Earthquakes

IntroductionFaults & EarthquakesSeismic WavesEffects of EarthquakesMeasurement of EarthquakesDistribution of EarthquakesEarthquake PredictionSummary

Diseased nature oftentimes breaks forthIn strange eruptions; oft the teeming earthIs with a kind of colic pinch'd and vex'dBy the imprisoning of unruly windWithin her womb; which, for enlargement striving,Shakes the old beldame earth and topples downSteeples and moss-grown towers.

William Shakespeare

2

Introduction• Earthquakes represent the vibration of Earth because of

movements on faults.• The focus is the point on the fault surface where motion

begins.• The epicenter is the point on Earth's surface directly above

the focus.

The deadly Izmit earthquake struck northwest Turkey onAugust 17, 1999, at 3 a.m. Over 14,000 residents of theregion were killed as poorly constructed apartmentcomplexes pancaked to the ground, each floor collapsing onthe one below (Fig. 1). The death toll from this single eventwas greater than the average annual loss of life from allearthquakes worldwide.

An earthquake occurs when Earth’s surface shakes becauseof the release of seismic energy following the rapidmovement of large blocks of the crust along a fault. Faults arebreaks in the crust that may be hundreds of kilometers longand extend downward 10 to 20 km (6-12 miles) into thecrust. The 1,200 km (750 miles) long San Andreas Fault thatseparates the North American and Pacific Plates inCalifornia is the most active fault system in the contiguousU.S. The Izmit earthquake occurred on the North AnatolianFault, a fault that is of similar length and sense of movementas the San Andreas Fault. Unraveling the movement historyof large faults that produce devastating but infrequentearthquakes can help predict the potential threat from similarfaults elsewhere.

The point on the fault surface where movement begins, theearthquake source, is termed the focus. Seismic wavesradiate outward from the focus. Earthquake foci (plural offocus) occur at a range of depths. The majority ofearthquakes occur at shallow depths that range from thesurface down to 70 km (44 miles). Less frequentintermediate (70-300 km; 44-188 miles) and deep (300-700km; 188-438 miles) earthquakes are generally associated withsubduction zones where plates descend into the mantle.Damage is greatest from shallow earthquakes because theseismic waves travel a shorter distance before reaching thesurface. The earthquake effects, the type of damageassociated with earthquakes, include changes in the naturalenvironment such as landslides but most attention is focused on

Figure 1. Collapsed structuresdestroyed by the shallow 1999 Izmit,Turkey (top), and 1994 Northridge,California (bottom), earthquakes.Images courtesy of USGS Expedition toTurkey and USGS Open-File Report 96-263 (Northridge).

3

the impact on constructed structures. Building codes are inplace in most earthquake-prone areas but they are of little use ifenforcement is lax, as was the case in Turkey. Following theearthquake it was discovered that some contractors had cutcorners in the construction of multistory apartment complexes.The poorly built structures were left as piles of rubble amongstother apartments that remained standing.

In contrast, on February 28, 2001, the strong Nisquallyearthquake (Fig. 2) occurred below western Washington 56 km(35 miles) south of Seattle. Buildings in Seattle and thesurrounding communities sustained relatively little structuraldamage, no one was killed, and only a handful of peoplereceived anything more than minor injuries. Seattle hasenforced a stringent building code over the last 30 years thatrequires new structures to be able to withstand largeearthquakes. In addition, over the last decade, many olderbuildings and bridges were retrofitted to ensure that they couldendure the big earthquake predicted for the region. Residents inwestern Washington were doubly fortunate, not only did theyhave well-built structures but the earthquake occurred muchfurther below the surface than the Izmit quake, further reducingthe resulting ground shaking.

Seismic waves are captured by a recorder known as aseismograph. The relative arrival times of different types ofseismic waves is used to determine the distance of theseismograph station from the origin of the earthquake. Three ormore records can be used to pinpoint the earthquake'sepicenter, the geographic location of the point on the earth’ssurface directly above the focus (Fig. 3). Earthquakes arenamed for the epicenter location, for example, the Nisquallyearthquake occurred 53 km (33 miles) below the mouth of theNisqually River in western Washington. Loss of life in theTurkish earthquake was greatest in the city of Izmit, locatedclose to the earthquake's epicenter. Earthquake distributionis far from random. Earthquakes occur on faults that arepreferentially located along plate boundaries. The largestearthquakes along convergent plate boundaries.

One method of earthquake measurement is to determine thelevel of destruction following an earthquake. However, as theTurkish earthquake so vividly illustrates, the degree of damageis often related more to human activity than the earthquakeitself. Consequently, more quantitative measures have beenadopted that measure the magnitude and timing of the vibration

Figure 3. The focus is thesource of the earthquake andthe epicenter is the point onthe surface directly above thefocus.

Figure 2. A smashed carburied under fallen bricksresulting from the 2001Nisqually earthquake nearSeattle, Washington. TheNorthridge (see Fig. 1) andNisqually earthquakes wereof similar magnitudes but thelatter occurred further belowthe surface, reducing thescale of the damagesresulting from the shaking.Image courtesy of FEMA NewsPhotos.

4

of sensitive instruments (seismographs) and the distance fromthe earthquake to calculate a value for the event. Thisinformation is often combined with data on the geologysurrounding the fault to generate an even more accuratemeasure.

Over two million people were killed during the previouscentury by earthquakes and associated phenomena. The threatof future earthquakes in heavily populated regions likeCalifornia (and Turkey) has spurred efforts in earthquakeprediction. Analysis of past earthquakes allows thedetermination of the potential size and location of futureevents, the problem is determining when such events will occurwith any degree of accuracy. The principal difficulty is thatlarge, dangerous earthquakes occur at intervals measured indecades or centuries yet our record of past earthquake activitystretches back only a few hundred years, making it inadequatefor rigorous predictions.

Earthquakes are expensive. Even the relatively minor damagesfrom the Nisqually earthquake cost $2 billion to repair and thesubstantial damages resulting from the 1994 Northridge quakehave been estimated to cost $30 billion, making it one of thecostliest disasters in U.S. history. People living in areas withthe potential for damaging earthquakes may seek to purchaseearthquake insurance to provide some financial protection fortheir property. Approximately 40% of residents in Northridgehad insurance and their insurance companies enduredsignificant losses in paying an estimated $15 billion in claims.The Northridge event caused insurers to drastically rethinkearthquake coverage. Deductibles rose to 10 to 15% andpolicies now exclude loss of building contents and reductionsin other costs. These less generous insurance policies wouldhave trimmed claims from the Northridge earthquake to about$4 billion.

