Environmental Geology - 106
Lecture 6 - Earthquakes
Lecture Outline
I. Earthquake Fatalities
The average annual U.S. death rate from earthquakes since 1811 has been 10 people. In contrast, 700,000 people die from heart disease, 50,000 from car accidents, 15,000 from simple falls, 200 from heat stroke, 18 from football, and 10 from skateboarding. Statistics thus tell us that earthquakes are not something to worry about. However, if one lives in a zone of seismic hazard, common sense tells us otherwise. Earthquakes occur infrequently and yet can be catastrophic, particularly if they occur at the wrong time of day or if buildings, dams, bridges, and other important parts of a society's infrastructure are not designed properly.
For example, in 1976, a Mag 8 earthquake in southern China killed over 650,000 people in a matter of minutes. In 1985, over 3000 buildings fell in Mexico City from ground-shaking induced by an earthquake that occurred over 200 miles away. In 1988, a M 6.8 earthquake in Armenia killed 55,000 people. The next year, an even larger M 7.1 earthquake in northern California killed only 67 people. In 1995, a moderate M7 earthquake struck Kobe, Japan. 5,500 people were killed, nearly 40,000 injured, 200,000 buildings were destroyed or damaged, and damage estimates were about $100 billion or 2-3% of the Japanese gross domestic product. Given these numbers, if an investment of even several hundred million dollars in seismic and engineering research had reduced the damage levels by even only 1-2%, the investment would have paid back for itself 10 times over in just this one earthquake!
Earthquakes are important because the most devastating earthquakes are focused around the edges of the heavily populated Pacific Rim. The list of major cities that could be damaged by a large earthquake is impressive. Most countries have now recognized that they cannot afford to ignore earthquake hazards.
II. What are earthquakes and what causes them?
An earthquake is the brittle, sudden failure of the earth's crust or mantle.
Earthquakes are caused by several factors; however, the common element is that stress builds in rocks until the yield strength of the rock is exceeded, at which point rupture occurs. The relative movement between the major tectonic plates is responsible for the stress build-up that causes the vast majority (e.g. more than 90%) of earthquakes.
To better understand the earthquake process, we must first define some simple physical quantities.
Stress - force per unit area.
Elastic deformation - A material changes shape when stressed, but after the stress is removed, it returns to its original shape. The bonds between the molecules and atoms of an elastically-behaving material, when stressed, stretch and bend, but retain memory of their original configuration. Once the stress is released, the stored energy is released and the material returns to its original shape. Examples: rubber band or super ball, rocks
Slide 1 - Faulting of a rocks in a road-cut
Brittle deformation - Rupture occurs in response when stress that is exerted on the material exceeds that material's yield strength. Example: glass, ceramic, rock
Slide 2 - Folding of a rocks in a road-cut
Plastic deformation - Flow occurs in response to stress - material does not return to original shape after applied stress is removed. Play Dough Some rocks (rock salt or halite) and other evaporites) flow when subjected to stress.
Demonstration of elastic and brittle deformation using a piece of wood
When the strip of plywood is subjected to a bending force, it deforms elastically at first. If the bending force is released, the wood returns to its original unbent shape. However, if the bending exceeds the yield strength of the weakest part of the wood, that part will rupture. Once the rupture occurs, we all hear the cracking noise - this represents the propagation of acoustic energy through the air due to the rupture. In other words, the rupture releases energy into the surrounding medium and this energy spreads away from the point of rupture. The farther that one is located from the point of the rupture, the softer the rupture noise is because that finite amount of energy released by the cracking wood is being spread over a larger volume as it moves away from the source.
OVERHEAD - Elastic Rebound
The overhead is a cartoon of what happens before and during an earthquake. In the earth's crust and in particular, within the fault zones that accommodate the motion between the rigid interiors of plates, the crust deforms elastically between earthquakes. The faults have geometric irregularities (bends) that prevent the crust on either side of the fault from slipping smoothly (i.e. creeping) in response to the steady-state motion of the plates on either side of the fault. Because friction prevents steady-state slip along a fault, rocks near the fault deform elastically in response to plate motion far from the fault. Once the amount of elastically-stored energy exceeds the strength of the weakest area of rocks along a fault, that patch of the fault ruptures. At the point of rupture, rocks on either side of the fault slip (within a few seconds) to their new location and in the process, release lots of stored energy that propagates away from the point of rupture. A small rupture in one area of a fault can place a sudden strain on a nearby, more strongly locked section of the fault and cause that section of the fault to rupture, too. Thus, one earthquake can trigger another. This cartoon clearly oversimplifies the earthquake process. In reality, faults often have bends; the rocks on the fault face can have different frictional and elastic properties; fluids may lubricate the fault; and other nearby faults may change the local stresses.
