In this article we shall look at what faults are, what causes them, and how we can recognize them. It will be helpful if the reader is familiar with the articles earlier in this textbook on the physical properties of rocks, on terranes, and on the principle of faunal succession.
Causes and appearance of faults
A geological fault is a planar fracture in a volume of rock caused by motion of one side with respect to the other. Motion along faults, and indeed the faults themselves, are caused by tectonic events; by the stretching or compression of the crust. This causes the rock to fracture when it is near enough to the surface to be brittle rather than ductile (as explained in the article on the physical properties of rocks).
Faults are classified according to the nature of the motion producing them. Dip-slip faults are those which involve vertical as well as horizontal motion; these can be classified as normal faults, where the landscape is being pulled apart, and reverse faults, where one part of it is being pushed over another. These are more easily illustrated than described, and are depicted in the block diagram to the right. The reader should note that there is nothing particularly normal or common about "normal" faults: the name is just a name. In dip-slip faults the rock lying above the fault is known as the hanging wall and that below the fault as the foot wall.
A thrust fault can be defined as a reverse fault in which the angle of the fault is more than 45° from the vertical. An interesting point to notice is that in a reverse fault, and especially in a thrust fault, older rocks end up directly above younger rocks, presenting geologists with an interesting stratigraphic puzzle; we shall return to this point later.
In a strike-slip fault, the blocks on either side of the fault move laterally but not vertically with respect to one another in a direction parallel to the fault: the San Andreas Fault is the most notorious example of this variety of fault. The transform faults discussed in the article on sea-floor spreading are a special case of strike-slip faults.
Strike-slip faults can be categorized as right (or dextral) and left (or sinistral) according to their direction of motion: in the case of a left fault, for example, anyone standing on one side and looking at the other when it was moving would see the other side moving to the left. The strike-slip fault depicted in the block diagram is a left fault.
An oblique fault combines elements of a dip-slip fault and a strike-slip fault.
In these block diagrams I have portrayed faults in which an originally continuous piece of landscape has been disturbed by a fault. However, as discussed in the article on terranes, faults will also arise when two previously disjoint land masses are forced together by plate motions. For more information about such faults, see the article on terranes.
We shall now turn to the question of how we would go about identifying inactive faults: that is, faults where the two sides of the fault are no longing moving relative to each other.
When a fault is still active then it is easy to spot it. The San Andreas Fault, for example, is hardly inconspicuous, and the motion along it is measurable.
In inactive faults, where motion has ceased, we will still, perhaps, be able to discern a crack in the rocks, perhaps accompanied by a sudden discontinuity of rock types: but perhaps the crack is just a crack, and the discontinuity is just an unconformity. What we would like is evidence that there was once motion along the suspected fault plane. There are a number of clues that point us in this direction.
In the block diagrams above, the fault itself is just shown as a straight line with no width. In reality, this is not the case. The two sides of the fault do not fit neatly together and slide smoothly past one another; instead they grind off fragments, often large ones, and crush and mill them, producing fault rocks such as fault breccia; the milling process will continue until the breccia no longer impedes motion along the fault, producing finer material filling the gaps between the coarser clasts. Fault gouge is a similar rock but with finer clasts. In the photograph to the right, courtesy of the U.S. Forest Service, fault gouge can clearly be seen interrupting the horizontal strata on either side of it.
Fault breccia and fault gouge can be found in active faults. For example, when engineers constructed the aqueduct between Owens Lake and Los Angeles, they were forced to tunnel through the San Andreas Fault, and found themselves tunneling through a thick sheet of fault breccia and gouge. So when we find something that looks like this in the geological record, but not associated with any present motion, then it is reasonable to conclude that this too was produced by motion along a fault; not just because of its similarity to rocks associated with modern faults, but because the nature of the fault rocks themselves unmistakeably indicate processes of fracturing, crushing, and grinding. When we also note that these rocks form a narrow sheet sandwiched between other rocks, and that the clasts of the breccia match in composition the rocks that they're sandwiched between, the conclusion becomes irresistible.
Another rock characteristic of faults is mylonite. This is formed at depths where deformation of rocks is more ductile. The effect of this on the fabric of the rock is to crush its component minerals and draw them out in streaks, producing a wood-like grain as shown in the photograph to the right, with the grain being parallel to the direction of motion of the fault.
Although mylonite is formed at depth, we can still see it in active fault zones because it is exhumed: that is, it is brought out of the Earth by the rising side of the fault if its motion has a dip-slip component. For example, the Alpine Fault in New Zealand is a dextral-reverse fault with the hanging wall rising at a rate of 6-9 mm/yr: the grain of the exhumed mylonite on the hanging wall is consistent with exhumation by the motion of the fault.
So when we find mylonite in the geological record in the shape of a fault and sandwiched between stratigraphic evidence for faulting (as will be discussed below) we can take this as an indication that we're looking at a fault, even if it is no longer active.
Slickensides are smoothed and striated surfaces produced by the friction between the two sides of the fault, or rather between the two sides of the fault and the breccia between them. These are not dissimilar to the smoothed and striated surfaces left by glaciers. However, slickensides often take on a much higher polish than rocks smoothed by glaciers, which is why they are sometimes called fault mirrors.
Faults and stratigraphy
In the case of a dip-slip or oblique fault, if the rocks are stratified and if we are able to look at the fault side-on, we are able to see discontinuities in the originally continuous layers at the fault plane, as illustrated in the block diagrams above. Sedimentary strata, lava flows, and volcanic sills will not join up across the fault. If we are lucky, we will be able to see how the strata originally joined up, and so figure out the extent and direction of motion; but if the motion has been sufficiently great and erosion has been sufficiently severe, or if the rocks do not have distinctive strata, or if we are not privileged to have a side view of the fault, this will not be possible.
In the case of strike-slip faults there is no vertical movement, and so not such a pronounced disturbance of lateral continuity. However, strike-slip faults disrupt linear features such as dikes and riverbeds.
When two large pieces of landscape are pulled past one another, there will inevitably be frictional resistance, and this can pull the material of the rocks backwards (i.e. in the opposite direction to their directions of motion) distorting the structure of the rock and producing drag folds. The phenomenon is illustrated in the block diagrams to the right.
In a reverse fault, particularly a thrust fault, there will be places where older fossils (according to their usual arrangement in the fossil record) appear above younger fossils, in apparent, though not real, violation of the principle of faunal succession. If this was the only sign indicating a reverse fault, then perhaps we might suspect that we weren't looking at a reverse fault, but at an actual violation of the principle. But of course we can look for the other signs of a thrust fault, as listed above: slickensides, drag folds, fault breccia, etc; and we can also check that when we subject the rocks to absolute dating, the sequence of ages suggested by the fossils is confirmed by the dates, and that we really are looking at older fossils pushed by a thrust fault over younger fossils.