Sunday, November 07, 2010

Tectonic Similarity

Spin a globe, tilt it, and center the South Atlantic Ocean, with the coast of South America to the west (and left), that of Africa on the east (and right): see how the South American coastline to the west seem somehow similar to that of Africa on the east. Tilt north and rotate slightly west, to the center of the North Atlantic Ocean: the North American coastline to the west is, with slight imagination, similarly similar to that of the North African Atlantic coast. One more time: east and south, to the center of the Indian Ocean: The facing coasts of East Africa, India, Antarctica, and Australia; rotated back and forth a bit and imagine the coast lines as edges of spherical puzzle pieces, with Madagascar a gap-filling fragment. Might all of these continents, and Eurasia too, have once formed a collective megacontinent? This question has been around in one form or another for not merely decades but two or even three centuries. But, it wasn’t until the mid-1960s that the solid earth scientific community reached near consensus: The answer: a resounding “Yes”.
However, to reach the point at which geologists (stratigraphers, petrologists, volcanologists…), geophysicists (seismologists, paleomagnetists…), and paleontologists (specializing in fossils of plants and both vertebrate and invertebrate animals) could all agree, multidisciplinary results from each field had to be shown as mutually consistent and integrated. 

Orson Anderson (1972): During the last decade, a scientific revolution has stirred the geological sciences. Scientists from the diverse specialties of micropaleontology, geomagnetism, marine geomorphology, seismology, petrology, geochemistry, and tectonophysics are working together in a convergence of disciplines. A stage of development has been reached in which the research results arising from one of the disciplines are likely to affect the course of research work in many sister disciplines. … It is commonly acknowledged that a number of extraordinary experiments on a grand scale have produced this revolution in the earth sciences. A bold concept, plate tectonics, has arisen from these experiments and from many others.. 
Le Pichon et al. (1973): Plate tectonics is a unifying working hypothesis which provides a kinematic model of the upper layer of the earth. It can be used to make quantitative predictions about most phenomena studied by the different disciplines of the earth sciences. This leads to a reexamination of the large amount of geological, geophysical and geochemical data already obtained, in order to elaborate a precise theory of evolution of the earth. 
Cox (1973, p. 2): The earth sciences are currently in an intellectual ferment as a result of recent advances in the study of magnetic reversals, sea-floor spreading, and plate tectonics. The last decade, in which most of these advances were made, was a most unusual one in the history of earth science. Previously, during the century following the great works of Charles Darwin, the earth sciences were characterized by increasing specialization and divergence. Paleontologists, seismologists, geomagnetists, geologists, and marine geophysicists became better and better at what they were doing, but at the same time they seemed to have less and less to say to each other. Then, in a remarkable series of articles written between 1962 and 1968, the trend toward divergence was reversed and many of the main threads of geologic research were brought together to form the fabric of plate tectonics.

The plate tectonics revolution in the earth sciences is now a half-century old. It affected virtually every solid Earth discipline. Prior to the late-1960s, most geologists and geophysicists assumed that Earth’s continents and ocean basins were stationary: for example, the Atlantic Ocean between the Americas and Africa was inferred to be a static feature for hundreds of millions of years. And, mountains such as the Andes, Himalayas, Rockies, and Appalachians “were” largely a result of vertical movements initiated by phase changes or convection currents a hundred or more kilometers beneath.

Plate tectonics consists of a few basic concepts supported in varying degrees by a widely diverse set of observations. In summary, continents and ocean basins are not static, but occupy thin spherical plates in motion relative to Earth’s interior and each other. Continent-ocean boundaries do not necessarily correspond with plate boundaries, although those that do not nevertheless originated as plate boundaries. Thus, as had been proposed decades before, one consequence was that continents are, indeed, “drifting” away from or towards each other. However, because ocean basins are also dynamic, the term “continental drift” is inadequate as a name for the new theory.

Looking back on the changes of perception that active geologists, geophysicists, and paleontologists experienced during the revolution, one of its contributors, Allan Cox saw a resemblance to Thomas Kuhn’s description of previous scientific revolutions; so the new ideas could, according to Cox, be described as the “plate tectonics paradigm,” utilizing Kuhn’s terminology.

Consider the typical features of a scientific revolution (or, less dramatically, a “breakthrough”) which results in a new paradigm (or foundational theory): (1) A set of observations (or experimental results) are observed to be inconsistent with or incapable of description by prevailing models, if such models even exist; in some cases this inadequacy may be described as a “crisis”, at least retrospectively. If those involved are conscious of the shortcomings of existing theory (such as the structure of DNA), they “race” against perceived competition: for example Crick and Watson against Pauling. (2) the scientific community may or may not be aware of the inconsistency, particularly if it involves more than one discipline. (3) A new model produces predictions that are consistent with existing and, in most cases, new observations or results (or, the new model is simpler and capable of more elegant and extended predictions); the new “paradigm” may emerge from outside the recognized disciplines and, perhaps, even redefine them, if not create new disciplines. In various ways, this framework can be applied to the Copernicus-Kepler solar system, Newtonian mechanics, Einstein’s relativity, quantum mechanics, and Crick & Watson’s DNA helix, as well as plate tectonics. The development and application of the new ideas that comprise a paradigm then represent “normal” scientific research. Of course, the “scientific method” is applicable non-scientifically as well, in virtually any field of human endeavor in which a mystery is to be revealed or a problem solved.

One can explore Kuhn’s contributions as a means of exploring the degree to which various revolutions or breakthroughs resemble each other. In a sense, whether one characterizes various advances in scientific understanding as revolutions, breakthroughs, or, quite simply, incremental progress, may be seen as a matter of degree. Previous models may be replaced by very different formulations (e.g., Newtonian by Einsteinian mechanics; the mathematics of the former can be viewed as an approximation of the latter under particular circumstances, however the meaning of each concept is quite different), or a progress with a particular formulation may involve only slight modifications of numerical values (e.g., measurement of the precise speed of light in a vacuum or the universal gravitational constant).

The scientific method, usually ascribed to Roger Bacon (more philosopher than scientist), is assumed to be iterative:
  1. Make observations (either in the context of an experiment or in characterization of some natural phenomenon),
  2. Produce possible hypotheses to explain the observations,
  3. Produce predictions based on the hypothesis and compare with existing observations and, if possible, new measurements,
  4. Assess the viability of the hypothesis, and
  5. Reject, modify, or refine the hypothesis.
  6. If the hypothesis is unacceptable, return to step 2.
  7. If the experiment or observations are inadequate to test the hypothesis, return to either step 1 or refine the hypothesis (step 2).
  8. If the hypothesis is accepted, it becomes “theory”. And, following Kuhn, if the implications of the theory are broad and transformational, then the term “paradigm” is applicable.
  9. If a theory or paradigm is acceptable, return to step 1 to refine and/or extend it.

Practically speaking, the working scientist may begin at step 3 (without really thinking about the method), with an existing hypothesis, theory, or even paradigm, in the hope of refining or extending the existing model. For plate tectonics, this has been the state of its application and development over the past few decades.

Some scientists concentrate on steps 2 and 3, leaving observations to others (Einstein was such a “theorist”; Stephen Hawking is a contemporary example). However, no working scientists really think in terms of the “scientific method”, treating it like a slavishly followed recipe. (And, there are practical components as well: obtaining funding for the research – all the way from salaries, to equipment, to computer time, to hardware and supplies, and equipment construction – and publication of the results in peer-reviewed journals – never assured).

There is a kind of psychological barrier that commonly prevents those engaged in “normal science” under a previous paradigm from jumping on the bandwagon of a new, grand or even grander, hypothesis. Part of the apparent mental block may be geographical: Southern hemisphere geologists, especially from South Africa, in the first half of the Twentieth Century were naturally more familiar with examples in their backyard geology and paleontology and with similar occurences across the South Atlantic and Indian Oceans in South America, India, and Australia. Conversely, the older Appalachians and younger Rockies of North America provided a remarkable enigma when compared with the Alps and Himalayas – the North American ranges occur adjacent to contemporary ocean basins, while the latter are flanked by continents or continental fragments on both sides. It was more difficult to see similarities between the Alleghenies and the Moroccan Atlas.

Northern hemisphere structural geologists of the early Twentieth Century constructed models of huge, linear troughs – “geosynclines” – in which thick sedimentary strata accumulated and then collapsed, producing mountains. Underlying slow-moving convection currents in the mantle, with hot rock rising here and cold rock sinking there were inferred to cause the down warps leading to the mountain-building collapse. Deeply buried rocks melted, producing intrusive granites in the core of the collapsing mountains. The collapse and the granitic intrusions along with deposits of coarse-grained sediments shed onto the flanks of the emerging mountains were the “evidence” of what they called “orogeny.”