Estimated annualcost of earthquakeinsurance for a$100,000 home inCalifornia: $280

Think about it . . .How frequently do earthquakes occur near where you live?Where do earthquakes occur in your state? Go to the USGSNational Earthquake Information Center’s website(http://wwwneic.cr.usgs.gov/neis/states/states.html) andanswer these basic questions.

Faults & Earthquakes• Earthquakes represent the vibration of Earth because of

movements on faults.• Faults can be identified by the offset of rock layers on

either side of the fault surface.• Normal and reverse faults are types of dip-slip faults.• Left-slip and right-slip faults are types of strike-slip faults.

An earthquake occurs when Earth shakes because of the releaseof seismic energy following the rapid movement of largeblocks of the crust along a fault. A fault is a fracture in thecrust. During the Izmit earthquake, the crust broke along theNorth Anatolian Fault in northern Turkey (Fig. 4). Whenfault movement occurs it may be slow and gradual andgenerate only small earthquakes, or it may be rapid andcatastrophic causing widespread destruction. Ground shakingassociated with most earthquakes is over in a matter of secondsbut it involves such large regions of Earth’s crust thattremendous amounts of energy are released. The ground shookfor 45 seconds in the Izmit earthquake and affected a region ofapproximately 100,000 square kilometers.

Faults may be hundreds of kilometers in length but only part oflonger faults typically break during an earthquake (Fig. 4).Fault segments that have not experienced a recent earthquakeare termed seismic gaps and are considered potential sites forfuture events. The Izmit earthquake occurred in a 150 km (94mile) long gap at the western end of the North Anatolian Fault.

Figure 4.Earthquakesequence alongthe NorthAnatolian fault,Turkey, 1939-1999. A series oflarge earthquakeshave occurred onthe fault system;each resulting fromonly one segmentof the faultbreaking at a time.Image courtesy ofUSGS.

5

6

Two adjacent fault segments to the east and west broke duringlarge earthquakes in 1963 and 1967. Eleven earthquakes ofmagnitude 6.7 or greater occurred along segments of the faultover the previous 60 years. Even though scientists can identifypotential seismic gaps, the faults may not cooperate to generatean earthquake. Earthquake specialists predicted an earthquakewould strike the region around the small California town ofParkfield before the end of 1993 but the quake still hasn'tshown up, despite the fact that there are millions of dollarsworth of instruments in the ground waiting for the big day toarrive.

Even the largest earthquakes require relatively small faultmovements because such large volumes of rock are involved.Offsets on faults for the largest of earthquakes are less than 10meters, and typically less than 5 meters per quake. Theaccumulated movement from hundreds of thousands ofearthquakes over millions of years results in the formation ofmountains in association with plate boundaries.

Fault ClassificationFaults are distinguished as dip-slip or strike-slip faults. Twotypes of dip-slip faults are identified on the basis of therelative motion of rocks across an inclined fault surface (Fig.5).The block of rocks above a fault is termed the hanging wall;the footwall lies below the fault. Miners working in shafts thatcrossed faults were able to hang their lanterns from the hangingwall while their feet remained below the fault in the footwall.Inclined faults can be identified by the offset of rock layers oneither side of the fault surface. The hanging wall moves downrelative to the footwall in a normal fault. In contrast, thehanging wall moves up relative to the footwall in a reversefault. Movement on a dip-slip fault often results in a break or

Figure 5. Top: Thehanging wall (hw) liesabove an inclinedfault; the footwall (fw)lies below the fault.Bottom: Normal fault.

Figure 6. Fault scarpformed during the HebgenLake earthquake,Montana, 1959. Person inforeground isapproximately 2 meterstall. Surface at bottom ofslope was at sameelevation as upper surfaceprior to movement on thenormal fault.

7

offset at the land surface. This break in slope is known as afault scarp (Fig. 6).

Two types of strike-slip faults, left-slip and right-slip faults,are identified on the basis of the motion on vertical faultsurfaces (Fig. 7). An observer, standing on one side of thefault, sees objects on the other side of the fault move to theright for a right-slip fault or to the left for a left-slip fault. The1,200 km long San Andreas Fault, California, is a famousexample of a right-slip fault. Areas of frequent earthquakeactivity are laced with faults. Maps of California show thatseveral faults make up the San Andreas Fault system. TheNorth Anatolian Fault that broke during the Izmit earthquake isalso a right-slip fault (Fig. 8) and is of similar length as the SanAndreas fault.

Faults and Plate BoundariesEarthquake distribution is far from random. Earthquakes occuron active faults and active faults are preferentially locatedalong plate boundaries (Fig. 9). Although, both dip-slip andstrike-slip faults are associated with all types of plate boundary,each type of boundary is characterized by a specific fault style.Strike-slip faults are common at transform plate boundaries

Figure 7. An exampleof a left-slip strike-slipfault.

Figure 8. Offset in a fenceas a result of the Izmitearthquake. Note therelatively small movementon the fault, even thoughthe earthquake was large.Can you classify the fault?Image courtesy of USGSExpedition to Turkey.

Figure 9. Plate tectonicsetting for the Izmitearthquake. The smallAnatolian Plate is movingwestward as it is wedgedbetween the convergingArabian and Eurasianplates. A subduction zonein the easternMediterranean Sea marksthe boundary with theAfrican Plate to the south.

where two plates move in opposite directions. Reverse faultsoccur most frequently at convergent boundaries where platescollide; and normal faults are most common at divergentboundaries such as oceanic ridges, where plates break apart.

Think about it . . .Finish the partially completed concept map for faults andearthquakes found at the end of the chapter. Print the pageand fill in the blanks with appropriate terms.

8

Seismic Waves• Seismic waves can be divided into surface waves that

travel on Earth's surface and body waves that travelthrough Earth.

• Body waves are further divided into S waves and P waves.• Seismic waves are recorded on a seismogram at a

seismograph station.• The distance of an earthquake epicenter from a

seismograph station is determined by the difference in thearrival times of P and S waves at a seismograph station.

