wobble significantly. I only want to massage the outer skin of the earth with the movement of water load and this seems to be a much more modest requirement than causing the whole enormous mass of the earth to gyrate erratically like a rag being shaken by a dog which is what the atmospheric mass clearly achieves every year.
The idea is this - in the southern hemisphere summer the sun shines on a great deal more ocean than it does in the northern hemisphere summer and is thereby able to remove more water by evaporation from ALL the worlds oceans than in the northen hemisphere summer. I have lived in Australia and at great expense and loss of skin worked under that sun. Seeing how fast the sun can change lakes into deserts while trying to also fry a few mad dogs and Englishmen creates a great impression. How much water? About 60 cm per year from the whole of the Pacific and presumably from all other oceans as they are all connected. Sure about half of this goes straight back into the oceans as rain but a significant mass gets locked up on the northen hemisphere continents as snow and ice and as atmospheric load on the Siberian area during that period. I can also attest to the snow and ice phase of this cycle having lived in Canada as well, were global warming sounds like a cruel joke when a -30 degree C day can kill you even faster than can that Australian sun (and such days seem no less common now than they did 26 years ago).
OK so we remove water from all the oceans during the southern hemisphere summer and this unloads all of the oceanic crust of the earth during that period. It should also be understood that the continents of the southern hemisphere are also being unloaded by the same sun so that the total area of crust that is unloaded is all the oceanic area plus the continental area in the southern hemisphere. Remember this because it is important later.
Now if crust is unloaded does it not rebound and rise? It certainly does, as seen in the Fennoscandian uplift of the Baltic area and the rise of the James Bay area of Canada. However these rebounding zones can not provide a model for what I am proposing other than to make us comfortable with the idea that the crust can rise at rates up to 10 cm per year. OK if a large area of crust rises will it not have to extend to match the greater area of the elevated spherical surface as in the following illustration? (pl6.gif)
see big copy
In fact the global total of lateral extension per year is about 10 cm and it would require no more than about 2 cm of vertical rise of the oceanic crust to achieve that amount of lateral extension.
The problem with the well known glacial rebound zones such as Fennoscandia is that those zones are compensating the rebound by means of the lateral inflow of material within the mantle. If we want to raise the whole of the Pacific plate by about 2 cm in one year there is no way that this uplift could be compensated by means of lateral flow within the mantle This would be like applying a gigantic toilet plunger to the pacific subcrust and the inflow of material at the edges would be truly awe inspiring. No that cant be it. If uplift is to be achieved over a large area such as the Pacific plate then compenstion for that uplift has to be achieved by an increase of volume of some part of the upper mantle. Could such a change of volume take place? Well how about the partially molten zone located at a depth of about 100 km and about 100 km thick. In this zone the rock is agreed to be in a state of transition from soild to liquid so that if the load on this zone decreases then more of it will become liquid. My education in basic physics suggests that a liquid usually requires more volume than the soild that was its origin so that this appears to be a way in which the upper mantle volume could increase and do it very quickly because the rocks in the partially molten zone already contain all the thermal energy required for melting. It seems reasonable that such a change of volume by fusion could elevate extensive crustal areas by as much as 2 cm within a year. And of course when the load returns this process is reversible so that the volume would decline and the crust would move down again but if during the uplifed phase the crust had achieved extension by fracture and infilling then it would no longer fit the smaller spherical surface onto which it would be forced and so would have to get rid of the extra length which is where subduction comes in. Im sure that by now you can see that in this model, spreading and subduction alternate in phase with the unloading and reloading of the oceanic crust.
Now how much water must be removed to achieve the desired uplift of 2 cm? Well I estimate - based on the densities of the materials in the upper mantle - that it would only require somewhere between 20 and 50 cm which is in the ball park of what is evaporated each year. But can this load removal provide the energy to lift a whole oceanic plate - no. But the thermal energy in the partially molten zone can provide that lift and the load simply moderates that energy. However as the process is cyclic and the parially molten zone would cycle between more and less melting, the thermal energy in the partially molten zone would also go through a cycle but with a net loss due to the need for energy to power the extension of the plates and to forcing the excess length down the subduction zones. OK there is a great deal of room for argument here.