Once an earthquake has occurred along a section of a fault, much of the stress on those rocks is relieved. However, since steady-state plate motion is still occurring, stress immediately begins to build again, leading to the earthquake cycle in which repeat earthquakes occur along sections of a fault.
The frequency and strength of earthquakes along a given fault depends on how quickly the stress builds, how weakly or strongly the fault is locked in a particular region, and interactions with other nearby faults that are also responding to the stress buildup. This makes it difficult to to model the earthquake cycle.
III. Rupture and propagation of seismic energy
OVERHEAD - Propagation of Seismic Energy Away from an Earthquake
The passage of seismic energy through the earth is both a blessing and curse. It is a curse to those whose lives and property are destroyed by earthquakes. For those who study the earth, seismic waves are the most valuable source of information about the composition and behavior of the earth's interior.
If we throw a stone into a pond, we all know that ripples on the pond's carry energy away from the point of impact. Although we don't see it, some of the energy is also carried down into the pond as sound, which we could hear if we were beneath water when the stone was thrown. In a similar fashion, during an earthquake rupture, two broad categories of seismic waves are generated.
Body waves, which carry seismic energy through the interior of the earth
Surface waves, which carry seismic energy along the surface.
Body waves can be further sub-divided as follows:
OVERHEAD - Particle displacements for P- and S- waves
P wave (primary) - Compressional - Particles displaced in direction of energy propagation
S wave (secondary) - Shear - Particles displaced perpendicular to direction of energy propagation.
Surface waves, which cause the earth's surface to roll as they pass by, are often responsible for the majority of earthquake damage. Surface wave amplitudes can reach several meters, meaning that during a large earthquake, one end of your house could be in the trough of a surface wave several meters beneath the other end of your house, which could be surfing on the crest of a surface wave! Surface waves travel slowly - often take several minutes or longer to travel tens of miles. Body waves arrive within seconds, but aren't as likely to cause major shaking.
Why are seismic waves useful?
Seismic waves are useful for locating earthquakes, determining the amount of energy that was released, and determining what type of fault slip occurred. Seismologists routinely exploit this information using a global network of seismographs that continuously feed their readings into several analysis centers. Earthquake locations (epicenters)and magnitudes are typically available less than an hour after an earthquake. The operation of seismic networks has been funded by the government, partly due to their need to monitor nuclear tests by the Soviet Union and China during the Cold War, and has lead to huge strides in our understanding of the earth's interior and the motions of plates.
Location - Three things required to completely describe the location of an earthquake - its latitude, its longitude, and its depth. These three together describe the earthquake hypocenter, which is the point within the earth where an earthquake started to rupture a fault. The point on the earth's surface directly above the hypocenter is called the epicenter. (READ TEXT TO SEE HOW P and S waves can be used to locate earthquakes).
Magnitude - The magnitude of an earthquake measures the total amount of ground shaking produced near the epicenter. Scientists use a variety of magnitude scales to measure earthquake size; however, the magnitude scale most widely cited in the press is the Richter magnitude scale. Richter magnitudes vary from 1 to about 9, with 1 being very small and 9 being enormous. In general, an increase of 1 point in the magnitude represents a 10x increase in the amount of ground motion and a 31X increase in the amount of energy release. Thus, a magnitude 8.3 earthquake, which is the size forecast for southern California, could generate significantly more ground shaking than the 1994 Northridge earthquake in the San Fernando Valley, whjch caused an estimated 15 billion dollars in damage.
Intensity - An alternative way to measure the size of an earthquake is by its effect on humans and surface features such as buildings. This technique has shortcomings because it depends on the often subjective observations of individuals. However, for earthquakes that occurred before regular instrumental recording made it possibly to routinely estimate earthquake magnitudes, estimates of intensity are the only way to locate epicenters and determine how large the earthquake was. For instance, a substantial earthquake hit Charleston, South Carolina in 1886 caused damage over a wide area, including southern Wisconsin. This damage was widely recorded in the press, and could be used to map how intensity varied by location. The heaviest damage occurred in Charleston, suggesting that the epicenter was near there. The most widely used intensity scale is called the modified Mercalli scale. The textbook supplies details about this.
III. Earthquake Risk factors
IV. Extreme Damage and Mitigation
The intensity of ground shaking is determined by:
OVERHEAD SHOWING AMPLIFICATION OF SURFACE WAVES DUE TO SOIL CONDITIONS.