Geology students even into the mid-1960’s were taught “all about” geosyclines, orogenic events, and converging convection currents; in retrospect, did anyone really accept these explanatory fantasies? Well, they had to accept something, didn’t they? A particular anomaly was evidence for shallowly inclined fault surfaces separating older rocks above from younger rocks below (younger sedimentary rocks are almost always deposited on top of older) and requiring displacements of the older rocks on the order of 10, 50, 100, even 150 kilometers. Either the crust on either side had to have contracted, or, the old rocks slid off of highly elevated mountains in the “hinterlands” – highly elevated mountains (much higher than Everest) which no longer exist and for which no clear evidence could be found. Could a new theory explain thick sedimentary rocks originally deposited in a marine environment and now comprising the crest of both the Matterhorn and Everest? If as, the Swiss suggested, the Alps represented a crunchy convergence of Europe and Africa (evidenced by the low-angle faults of large displacement) shouldn’t there be a contemporaneous hole in either Europe or Africa to accommodate the crunch?

Earth scientists in the 1950’s were particularly challenged when the California fault responsible for the 1906 earthquake that devastated San Francisco was more closely mapped and studied by field geologists Mason Hill and Thomas Dibblee. Rocks to the west of the San Andreas fault north of San Luis Obispo are completely unlike those immediately to the east of the discontinuity. However, far to the south on the east side, rocks virtually identical in age and mineral composition with those to the north on the west side were correlated. More recent study refines the profound implication: 300 kilometers of net displacement of the San Luis Obispo side to the north-northwest of corresponding rocks in Southern California over the past 30 million years and continuing today (with a big earthquake tomorrow?). Similar “strike-slip” faults were documented from New Zealand and Turkey. How could such displacements be accommodated? Until the mid-1960’s no one had a satisfactory answer. In the meantime, the implications were clear – allow the motion to continue for a few tens of millions of years more and Los Angeles would have a new set of suburbs: the cities of the Bay Area.

An additional effect came from the development of new technology in one part of the world – the ability to measure rock magnetic properties that revealed new evidence. Paleomagnetic research was initiated in England first, and revealed evidence for closer proximity of Europe and North America in the geological past; the magnetic properties of rocks of known age point to the location of the magnetic pole at the time the rock acquired its magnetism. Rocks of about 180 million years old from England point to a different north magnetic pole than do rocks of similar ages from North America; and neither location corresponds with the contemporary average magnetic pole (indistinguishable from the Earth’s rotational North Pole). However, the uncertainties in the measurements and the ages were large enough for the non-paleomagnetists to view the results as interesting but not persuasive. Even some paleomagnetists were more interested in the progressive departure of measured magnetic poles from Earth’s spin axis and postulated what they called “true polar wander” – that planet’s entire crust shifted over the deeper interior over geologic time.

Two more new and critical technologies emerged out of the Second World War. In an effort to detect enemy submarines, instruments for measuring variations in Earth’s magnetic field, especially perturbations from the steel hulls of the vessels, were developed. Following the war, as these instruments were applied to systematic surveys of parts of the ocean basins, together with continuous sonic depth sounding, some very curious observations emerged. Isolated mountains rising from the seafloor (as detected by the continuous sounding) were often accompanied by increases in the strength of the magnetic field, as expected. However, in some cases, decreases in the field were observed to correlate with some seamounts.

More critically, as maps of the magnetic field variations were made (after removal of the effect of the present-day magnetic field) beginning with offshore California (by Arthur Raff and Ronald Mason), linear “stripes” in the magnetic variations began to emerge. Further, the stripes typically terminated against troughs in the seafloor at near right-angles. As more and more of the seafloor was mapped. Similar patterns in the stripes were recognized across the troughs with displacements on the order of hundreds even thousands of kilometers – even in excess of the documented displacement across the San Andreas Fault. The troughs were readily interpreted as faults like the San Andreas. However, if projected into the adjacent North American continent, no comparable features were apparent. Like the San Andreas Fault, the submarine Murray, Pioneer, and Mendocino fracture zones (as they came to be known) terminated in mystery.

Magnetic surveys of the North Atlantic and Southern Pacific, across mid-ocean ridges only dimly known prior to the war, produced variations like those of offshore California, but, more strikingly, showed a kind of mirror-image symmetry across the ridges. Whatever was responsible for the magnetic anomalies involved a symmetrical process centered on the Mid-Atlantic and South Pacific ridges.