We now come to the outstanding consequence of this model which is illustrated in the following diagram. (pl7.gif)
see big copy
This consequence is that rate of spreading will be controlled by the length of crust being unloaded to achieve spreading. So naturally the Pacific spreads 4 times faster than the Atlantic because it is 4 times wider and Rule #6 is obeyed. Show me any other plate drive model that comes even close to matching the elegant simplicity of this explanation of relative spreading.
Consider that the greater part of what might be called classical geological processes such as errosion and deposition derive their energy from the sun so that while this plate drive concept does not derive its energy from the sun it can not operate without solar input so that it is a classical or surface oriented model which allows the crust to play a major roll in its own destiny rather then being no more than the scum floating on Authur's porridge. To me this is more in line with the words of William of Occam than to ascribe plate motion to the whim of the supposedly convecting mantle.
But now you should be asking what about the Southern Ocean between Australia and Antarctica which is no wider than the Atlantic but which is apparently spreading as fast as the main spreading zone in the Pacific? No problem, because as noted above, in that hemisphere the unloaded crust also includes the continental areas (i.e. they are being subjected to the same solar radiation as the unloading ocean and the load is being placed only on the northern hemisphere continents) so that the extension seen in the Southern Ocean is a reflection of the extension if the unloading of an area which effectively includes the WHOLE of the Southern Hemisphere which is an area comparable to the area of the Pacific and therefore spreads at a rate comparable to the Pacific.
Still not convinced? Well how about this. If spreading and subduction alternate seasonally with spreading taking place globally during the southern hemisphere summer due to the unloading of all the worlds oceans in that phase, and the global reloading of those oceans during the nothern hemishere summer as snow and ice runs off the northern hemisphere continents then the earthquake activity associated with subduction should peak in the nothern hemisphere summer. (I do love long sentences) The following illustration my surprise you. (pl8.gif)
see big copy
I took two independant compilations of earthquakes and added them up on a monthly basis. The results shown here demonstrate that if you want to go to LA or San Fransico the best month is February because both sets of data show that there is indeed a seasonal variation of earthquake activity and that it is at a minimum in February and it peaks in the middle of the Northern Hemisphere summer as predicted by the Retician oscillatory model. Notice that these compilations relate to consecutive periods 1904-1952 and 1953-1965 and that curiously the more recent compilation found more events in 13 years than did the earlier one in 48 years so that clearly the author (Rothe) was using different criteria for including events than were Gutenberg and Richter, but even so, both show a seasonal effect peaking exactly where this model says it should be. It might also be noted that the rise to the peak is sharper than the decline (if you agree with my smoothing of the curve) after the peak possibly due to the more rapid nature of the reloading as the snow and ice melts off the northen hemisphere continents than the deposition of that load on the continents. Obviously there is a great deal more that could be done to analyse this possible seasonal effect in earthquakes but I find it very satisfying that the results come out exactly as the model predicts
So in this model we lift large areas of the lithosphere by 2 cm and thereby extend the lithosphere. If this lift also causes each point on the plate to move laterally as well as radially then we have the lateral movement of the whole plate but this part of the whole idea is certainly its weakest point and it still appears that a lateral drive is needed.
However, the model has a number of other interesting consequences the first of which is that the flow of material in the lithosphere must be compensated by a return flow of material to the ridge. This has to happen in any model. In the Retician model, the lithosphere being injected into the upper mantle at each subduction zone is added to the partially molten zone and this same zone is also the source of supply of material to each ridge so that the partially molten zone forms a conduit by which material moves towards the ridges from the subduction zones in addition to its other roll of cycling or flapping the overlying plate up and down. As this material moves through the partially molten zone towards the spreading ridge it flows over the underlying mantle and gains heat from that mantle. This heat will then be carried to the ridges by the material flow. This is just the same as pouring a bucket of cold water on the crest of your house roof on a hot day and by the time it reaches the rain gutters that water will be notably hot having gathered heat from the whole roof. The flow of material towards each ridge will be gathering a part of the geothermal flux from the whole of the mantle over which it flows. The convergence of two such material flows at the ridge will cause a corresponding convergence of thermal flows with the consequence that the ridge will inevitably be the focus of unusually high concentrations of heat. As the ridge provides the sink for the material it ensures that the material and the heat continue to move towards the ridge and even if the ridge moves the supply of both material and heat will move with it so that Retic's Rule #1 is obeyed. In the Retician model Iceland is just a location where the ridge is sucking in more material and consequently more heat. This return flow though the partially molten zone can be summarized by the following illistration. (pl9.gif)
see big copy
This illustration might cause concern because this model has the two flows of material i.e. the overlying plate and the flow in the parially molten zone in close proximity but flowing counter to each other where the porridge stirrers league would have the mantle flowing in the same direction as the overlying plate. Well the porridge model still has to have a counter flow in it somewhere and if that flow is in the upper mantle the conflict between the flows would be essentially the same.
