.New Frontiers in Integrated Solid Earth Sciences Phần 7 docx

Recent Developments in Earthquake Hazards Studies 247Fig.. GPS data are important for studies of the earthquake source process since the measurement of surface displace-ment is mathemati

Recent Developments in Earthquake Hazards Studies 247

Fig 8 The Kashiwazaki-Kariwa nuclear power plant (KKNPP), located about 10–20 km from the epicenter in the Niigata

prefec-ture This power plant was shut down after the July 16, 2007, earthquake caused damage to the plant

over thirty percent of the nation’s power All Japanese

nuclear facilities have been engineered to withstand

earthquakes of up to Mw= 6.5 In this instance,

imple-mentation of earthquake building codes in Japan’s

nuclear facilities almost certainly saved lives

• Tsunamis are another secondary effect of

earth-quakes In one well known case, the Mw = 9.2

earthquake that struck the coast of Sumatra,

Indone-sia, in December of 2004 triggered an Indian Ocean

tsunami that devastated several countries

sepa-rated by more than 4,000 miles, from Southeast

Asia to Africa The tsunami death toll exceeded

230,000 and led to the displacement of millions of

people

• A Mw = 7.9 earthquake struck eastern Sichuan,

China, on May 12, 2008, and resulted in the death

of some 89,000 people and left over a million

home-less This earthquake occurred within the

Long-men Shan region which is located at the

bound-ary between the high topography of the TibetanPlateau to the west and the relatively stable SichuanBasin to the east (Fig 9; Burchfiel et al., 1995) Theground shaking was felt over much of central, east-ern, and southern China (Fig 9) The earthquakeled to numerous landslides that buried villages andcomplicated rescue efforts by blocking transporta-tion routes Medical supplies, water, and food maynot reach isolated communities affected by the dis-aster and the inability to distribute critical suppliesmay dramatically increase the casualties

Earthquake Engineering and Building Codes

The design of buildings to sustain earthquake strongground motions is a critical step in reducing the loss

b

Fig 9 A: Location map of China and neighboring countries.

Star in center of map marks the location of the Mw = 7.9

Wenchuan (Sichuan Province) earthquake The epicenter is on

the eastern flank of the Tibetan Plateau Black line near star

marks the location of cross-section in part B: Crustal cross

sec-tion at the hypocentral locasec-tion of the Wenchuan, China, quake The thicker crust of the Tibetan Plateau is being thrust eastward over the neighboring Sichuan basin

earth-Recent Developments in Earthquake Hazards Studies 249

Fig 10 Rescue workers and

local residents search for

survivors in the rubble

following the August 15,

2007, Mw = 8.0 Pisco, Peru

earthquake Many of the

deaths and injuries occurred

in homes constructed with

highly vulnerable adobe

bricks

of life The importance of building codes was

high-lighted by the August 15, 2007, earthquake in Pisco,

Peru (USGS, 2007) Peru is a country where

tradi-tional and modern building designs are found in close

proximity Adobe buildings account for 65% of all

buildings in rural areas and nearly 35% of all

build-ings in urban areas Adobe bricks are indigenous,

sun-dried building materials consisting of sand (50–70%),

clay (15–30%), and silt (0–30%), that are often mixed

with a binding material, such as straw Adobe brick

walls are highly vulnerable to collapse when

sub-jected to severe ground shaking When the Mw= 7.9

Pisco earthquake struck, many of the adobe houses

in Pisco and Ica collapsed, whereas the modern

rein-forced concrete buildings were only superficially

dam-aged (Fig 10) There were more than 500 fatalities

due to the Pisco earthquake, and an estimated 58,000

homes (80% within the city of Pisco) were destroyed,

leaving more than 250,000 people without shelter

(Fig 10)

Disaster struck Iran in 2003, when a Mw = 6.6

earthquake ruptured along the Bam Fault in central

Iran The earthquake caused 43,000 fatalities, most of

these due to building collapse (Eshghi and Zaré, 2004)