As paleomagnetists acquired their measurements of rock magnetism on land, they observed not only apparent wander of the magnetic poles (and divergence between continents over hundreds of millions of years), but evidence for reversal of the polarity of Earth’s magnetic field at irregular intervals. At the same time isotope geochemists were refining the ability to determine the age of rocks from the decay of an isotope of potassium to an isotope of the gas argon trapped in potassium-bearing rocks. Fresh, unaltered volcanic rock samples produced reproducible ages with precisions on the order of 5 to 10 per cent. When such dating was applied to diverse volcanic rock sequences whose magnetic polarity had also been measured, the reversals in the magnetic field were demonstrated to be globally synchronous by Allan Cox, Brent Dalrymple, Ian McIntyre and their associates. Combined age and polarity measurements from California, Hawaii, Iceland, and Australia were readily shown to preserve the same polarity sequence extending back as far as 5 or 10 million years ago.

Almost simultaneously, geophysicists in three different research institutions recognized the possible explanation for the marine magnetic variations. Following a hypothesis of Princeton’s Harry Hess, Canadian Lawrence Morley, Brits Fred Vine and Drummond Matthews, and American Robert Higgs saw the correspondence of the youngest portion of the magnetic reversal pattern and the developing geomagnetic time scale (Vine and Matthews were the first to publish the idea; Morley’s initial submissions were rejected; Higgs’ contribution was hidden in a US Navy report). Subsequently, Americans Walter Pitman, James Heirtzler, and associates correlated magnetic surveys from each of the major oceans with each other and with the developing geomagnetic time scale.

If new seafloor is being created at mid-ocean ridges, either the earth must be expanding, or the growing new crust must be consumed somewhere else. Enter the seismic evidence: The onset of global seismic monitoring as part of the nuclear test-ban treaty produced an auxiliary bumper crop of precisely-located earthquakes – and the zones of earthquakes around much of the Pacific, in the northeastern Indian Ocean, and parts of the Mediterranean, Caribbean, and southernmost South Atlantic were shown to extend from oceanic trenches as tilted slabs to as deep as 700 km in just the right places to accommodate the growing plate at the ridges by sinking of oceanic plate back into the earth. Shallow earthquakes were observed to correspond with the crest of the inferred seafloor spreading. The restriction of the most highly energetic earthquake zones to narrow regions around the Pacific and from the northern Indian Ocean, beneath the Himalayas to the Mediterranean, with lesser narrow regions along the mid-ocean ridges led to the recognition of and definition of lithospheric plates – thin, spherical segments of crust and uppermost mantle which do not deform internally, but only where they come in contact with other, separate plates.

Fracture zone offsets, both on land and submarine, form the third kind of plate boundary: transform, along with divergent (ocean ridges) and convergent (trenches and inclined seismic zones). As recognized by Canadian J. Tuzo Wilson, offsets of magnetic stripes across submarine fracture zones do not necessarily represent displacements; rather, the apparent offsets form that way in most cases. More complicated transform faults, such as the San Andreas fault, may also connect spreading centers with convergent and convergent with convergent plate boundaries.

Frosting on the plate tectonics cake was spread by the initial results of then new Deep Sea Drilling Project in the late 1960’s: cores drilled through the seafloor and underlying sediments to the top of the volcanic crust produced fossil ages for the oldest sediments and isotopic dates of the basaltic crust entirely consistent with Heirtzler and company’s projected geomagnetic timescale. A classic scientific hypothesis, based on multiple lines of evidence and reasoning, was confirmed by a grand experiment.

In summary, a speculative hypothesis, continental drift, was advanced based on (1) similar shape and scale of Atlantic and Indian Ocean coastlines, (2) corresponding fossils of land-dwelling organisms across wide stretches of ocean, (3) comparable old crystalline continental rocks on corresponding continental edges, (4) discrepancies between apparent polar wander curves from different continents, and (5) evaluation of goodness-of-fit of continental margins by application of spherical reconstructions.

A complementary hypothesis, sea-floor spreading, acquired documentary support from (1) thinner oceanic compared with continental crust, (2) near-global mid-ocean ridge system, (3) higher heat flow near ridges, (4) magnetic stripes with spacing corresponding with the developing geomagnetic time scale, and (5) variable offset of stripes along fracture zones (transform faults).

Integration of continental drift and sea-floor spreading into plate tectonics was completed by (1) recognition of the principal seismic zones along divergent (mid-ocean ridge), convergent, and transverse boundaries, (2) demonstrated earthquake displacement along the boundaries in the expected sense, (3) successful application of spherical kinematics to seismic, magnetic stripe, and fault zones, and (4) consistency of deep-sea drilling results with predicted ages from magnetic stripes.

No comments: