Restraints caused by boundary walls and internal transform faults
These restraints cannot be quantified with accuracy. The number of boundary walls in a simple model are, of course, two, with the plate bounded by a spreading-ridge at one end and by a subduction trench at the other. Real plates, as will be seen from Figure 2.2, are much more irregular and complex. Also, real plates contain a number of internal vertical, strike-slip, transform faults which off-set the spreading-ridge boundary to a plate. These fractures usually trend normal to the pole of rotation of the plate. How these strike-slip faults develop is not yet clearly understood. We consider the most feasible explanation is presented by Gudmundsson (1995), to which we direct the interested reader.
The structure of transform faults is known to be far from uniform (the reader may wish to refer to various papers on this topic Oceanic Fracture Zones, Journ. Geol. Soc. London, 143, 5, 1986). However, Garfunkel (ibid) argues that the great majority of such ridge-to-ridge transforms are leaky. At such 'leaking' faults (which, from time to time, extrude lava at the surface) the resistance to strike-slip motion will be relatively small.
The length of such features varies from place to place. For example, the Clipperton and Clarion Transforms, in the Pacific, extend for about 5000 km, with several others in the E Pacific having a length of 2000-3000 km. However, transform faults in the Atlantic and other oceans are usually less than 1500 km, but the portion of the fault that moves is largely, or wholly, restricted to the distance between the off-set, spreading-ridge. Consequently, one can infer that transform faults develop in areas where the oceanic lithosphere is relatively thin. Hence, we suggest that they will constitute a relatively small percentage of the total restraint to movement of a major plate.
Plate boundary faults may be non-leaking and, of course, deeper and also of greater lateral extent than transform faults. In a large, mature oceanic plate, the average depth of the boundary surfaces may be about 60 km, and their total length reaches 50,000 km, so that the area of such features is about 3*106 km2, with perhaps only half this area cutting the elastic region of the oceanic lithosphere. However, the area of the Pacific plate is of the order of 100*106 km2, so that the boundary and transform fault areas are likely to constitute only about 2 per cent of the total restraining areas of a major plate.
Hence, when one compares the total area of the major boundary plates combined with that of the transform faults, with that of the basal area of the plates, it is likely that the resistance to movement produced by the boundary faults will be only a small fraction of that produced by the basal slip plane on a large plate. The ratio of the boundary-fault/retardation may increase somewhat for small plates.
However, if small sections of the boundary faults, on opposite sides of the plate, converge slightly, this would increase the retardation effect. Such converging boundaries would locally build up large stresses. If the stresses become too large, then, of course, the converging part of the boundary will be 'sheared-off', hence, there is likely to be a 'feed-back mechanism' which tends to set a limit on this 'convergence' effect (cf. the San Andreas fault complex).
Clearly, precise evaluation of the various retarding effects on boundary and internal strike-slip faults is constrained by our lack of knowledge regarding these features. However, it is probably prudent not to underestimate the contribution that these boundary and internal strike-slip faults make to plate movements.
From the above discussions, it would follow that the total combined push-pull stresses that can be accounted for by using the simple models described above are not more than about 1.0 kb (and may be as little as a few hundred bar). It is not reasonable to expect this magnitude of stress to overcome all the elements that resist plate motion, and also account for the deformation that results from continent/continent collision and the attendant mountain building, such as the Himalayas. As we have already noted, it has been inferred that deformation which develops as the result of such tectonic processes must sometimes attain a magnitude of at least 4-6 kb (Sibson, 1975). In this assessment we have initially given slab-pull equal weighting with ridge-push. However, the S American plate is not influenced by slab-pull, yet is still able to move westward at a respectable plate velocity. Consequently, we may infer that if this plate lacks slab-pull and yet moves and also sustains a mountain chain, then the dominant element in sustaining plate motion must be push and, moreover, it requires a push many times larger than can be accounted for by the model of ridge-push outlined above. A mechanism capable of generating a push-dominated, differential stress with a magnitude of several, perhaps ten, kilobars must be missing from our assessment.
Was this article helpful?