Stories of the Lithosphere

Volcanos and Earthquakes--Part 2, Earthquakes

Earth Science Essentials

by Russ Colson

 

Lecture Recap

  

 

 

A source of energy and motion

No, the source of energy is not 'friction.'   Friction uses energy, converting it into heat, and so is not a source of energy.

 

Heat (energy) is being generated inside the Earth (and inside many other planets) by natural radioactive decay.   Elements such as Uranium, Thorium, and Potassium include radioactive isotopes that decay to form other elements plus heat.   That heat causes convection in the mantle of the Earth as 1) warmer (and less dense) rock rises (slowly—it is solid after all), 2) releases its heat into the crust (and from there into space), and then 3) the cooled (and denser) rock sinks.   This convection provides a way to convert the heat energy into motion.  

 

On Earth, the convection in the mantle is linked to movement of large chunks of rigid lithosphere called plates.   When rock moves, we call this process tectonics.   So, on Earth, one of the main kinds of tectonics is plate tectonics.   At the boundaries between these plates there is relative movement—one plate is moving in a different direction or at a different speed from the adjacent plate—and that relative motion at plate boundaries provides the source of energy and motion for most (but not all) earthquakes on Earth.   Thus, most earthquakes occur along narrow bands that mark the boundaries between tectonic plates.

 

Science literacy question:  

 

 

A means to store slow-motion energy over time

As we learned in lecture, energy is stored in rock when the rock bends elastically like a spring.   When the stretched rock breaks or slips, this energy accumulated over a long period of time is released in a short period of time, causing an earthquake as the rock 'snaps back'.   The offset that occurs when the rock snaps back can be on the order of meters.   For example, the offset for the famous San Francisco earthquake of 1906, an earthquake with a magnitude estimated in the range of 7.7 to 8.3, was 3-10 meters at depth (somewhat less at the surface) and extended for nearly 300 miles.   Smaller earthquakes have typically smaller offsets and extend for shorter distances.

 

Elastic behavior occurs when the deformation (strain) of a material is proportional to the force (stress) acting on it.   Proportional means that strain = constant * stress.   For example, if we pull on a spring twice as hard, it will stretch twice as far, but the amount of stretching depends on the strength of the spring—the constant in the equation above.  

 

On a graph, this looks like the following (strain increases to the right and stress increases upward).

Real materials, like rocks and steel girders for bridges, are usually not elastic for all values of strain.   Rather, once the stress becomes great enough, dislocations, fractures, and other defects begin to form in the material and the material begins to act somewhat plastically.   The typical behavior of a material is shown below.

 

At first, the defects make the material stronger (called work hardening), but eventually the defects become numerous enough that the material weakens and eventually breaks.   When the material breaks, that portion of elastic energy stored in the material is released, as shown below.

 

When bridges fail, forensic engineers examine the metal to see if there are signs of defects in the structure of the metal—a sign that the stress on the metal exceeded its design limits, which should not exceed the elastic limit.

 

Rocks behave elastically when they are cold and under low pressure.   As pressure and temperature increase, the rocks behave more plastically and thus don't store energy.   This dependence on temperature is true for other materials as well, blacksmiths heat up steel to bend it plastically so it doesn't spring back.   If you heat up a candy cane over a candle, it will become plastic.

 

Remember the closing question from the lecture about whether we expect more earthquakes deep in the earth where the rock is hot and under high pressure, or near the surface where rocks are cold?   Because hot rock deforms plastically and doesn't store energy, most earthquakes are shallow--less than 40 miles deep.  

 

The deepest earthquakes can be 400 miles deep.   The deep earthquakes occur at plate tectonic subduction zones where a cold slab of oceanic plate is pushed down into the mantle.   The key word is 'cold'.   The slab is moving at a few centimeters per year.   It takes a while for the thick rock to heat up, therefore it is colder than the rest of the mantle at the same depth.   Because it is colder, it remains elastic and can store energy for earthquakes.   However, deep earthquakes are farther from the surface and so they have less impact on people than shallow earthquakes.

 

Graph reading questions:  

 

 

 

 

A way to release stored up energy suddenly

To release its stored up energy, a rock has to break or slip along a surface (fault).   This allows the spring-loaded rock to snap back to its un-deformed state, releasing its energy.   This leaves an offset along the fault surface.  

 

Earthquakes can sometimes be made more likely by people pumping water into the ground.   This water pressurizes fluid along the fault surface and 'lubricates' the rock, making it easier to slip.   Although this might decrease the strength of the rock, making for weaker earthquakes, it also makes earthquakes more likely.  