I will grant that this conflict of the two flows will tend to work against the overlying plate moving away from the ridge.
OK but here is another interesting consequence of this model. As the model suggests that spreading and subduction alternate rather than occuring at the same time then we may feel a good deal more comfortable about having a back arc spreading zone in close proximity to a subduction zone as the activity on each will occur at different times of the year with the spreading going with the global spreading in mid southern hemisphere summer and subduction occurring in the northern hemisphere summer. So why dosnt somebody test this? It should be relatively easy to observe.
Here is another spreading observation that deserves to be explained. The Red Sea shows two distinct phases of spreading about which the porridge model says nothing. Well its obvious, the Red Sea started life in the southern hemisphere and along with the adjacent continental land masses has now moved into the northern hemisphere. The Red Sea is too narrow to be spreading as an ocean so that its spreading will go with the unloading and extension of the adjacent continental crust. OK continental rifting in the southern hemisphere must occur in the southern hemisphere summer while a northern hemisphere rift must spread in the northern hemisphere summer as the load is removed from the continental northern hemisphere surface and put back into the ocean. So a rift which starts in the southern hemisphere and which is carried by plate motion into the northern hemisphere must undergo a reversal of the phasing of its spreading and must therfore undergo a period when it cant make up its mind which to do when it lies astride the equator. This may explain why the East African Rift has not yet developed into a rift which the ocean can invade because it is in transition between hemispheres and therefore relatively inactive at this time.
Of course this whole concept is based upon the present biasing of the distributions of continents and oceans between the hemispheres and clearly the movements of the plates changes that distribution. If the continental areas were symmetrically distributed between the hemispheres this would certainly reduce this effect. Correspondingly, if the continental distribution was tending towards higher concentration in one hemisphere this would enhance this effect and cause an acceleration of the movement of plates. It seems probable that this feedback would lead to vigorous but relatively brief periods of very rapid spreading interspersed with much longer periods of low spreading rates as the asymmetry of continental distribution waxed and waned which to me sounds a lot like the well established orogenic behaviour of the earth seen in the gologic record. It is also clear that changes in solar output and of the ecliptic angle would affect this concept. Another possible influence would be major glacial periods in which the load transfer process might be altered.
Looking back at the Rules we have
NUMBER 2: ANY PLATE MECHANISM HAS TO EXPLAIN HAWAII.
NUMBER 3: SUBDUCTION ZONES ARE NOT SELF SUSTAINING.
NUMBER 4: DOWNSLOPE PUSH BY RIDGES MAKES NO SENSE IN THE SOUTH ALANTIC/SOUTH AMERICAN PLATE.
NUMBER 5: ICELAND MUST BE MOVING OVER THE MANTLE SO HOW CAN IT BE A DEEP SEATED HOT SPOT?
NUMBER 6: ANY PLATE DRIVE MECHANISM MUST EXPLAIN THE DIFFERENCE IN SPREADING RATES BETWEEN THE ATLANTIC AND THE PACIFIC.
So it seems to methat #1 is taken care of by having the material flow carry the heat to the ridge. #2 Is dealt with by having the Mantle be a rigid non convecting body which can provide stable vertical conduits. #3 is no bother as I dont ask the subduction zone to suck the plate in behind it. #4 Perhaps the lateral motion could be achieved by the vertical uplift being greatest at the young end of a plate so that each point on the plate would experience some lateral motion as it rose. #5 Iceland is just a location where north and south flows under the ridge converge. #6 This follows naturally from the Retician model.
No doubt you are by now either a dedicated convert to the Retician model or you are complaining that the World Wide Web has a lot to answer for allowing such drivel to drift through cyberspace. Which ever view you take I hope you enjoyed this one sided conversation.