PLATES L: AULACOGENS

The Congo Aulacogen.  Note abrupt change in trend of continental margin, precisely where the Congo River flows into the sea

Okay, so – yes – I DID say that my last effort (Triple Junctions) was la ultima; the last.  But then I got to thinking about how I had completely ignored one aspect of RRR triple junctions that has some fairly important geological significance, and moreover provides a neat name to drop at cocktail parties – aulacogen.

In plain English, aulacogen translates into “failed arm”.

Consider the Afar TJ, featured in Plates K.  I mentioned that its southern arm, the East African Rift Zone, might someday separate east Africa from the rest of the continent.  But what if it doesn’t?  What if the Afar RRR TJ runs out of steam before rifting is complete?  Then we will be left with a long, linear valley that after a suitable period of time will fill with sediment.  Also, while filling up it may, probably would, be, the locus of a major river system.

So, are there any now?  You bet.

Aulacogen was first used by a Soviet scientist, S. N. Shatski, as a distinctive term for long, linear accumulations of sedimentary rocks; as such of much interest in petroleum geology.  This was about 1960, and Shatski hadn’t a clue about plate tectonics, much less about RRR triple junctions.  Subsequently, however, geological mapping has revealed many such linear features; Google “aulacogen” and you will see.  Most are inactive; prominent modern examples include the East African rift zone and, perhaps, the Congo River basin.

As an aside – why do RRR triple junctions form?  Well, pretty clearly, they are a response to crustal stretching, sometimes possibly caused by initiation of a new mantle hotspot.  As to why such stretching results in three cracks – not four, or more – better ask a physicist.  I conjecture that it has something to do with minimum work; it is “harder”, in an energy sense, to form more cracks than just three – and a single crack won’t do the trick.  Consider cooling lava flows and ask yourself “why are so many basalt columns six-sided”?  Same reason, I suspect.  

 

PLATES K: Triple Junctions

                              AFAR:  AN RRR TRIPLE JUNCTION

I have left this topic for dead last.  I might pretend that it is of relatively little importance, but that wouldn’t be strictly true.  The real treason for putting it off  probably is that essential details of  this topic – triple junctions – have been the hardest plate tectonic concepts I have ever been forced to to wrap my brain around.  The heroes here – I almost said villains – are Dan McKenzie (Cambridge University, UK) and W. Jason Morgan (Princeton).  Together they wrote a seminal paper entitled:

McKenzie, Dan P. and Morgan, W. Jason, Evolution of triple junctions; Nature, v. 224, pp. 125-133, 1969

which I have seriously tackled at least three times over the past five decades.  I am still not certain I am quite up to speed but – what the heck – let’s go.

Okay, so a triple junction is a point at which three lithospheric plates touch one another.  Expressed otherwise, a TJ is a spot at which three plate-bounding structures meet.  Following Mc & M, we will call the plate boundaries R (a spreading center, as in the mid-Atlantic Ridge), T (an oceanic trench, our subduction zone) and F (a transform fault).  Thus, an FFF triple junction is a point on the earth’s surface at which three transform faults intersect, RRR is the juncture of three spreading centers; TFR then would denote the intersection of a subduction zone, a  transform fault, and a spreading center. 

Obviously, TJs are relatively rare, yet there are a few dozen of them currently scattered about.  No stable examples of FFF exist nor can exist; Mc& M demonstrate that, because of brute geometry, they are impossible.  RRR, on the other hand, is always stable; stays put and keeps chugging away.  Other species of TJ – there are a dozen of them - are stable under certain restrictive circumstances, unstable otherwise.

By stable, above, I mean that they maintain their geometric configuration.  For instance the famous Afar triangle is an RRR triple junction where spreading centers forming the Red Sea, the Gulf of Aden, and the East African Rift Zone meet at a point.  Being RRR, it will be around for a long time, and may succeed in prizing a chunk of East Africa away from the rest of the continent.