 

The only earthquake that I experienced was one of these 'human made' earthquakes, a small temblor in Denver Colorado in the late 1960s.   Denver had been free of earthquakes for decades, but in 1962 the military began disposing of contaminated water by pumping it underground.     Frequent earthquakes began and continued until the practice was discontinued.   Sadly, I slept through my only experience with an earthquake—I was only a little kid!

 

More recently, concerns have been raised about the practice of hydrofracking to shatter impermeable rock so that oil can be extracted from it.   The process involves injecting water underground under high pressure.   Under the right conditions, this could lubricate faults and cause earthquakes.

 

Science modeling question:

 

 

A way to move the energy from the area of the earthquake into surrounding areas

 

Energy travels outward from an earthquake via seismic waves.  

 

Sound is a kind of seismic wave.   Here's a thought question for you: when sound travels outward from a broken pencil, what is it that is moving?   Is the air moving, blowing past you as the sound passes?

 

Another thought question:   You're out fishing in the lake.   Your bobber is floating on the water and waves pass it.   As the waves pass, what is moving?   Is the water moving?   Do the waves carry the bobber with like a current?

 

Well, your bobber bobs up and down in the waves, but it does not travel with the wave.   It's not the water, or the bobber, or the air that is travelling with a wave.

 

Energy, not matter, is moving with a wave.  

 

That's not to say the molecules of air or water, or the bobber, don't move as the wave passes.   But the molecules or bobber don't continue with the wave.

 

There are two kinds of seismic waves and two varieties of each kind, making four different types of waves.

 

The first kind are body waves—so called because they travel through the body of a material, not along a surface between two materials.

 

The second kind are surface waves.   These kind of waves can only exist at the boundary between two types of materials.

 

What do you think—is sound a body wave or a surface wave?   If someone is talking to you, does the sound come to your ear along surfaces only, for example by following the surface of their body down to the floor, then across the floor, and up your body to your ear?   Or does it travel through the body of the air?

 

Body wave type 1—Compressional wave

Sound is a kind of body wave called a compressional wave.   Sound energy compresses molecules of air, which then push against the next batch of air molecules to move the wave along.   The air is "springy" when compressed, allowing the air molecules to spring back and forth as the waves pass, as shown below.

Sonar uses underwater sound waves to locate objects in the sea.   Sonar can help the little fish avoid being eaten!

Compressional seismic waves are called Primary waves (P-waves for short), because they travel the fastest of any type of seismic wave and so arrive first from an earthquake.   Compressional waves can travel through gas, liquid, or solid because all of these materials resist compression.   The material's resistance to compression provides a force to make the molecules 'spring back' to their original position as the wave passes.

Body wave type 2—Shear wave

The second-fasted kind of seismic wave is called a Secondary wave (S-wave for short).   It is a kind of shear wave, meaning that the particle motion is perpendicular to (shear to) the direction of wave propagation.

 

 

Gases and liquids have no shear strength, meaning there is no force to make the particles 'spring back' when the wave moves the particles sideways to the direction of propagation.   Thus, S-waves cannot pass through gases or liquids.   Solids have shear strength, and so shear waves can pass through solids.

 

Surface wave type 1—Rayleigh Wave

For a Rayleigh wave, the direction of particle motion is roughly perpendicular to both the surface that the wave is traveling along, and to the direction of wave propagation.   An example is a wave on the water.   As the wave passes, particles move perpendicular to the direction the wave is traveling and perpendicular to the horizontal surface of the water.   Rayleigh waves can travel at the boundary between liquids and gases as long as there is a force, such as gravity, to make the particles 'spring back' as the wave passes.  

 

Rayleigh waves are important not only on oceans and lakes, but they occur between different layers in the atmosphere where they affect cloud formation (clouds form in bands where the air molecules move upward, while cloudless bands from where the molecules move downward).  

 

In earthquakes, Rayleigh waves make the rock at the surface of the earth roll like a wave on the ocean, which has the effect of tilting buildings sideways.   That makes this kind of wave one of the most destructive of an earthquake.  

Surface wave type 2—Love Wave

In a Love wave, the direction of particle motion is perpendicular to the direction of wave propagation but is parallel to the surface.   This means that the surface must have a shear strength for the wave to propagate.   Thus, this type of wave can only travel along a surface where one side is a solid.    

 

 

Applying what you know question:  

 

 

Strength of Earthquakes

 

The magnitude for earthquakes was formerly reported on the Richter scale, named after Charles Richter who developed it.   This scale was based on the initial amplitude of seismic waves as determined from ground motion measured by seismographs.   In this scale, each increase of one unit on the Richter scale corresponded to an increase of a factor of 10 in ground motion.   Each increase of one unit corresponded to a factor of 32 increase in energy released.

 

We now use a slightly different magnitude scale that is more directly tied to the energy released rather than seismic wave amplitude, however it is scaled to maintain comparability to the old Richter scale, thus energy released increases by x32 for each unit increase in magnitude.