Like Peru, the Bam area of Iran also utilizes traditional

housing constructed from adobe The tectonic setting

of the Bam, Iran, earthquake is crustal compression

and reverse faulting, as confirmed by earthquake focal

mechanisms and analogue stress models of this nental collision zone (Fig 11; Eshghi and Zaré, 2004;Sokoutis et al., 2003)

conti-It is not always the case that traditional structuresare weaker than modern designs In the 2005 Mw =7.6 Kashmir earthquake in Pakistan, western-style con-struction such as concrete block and brick masonrystructures suffered more intense damage than the tra-ditional timber-brick masonry typically used in thisregion (Naseem et al., 2005) In this case, buildingsconstructed using traditional styles and timber materi-als responded much better to ground shaking than allother building types Traditional wood-framed build-ings in Indonesia also perform much better than mod-ern brick or unreinforced concrete building A compar-ison of the 2005 Kashmir earthquake to the Pisco andBam earthquakes indicates the importance of creating

a building code appropriate for each specific region

Future Directions in Earthquake Science

Enhanced Seismic Monitoring

Seismic monitoring systems have undergone dous growth during the past twenty-five years TheGlobal Seismic Network (GSN) was initiated by the

tremen-Fig 11 Seismicity map of

Iran, with location of the

Mw = 6.6 Bam earthquake

(red star) of 2003 that caused

some 43,000 fatalities The

recurrence interval for large

earthquakes in this region is

estimated to be more than

1,000 years However, even

regions with long recurrence

intervals may be highly

vulnerable to earthquake

disasters

Incorporated Research Institutions for Seismiology

(IRIS) and now has more than 150 high-quality,

broad-band seismic stations (Fig 12) This system is operated

in collaboration with the US Geological Survey and

the University of California-San Diego Some 75% of

these stations are available in realtime using satellite

telemetry systems

Many national seismographic systems have also

been upgraded The disastrous 1995 Kobe earthquake

in Japan led to major upgrades in the seismic

moni-toring systems in that country These include a

high-sensitivity seismic array with 698 stations, a

broad-band array with 74 stations (F-net) called Hi-net and

a strong-motion network with 1,043 accelerometers

The high-sensitivity array can rapidly and accurately

locate earthquakes; the broadband array provides data

on the earthquake source; and the strong motion array

provides earthquake engineering data (as well as

infor-mation about the source) A similar program of

net-work upgrades has been completed in Taiwan In

main-land China, there are more than two thousand

short-period seismographs, two hundred broadband stations

and more than four hundred accelerometers In Europe,

a federation of national seismic systems, and national data collection program (e.g., ORFEUS andGEOSCOPE) provide abundant realtime data In theUnited States, the Advanced National Seismic Sys-tem (ANSS) is a comprehensive system that providesrealtime seismic data from seismic sensors located

inter-in the free field and inter-in buildinter-ings Similar to othernational networks, instrumentation includes a network

of broadband sensors, accelerometers and high-gainseismic stations The total number of sensors exceeds7,000 in number, and the system automatically broad-casts information when a significant event occurs Sig-nificant network upgrades have taken place in Mexico,Thailand, and Malaysia

Global Positioning Systems (GPS)

Global Positioning Satellite (GPS) technology candetect minute motions of the Earth’s crust that increasethe stress on active faults and eventually leads toearthquakes (Segall and Davis, 1997) This technology

Recent Developments in Earthquake Hazards Studies 251

provides an excellent picture of how slip (or ground

displacement) can accumulate on faults throughout

the earthquake cycle (e.g., Bakun et al., 2005)