Key points about overhead: Seismograms from Oct. 17, 1989 Loma Prieta earthquake near San Francisco show how ground shaking is related to characteristics of soil/rock column at site.
All three sites in Oakland are about same distance from epicenter. If each of these sites had similar near-surface rocks, the ground shaking would be similar. Instead, we can see that the ground shaking varies dramatically, depending on whether the site is in mud, sediment, or on bedrock.
Case Study: 19 September, 1985 Michoacan Earthquake
OVERHEAD Location of Earthquake relative to Mexico City
On Sept. 19, 1985, a M 8.1 earthquake shook coast of Mexico ~100 miles from Acapulco. In Acapulco, about 100 miles from the epicenter, so little shaking occurred that some people slept through the earthquake. In Mexico City, 220 miles away, over 5,000 people were killed and 3,000 structures collapsed. Why?
Buildings in Acapulco sit on bedrock, and thus experience little ground shaking. In contrast, sections of Mexico City are built upon water rich lake sediments that tend to amplify surface waves as they pass through.
OVERHEAD "Jello Bowl" Fig. 4.31, Page 85, Pipkin
Buildings in Mexico City were engineered to withstand strong seismic shaking, but in this earthquake, shaking was more intense and lasted longer than had been anticipated by seismologists and building engineers. Ground roll lasted nearly 5 minutes within the lake bed sediments. Worse yet, the frequency of the surface waves caused what is known as resonance within structures of intermediate heights, 15-20 stories. Resonance means that each time a building sways due to one wave, they would get an extra push by the next wave, further increasing the swaying. Eventually, the buildings swayed past their design limit and either fell over or collapsed. Buildings that were taller or shorter than this didn't resonate and were relatively undamaged!
OVERHEAD - "Collapsed and Undamaged Buildings" Fig. 4.30, P. 84 and 86, Pipkin
Seismic Risk Zones in the U.S
By the mid-1970s, enough basic seismological research had been done in this country to demonstrate that large areas of the United States have significant seismic risk. In response, the federal government initiated a program call National Earthquake Hazards Reduction Program (NEHRP) that operated on several levels, including funding of basic seismological research including efforts at earthquake prediction, funding of earthquake engineering studies, and coordination of federal disaster response. In addition to this program, the state of California implemented a number of programs that were designed to upgrade construction codes and disaster preparation in the event of a major earthquake in California. In the last year, President Clinton has initiated an additional earthquake program called National Earthquake loss reduction program (NEP) with goals complementary to those of NEHRP.
What is being done?
(1) Construction codes - Engineering and seismic research under NEHRP has demonstrated that for about 2% of the cost of a building's original cost, a structure can be designed to minimize the possibility of catastrophic collapse in an earthquake. Typical design parameters are inter-connected structural elements, and design of buildings that do not resonate at the frequency of the expected seismic waves.
Structures not designed to withstand earthquakes often react poorly to the passage of seismic waves, either because the weight-bearing members of the building are not tied together, or because the building as a whole vibrates at some natural frequency that tends to tear apart the building (or both).
Damage suffered by buildings separated by only a few blocks or even a few feet can differ dramatically depending on the local soil conditions and construction characteristics of the building.
LESSONS FOR YOUR OWN HOME IN A ZONE OF SEISMIC HAZARD:
(1.1) - Don't locate on land that fills in a pre-existing topographic depression, particularly if it is near or in a previously existing body of water.
(1.1) - Build a wood frame house (not brick). A WF house flexes during the passage of surface waves, and typically preserves the occupants!
(1.1) - Avoid placing heavy hanging objects or furniture in places where it can fall or tip onto sleeping occupants or block passageways.
Problem of Retro-fit - costs lots of money to retrofit buildings built before implementation of earthquake construction codes. Not even California, which is at the forefront of earthquake awareness, has retrofitted old bridges or required retrofitting of many older buildings. They have instead focused on retro-fitting buildings and structures such as important bridges and emergency/government buildings/schools/hospitals first, and they've delayed retrofitting lower-priority structures.
(2) Seismic Risk Studies - There is a need to assess which regions are prone to earthquakes, how often earthquakes recur, and their expected magnitudes. Geological and seismological research often provides this information, enabling regional planners to know whether changes in construction codes are required, and whether an emergency plan should be adopted. Earthquake Forecasts assign a probability that an earthquake will occur within a given number of years within a particular region or on a specific fault. Can be useful for preparedness.
(3) Earthquake Prediction - conclusion of one of world's most eminent seismologists in recent paper about earthquake prediction - there is not yet one single instance of a successful prediction of an earthquake time, location, and magnitude. Earthquake prediction has proven difficult for the following reasons: