The movement of the tectonic plates produces forces that deform Earth's crust. These forces are greatest at the plate boundaries, but can be transmitted throughout any plate. Recall that a stress is a force acting on a specified area. The strain is a measure of deformation or change in shape of a material (often caused by a stress). For example, hanging a weight on a spring causes the spring to stretch - the stress is the weight, the strain is the change in length.
We classify the type of deformation a material experiences based on the relationship of stress and strain. If the strain is directly proportional to strain and if after the stress is released, the material returns to its original shape, we have an elastic deformation. If the relationship of stress to strain is more complicated, we call the deformation ductile. If the stress causes the material to break, the deformation is brittle.
Recall that the ratio of stress to strain was called the elastic modulus and that the modulus depended on the type of material. Thus, the deformation produced by a stress depends on the rock type experiencing that stress. If a rock undergoes brittle failure - a fracture forms. We call these fractures faults. If a rock undergoes ductile deformation, it may thin if is extended, or form a fold if compressed.
Two types of factors affect how a rock will respond to a stress:
Several aspects of mineral affect the "strength" of a rock. Rocks with strong minerals are strong, rocks with weak minerals are weak, but a rock is only as strong as its weakest mineral.
Igneous rocks tend to be strong, particularly those with large crystals. Fine-grained rocks have weaker grain boundaries and tend to deform more easily. The common minerals in igneous rocks, quartz, feldspar, pyroxene, and olivine are strong.
The texture of a rock can also influence the rock strength. For example a schist is relatively weak in the direction parallel to the thin sheets comprising the rock.
Sedimentary rocks vary greatly in strength. Quartz-rich, well cemented sandstones are often strong, clay-rich mudstones are usually weak and deform easily.
Any rock will become more ductile if it in a high-pressure and/or high-temperature environment. Since pressure and temperature increase with depth into Earth, at some level, the dominant type of deformation is ductile.
We call the depth at which the style of deformation changes from brittle to ductile, the brittle-ductile transition. In general, the brittle-ductile transition is not sharp, ductile deformation does occur above that depth, and in some instances, brittle deformation occurs below. Usually, the transition occurs at a depth of approximately 15 km beneath the continents. The precise depth depends on the geotherm and mineralogy of the rocks in a given region, but in general, the lowest part of the crust is ductile.
Because the mantle is composed of a different rock type than the crust, the upper mantle often behaves in a brittle manner until the temperature rises and causes mantle minerals to deform by ductile processes.
As always, once we know the causes of rock deformation, the way we use the information is to observe rock deformation and piece together the history of the rocks. To describe rock deformation, we need some terminology.
Joints are fractures in the rocks across which no movement has occurred. They can be caused by thermal cracking, the affects of fluids or weathering, but most are the result of the release of confining pressure on rocks as they are uplifted by erosion. Joints are common in all rock types exposed at the surface. Generally, the orientation of the joints is planar. The planes form in the direction of the weakest stress (and thus indicate the stress orientation at the time the joint formed).
Faults are also fractures in rock, but they are usually spaced more than joints, and the rocks on both sides of a fault have moved relative to each other. Faults are agents of brittle deformation and generally form above the brittle-ductile transition. Often fault surfaces are nearly planar, so we often talk about the "fault plane".
The type of faulting occurring in a given region indicates the stress orientation and thus identifying and mapping faults is very important for understanding a region's geology.
Many faults are slanted or "dip" into Earth at an angle. We call the block of Earth above the fault plane the hanging wall, and the block below the footwall. The throw on a fault is the amount of offset across the fault. To describe the orientation of the fault, we use the angles strike and dip. Strike is the orientation of the fault's intersection with Earth's surface with reference to north. Dip is the angle that the fault plane makes with Earth's surface measured perpendicular to strike.
There are three general types of faults, but any fault may in fact be a combination of these three.
A normal fault is common in extending regions - the hanging wall slides down the footwall (the "normal" situation). Normal faults are common at mid-ocean ridges and anywhere are plate is rifting apart.
A reverse fault is common in a region under compression, such as when two plates collide. In a reverse fault, the hanging wall is pushed up over the footwall. We call shallow-dipping reverse faults thrust faults. They are common in subduction zones and are the types of faults involved in the largest earthquakes.
A strike-slip fault is common in regions where plates are sliding past one another (transform margins). The motion of a strike slip fault can be right-lateral (the opposite side moves to the right) or left-lateral (the opposite side moves to the left).
Often the slip on a fault is not perfectly dip slip or strike slip. We call such faults oblique slip.
Movement of rocks across faults is usually not smooth. The movement occurs during earthquakes and for large faults, the cumulative "slip" can be hundreds of km.
"In the field", faults are often identified by the juxtaposition of different rock types along a narrow zone (the fault trace). Mature faults contain a fine-grain rock or powder called fault gouge, the result of grinding rocks along the fault plane. Often, when on of the blocks on either side of the fault are exposed, we can see slickensides, which are scratches that indicate the direction of movement on the fault. In some regions, fault scarps are evident. A fault scarp is a topographic offset between two blocks. Scarps are very noticeable in Nevada along the bases of some of the mountain ranges. In some climates, scarps erode quickly and not all faults rupture the surface producing scarps.
Not all faults are active. Some are inactive and are no longer slipping. Although they pose no immediate hazard, inactive faults are important for understanding the geologic evolution of a region.
Rocks can also be folded into a variety of shapes. The most common types of folds are anticlines and synclines.
The detailed geometry of the folds provides information on the types of stresses that produced the folds. Folds can take on many different geometries. They can be open with wide folds, or isoclinal with very narrow folds. The tighter the folds, the more intense the stress that caused folding. Folds can also be symmetric or asymmetric, upright or overturned, curved or cornered. A fold "knocked" on its side is called recumbent. Folds don't have to be perfectly horizontal - they can plunge into Earth at an angle.
When we slice into a fold such as at a "road cut" for a highway, folds are often easily identified. The pattern of rocks observed along the surface is also a clue to the existence and type of fold that may exist beneath the surface.
Rocks can fold through three primary mechanisms:
In cylindrical folding, distinct rock layers slide along layer boundaries to accommodate the strain.
In Shear folding, small fractures cleave and the rocks in the layers are displaced along the fractures to produce the fold. This is a brittle deformation mechanism.
At high enough temperatures, the rocks will flow, producing folds by ductile deformation.
Sedimentary rocks are easily folded and the strength is low between layers. Crystalline rock is more likely to fracture than fold, but at a high temperature, any rock will flow.
In general, earthquakes occur on faults (the only exception are the deepest earthquakes, which may represent a different type of shear failure).
In fact, active faults are usually mapped by recording and "locating" earthquakes. The point at which the rupture of a fault during an earthquake begins is called the hypocenter or focus. The epicenter, often reported in the news, is the point on Earth's surface directly above the hypocenter.
Although often shown as "points" on a map, earthquakes are actually rupture areas of a fault. The amount of fault that slips is larger for larger earthquakes. The length of a large earthquake rupture can be hundreds of kilometers, the width is usually less than 60 or 70 km, depending on the fault.
By far, most earthquakes are caused by plate movement and occur along plate boundaries. Some are a result of magma migration, and a very minor amount are induced by human activities such as the building of large damns, etc..
Most earthquake damage and hazard is a result of tectonic (plate-motion caused) earthquakes.
Most earthquakes are a consequence of the movement of plates and the process of plate movement produces an earthquake cycle.
Form most of the time, the two blocks on either side of a fault are held together by friction. During that time, the two plates continue to move, and the blocks near the fault "fall behind" the rest of the plate.
Eventually (over hundreds, thousands, or tens of thousands of years), the plates get so far ahead of the near-fault material that the strain near the fault is to great for the fault friction to hold, and the fault fails suddenly. The sudden failure (which may take a few seconds to a few minutes, depending on the size of the region of the fault that is failing) is called an earthquake.
