Earthquake Seismology

LAST MODIFIED: earthquake size vs occurrence moved from Topic 5; induced seismicity moved to Topic 5

POSTED: 10 June 2020

• Earthquake Statistics: How often do Earthquakes Occur?

• about 3.5 Mio earthquake/year worldwide (Table 5.5)
• majority of EQ are very small; large EQ are rare!
• e.g. 3 Mio mag. 1 EQ/year vs. 0.3 mag. 8.5 EQ
• this means that, to see all number in a graph, the numbers have to be plotted on a logarithmic scale, not on a linear scale. On a linear scale, the small large earthquakes don't show up

linear scale log scale

magnitude # of quakes/year 8.5 and up 0.3 8.0-8.4 1 7.5-7.9 3 7.0-7.4 15 6.6-6.9 56 6.0-6.5 210 5.0-5.9 800 4.0-4.9 6200 3.0-3.9 49,000 2.0-2.9 350,000 0.0-1.9 3,000,000

example for seismicity for 1 week in Southern California

magnitude # of quakes 3.8-4.2 3 3.3-3.7 1 2.8-3.2 12 2.3-2.7 37 1.8-2.2 119 1.3-1.7 201 0.8-1.2 242 0.0-0.7 125

• get recurrence rate of EQ of certain magnitude from number of EQ/year
  • e.g. there are 210 mag. 6.0-6.5 EQs per year globally; so an EQ of this magnitude occurs every 1/210 years = 0.00476 years, or 1.74 days (0.00476*365)
  • => a magnitude 0.0=1.9 EQ happens every 10.5s; a mag 8.5 or larger EQ happens only every 3.3 years
• EQs do not occur regularly (i.e. time interval between EQ is not regular)
• the recurrence time of an EQ of a certain magnitude has a certain probability
• scientists use these probabilities to forecast the chance for an EQ along a fault; e.g. the chance of a magnitude 7 EQ to happen before 2032 is greater than 30% in many places along SAF, including southern CA



• Earthquake Physics: Stress and Strain

enlarge

review lecture slide addendum from Lecture 2 loading a fault is analogous to loading a spring stress: forces (shear, tension and compression) acting on blocks across a fault strain: deformation blocks undergo due to acting stresses; strain along an earthquake fault may be elastic (i.e. reversible) until an earthquake happens stress causes strain brittle failure: when force/stress is so large that material breaks stress drop: the drop in stresses acting along a fault during an earthquake

• Earthquake Physics: Asperities and Friction

If there is already a fault along which EQs can happen and stressed make the blocks move past each other, why do these blocks not move continuously? The answer is because friction between the blocks makes them stick together until a certain threshold is passed. The friction is caused by the uneven edges of the blocks.

enlarge

protrusions of surfaces across the fault plane that increase friction in a fault asperities act like anchors hold a ship in port friction is a force that counteracts the stress an EQ happens when accumulated stress is higher than friction and blocks break free during an EQ or during aseismic slip (movement without EQ) some of the asperities are crushed and ground up by the moving blocks. The crushed material is the "fault breccia" (coarser material) or "fault gouge" (finer, claylike material). The fault gouge zone can be several meters wide.

• Earthquake Physics: Stick-Slip Behavior

enlarge

stick-slip behavior: repeated accumulation of stresses between earthquakes, then stress drop during an earthquake blocks first stick together stress and strain build up (block still stick) stress becomes larger than friction slip (i.e. EQ) and stress drop occur; strain continues to grow repeat of cycle explains the repeated rupture along an earthquake fault; if this earthquake model was right, an earthquake along a certain fault would have a certain recurrence time which would help forecast the next one

• Different Types of Seismic Waves

download lecture slides
• P Waves
• primary waves; first to arrive at a seismic station
• compressional waves
• a type of seismic wave that travels straight through Earth from EQ to station (body wave)
• particle motion in the direction of wave propagation
• somewhat equivalent to sound waves (acoustic waves)