Satel-lites deployed across the globe emit precisely timed

radio signals to tracking stations on the ground that

record both gradual, aseismic motion as well as

sudden displacements during earthquakes GPS

net-works may be deployed in campaign (temporary)

and permanent modes, but the decreasing cost and

widespread use of this technology has been

shift-ing more deployments to permanent status (Jordan,

2003) These data help in estimating earthquake

poten-tial, identifying active blind thrust faults and

deter-mining the potential response of major faults to the

regional change in strain As well, the ability of GPS

technology to provide a measurement of the total

slip caused by an earthquake complements traditional

seismological methods of determining earthquake

magnitude

GPS measurements of crustal deformation are

avail-able for nearly all active tectonic environments These

data provide new and more accurate maps of the

present crustal deformation field, a fundamental

mea-surement of active continental tectonics GPS data

are important for studies of the earthquake source

process since the measurement of surface

displace-ment is mathematically related to a dislocation on a

fault in an elastic medium This relation permits the

inversion of the geometry of the earthquake rupture

Such an inversion is more reliable when performed

using near-field strong motion data (e.g., Bakun et al.,

2005)

GPS data are also useful for the study of postseismic

processes The 1989 Loma Prieta, California,

earth-quake showed postseismic strain with a

characteris-tic decay transient of 1.4 years (Savage et al., 1994)

These authors report, contrary to expectations, that

the transient parallel to the fault is smaller than the

transient perpendicular to the fault The interpretation

of this observation is still debated

GPS and older geodetic data have been used in a

search for precursory crustal deformation prior to large

earthquakes Slow precursors were found for eight

con-vergent margin earthquakes, including the 1960 9.2 M

Chile, 1964 9.2 M Prince William Sound, Alaska,

and the 1,700 Cascadian earthquakes (Roeloffs, 2006)

On the other hand, no pre-seismic deformation was

detected for the following terrestrial earthquakes: 2004

6.0 M Parkfield, 1992 7.3 M Landers, 2003 8.1 M

Tokachi-oki (Irwan et al., 2004), and 1999 7.1 M tor Mine earthquakes (Mellors et al., 2002) Since slowcreep can go entirely undetected unless high qualityGPS array data are available, it is presently inconclu-sive how often earthquakes are preceded by slow aseis-mic slip This is an important research topic

Hec-Interferometric Synthetic Aperture Radar (InSAR)

InSAR is a recent, innovative technology that permitsthe imaging of earthquake (crustal) deformation down

to the millimeter scale (Wright et al., 2001a) Similar

to GPS measurements, radar waves are emitted fromsatellites across the globe to the Earth’s surface In thecase of InSAR, these radio waves are reflected from theground surface and returned to the satellite The satel-lite is sensitive to both: (1) the intensity of the returningelectromagnetic wave, which has a different signaturedepending on the nature of the ground material, and (2)the phase of the returning wave, which will have beenaltered if ground displacement has taken place betweensuccessive passes of the satellite over the same loca-tion This technology opens the door to continuouslymapping deformation along active plate boundariesover larger areas and in greater detail than can prac-tically be monitored by GPS measurements InSARderived interferograms have successfully been used toacquire a rapid map of surface deformation after anearthquake, such as the 1999 Izmit earthquake and intracking interseismic strain accumulation along a largesection of the Northern Anatolian Fault through minutemeasurements of surface displacement over a nearlydecadal timescale (Wright et al., 2001b; Fig 13).InSAR techniques are also effective in measur-ing deformation on active volcanoes and landslides,both of which are significant geological hazards Forexample, magma movement can be detected at other-wise apparently dormant volcanoes As more InSARsatellites come into orbit, the capability has emerged

to make measurements more frequently, and therebymake greater use of the technique as a monitor-ing tool InSAR measurements of fault slip comple-ment determinations made using seismic and GPSmeasurements, and generally cover a wider geographicarea

Recent Developments in Earthquake Hazards Studies 253

Fig 13 Radar interferogram

for the Izmit earthquake (data

copyright ESA) revealing the

surface displacements,

measured in the satellite’s

line-of-sight, in the 35-day

period between the two image

caused by pure horizontal

motion Red lines are the

mapped surface rupture

[Barka, 1999] and the dashed

lines are previously mapped

segments of the North

Anatolian Fault [ ¸Saroglu

et al., 1992] (after Wright

et al., 2001a)