During an earthquake, part of the strain energy previously stored in the rocks is released as seismic waves. These waves travel outward from the fault and are large and dangerous near the fault, often knocking down buildings and structures. Farther away, these waves are used to study Earth's structure and the details of the particularly fault rupture that generated the waves.
After the fault slips, it sticks once again and the cycle of strain buildup begins once more. This is a simplification of the earthquake cycle, and is called the elastic rebound model of earthquakes.
Most plate boundaries have a combination of faulting styles, usually only one or two is the dominant mode of deformation.
Divergent boundaries are generally classified by normal (extensional faulting) and strike slip faulting. The extension occurs where new oceanic crust is separating. The strike slip faulting occurs along faults that offset the ridges. The depths of these earthquakes is usually shallow, no deeper than 15-20 km.
Transform margins are dominated by strike-slip faulting. The San Andreas fault system in California is the best example of this style of faulting. Westernmost California is part of the Pacific plate, which is moving to the northwest relative to North America. Los Angeles is moving towards San Francisco at a rate of about 5 cm/year. The depth of faulting in California (along the San Andreas system) is usually less than 10-20 km
Convergent boundaries are the source regions for the largest earthquakes. The interface between the overriding and underthrusting plates is large, has a shallow dip, and thus has a very large area capable of brittle failure. Even deeper, the oceanic lithosphere remains cold for some time and can host earthquakes to depths of 700 km.
We live in a region of intraplate earthquakes - the New Madrid Seismic Zone near southeastern Missouri was the source of some of the largest earthquakes in the lower 48 states.
In 1811-12, three very large earthquakes occurred in that region. Because of geologic differences between eastern and western North America, the energy released in these earthquakes was efficiently propagated throughout the young United States.
Bells rang in Boston and the damage in the immediate vicinity of the shocks was great. St. Louis suffered some damage, particularly in regions of thick soil cover.
Intraplate earthquakes are rare and generally associated with weaknesses in the plates. In the case of New Madrid, a 500 million year old failed rift makes the region relatively weak and the focus of seismic activity in the central United States.
Earthquakes radiate seismic waves in all directions. These waves can be detected and recorded by seismometers and seismographs. We operate continuously-recording seismometers all over the continents and on many islands. Thus we record seismograms from all large earthquakes.
We can "locate" (find out where they occur) most large earthquakes that occur using a generalized form of triangulation. To triangulate an earthquake location, we must have at least three observations of the arrival times of seismic waves. That is, we need three seismometers. Generally, we use dozens of seismometers for each earthquake to insure better estimates of location.
The size of an earthquake is usually quantified using the magnitude. There are many different types of earthquake magnitude, the one you may have heard of is called the Richter Magnitude. Earthquake magnitude is an instrumental quantity - you must have a seismometer to make the measurement. Usually, we use the average magnitude value measured from many seismometers.
All magnitude scales are logarithmic - a unit increase in magnitude corresponds to an increase of a factor of 10 in ground shaking amplitude. That is, the ground shaking is 10 time larger for a magnitude 5 than a magnitude 4. Thus shaking from a magnitude 8 is 10,000 times larger than a magnitude 4. Magnitudes can range from negative numbers (for very small earthquakes) to numbers around 9. The largest instrumentally recorded earthquake occurred in 1960 along the west coast of Chile.
We have only had seismometers for about 100-120 years. Before these instruments were developed, we used a shaking-intensity scale to assess the size of earthquakes. The most common, recent intensity scale is called the Modified Mercalli Intensity Scale. Intensity values range from I to XII. The intensity is measured by the effects of ground shaking on structures (human-constructed or natural).
For many historical earthquakes, information for intensity maps were constructed by visitors to the epicentral region (if they could get there). For even older earthquakes, records from newspapers, journals, letters are used to estimate the intensity level.