RRR is always stable, but another important example – the Mendocino TJ -  is stable only under specific geometric conditions.  The MTJ formed about 30 Ma years ago when a transform fault between the Farallon and Pacific plates intersected the margin of North America, which previously had been the locus of Farallon-North America subduction.  (See Tanya Atwater*).  This brought the Pacific plate into contact with North America, and resulted in the creation of an FFT triple junction.  One stability condition of FFT is that one F be collinear with T.  This condition was satisfied here because the newly created F (our San Andreas fault) was collinear with the trench still existing to the north.  (Too many acronyms, I know – draw yourself a diagram!)  A geometric consequence of this configuration is that the TJ migrates along the direction of the common, collinear boundary.  In the Mendocino example, the TJ will appear – seen from North America – to slide steadily northward, thereby lengthening the San Andreas system.  Now, however, at Cape Mendocino the smooth, this linear plate boundary has abruptly disappeared.  My guess is that the MTJ has become unstable and is in the process of breaking up; hence all the seismic excitement around Eureka, CA.

Don’t let me leave you with the notion that triple junctions are dynamic elements that move blocks of crust around willy-nilly.  They aren’t; like Euler poles, they are merely useful descriptive devises.  Geophysical forces, ultimately arising from  geothermal energy and gravity, are what move plates.  Triple junctions, like  Euler poles, are simply mathematical constructs that help us understand geological history.  

*Atwater, Tanya, Implications of plate tectonics for the Cenozoic tectonic evolution of western North America, GSA Bulletin, v. 81, pp. 3513-3536, 1970.

Plates J. PALEOMAGNETIC EULER POLES

 

There is an important aspect of Apparent Polar Wander and Euler Poles that I have not yet addressed, and I will attack it right now.  You might call this essay Euler APW, part 2.

Let us assume that continents (or tectonic terranes and  other scraps of crust) move in response to the sum of all forces applied to them, and that such sum of forces remains reasonably constant for geologically significant blocks of time.  Knowing as we do that APW is the result of the motion of crustal blocks rather than movement of the pole itself, it follows that APW paths should trace arcs of circles centered upon an appropriate Euler pole.  It also follows that abrupt changes in APW should be evidence of major changes in the sum of driving forces*, which ought to be evident in the geological record.  That is the reasoning upon which a paper by Bernie Housen and myself is based**.

I should emphasize that Bernie and I did not first come up with this rather clever idea; Richard Gordon and Allan Cox (both Stanford at the time, I think) get that credit.  Bernie and I simply tried to ride it to its logical geological conclusion.

Much to my consternation, when Bernie and I first ran this idea up the flagpole, damned few geologists saluted it!  I have been out of the field for a long time, so I can’t do an adequate job of even speculating about why this should be so.  Perhaps the sum of applied external forces acting on your typical crustal fragment undergoes very frequent and very important changes in some random manner. This would throw a large monkey wrench into the works, at least insofar as the correlation between APW and important geotectonic events are concerned.

*Driving forces?  Well, for instance “ridge push” and “slab pull”, occurring at spreading centers and some subduction zones, respectfully, act to move a plate in what we may call a positive direction, whereas friction along transform plate boundaries retard such motion.  When I last studied this stuff, the effect of the mantle underlying the plate was under debate.

** Beck and Housen, Absolute velocity of North America during the Mesozoic from paleomagnetic data, Tectonophysics, v. 203, pp. 33-54, 2003.

So, anyway – I thought that this line of reasoning is basic enough to be worth expounding – even if it turns out to be incorrect!

Next time, triple junctions.

PLATES I: Euler poles

After sober reflection, aided by several bottles of my favorite beer and a long patch of gloomy weather, I have decided to finish off my account of being in on – although contributing little to –  the plate tectonics revolution,  by attempting to describe some of its more esoteric and certainly least intuitive aspects.  Today I am going to describe something called an Euler pole. 

That’s Euler, as in someone who applies oil to things – not “Youler”, as its spelling seems to imply.  It is named after a famous 18^th century Swiss mathematician who first described it, Leonhard Euler.

In passing, this is not really a proper part of my history as a student watching the birth of plate theory with open-mouthed amazement.  I learned about Euler poles long after my college years; by that time I was with the USGS or maybe already at WWU.  But, still …..