• Earthquake magnitude is calculated using the amplitude(height) of the S wave recorded on a seismogram.

Seismic waves represent the energy released from theearthquake focus. There are two types of seismic waves:

• Surface waves travel on Earth’s surface and cause muchof the destruction associated with earthquakes.Undulations of the land surface during an earthquake are arepresentation of surface waves (Fig. 10). Surface wavesmay result in vertical motions (Rayleigh waves), muchlike waves traveling through water, or sideways motions(Love waves) with no vertical component of movement.

• Body waves travel through Earth’s interior. These arefurther subdivided into P (primary) waves and S (secondaryor shear) waves based upon their vibration direction andvelocity. Variations in seismic wave velocity are used toinfer the properties of Earth’s interior.

Figure 10. Rayleigh (top) andLove waves (bottom) are surfacewaves with contrasting motiondirections generated during anearthquake.

9

P waves vibrate parallel to their travel direction in the sameway a vibration passes along a slinky toy (Fig. 11). P wavestravel at speeds of 4 to 6 km per second (2.5-4 miles persecond) in the uppermost part of the crust. S waves vibrateperpendicular to their travel direction, like the wave that passesalong a rope when it is given a sharp jerk (Fig. 11). S wavevelocity is 3 to 4 km per second (2-2.5 miles per second) in theshallow crust.

The velocity of seismic waves is lower in loose, unconsolidatedmaterials (sand, partially melted rock) and higher in solidmaterials (rock).

Both P and S waves are generated at an earthquake focus as aresult of movement on a fault. P waves will arrive at arecording station (seismograph station) first because of theirgreater velocity. Surface waves are the last to arrive because Pand S waves travel a more direct route through the earth (Fig.12).

The record of an earthquake at a seismograph station is aseismogram (Fig. 13). The principal elements of a seismogramthat interest seismologists (scientists who study earthquakes)are the relative size of the recorded waves and the difference intime that the first P and S waves were recorded. The amplitudeof the recorded wave is proportional to the magnitude ofshaking associated with the earthquake but shaking may varywith the character of the material underlying the seismographstation. Loose, unconsolidated materials (e.g., mud, sediment)may exaggerate the shaking whereas solid bedrock may resultin smaller vibrations.

Figure 11. Analogs of Pwave (left) and S wavemotion (right). P wavesare similar to thepassage of a vibrationthrough a slinky. Thevibration occurs in thesame direction as thewave travels. S wavemotion is analogous to avibration moving along arope. The vibrationoccurs perpendicular tothe direction in which thewave travels.

Figure 12. Contrastingtravel paths for surfacewaves and body wavesfollowing an earthquake.

10

The difference in arrival time between P and S waves on aseismogram can be used to determine the distance of the stationfrom the earthquake source and the amplitude (height) of the Swave recorded at the station can be used to determineearthquake magnitude (see Measurement of Earthquakes).

The time interval between the recorded arrival of P and Swaves increases the further the seismograph station is locatedfrom the epicenter. Seismologists match the time differencewith standard curves (Fig. 14) to determine distance from theearthquake.

Figure 13. An idealizedseismogram illustrating thesequential arrival of seismicwaves. Determination ofthe distance from anepicenter requirecalculating the difference inarrival time of P and Swaves (~14 seconds).Earthquake magnitude isrelated to the amplitude ofthe recorded S wave.

Figure 14. Graph ofdistance from the epicenterand time for seismic wavesto reach seismographstation. The time intervalbetween the arrival of Pand S waves increases withincreasing distance fromthe epicenter.

11

Data at a single seismograph station are insufficient to pin-point the earthquake epicenter because a seismogram yieldsonly the distance from the earthquake source. The epicentercould be located anywhere along a circular arc of the calculateddistance from the seismograph station. Seismologists must usedata from multiple recording stations to learn the location ofthe event. The common intersection point for several circlesplotted relative to different stations represents the point on thesurface above the earthquake source (Fig. 15).

Figure 15. An earthquakeoriginating in the PacificNorthwest would berecorded at seismographstations in Denver,Quebec, and Lima (Peru).The difference in arrivaltimes between P and Swaves would be least atDenver and greatest atLima. Circles plotted ateach station reflect thedistance from theepicenter but the directioncan only be determined byidentifying the intersectionpoint for three or morecircles.

Think about it . . .Try the Virtual Earthquake exercise that guides users through thedetermination of the location of an earthquake epicenter andearthquake magnitude using records of seismic waves recordedat three seismograph stations. Print the "Virtual Seismologist"certificate upon completion of the exercise.http:// vcourseware4.calstatela.edu/VirtualEarthquake/VQuakeIntro.html

12

Effects of Earthquakes• A major earthquake under a heavily populated area in the

U.S. could result in thousands of deaths.• Several effects of earthquakes could result in extensive

damages.• Ground shaking can collapse buildings.• Uplift may raise or lower large areas of Earth's surface.• Liquefaction in water-saturated sediment can result in the

collapse and subsidence of the ground surface.• Landslides are a potential hazard on steep slopes in seismic

zones.• Tsunamis, giant sea waves, are dangerous to coastal

communities, especially around the Pacific Ocean.

A magnitude 6.7 earthquake struck the Northridge suburb ofLos Angeles on January 17, 1994. The earthquake resulted inthe deaths of 57 people and injured over 9,000 more. There areabout 150 earthquakes of this magnitude worldwide each yearbut this was the first time a quake of this size occurred in aheavily developed area of the U.S. An earthquake of similarsize killed over 50,000 people in Iran in 1993.

The Elysian Park fault was recently discovered belowdowntown Los Angeles and may produce substantial futureearthquakes. Movement on the 55 km (35 miles) long faultcould result in up to 5,000 deaths, leave 750,000 homeless, andcause $100 billion in damages in Los Angeles. A comparableearthquake in Kobe, Japan (exactly one year after theNorthridge quake), killed over 6,000 people.

The images in the following figures show damage from thelargest recorded U.S. earthquake (Alaska, 1964) and illustratethe effects of earthquakes. All images taken from USGSNational Earthquake Information Center (NEIC).

Sudden changes on or near the earth’s surface result fromearthquakes and may include:

Ground Shaking: Rapid horizontal movements associatedwith earthquakes may shift homes off their foundations andcause tall buildings to collapse or "pancake" as floors collapsedown onto one another. Shaking is exaggerated in areas wherethe underlying sediment is weak or saturated with water (Figs.16, 17).