Shakemaps of Seismic Intensities

Seismic intensity is a measurement of the severity of

an earthquake’s effects at different sites The Modified

Mercalli Intensity (MMI) scale ranges from Roman

numeral I to XII, the highest level being total

destruc-tion The MMI scale predates instrumental recordings,

and is derived from field observations of damage The

intensity for historical earthquakes can also be

deter-mined from newspaper accounts, diaries, and other

documents The local intensity of an earthquake is of

greater importance than the earthquake magnitude to

those who manage emergency response because the

intensity directly relates to damage effects

A recent key development by the U.S

Geologi-cal Survey and its partners is an online system that

provides near-real-time post-earthquake information

regarding ground shaking Shakemap (Wald et al.,

2003) provides a map view of the ground shakingintensity in the region of an earthquake based on mea-surements from seismometers Whereas an earthquakehas a unique location and magnitude, the intensity ofground shaking it produces depends on such factors asthe distance from earthquake, local site conditions andseismic wave propagation effects due to complexities

in the structure of the Earth’s crust Shakemap ware produces near real-time intensity maps for earth-quakes, such as the May 12, 2008 Mw 7.9 EasternSichuan earthquake in China (Fig 14) The widespreadavailability of such maps through the internet is valu-able for the coordination of emergency response teams.The ground-shaking of hypothetical future earthquakescan also be evaluated, as well as the damage thatwould be associated with them today ShakeMapthus serves as a useful, predictive tool by simulat-ing the seismic intensity related to hypothetical futureearthquakes

soft-Fig 14 Seismic shaking

intensity map produced by the

USGS shortly after the

Mw = 7.9 May 12, 2008,

Wenchuan, China The map

correctly indicated that a high

population density NW of the

epicenter were subjected to

violent-to-extreme ground

shaking intensities Such

maps, which are produced by

processing data from local

seismographs, are useful in

planning earthquake

emergency response

Earthquake Forecasting vs Earthquake

Prediction

Earthquake prediction refers to the ability to calculate

the specific magnitude, place and time for a particular

future earthquake, similar to how meteorologists can

now forecast an oncoming hurricane or tornado on a

short timescale The current state of earthquake science

precludes any ability to truly predict specific future

earthquakes Earthquake forecasting, refers to

model-ing the probabilities that earthquakes of specified

mag-nitudes, and faulting types will occur during a

speci-fied time interval (usually several years) on a specific

fault segment Such probability estimates, when lated over a specific time interval, are known as time-

calcu-dependent earthquake forecasting Time-incalcu-dependent

forecasting, also known as long-term forecasting, is

a general assessment of the likelihood of faults torupture, not over a specific timeframe, and does nottake into account whether earthquakes have occurredrecently on particular faults Time-independent fore-casting is frequently used to evaluate building codesand developments or projects that must be sustainable

in the long-term A comparison between short-termand time-independent forecasting models can be found

in Helmstetter et al (2006) In order to calculate eithertype of earthquake probability forecast, a variety of

Recent Developments in Earthquake Hazards Studies 255data are assembled and analyzed, including earthquake

recurrence intervals from paleoseismic, historical

and instrumental records, deformation and slip rates

from GPS and InSAR and long-term plate-tectonic

models

In 2007, the Working Group on California

Earth-quake Probabilities (WGCEP) developed a state-wide

rupture (time-dependent) forecast called the Uniform

California Earthquake Rupture Forecast (UCERF)

This probability map specifies the likelihood of a Mw

> 6.7 earthquake striking California over the next

30 years (Field et al., 2008; Fig 15) Such

prob-ability maps are critical to ensure public safety in

regions of high seismic hazard such as California or

Alaska The UCERF forecast will be used by the

California Earthquake Authority (CEA) to analyzepotential earthquake losses, set earthquake insurancepremiums and develop new building codes

Earthquake Early Warning

It is evident from the preceding review that muchprogress has been made in understanding earthquakes.Nevertheless, routine short-term earthquake predictionhas not been achieved Indeed, it will likely requiremany decades of additional research to address thisproblem Therefore, it is useful to ask if it is feasible toprovide an early warning of impending strong ground