• S Waves
• secondary waves; second to arrive at a seismic station
• shear waves
• a type of seismic wave that travels straight through Earth from EQ to station (body wave)
• particle motion perpendicular to the direction of wave propagation
• equivalent to waves on a string or electromagnetic waves (e.g. light waves)


• Love Waves
• a type of seismic wave that travels along the Earth surface between EQ and station (surface wave)
• superposition of multiply reflecting shear waves
• slower than shear waves
• particle motion perpendicular to the direction of wave propagation
• particle motion decreases exponentially with depth
• equivalent to waves in a wheat field when a passing car causes spikes to sway


• Rayleigh Waves
• a type of seismic wave that travels along the Earth surface between EQ and station (surface wave)
• complicated superposition of multiply reflecting body waves
• typically slower than Love waves
• retrograde elliptical particle motion (orbits against propagation direction)
• particle motion decreases exponentially with depth
• equivalent to water waves except that particle motion of water waves is prograde and circular
• typically the longest phase with the highest amplitudes -> most damaging ground motion



• The Phases in a Seismogram

body waves arrive first the P phase is rarely felt, except close to the epicenter the S phase is typically felt as the first sideways or forth-and-back jolt surface waves arrive later and are felt as complicated ground motion in all directions for large earthquakes, shaking near the source can last up to a minute; for very large earthquakes, shaking can last several minutes


NB: the seismogram above is a vertical-component seismogram. The arrival time of the Love wave is indicated but there is actually hardly any signal. Since Love waves consist of purely horizontal motion, a horizontal seismometer is needed to record that motion. A full seismic record consists of 3 seismograms that record the vertical motion as well as that in the North-South direction and the East-West direction. The whole record, consisting of 3 seismograms, is then called a 3-component record.


• Locating an Earthquake

Travel-Time plot

Seismologists use the P and S arrival times in seismic records to estimate the distance to an earthquake. Travel-time tables or computer programs help to determine this distance. With the seismograms of at least 3 stations, they can determine the location of the earthquake.

Example for the 1811 New Madrid event using 3 fictitious stations

since absolute source time is unknown, triangulation is performed with P - S differential times (the time intervals between the P and the S arrival) infer distance to earthquake (epicentral distance) for each station draw circle with inferred distance times around each station the center of the common area of all circles is the epicenter



• need at least 3 seismograms/stations to find epicenter
• need at least 4 seismograms/stations to find hypocenter
  (one additional station is needed to find one additional parameter, the source depth)

NB: to determine the hypocenter, one would have to draw spheres around 4 stations. The point at which the 4 spheres intersect is the hypocenter. In practice, this is, of course, not drawn on paper but calculated on a computer.

• Earthquake Prediction: The Case Parkfield

• prior to 2004, Parkfield had a 90% chance for a mag. 6 EQ occurring before 2032
• segment located between ruptured segments of S.F. 1906 EQ to north and Fort Tejon 1857 EQ to south
• 6 large EQ between 1857 and 1966; almost regularly 1 EQ every 22 years
• next EQ predicted for window between 1975 and 1993
• seismologists deployed instruments in Parkfield
• EQ did not happen until Sept. 28, 2004
• Question: was EQ#7 11 years late (case 1) or did #7 never happen and EQ #8 was right at beginning of predicted time window (case 2)
• reason for case 1: unknown
• reason for case 2: unknown why #7 did not happen, but San Simeon EQ (Dec 22, 2003) on fault to west, parallel to SAF increased stress on SAF, perhaps making #8 more likely to happen
• last EQ puzzles seismologists because rupture propagated south, not north as most other EQ did



• Long-term Earthquake prediction

• extremely difficult (see Parkfield)
• each fault system is unique, has different characteristics
• use past seismic records (historic and instrumental) to obtain recurrence time of certain magnitude
• use paleoseismology/trench work where seismic records are sparse (e.g. San Diego) (Fig. 5.19/4th 4.21)
• some faults are hidden beneath unperturbed sediments (blind fault) and only discovered after an EQ (e.g. the fault that ruptured in the 1994 Northridge in L.A. was previously unknown)



• Short-term Earthquake prediction (seismic)

• also very difficult
• seismometers, which are designed to record motion during an earthquake, cannot record what happens between earthquakes
• use GPS and laser strain meter to determine continuous movement along fault, also at aseismic times (time when no EQ happens); advantage: continuous data; disadvantage: only spot measurements where instruments are deployed
• determine stress pattern after EQ: rupture on one fault can change stress level on nearby fault (e.g. San Simeon Dec 22, 2003 EQ may have made Parkfield Sep 28, 2004 EQ more likely)
• How much is aseismic (creep, no EQ)?
• Is there significant increase in strain rate just before EQ?
• How long before an EQ?
• studies still in its infancy

INSAR image for the 17 August 1999 Izmit Turkey EQ. The MW 6.7 quake killed over 17,000 people. Large fringes indicate little movement away from the fault. Small fringes indicate large movements closer to the fault.

InSAR (Interferometric Synthetic Aperture Radar): a relatively new field that was originally used to image displacement caused by an earthquake. Makes use of data collected by a satellite that passes an area every few days. Complete images are composed of swath data measuring the distance between Earth's surface and the satellite (so the images are essentially elevation maps). Images are compared, e.g. one before and one after an EQ, and are overlaid. At places where displacement occurred, the images are out of sync. Mathematically, the "out-of-sync" displacement manifests itself as a phase. Remember that a pointer or a vector in a circle is defined by its length (though we assume a unit length here) and its phase measured clockwise. A phase is only defined between 0 and 2pi, and to keep track of the displacement, we have to count how often the pointer goes around the circle. Displacements therefore can be measured by counting the number of fringes by which the images are out of phase. Wide fringes imply relatively small displacements but very narrow fringes indicate large displacements. The big advantage on this technique over observations using seismometers is that this technique can also be used at times when no EQ occurs, e.g. along the southern segment of SAF east of the Salton Sea. Recent studies have found that the far field (area away from the fault) has moved as much as it would take to generate a magnitude 8 EQ, while no movement was recorded close to the SAF. This implies that the southern segment of the SAF is about ready for a magnitude 8 EQ (see previous lectures on loading a fault)




• Unlikely Prediction Tools (Extremely questionable!!)

• comets (unless they actually hit the ground)
• alignment of planets (if Earth tides don't work, planets don't work!!!)
• weather (hurricanes may influence volcanic eruptions but no influence found for EQ!)
• stock market (yeah, right!)
• grandma's arthritis (probably more weather related)



• Earthquake Hazard Mitigation: What Scientists can do

• (seismic) monitoring (also GPS, possibly EM, ground water table in wells)
• Tsunami warning system
• educate residents how to prepare and react; see EQ Safety Booklet published by Southern California Earthquake Center (SCEC), USGS and others
• recommend building codes
• develop effective ways to build EQ-safe structures and retrofit existing structures (e.g. external braces for high-rise buildings; ball bearings for footing of bridges and freeway overpasses)



site amplification maps: get info on local geology and predict how local (thick) sediments enhance shaking (e.g. L.A. Sedimentary Basin vs. hard-lava-rock volcanic islands of Hawaii)

make shake probability maps: get info on fault type and geometry, seismicity patterns and calculate probability of of shaking of certain strength




TERASHAKE still image

TERASHAKE: A large group of scientists affiliated with SCEC have recently collected all know information on earthquakes in southern California, such as faults, seismicity and local geology. Together with colleagues at the San Diego Supercomputer Center, they created a simulation of what would happen if the "Big One" on the southern end of the SAF would occur. The simulation predicts that, if the rupture of a magnitude 8 earthquake would start at the southern end and then propagate northward, then the seismic energy would be handed over onto neighboring faults that reach into the Los Angeles basin and cause major shaking there. NB: As of 10/15/18, the links to the original Terashake movies at SCEC (Southern California Earthquake Center) no longer exist, about 14 years after they were made. We provide a converted quicktime movie here on our own repository.

PAGER population exposure map for 2010 Haiti EQ

immediately after an earthquake, determine (preliminary) fast-response shake maps, depending on population densities and direct relief teams to likely worst-hit areas; this is a relatively new program by the U.S. Geological Service called PAGER (Prompt Assessment of Global Earthquakes for Response). Check it out




• Earthquake Hazard Mitigation: What Governments can do

• control building in hazardous area
• establish building codes
  (hard to enforce in less developed countries)
• construct lifelines according to building codes
  (e.g. re-enforced bridges)
  (hard to enforce in less developed countries)
• secure life lines/make disaster plans
• provide funding for disaster prevention and relief
• provide agencies to coordinate disaster prevention and relief (e.g. FEMA) and early warning (e.g. USGS and Pacific Tsunami Warning Center)



• Earthquake Hazard Mitigation: What we can do

• don't build in hazardous area
• build structures according to code
  (e.g. use flexible wood frames, not bricks-and-mortar technology)
• construct lifelines to earthquake building codes
• landscaping and grading according to code (reduce landslide risk!)
• make disaster plans with family and/or friends/colleagues and/or employer
• have an emergency preparedness kit:
  • first aid kit
  • three-day survival kit including drinking water (1 gallon per person per day) and ready-to-eat canned food; dried fruit/food/juices; high-energy (granola bars) and comfort food;
  • pet food; pet carrier
  • medication
  • extra clothing
  • toiletry (e.g. toothbrush, toothpaste, soap, toilet paper, personal hygiene items and the like)
  • flashlight and spare batteries; spare cellphone battery
  • battery-operated radio and spare batteries; listen to KOKO/AM 600
  • cash
  • copies of essential paperwork (ID's, insurance policies etc.)

"did-you-feel-it" 4 Apr 2010 El Mayor EQ

citizens can also help scientists to better understand what effects earthquake shaking has, i.e. where shaking occurs, how strong it is and what factors play a roll. The U.S. Geological Survey (USGS) maintains the "Did You Feel It?" website where you can fill out a questionnaire about a particular earthquake (choose from a list) or a new one (before you choose a new one, make sure it's not in the list). You can enter things like whether you were asleep or awake, whether you were outside or indoors (if indoors which type of building which story), whether everybody around you felt it or only you. After you are done, the website will return a shake intensity that corresponds to your descriptions. You will also see a shake-intensity and damage map for your earthquake. Example: USGS "did-you-feel-it" page" for the 4 Apr 2010 El Mayor, Baja California Easter Quake.




• Recommended Reading

• "Earthquakes" by Bruce A. Bolt; ISBN: 0-7167-7548-4; price at amazon: $44.04. This book describes the basics of earthquakes, earthquake hazards and gives plenty of examples and historical accounts. This should be an easy read for anybody interested in earthquakes and earthquake research.
• "Earthshaking science" by Susan E. Hough; ISBN: 0-691-11819-1; price at amazon: $25.94. Also an introductory book on earthquakes but concentrates more on earthquake science and is a bit more specialized toward seismology.

1. true
2. false

1. P Waves
2. S Waves
3. Love Waves
4. Rayleigh Waves

1. true
2. false

1. 1
2. 2
3. 3
4. 4

1. the Pacific has smaller earthquakes
2. the Pacific has less earthquakes
3. less people live along Pacific coasts
4. the Pacific Ocean has a well-functioning tsunami warning system
5. the Pacific has only deep earthquakes

The 2002 Denali, Alaska Earthquake

• The 2002 Denali EQ, the Trans-Alaska Oil Pipeline and EQ Hazard Mitigation
  On Nov 3, 2002 a large magnitude 7.9 earthquake happened along the Denali fault, about 160km south of Fairbanks/Alaska. There have been smaller earthquakes in the past (e.g M 7.2 in 1912, M 6.2 in 1958). The predominantly right-lateral strike-slip event caused the surface to rupture for a total length of 320km. The earthquake caused numerous debris flows, avalanches, snow slides and rock falls along the fault and significant horizontal displacement. For example, scientists measured an offset of 2.5m on the Richardson Highway while a section of the Trans-Alaska Pipeline did not break. The pipeline is built above ground and sits on a frame. During transport the oil is to be kept warm to flow and burying the pipeline in permafrost would melt the ground and make it unstable. The pipeline was designed to withstand magnitude 8 earthquakes and though the pipeline was shaken off its frame, it did not burst.
• The 2002 Denali EQ and Geyser Activity in Yellowstone National Park
  The Denali earthquake was large enough to trigger small earthquakes in Yellowstone National Park almost 2000 miles away. Scientist speculate that the strong Rayleigh waves temporarily changed the stress field in Yellowstone enough to trigger these quakes. More recent research also found that this earthquake changed the timing and behavior of some of Yellowstone's geysers and hot springs. Some small hot springs suddenly started to boil with eruptions as high as 1 m. The temperature if some spring changed from 42°C to 93°C (108°F to 199°F) and became less acidic. Some geysers erupted more frequently while other erupted less frequently. Check out this story by scientists at the University of Utah.

Induced Seismicity by Pumping Water into Deep Holes (see also course book)

• 2006 Induced Seismicity in Basel, Switzerland by Geothermal Prospecting
  Deep Heat Mining drilled an over 4-km-deep exploratory hole in a known earthquake zone near Basel to explore the possibility of geothermal power generation. The company had not done a hazard assessment study prior to their activity. On 2 December 2006, the company started to pump water at high pressure into the borehole. Six days later, people in the area felt shaking from a magnitude 2.9 earthquake. The company stopped pumping but earthquakes continued, culminating in a magnitude 3.4 quake that day. More magnitude 3 or larger quakes occurred on 6 January 2007, 16 January and 2 February. The public protested strongly and 2700 claims of damage on homes were processed. After three years of study, the company canceled the project in December 2009. The study revealed that the area would have to endure more such earthquakes if operations were to continue.
  In a twist of events, the company planned similar operations in Central California but the U.S. reacted with new regulations on deep geothermal energy projects.
• Wastewater Pumping and the 1962 Denver Earthquake Series
  There had been some historical earthquakes in the area but at the time, seismicity was relatively low. Beginning in April 1962, there was a sudden increase, with 700 events in the Richter magnitude range 0.7 - 4.3 occurring by September 1963. Most of these event were within a 8-km-radius of the Rocky Mountain arsenal northeast of Denver were weapons were manufactured. The waste water from this operation was initially evaporated but then pumped down a 3670-m borehole. The water was injected under pressure from 8 March 1962 through 30 September 1963. When the pumping ceased, so did the earthquakes. After pumping resumed in September 1964, new seismicity unsettled the population in the area until the pumping stopped. In 1969, the U.S. Geological Survey conducted a controlled study to find out how pumping water into a borehole induces seismicity. This was the first study to establish a connecting between the two.
• Large Reservoirs and Seismicity
  Bolt (1993) gives examples of induced seismicity at large reservoirs. Even though the added weight of the water in a reservoir exerts a large amount of pressure on the bottom, this pressure is much smaller than the pressure exerted by natural processes within the Earth's crust. It is therefore thought that water leaking from the reservoirs penetrates into existing faults, lubricating them and thereby decreasing the friction along these fault, thereby making earthquakes more likely to occur. It could also be that the pressure of the water causes new microcracks through which the water can migrate to reach existing faults. Another factor increasing the seismic hazard is a dramatically changing water table in the reservoir. This causes changes in the underlying pressure and can weaken the underlying rock faster than if the pressure stays content. It is thought that annual changes in Lake Nasser, Egypt the river Nile is dammed by the Aswan Dam caused a magnitude 5.6 earthquake on 14 November 1981. The water level in the lake changes by as much as 5 m throughout the year.