Figure 16. Part of a railroadbridge over the Copper Riverwas shaken loose by the 1964Anchorage earthquake. Imagecourtesy of USGS.

Figure 17. This map shows incolor those parts of thecontiguous 48 states that have a10% chance of experiencing anearthquake strong enough tocause appreciable damage in a50-year period. In the yellowareas, maximum groundshaking would be strong enoughto damage unreinforcedmasonry buildings, even thosebuilt on bedrock. Darker colorsare at the same risk for moreintense shaking, and areas leftblank would have less intenseshaking. Image courtesy of USGS.

13

Fault Rupture and Uplift: Break of the ground surface by thefault plane may form a step in the surface known as a faultscarp (Fig. 6). Large sections of Earth’s surface may changeelevation as a result of uplift on an earthquake fault (Fig. 18).Mountains east of Los Angeles were uplifted 0.3 meters (1foot) by the 1994 Northridge earthquake.

Liquefaction: Liquefaction occurs when water-saturatedsediment is reorganized because of violent shaking. Thesediment collapses, expelling the water, and causing the groundsurface to subside.

Landslides: Earthquakes are often associated with mountainsformed along convergent plate boundaries. The steep slopespresent in these environments are prone to landslides whenshaken (Fig. 18). Landslides are common followingearthquakes in California.

Tsunamis: Giant sea waves are generated by submarineearthquakes, especially noted from the Pacific Ocean.Tsunamis caused by earthquakes around the ocean’s perimetermay travel thousands of miles to destroy coastal property inHawaii. Tsunami waves may reach heights of 15 meters (50feet) near shore and travel at speeds up to 960 km/hr (600mph). Many casualties associated with the 1964 Alaskaearthquake resulted from tsunamis.

Figure 18. Top: Uplifted seafloor in Prince William Sound,Alaska. The 400-meter-widewhite surface was raisedabove sea level. Bottom:Diagonal crack represents theupper part of a landslide in anAnchorage residential districtassociated with the 1964earthquake. Images courtesy ofUSGS.

14

The Pacific Tsunami Warning System (PTWS) is a network ofstations that attempt to identify potentially damaging tsunamisfrom earthquakes in or around the Pacific Ocean. The PTWSissues warnings or watches that predict tsunami arrival timesfor coastal areas.

Measurement of Earthquakes• There are three methods used for measuring earthquakes.• The Modified Mercalli scale measures intensity and is often

used to rank the cultural effects of historical earthquakes.• Mercalli scale values vary with distance from epicenter,

building materials used, and population density.• The Richter scale is the most well known and measures

earthquake magnitude using the amplitude (height) of the Swave recorded on a seismogram.

• Each division in the Richter scale represents a 10-foldincrease in amplitude and an approximate 30-times increasein energy released.

• The moment-magnitude scale has recently found favor as amethod that more accurately measures energy release onlarge faults.

There are three principal methods of measuring the effects ofearthquakes.

Think about it . . .1. Review the possible effects of earthquakes and examine

a description of the 1989 Loma Prieta earthquake and/orthe 1906 San Francisco earthquake and suggest whatcould be done to diminish the potential for damages andloss of life resulting from earthquakes.

2. Use the Venn diagram located at the end of the chapter tocompare and contrast the characteristics and effects ofthe 1989 Loma Prieta and 1906 San Franciscoearthquakes.Loma Prieta information available here:http://www.es.ucsc.edu/~jsr/EART10/Trips/FT3/index.htmlSan Francisco earthquake information available here:http://quake.wr.usgs.gov/info/1906/index.html

10 largest U.S.Earthquakes(with moment-magnitude values)1. Prince William Sound,Alaska 1964 (9.2)2. Andreanof Islands,Alaska 1957 (8.8)3. Rat Islands, Alaska1965 (8.7)4. Shumagin Islands,Alaska 1938 (8.3)5. Lituya Bay, Alaska1958 (8.3)6. Yakutat Bay, Alaska1899 (8.2)7. Cape Yakataga, Alaska1899 (8.2)8. Andreanof Islands,Alaska, 1986, (8.0)9. New Madrid, Missouri,1812 (7.9)10. Fort Tejon, California,1857 (7.9)

15

• Modified Mercalli scale is used to measure damage andhuman perception of an earthquake.

• Richter scale is the most familiar and measures the size ofthe seismic waves recorded at a seismogram.

• Moment-magnitude scale has replaced the Richter scale inpopularity among geophysicists because it gives a moreaccurate interpretation of the amount of energy released byan earthquake.

Modified Mercalli ScaleThe Mercalli scale measures earthquake intensity: the level ofdestruction of the earthquake (higher values) and the effect ofthe event on people (lower values). The scale ranks intensityfrom I to XII (1-12) using Roman numerals. The table belowsummarizes the characteristics of the Mercalli scale.

Index Effects of Earthquake on People and Structures

I Not felt by people.II Felt by people at rest on upper floors of buildings.III May be felt by people indoors. Vibrations similar to the

passing of a truck.IV Felt indoors by many, outdoors by few. Dishes,

windows, doors disturbed; walls make cracking sound.Sensation like heavy truck striking building.

V Felt by nearly everyone; many awakened. Some dishes,windows broken. Unstable objects overturned.

VI Felt by all; many frightened. Some heavy furnituremoved; a few instances of fallen plaster. Damage slight.

VII Slight to moderate damage in ordinary structures;considerable damage in poorly built or badly designedstructures; some chimneys broken.

VIII Slight damage in buildings designed to withstandearthquakes; heavy damage in poorly constructedstructures. Chimneys, columns, walls may fall.

IX Considerable damage in specially designed structures.Damage great in substantial buildings, partial collapse.Buildings shifted off foundations.

X Some well-built wooden structures destroyed; mostmasonry and frame structures destroyed withfoundations.

XI Few, if any (masonry) structures remain standing.Bridges destroyed. Rails bent greatly.

XII Total damage, objects thrown into air.

Five mostdestructive historical

earthquakes(number of deaths)

1556Shansi, China

(830,000)

1737Tangshan, China

(255,000)

1138Aleppo, Syria

(230,000)

1927Xining, China

(200,000)

856Damghan, Iran

(200,000)

16

The Mercalli scale is relatively easy to use but it is not widelyapplicable to modern earthquakes because its interpretation isdependent upon:

• Variations in population density: earthquake intensitywould be underestimated in sparsely populated areas.

• Building materials and methods: earthquakes of similar sizecould give different values depending upon building codes.

• Distance from the epicenter: values decrease withincreasing distance from the epicenter. Each earthquake hasseveral different intensity values making it difficult tocompare individual events.

Some of the largest historical U.S. earthquakes occurred inthe eastern half of the country (Fig. 19). Three majorearthquakes were centered in southeastern Missouri (NewMadrid) over a three-month period from December 1811 toFebruary 1812.

The Mercalli scale is useful in ranking historical earthquakesthat occurred before the widespread use of seismographs(after World War II). Notice that the map above containsmany historical earthquakes in the eastern half of the U.S.,some equally severe as those in California. Isoseismal mapscan be created that show areas of equal earthquake intensity(Fig. 20). A comparison of isoseismal maps for earthquakes ofsimilar size from the eastern (New Madrid, Missouri) andwestern (San Francisco, California) U.S. illustrates that theeastern event was felt over a much larger area.

Figure 20. Isoseismal map of1964 Alaska earthquakeshowing areas with equivalentdamages following the largestrecorded U.S. earthquake.

Figure 19. Locationsof U.S. earthquakescausing damage1750-1996, Mercalliintensity VI to XII.Large red squaresrepresent locations oflargest earthquakes(intensity XII). Notesquares in southeastMissouri from NewMadrid (1811-1812)earthquakes. SourceNEIC.

17

Richter ScaleThe Richter scale measures earthquake magnitude, theamplitude of seismic waves recorded on a seismographfollowing an earthquake. (See the Virtual Earthquake exerciseon page 11).

Charles Richter developed the scale in the 1930s to measureshallow earthquakes in California. These early measurementsof magnitude (ML- local magnitude) simply relied on using twofactors (the difference in P- and S-wave arrival times and S-wave amplitude). The measured earthquakes were less than600 km (375 miles) from the seismograph stations andoccurred at similar depths in the crust.

More complex formulae to determine magnitude from seismicbody waves or surface waves were developed as the number ofseismograph stations increased and it was recognized thatearthquakes occurred at a range of depths.

The Richter scale is logarithmic, each division represents a 10-fold increase in the ground motion associated with theearthquake, and ~30-times increase in energy released. Forexample, a magnitude 7 earthquake has ten times as muchground motion (and releases over 30-times the energy) as amagnitude 6, 100 times as much motion (900 times the energy)as a magnitude 5, 1,000 times the motion of a magnitude 4, etc.

Magnitude Ground Motion Energy1 1 12 10 303 100 9004 1,000 27,0005 10,000 810,0006 100,000 24,300,0007 1,000,000 729,000,0008 10,000,000 21,870,000,000

Unlike the Mercalli scale, the Richter scale does not have amaximum value; it is open-ended. The largest earthquakesmeasured with the Richter scale have magnitudes between 8and 9. It is probable that rocks in Earth’s crust are unable towithstand stresses necessary to generate earthquakes ofmagnitude 9 or more.

Mb = log10(A/T) + QFormula to determinemagnitude from body

waves (Mb) where A is theamplitude of ground motion(microns); T is time taken

for motion (seconds); and Qis a correction for distancefrom the epicenter and thefocal depth (kilometers).

18

The terminology used to describe earthquakes is dependentupon their magnitude.

Description Magnitude EquivalentIntensity

Number perYear

Great 8+ XI-XII 1Major 7-7.9 IX-X 18Strong 6-6.9 VII-VIII 120Moderate 5-5.9 VI-VII 800

Moment-Magnitude ScaleThe moment-magnitude (Mw) scale measures the energyreleased by the earthquake more accurately than the Richterscale. The amount of energy released is related to rockproperties such as the rock rigidity, area of the fault surface andamount of movement on the fault. It provides the most accuratemeans of comparison of large earthquakes.

Distribution of Earthquakes• Earthquakes are most frequent along plate boundaries.• The largest earthquakes are associated with convergent

plate boundaries.• Oceanic ridges are characterized by shallow earthquakes.• Deep earthquakes (to depths of ~700 km) occur within

subduction zones along convergent plate boundaries.

Mw = 2/3 log10(Mo)-10.7Formula to determine

magnitude where Mo = mSd,where m is shear strength ofthe faulted rock, S is the area

of the fault, and d is faultdisplacement.

Think about it . . .Answer the conceptest question below.Three sites (L1, L2, L3) record earthquake intensity andearthquake magnitude for the same earthquake. L1 is locatedclosest to the earthquake focus and L3 is farthest away.The intensity values are greatest at _____ and the earthquakemagnitude (calculated using seismograms) _______________.a) L1; is the same at each site.b) L3; is the same for each site.c) L1; decreases with distance from the focus.d) L3; decreases with distance from the focus.

• The most devastating earthquakes are typically shallowearthquakes (0-33 km depth).

• Alaska has the largest U.S. earthquakes but California hasthe most damages because of a larger population.

• Some large historical earthquakes have been identified inthe eastern U.S., in particular a swarm of three majorquakes which occurred at New Madrid, Missouri, 1811-1812.

Where Do Earthquakes Occur?Earthquakes occur at many sites around the world butseismicity is concentrated in specific locations. A map of thePacific Ocean basin showing the location of large earthquakesover a 20-year period (1975-1995) is presented below.Compare the map with a map of plate boundaries for the samearea (Fig. 21).

Earthquake distributions have several characteristics:• There is a strong correlation between earthquake foci and

plate boundaries (Fig. 21).• Swarms of earthquakes resulting from collisions of

Figure 21. Top: Distributionof earthquake focal depthsaround the Pacific Ocean.Orange and yellow dotsrepresent shallow focaldepths (0-70 km); green andblue focal depths are 71 to300 km; purple and red dotsrepresent focal depths of 301km or greater. Bottom: Plateboundaries (yellow lines)within and around the PacificOcean. Notice thecorrelation between plateboundaries and thedistribution of earthquakefoci. Images courtesy of USGSNEIC.