Seems that Dr. Euler once showed that any motion on a sphere could be described as a rotation about a fixed point located somewhere else on the same sphere.  Thus, the opening of the northern Atlantic Ocean can be modeled as rotations, in opposite directions, of the North American and Eurasian plates about a pole located in the artic region. This explains why (at one time a puzzle) the spreading rate on the Mid Atlantic ridge is not constant, but rather increases from negligible in the Arctic region (where the Euler pole is located) to a maximum of several cm/year near the equator, then decreases as the plate boundary is traced further south.  As another example, the San Andreas Fault, which is the boundary between the Pacific and North American plates,  mathematically can be described  rotation about an Euler pole located in the north Pacific, off Labrador.  Every moving chunk of crust has its own Euler pole and, believe it or not, all that confusion turns out to be useful!

Don’t fall into the trap of regarding Euler poles as in any way active geodynamic elements; they are passive - merely useful descriptions.  Plates move in response to the sum of forces acting on them – not because of something an Euler pole somehow tell them to do!

As if this little essay were not confusing enough already, let me dump one more non-intuitive concept on you.  It transpires that Euler poles can be treated as vectors!  Hence the, Pacific-North American relative motion can be described thusly:

                                   _LatR_Long

where Lat and Long describe the location of the pole and R represents the relevant angular velocity (say, in degrees/million years).  This turns out to be handy:  if, for some reason you wished to know the relative motion between North America and Antarctica you could calculate it as the sum of the NA/Africa and Africa/Antarctica Euler vectors.

Sorry.  There is a bit more of this kind of stuff.  If I should quit now, please let me know!

 

 

PLATES H: Transform faults

 

I was all set to conclude this series of blogs about “growing up” (geologically) in the formative stages of the plate tectonics revolution with an attempt to describe an advanced stage – wherein it got all mathematical.  I have been dreading this task for days – not that the subject matter is at all boring; quite the contrary.  Fascinating and important it is, for sure, but darned hard to explain.  Maybe I can get Nick to do a lecture on it someday,  with pictures.  But in the meantime . . . .

I have inexcusably failed to pay tribute to one of the most important figures of the time: J. Tuzo Wilson, of the University of Toronto.  Tuzo not only named, described and analyzed the third kind of plate boundary, the transform fault, he was a nice man who took the trouble to explain it all to a sniveling graduate student.  Me.

So what, then, is a transform fault?  The short answer is that it is a plate boundary in the form of a strike-slip fault that connects two other plate boundaries; see the illustration above.  Not all strike slip faults are proper transforms, but the most famous are.  For instance, the San Andreas Fault can be understood as a transform that connects the East Pacific Rise at its southern terminus with the fracture zone delimiting the Gorda Plate to the north.  

Also, if you look carefully at maps of oceanic fracture zones you will see that in large part they consist of offset segments of ridge (spreading centers) connected by short segments if strike-slip faults.  An important aspect of this configuration, pointed out by Wilson (and others) is that strike-slip earthquake activity is limited to the segments between the bits of ridge, and is in the opposite sense that one would predict from the relative position  of the ridge segments!  For instance, the upper pair of apparently offset ridge segments appears to be right-lateral, whereas the seismic activity all is left lateral.

Try to figure out why, Patrick.

So, now I must attempt to explain such non-intuitive concepts as Euler poles, triple junctions, and plate motions as vectors in three-dimensional space.  Give me a few days!

PLATES G: Apparent Polar Wander

    Apparent polar wander and continental drift

Okay, here I am again, at Stanford, eagerly absorbing all the new stuff that eventually brought us to Plate Tectonics.  Possibly the most conclusive bit that came along at that time was what was (and still is) known as apparent polar wandering.  I started to explain it last time, but didn’t get very far.

By now you are comfortable with the notion that the magnetic direction locked in some types of rock formations can tell us where the pole was at that time, with respect to that rock unit.  Now, for reasons best left to Physics 101, it appears extremely unlikely that the pole itself moves much; consider the dynamics of a spinning top – or a gyroscope.  Thus, if paleomagnetic poles for a given continent or tectonic terrane appears to move, it is the rock mass – not the pole – that must be in motion.  Ergo, curves of apparent polar wander.  The rock mass is moving, not the pole.  We abbreviate this as APW. 