Fortunately, large earthquakes are much less frequent than small earthquakes. Each day, Earth has thousands of small earthquakes. They are imperceptible and have little affect on our society (they do provide useful information on the location of active faults). Each year, Earth has about:
The relationship between the frequency of earthquake occurrence and magnitude is called the Gutenberg-Richter relationship. It has the form
where N is the number of earthquake each year, M is the magnitude, and "a" and "b" are constants which depend on the region you are studying (most regions are very similar). For the entire Earth, "a" ~ 6.7 and "b" ~ 0.9.
If you look at a active plate margin, earthquakes of magnitude 3 occur on a daily time frame, those of magnitude 5 on a yearly time frame, and very large earthquakes occur on a century time frame.
Magnitude is a useful way to compare earthquake size, but it has some limitations for comparing large earthquakes. Most magnitude scales "saturate" and fail to recognize very large earthquakes. Magnitude is also not related to the physical processes that operate during earthquakes. These inadequacies in magnitude lead to the development another measure of earthquake size called the seismic moment.
Since earthquakes are the result of the "rupture" of part of a fault, and the movement of the rocks on either side of the fault, you would expect that larger earthquakes rupture larger parts of a fault, and the offset during the rupture is larger for larger earthquakes. The seismic moment is a measure of those aspects of the earthquake, and includes a component that is related to the strength of the rock that ruptures.
The seismic moment can be estimated using seismograms and parameters such as the fault area can be estimated from aftershock locations. Thus we can estimate the amount of slip that occurs in earthquakes that don't rupture Earth's surface.
Large earthquakes generally account for most of the slip, or offset of the plates that occurs along faults. For a large earthquake, the slip may be on the order of 5 to 10 meters. We can use this information and the rates of plate motion to estimate the recurrence interval of earthquakes. In the 1906 San Francisco earthquake, the rocks across the San Andreas fault moved about 6 meters. The average motion of the Pacific Plate relative to the North American Plate is about 5 cm/yr., or 0.05 meters/yr.
Thus, to build up the amount of strain released in 1906 takes:
Thus, we would expect a large earthquake is possible in the next century.
Unfortunately, earthquake behavior is tricky enough that such simple calculations can only be used as rough guides to earthquake occurrence, other factors complicate the precise time of an earthquake, and precise predictions are often unavailable. Although our record is short, we have evidence that the time between earthquakes in any given region varies. Only the inevitability of an earthquake is certain.
Technology has allowed us to directly measure plate motions. We use land-based geodetic observations (very similar to surveying). We also use two space-based approaches - VLBI (Very Long-Baseline Interferometry) and the Global Positioning System (GPS). VLBI takes more "work" and provides measurements referenced to quasars - astronomical radio sources located very far away. GPS uses 24 satellites to triangulate on a position at Earth's surface.
How do the results from geology and VLBI compare?
Most earthquakes cluster in time and occur as part of a sequence of activity.
The mainshock is by definition, the largest earthquake in a cluster. Smaller events that precede the mainshock are called foreshocks. Not all earthquakes have foreshocks - and it is impossible to tell a foreshock is a foreshock, until you have a mainshock. Most earthquakes have aftershocks - smaller events that follow a mainshock.
Typically, the largest aftershock is about one magnitude unit smaller than the mainshock. The number of aftershocks decreases with time following the mainshock, i.e. aftershocks are more frequent immediately following an earthquake and then become less frequent with time. Aftershock sequences for small earthquakes end quickly, for large earthquakes they can continue for years. Following a magnitude-6 mainshock, aftershocks may continue for several weeks. Following a great earthquake, aftershocks can continue for years.
We are much better at predicting the imminent eruption of volcanoes than we are at earthquakes. Near volcanoes, we measure the tilt of the ground as well as the higher-frequency shaking associated with small volcanic earthquakes. As magma migrates beneath the surface, it usually generates small earthquakes. With modern technology, we can monitor this activity for time of increased magma movement. Additionally, using tiltmeters, we can watch the surface of the volcano tilt as a magma chamber is inflated with magma. The mountain tilt will decrease as the magma moves out of the chamber, either into dikes and sills or erupts out of the surface.