19

20

continental plates form a belt across central Asia, throughthe Middle East and southern Europe.

• Continental interiors that are far removed from plateboundaries (e.g., Canada) have few earthquakes.

• A belt of shallow earthquakes can be traced along theglobal oceanic ridge system from the center of the Atlanticocean, through the Indian Ocean, around the southernPacific Ocean, and into the East Pacific.

• Earthquakes are present under hot spots such as theHawaiian Islands in the central Pacific Ocean.

• The largest earthquakes are associated with convergentplate boundaries (Fig. 22).

How Does Focal Depth Vary with Location?Epicenter locations on the maps above are colored based uponthe depths of the earthquake foci. Several patterns are obviousfrom the map:• Deep earthquakes (focal depth >300 km) are present in

association with subduction zones along convergent plateboundaries such as western South America (Nazca/SouthAmerican Plates), southern Alaska (Pacific/NorthAmerican Plates), and the southwest Pacific(Pacific/Australian Plates).

• The focal depths increase below overriding plates atconvergent boundaries in the direction of inclination of

Figure 22. Locationsand focal depths ofearthquakes ofmagnitude 7 andgreater from 1975-1998. Colorscorrespond to focaldepths (seescale). Note that themajority of largeearthquakes arelocated around therim of the PacificOcean, an areacharacterized byconvergent plateboundaries. Source:USGS NEIC.

21

subduction zones. For example, the Nazca Plate descendsbelow South America and foci increase in depth toward theinterior of the continent.

• The only area where deep earthquakes are not present alongthe Pacific Rim is in the western U.S. where a transformplate boundary exists.

• The largest earthquakes are typically shallow earthquakeswhere seismic energy is released closer to Earth's surface.

• Divergent plate boundaries such as the oceanic ridgesystems in the north Atlantic Ocean and continental riftvalleys (East Africa) are characterized by earthquake focaldepths of less than 33 km.

Where Are the Most Seismically Active Areas inNorth America?The most seismically active states are along the western marginof the continent (Fig. 23). In the U.S., Alaska and California, inthat order, experience the most earthquakes. Damage caused byearthquake activity is greatest in California because of itslarger population. Most of the largest earthquakes in U.S.history occurred on the southern coast of Alaska, along theconvergent boundary between the Pacific and North Americanplates.

The effects of earthquakes in eastern North America are feltfurther from their sources because the crust is less fractured(more rigid) than in the west. Earthquakes of comparable size

Figure 23.Seismicity in theconterminousU.S. reflected byearthquakesbetween 1977-1997.

22

in California affected a much smaller area (compare isoseismalmaps for the San Francisco and New Madrid earthquakes).Some of the largest historical earthquakes occurred in theeastern half of the continent. For example, three majorearthquakes were centered in southeastern Missouri (NewMadrid) over a three-month period from December 1811 toFebruary 1812.

States in the northern Great Plains of the U.S., such as NorthDakota, and adjacent provinces in central Canada (Manitoba,Saskatchewan) have experienced the fewest significantearthquakes.

Earthquake Prediction• Earthquakes represent the deadliest of natural hazards.• Earthquakes typically occur in areas of active faults,

especially along plate boundaries.• Earthquake magnitude increases with fault length.• Various instruments and satellite observations can be used

to measure the buildup of strain in rocks.• Scientists predict the long-term probability of earthquakes

for specific locations on the basis of information aboutstrain accumulations and recurrence interval.

• Short-term prediction, days or weeks before an earthquake,is still a long way off.

Earthquakes represent the most deadly natural hazard. Overtwo million people have been killed this century alone byearthquakes and associated phenomena. The threat of futureearthquakes in heavily populated regions like California hasspurred efforts to discover ways to predict future earthquake

Think about it . . .Examine the world map at the end of the chapter and predictwhich locations are most likely to have experienced recentearthquake activity then go to online maps (URL below) of currentseismicity to check your predictions.

http://wwwneic.cr.usgs.gov/neis/general/seismicity/seismicity.html

activity. The basic questions in earthquake prediction areWhen? Where? and How big?

A recent report by the Federal Emergency ManagementAgency (FEMA) estimated that the average annual propertydamage from U.S. earthquakes totaled $4.4 billion. Californiaalone accounted for 75% of this total. Several years may passwith few large events and little associated damage but a single

Figurereport damagearthqWest Chavingcombinfaults apopulathat magreatefrom eThe smoccursDakotaMinnesearthqwould less thannual

l

Average annual

osses from floods:$5.2 billion

Average annuallosses fromhurricanes:$5.4 billion

23

big earthquake in a large city can have a price tag of as muchas $30 billion. Population density and active seismicity havethe greatest influence over estimates of potential damages.When averaged over several decades, the potential cost ofearthquake damages for the populous eastern U.S. ranksalongside that of the more seismically active Rocky Mountainstates where population density is much lower (Fig. 24). Theupper Midwest and Great Plains states have the least risk forsignificant earthquake-related damages.

Where? How big?Answers to the Where? and How big? questions are alreadyknown in regions of frequent seismic activity. The answers tothese questions depend on an understanding of the earthquakemechanism. We have already discussed the fact thatearthquakes occur on faults. Many active faults have alreadybeen discovered but some questions remain about the potentialsize of earthquakes on faults that have no associated historicalearthquakes. For such faults, scientists attempt to estimatefuture earthquake magnitudes from fault size. Earthquakemagnitude is directly related to fault length - the longer thefault the bigger the earthquake (Fig. 25). The 1906 SanFrancisco earthquake (Mw 7.7) was caused by rupture of 400km (250 miles) of the San Andreas Fault and shaking lasted for

24. The FEMAon the potentiales fromuakes identifiedoast states as

theation of activend largetion centersy result in the

st damagesarthquakes.allest risk

in North andota where

uake damagesaccount foran $10,000ly.

24

nearly two minutes. In contrast the magnitude 6.7 Northridgeearthquake was caused by displacement on a 14 km (9 miles)long fault segment and the duration of shaking was just 7seconds.

When?Displacement on faults is related to crustal deformationassociated with plate tectonics and is concentrated in relativelynarrow zones along plate margins. Stresses build up in rockswhere plates interact. Faults exhibit movement when stressesreach sufficient levels. Rocks adjacent to the fault may bedeformed prior to fault movement. Stresses cause deformationof rocks (strain) and geologists can measure the accumulationof strain in deforming rocks in an effort to predict the timing offuture earthquakes.

Strain can be measured in the vicinity of active faults using avariety of instruments including creepmeters, strainmeters, andsatellite positioning systems. Creepmeters surveydisplacement between two points on opposite sides of a fault.As strain increases the distance between points increases.Strainmeters measure the distortion of the originally circularprofile of cylindrical boreholes as a result of deformation.Boreholes are distorted to an increasingly elliptical shape insection as strain accumulates. Satellites of the GlobalPositioning System (GPS) can be used to continually monitorthe location of receivers on the ground on either side of a fault.Distances between stations distributed over an area of hundredsof square kilometers can be determined to within a fewcentimeters. Monitoring of stations over months or yearsreveals changes in the relative positions of receivers related tothe buildup of strain along the fault.

Figure 25. Relationshipbetween earthquakemagnitude and faultsize for a series ofCalifornia earthquakes.Earthquake magnitudeincreases with faultlength.

25

Scientists can establish an average recurrence interval - thetime between earthquakes of similar magnitude - for individualfaults by determining the ages of offset layers of rocks and/orsediment. Analysis of how much time has elapsed since the lastearthquake and the amount of energy that was released(magnitude) help reveal which faults may be storing upsufficient strain for earthquakes in the relatively near future.

Probability TheoryResearchers have used statistical methods to predict theprobability of future damaging earthquakes on particularfaults with sufficient record of seismicity. Faults with a highprobability of an earthquake exhibit a lot of stored strain and along time interval without fault movement.

In 1990 a panel of experts convened by the NationalEarthquake Prediction Evaluation Council estimated a 67%probability for a major earthquake on one of four segments ofthe San Andreas fault in the San Francisco Bay area between1990 and 2020 (Fig. 26). Scientists predicted the near certainty(95% probability) of an earthquake at Parkfield, California,between 1986-1993. Parkfield, located on the San AndreasFault, averaged a magnitude 6 earthquake every 22 years since1857. Geophysicists distributed an array of monitoringinstruments around Parkfield in the 1980s hoping to pick up asignal that would aid in predicting future earthquakes.

Figure 26. Theprobability of faultmovement varies alongthe San Andreas Fault.Segments along thesouthern half of thefault system are mostlikely to break,especially at Parkfield.

2

However, the earthquake has still not occurred illustrating apotential pitfall of prediction by probability.

Probability theory assumes a random occurrence ofearthquakes but recent analyses suggest that earthquakescluster together in groups of events. For example, onemagnitude 6 or larger earthquake occurred every four years onaverage between 1836 and 1911 in and around San Francisco.There were no more earthquakes of that magnitude in the 68years that followed. However, since 1979 there have been fourmore magnitude 6 events. Scientists are now concerned that therelease of strain on one fault may increase the potential formovement on an adjacent fault in ways that cannot beaccounted for in traditional probability theory.

Even if it becomes possible to accurately predict earthquakeactivity to within a specific year, it is unlikely that individualevents can be pinpointed to within a few months, let aloneweeks or days. Furthermore, it is unlikely that we would beable to collect sufficient data to predict earthquakes in areas ofinfrequent seismic activity. Given the difficulty in predictingthe timing of future earthquakes we would be well advised tofocus instead on engineering solutions that attempt toearthquake-proof key structures.

S1Vr

Think about it . . .Following graduation you get a job working for a countyplanning task force in California. The task force must examinethe setting of several different cities and identify which is atgreatest risk for future earthquake damages from movementon known faults. You are given the assignment to create anevaluation rubric to rank the relative dangers for differentcities. Go to the evaluation rubric frame at the end of thechapter to complete the exercise.

6

ummary. What is an earthquake?ibration of Earth due to a rapid release of energy. Energy is

eleased because of rapid movement on a fault.

27

2. What is a fault?A fracture on which movement has occurred. Rapid movementof 1 to 10 meters is typically necessary to generate a significantearthquake. Faults are distinguished as dip-slip or strike-slipfaults.

3. What is the earthquake focus?The focus is the point on the fault surface where movementbegins, the earthquake source. Seismic waves radiate outwardfrom the focus. Earthquake foci occur at a range of depths;shallow (0-70 km), intermediate (70-300 km), and deep (300-700 km). Shallow earthquakes are the most common.

4. What is the earthquake epicenter?The epicenter is the geographic location of the point on Earth’ssurface directly above the focus. Earthquakes are named for theepicenter location, for example the 1994 Northridge earthquakeoccurred several kilometers below the city of Northridge inmetropolitan Los Angeles.

5. What are the differences between body waves and surfacewaves?

Seismic waves represent the energy released from theearthquake focus. There are two types of seismic waves.Surface waves travel on Earth’s surface. Undulations of theland surface during an earthquake are a representation ofsurface waves. Body waves travel through Earth’s interior.These are further subdivided into P (primary) waves and S(secondary or shear) waves on the basis of their vibrationdirection and velocity.

6. How do P and S waves differ?P waves vibrate parallel to their travel direction in the sameway a vibration passes along a slinky toy. P waves travel atspeeds of 4 to 6 km per second. S waves vibrate perpendicularto their travel direction, like the wave that passes along a ropewhen it is given a sharp jerk at one end. S wave velocity is 3 to4 km per second.

7. What is a seismogram?The record of an earthquake at a seismograph station is aseismogram. The difference in arrival time between P and Swaves on a seismogram can be used to determine the distanceof the station from the earthquake source. Furthermore, theamplitude (height) of the S wave recorded at the station can beused to determine earthquake magnitude.

28

8. What are the principal effects of an earthquake?Ground Shaking: Rapid horizontal movements associated withearthquakes. Shaking is exaggerated in areas where theunderlying sediment is weak or saturated with water. FaultUplift: Large sections of the earth’s surface (thousands ofsquare kilometers) may change elevation as a result of uplift onan earthquake fault. Liquefaction occurs when water-saturatedsediment is collapses due to violent shaking. Landslides:Earthquakes are often associated with mountains formed alongconvergent plate boundaries. The steep slopes present in theseenvironments are prone to landslides when shaken. Tsunamisare giant sea waves generated by submarine earthquakes,especially noted from the Pacific Ocean.

9. What methods can be used to measure an earthquake?There are three methods used for measuring earthquakes. TheModified Mercalli scale measures earthquake intensityrepresented by damages associated with earthquakes. TheRichter scale is the most well known and measures earthquakemagnitude using the amplitude (height) of the S-wave recordedon a seismogram. The moment-magnitude scale has recentlyfound favor as a method that more accurately measures energyrelease on large faults.

10. How is the Modified Mercalli scale used?The Mercalli scale measures earthquake intensity: the level ofdestruction of the earthquake (higher values) and the effect ofthe event on people (lower values). The scale ranks intensityfrom I to XII (1-12) using Roman numerals. Values of I to VIrepresent increasing awareness of people; VII to XII involveincreasing damages associated with the event. The Mercalliscale is not widely used for modern earthquakes because it isinaccurate in areas of low population density and in citieswhich lack stringent building codes, and has a variety of valueswith distance from the epicenter.

11. What regions of the U.S. have a history of earthquakeactivity?

Earthquakes are common in states along present-day plateboundaries (California, Alaska) and are least common in thecontinental interior (North Dakota, Minnesota). However,some ancient fault zones in Missouri (New Madrid) and SouthCarolina (Charleston) have experienced major infrequentearthquake events.

29

12. What is the difference between great, major, and strongearthquakes?

Great, major, and strong earthquakes are differentiated byRichter magnitude. Great earthquakes (magnitude 8+) are rare(average 1 per year); an average of 18 major earthquakes occurannually with a magnitude of 7 to 7.9; strong earthquakes aremore common (120 per year) with a magnitude of 6 to 6.9.

13. How is the Richter scale used?The Richter scale measures earthquake magnitude, theamplitude of S waves recorded on a seismograph following anearthquake. The Richter scale is logarithmic, each divisionrepresents a ten-fold increase in the ground motion associatedwith the earthquake, and ~30-times increase in energy released.For example, a magnitude 7 earthquake has 10-times as muchground motion (and releases over 30-times the energy) as amagnitude 6, 100 times as much motion (900 times the energy)as a magnitude 5, 1,000 times the motion of a magnitude 4, etc.

14. What controls the distribution of earthquakes?Earthquakes are concentrated in narrow seismic belts alongplate boundaries. The largest earthquakes are typicallyassociated with convergent boundaries.

15. Is there a difference in the distribution of deep and shallowearthquakes?

Deep earthquakes (to depths of 800 km) occur only inassociation with subduction zones along convergent plateboundaries. Shallow earthquakes occur along all plateboundaries.

16. Are all U.S. earthquakes confined to the active plateboundary along the western U.S.?

Most U.S. earthquakes occur in Alaska and California butseveral smaller quakes occur along old fault zones in thecontinental interior. A swarm of major earthquakes ofmagnitude 7 to 8 occurred near New Madrid, Missouri, in athree-month span from December 1811 to February 1812.

17. What factors control the size of future earthquakes?Earthquake magnitude is related to fault length. Longer faultsyield larger earthquakes that shake the ground for longerperiods. Future large earthquakes are anticipated where strainhas accumulated along faults that have not experienced recent

30

seismic activity. Scientists predict the long-term probability ofearthquakes for specific locations on the basis of informationabout strain accumulation and recurrence interval.

18. How do scientists determine the time between earthquakes?Scientists estimate the recurrence interval - time betweenearthquakes of similar magnitude - for individual faults bydetermining the ages of offset layers of rocks and/or sediment.Analysis of how much time has elapsed since the lastearthquake and the amount of energy that was released(magnitude) help reveal which faults may be storing upsufficient strain for earthquakes in the relatively near future.

31

Concept Map: Faults and Earthquakes

Finish the partially completed concept map for faults and earthquakes below. Print thepage and fill in the blanks with appropriate terms. Try to complete the map after readingthe section on faults in the this chapter. If you need some help, use some of the terms inthe list below to complete the concept map. There are more terms than spaces available.

strike-slip1-10 metersfault scarpdip-slipNew Madridhorizontallyhot spotsCaliforniaplate boundaries

volcanoes1-10 kilometersstream valleysSan Andreas fault1,000 metersfaultsAlaska1,000 kilometerssegments

32

Venn Diagram: Loma Prieta vs. San Francisco Earthquakes

Use the Venn diagram, below, to compare and contrast the similarities anddifferences between the 1989 Loma Prieta and 1906 San Francisco earthquakes.Both events occurred in the same region. Print this page and write features unique toeither group in the larger areas of the left and right circles; note features that theyshare in the overlap area in the center of the image.

o

Loma Prieta San Francisc

33

Earthquake Locations

Examine the map below and answer the following questions.1. Which location is likely to have experienced the largest number of recent

earthquakes?a) A b) B c) C d) D e) E

2. Which location is likely to have experienced the deepest recent earthquake?a) A b) B c) C d) D e) E

Go to St. Louis University’s (SLU) site to examine the distribution of earthquakesover the last 14 days or view maps of current seismicity of the world from the USGSNational Earthquake Information Center (NEIC) to check your predictions.

SLU: http://www.eas.slu.edu/Earthquake_Center/quakemaps.htmlNEIC: http://wwwneic.cr.usgs.gov/neis/current/world.html

34

Earthquake Risk Evaluation Rubric

Following graduation you get a job working for a county planning task force inCalifornia. The task force must examine the setting of several different cities and identifywhich is at greatest risk for future earthquake damages from movements on known faults.

You are given the assignment to create an evaluation rubric to assess factors that willinfluence the risk of potential damage from a future earthquake. The city that scores thehighest using the rubric will receive additional county funds to protect key structuresfrom earthquake damage. One factor is included as an example in the table below,identify four more. Consider the relationship between faults and earthquakes, thegeologic properties of the location, and cultural factors when developing your rubric.

Factors Low Risk(1 point)

Moderate Risk(2 points)

High Risk(3 points)

Proximity tofault

Far(more than 100 km)

Intermediate(20-100 km)

Close(less than 20 km)

Reviewing your evaluation rubric you realize that some factors are more significant thanothers. Your team decides to double the score of the most important factor. Which dothey choose? Why?

35

A map of the county showing the locations and characteristicsof four cities is provided below. Use your rubric to decidewhich site will receive funding to retrofit key buildings andother structures.