As I tried to convey last time, if there had never been any change in the configuration of the earth’s surface there would be no apw – every proper paleomagnetic study would return a paleomagnetic pole very near the spin axis – the geographic pole.  Polarity reversals were not precluded – but there would have been no apw.

However, it might happen that the whole earth moved, as a rigid mass, with respect to the spin axis.  This can be illustrated by setting a slightly asymmetric basketball to spinning on a table top - it will orient itself so as to position any excess mass on the equator.  Try it yourself.  This whole-earth repositioning, which might actually involve only what we now call the lithosphere, apparently does occur, but seems to be minor.  It would reveal itself as identical apw on every continent.

Well, of course, that’s not at all what was found.  Each continent as studied showed evidence of apw, and none of the resulting apw paths came at all close to coinciding.  The evidence was clear and, to my mind at least, completely irrefutable – the continents had moved with respect to one another, as well as with respect to the spin axis.  Moreover, if one graphically replaced the continents to what already was known as the Pangaea fit, these disparate apw paths tended to merge – to coincide.  That was it, end of story.  Wegener was right!

I should mention that much – most – of this work was done by people trained at the university of Newcastle upon Tyne, by a brilliant, enigmatic man named S.K. Runcorn.  Ted Irving (Australia, later North America), Ken Creer (South America, later Scotland), Ernie Hailwood (Africa, later England), Don Tarling (mainly Europe), and a bunch of others I can’t remember all were trained by Runcorn.  In the meantime, the USA was lagging far behind.  But in the next phase, we caught up.  On to (mathematical) plate tectonics!

 

PLATES F: Paleomagnetic poles

 

Continuing my recollections of the early days of plate tectonics…..

We are talking here about the state of MY comprehension circa 1960.  Allan Cox and Dick Doell had just published a Review paper in the GSA Bulletin, in which they did a fair job of outlining nuts and bolts, but were (in my later view) too reluctant to embrace some obvious but less orthodox conclusions.  Ted Irving’s magnificent book, which put it all right, was still a few years in the future.  Thus I was left to confront apparent polar wander more or less on my own.

Here, generally, is what I knew (c. 1960) to be true:

1)     1) To a first approximation the earth’s magnetic field, however  generated, has the same configuration as if it were generated by a very short, very powerful dipole magnet located at the center of the earth and oriented along the axis of rotation.  This was called the axial dipole field.  Temporary departures from this configuration were numerous, but small and – importantly – random.  This was known as the secular variation.

2)     2) Many rock types had the ability to acquire a permanent magnetic direction as they formed, then of retaining that direction essentially forever.

3) Note that magnetic directions are directions (vectors)  in three-dimensional space. They are described as follows.  The angle of the vector above or below the horizontal is known as its inclination; the angle of its horizontal projection  makes with due north is the declination.  Declinations are measured clockwise, hence range from zero (due north), through 90 (due east), to 180 (south) on around 360 degrees back to north.  Inclinations range from negative 90 degrees (straight up at the south magnetic pole), through horizontal (at the magnetic equator), to positive 90 degrees (straight down, at the north magnetic pole).  If this confuses you, check Butler's on line textbook.

4) Given D and I for a given rock unit, it becomes simple (given a dipole field)  to calculate where the pole would have to have been to produce that direction of permanent (we called it remanent) magnetization in that particular rock body.  Simply draw a great circle path from the rock’s location on a sphere in the direction indicated by D, go out an angular distance p, given by the formula cot (p) = tan (I)/2, and plop – there’s your paleopole.  If the D,I used represented an average direction representing a decent average of the normal variability of the geomagnetic, the pole thus determined would be regarded as a best estimate of the location of the geographic pole with respect to the continent on which the sampled resided at that particular time.  Such a pole position was called a paleomagnetic pole.  If it was pretty certain that the variability (called the secular variation) had not been properly averaged out, we called the thing a virtual geomagnetic pole. 

All this stuff was being hashed out as I watched from the bleachers, notebook a’tremble.  Obviously, it was fundamental stuff.  If all the paleomagnetic poles, of whatever age or location, consistently clustered near the present spin axis then the fixists had won; nothing interesting, like continental drift, could ever have occurred – and I might as well go home and work in the family lumber yard.  Another possibility was that the pole appeared to have moved – paleomagnetic poles tracing a path away from the present pole as the rocks sampled increased in age – but that the paths from all the continents coincided.  That would indicate a phenomenon known as true polar wander, which still is investigated today.

Naturally there was a third possibility, which I will take up next time.

PLATES E: Magnetostratigraphy

                  The black and white bars are the magstrat record

 

Let’s say that (preposterously) it was of the utmost importance for you to determine if a limestone in Kentucky, a lava flow in Iceland, a young pluton in the Cascades, and a shale in Wales were or were not of precisely the same age.  How to go about it?  Well, as you have guessed the old saying “no way, Jose” applies here, in spades.

 You might get lucky and discover that the two sedimentary units both contained fossils known to represent organisms that flourished briefly, spread all over the place, and then suddenly died out.  Such “index fossils” could establish the approximate age equivalence of the two, but of course says nothing at all about the lava and pluton.  The latter units might be of a sort that can be “dated” radiometrically, but of course the standard workhorse methods of absolute age determination don’t work so well – at all, usually – on sedimentary rocks.  What is needed is a time scale that effects the entire earth, is sensitive and precise,  and works on all types of rock.

Well, we’re lucky – such a scale does exist, courtesy of the geomagnetic field.  When the field undergoes polarity transition it (potentially) leaves a precise mark on every rock body forming at the time.  The only problem is that this new (in the 60s) tool has the property of a berserk traffic light: red, or green, for unpredictable lengths of time, separated by very short intervals of orange.  (Orange representing brief transition intervals.)

Thus:  If your two igneous units have the same geomagnetic polarity they might be contemporaneous, but, then again, they might not be.  However, if they have opposite polarities they certainly are not of the same precise age.  Given modern lab gadgets magnetostratigraphy can be applied to most sedimentary rocks, too.  Thus, given fossils, radiometric dating, and rock magnetism you might be able to establish the age equivalence of your four rock units quite precisely - +/- 5%.  As I said earlier; no way, Jose.   

Well, as you probably guessed already, the phenomenon of geomagnetic reversals stirred up a lot of excitement.  In the United States, the U.S.G.S. under Allan Cox and Dick Doell began a well-funded effort to place absolute ages on each of the latest dozen or so transitions.  They did the obvious: sampling young, fresh volcanic exposures wherever they might lie, and determining their polarity.  Then a Survey colleague, Brent Dalrymple, measured their absolute age.  Plot ‘em on a time chart, and a useful stratigraphic tool emerges.  At the same time, Don Tarling (Newcastle, England) was perusing the same goal.  Fortunately, the two trans-Atlantic data sets were easily merged and complementary.  After not much time the absolute date of perhaps the latest several dozen polarity transitions became very well established.

One exciting result of this work was that it suddenly became possible to measure the velocity of seafloor spreading.  Vine and Mathews had shown that the spreading seafloor was a type of magnetic tape recording of the history of geomagnetic polarity transitions.  Now we could determine when those transitions had taken place!  Thus, for instance: if a reverse-to-normal transition was located, say, 100 km from the ridge at which it had originated and was now known to be 5 Ma old, the seafloor had evidentially moved 10⁷ cm in 5X10⁶ years, or an average of 2 cm/year.  Neat stuff, eh!

So, what has all this have to do with my experience of the early days of plate tectonics?  Quite a lot, actually.  From George Thompson’s wonderful class and my own mulling things about I had decided to do an MS project in paleomagnetism.  It just so happened that at the time I was living about 200 yards from the back door of the Cox and Doell lab!  Through the good offices of George I obtained a key to that back door, and most every night – after everyone had gone home – I made paleomagnetic measurements.  How did I know what to do?  Well, Allan was a bachelor with time on his hands; often he would come into the lab late at night just to fool around.  So, I learned technique from Allan.

My first project was a lava flow, known as Table Mountain, Pliocene in age, in the west Sierra foothills.  It turned out to be Reverse in polarity.  I had verified polarity transitions!  I was excited!  I remember shouting at Allan (doing something elsewhere in the lab):  It’s R!  He smiled indulgently.

So next time, polar wander curves.  Again, time for my nap.