 

A magnitude 3 earthquake is just strong enough to feel.   A magnitude 4 earthquake will release 32 times a much energy.   A magnitude 5 earthquake releases 1000 times as much energy (32*32).   A magnitude 8 earthquake releases 34 million times as much energy (32⁵).

 

Major destruction begins with about magnitude 6 earthquakes (although damage can be extensive with weaker earthquakes, particularly in regions with architecture not designed to withstand earthquakes).   Earthquakes of magnitude 7, 8, or 9 can cause widespread devastation even in regions built to withstand earthquakes.

 

Sometimes people wonder if a 'super' earthquake is possible, thousands of times stronger than any known earthquake.

 

What do you think?   You have the tools to figure out what influences the strength of an earthquake.   Consider the analogy with a breaking pencil from our lecture.   Loudness of the crack is a measure of the energy released as waves.   Before reading on, take a few minutes to think about what factors will affect the loudness of the crack and which factors won't affect it much.   Jot down some of your thoughts in your notes (being told the answer is rather meaningless as a science lesson—to learn science, you need to understand how to figure out the answers on your own.)

 

Length of pencil isn't going to have much effect on the loudness, but the cross-sectional area of the break will affect it.   The speed with which we store up energy (faster motion) will have little effect, but the strength of the pencil,--that is, how much energy it can store up before breaking--will affect it.

 

So, the loudness of the crack depends on the strength of the pencil material and the thickness of the pencil.

 

Likewise, earthquake energy depends on the strength of rock and the cross-sectional area of the break.   The strength of rock is limited by the material properties of rock.   The cross-sectional area is limited by the thickness of the part of the earth that behaves elastically (mainly the crust) and by the length of rock that can break at one time.   Thus, you can't have earthquakes of any arbitrary strength.

 

On Earth, the strongest earthquakes, reflecting the maximum strength of rock and cross-sectional area of break, are around magnitude 9.

 

This is no small earthquake.   A magnitude 9 earthquake generates seismic waves with energy equivalent to nearly 500 million tons of TNT, or over 160 times all the explosives used in World War II combined, including the nuclear bombs.

 

Applying math question

 

Predicting Earthquakes

Like volcanic predictions, we can't yet predict earthquakes, but we continue to try.   The two basic approaches are 1) to take note of changes that might signal that strain has accumulated to the point of causing failure in the rock and 2) consider past patterns and frequency of earthquakes to project future patterns and frequencies.

 

As deformation of the rock approaches the point where the rock will fail, changes in the character of the rock (for example, formation of microfractures) can result in measurable changes at the surface.   For example, well water levels might change due to the changed porosity or permeability of the rock.   Changes in the permeability might cause a release of radon gas into the atmosphere.   The change in rock character might result in new velocities for seismic waves through the underlying rock.   Deformation of the rocks might even cause the generation of electromagnetic fields that can be measured at the surface.   Although these approaches have resulted in some apparently successful predictions, their failure for other earthquakes means that they can't be widely applied and, statistically, we can't tell them apart from fluke predictions.

 

Since earthquakes result from the slow accumulation of strain until the rock fails, we might expect a certain periodicity to earthquakes—it takes a fixed amount of time to stretch the rock enough to 'break'.   Some famous predictions have been made based on this approach.   For example, in 1985 scientists predicted a 95% chance that a major earthquake would occur along the Parkfield segment of the San Andreas Fault before 1993.   This prediction was based on the observation that earthquakes had happened, on average, every 22 years since the mid-1800s.   Everyone got ready for the earthquake, but it didn't happen until 2004, 11 years late.   Not a bad prediction, geologically speaking--a mere 11 years off!   But such predictions are not sufficiently precise to be helpful at a public policy level.

 

Another statistically-successful approach that lacks helpful public policy benefits is the recognition that a major earthquake is somewhat more likely to happen after a smaller 'foreshock'.   This is because the shifting of rock along one fault can increase the stress on another fault and cause failure.   This is also the reason for aftershocks—smaller quakes that follow a major quake.   However, a major earthquake does not always follow a small one, nor are all big quakes preceded by small quakes, and so this increases our predictive ability by only about 5%, not enough to make a sure prediction.

 

The take-away message is that, in practice, the region of a fault zone is cut by many faults, and the deformation is distributed along those different faults in complex ways that have so far defied efforts to predict earthquakes on human timeframes.

 

What we can say is that regions that lie along plate tectonic boundaries where relative motion is occurring are more likely to have earthquakes than regions distant from those boundaries.   When will those earthquakes happen?   Well, that's a harder question.

 

 Last updated June 25, 2015.   All text and pictures are the property of Russ Colson.

.