Fig 15 Probabilistic earthquake hazard map for the State of

California, USA, showing in yellow and orange those regions

with higher probabilities for an earthquake with Mw ≥ 6.7 in

the next 30 years Boxes outlined in white located the Greater San Francisco and Greater Los Angeles areas with high seismic risk

motion based on an automated earthquake monitoring

system, much as the lowered gates and flashing red

lights at a railroad crossing announce the imminent

arrival of a train

Rapid earthquake notification is distinct from early

warning systems The former is a broadcast system that

exists in many seismic networks that provides

earth-quake information within minutes after an earthearth-quake

occurs In contrast, an early warning system provides

an alert within seconds of the initial rupture of a

sig-nificant (Mw ≥ 5) earthquake, indicating that strong

ground shaking can be expected A warning that

pro-vides only some tens of seconds of advanced notice

of incoming strong ground motion may appear

incon-sequential, but in fact it would allow enough

warn-ing time for critical systems (e.g., high-speed trains)

to be shut down as well as for mobile individuals to

take protective cover from falling objects The

phys-ical basis for such a system, first realized by Cooper

(1868), is the fact that electromagnetic signals (radio

and internet communications) travel faster than

elas-tic waves Additionally, the first arriving P-waves have

much lower ground motions than the later arriving

sur-face waves

Earthquake early warning systems need to estimate

the potential magnitude of an earthquake within the

first few seconds of the rupture process (Ellsworth and

Beroza, 1995; Beroza and Ellsworth, 1996) The

fea-sibility of such a system requires that there be

suffi-cient information in the first-arriving

compressional-wave (P-compressional-wave) at local seismic stations to estimate the

potential size of the earthquake using empirical

rela-tions (Allen and Kanamori, 2003; Kanamori, 2005)

Test cases show that there is a strong correlation

between earthquake magnitude and the frequency

con-tent of the initial few seconds of the seismogram Early

warning systems use the information contained in the

initial portions of the seismic waveforms (the P-wave

arrival) to estimate the eventual magnitude of the

earth-quake This method of waveform analysis, as well as

other methods (Cua and Heaton, 2003), can provide

robust earthquake early warnings, especially in densely

instrumented regions, such as Japan, Taiwan, Europe,

and California

Earthquake early warning systems have already

been successfully operated Mexico successfully

issued an early warning to the public with their Seismic

Alert System (SAS) during the Mw= 7.3 September

14, 1995 Copala earthquake that occurred on the

sub-duction zone at the west coast, some 300 km fromMexico City Over 4 million people in the city werewarned The success was in part due to the fact that theearthquake occurred during the day, when the majority

of people were awake and had access to radios (Leeand Espinosa-Aranda, 1998) The Mexican SeismicAlert System consists of four units: seismic detection,telecommunications, central control, and early warn-ing The field stations are located 25 km apart, eachmonitor a region 100 km in diameter, and can estimatethe magnitude of an earthquake within 10 s of its initi-ation Other early warning systems have been installed

in several other countries, including Japan, Taiwan,and Turkey (Lee et al., 1998) In view of the difficulty

of achieving short-term earthquake predictions, quake early warning, like improved building codes, can

earth-be expected to play an increasingly important role inmitigating earthquake affects

Exam-in Tangshan, ChExam-ina; and 2004 Exam-in Sumatra-AndamanIslands in Indonesia and India We have highlightedsome key concepts such as the earthquake cycle andrecurrence intervals that are used in describing theunderlying cause of earthquake We have summarizedsome lessons learned from the earthquake record, such

as the larger geographical area that experiences highseismic intensities for earthquakes that occur in conti-nental interiors Finally, we have described five impor-tant advances that have been made that have greatlyenhanced our ability to monitor, report, and respond tolarge, damaging earthquakes These five advances are:(1) enhanced seismic monitoring and notification; (2)GPS and (3) InSar monitoring; (4) the introduction ofShakemaps, and (5) progress in earthquake forecastingand early warning

Have these steps succeeded in reducing earthquakehazards? The answer is certainly “Yes” Will thesesteps ensure a reduction in worldwide losses for theforeseeable future? The answer to this question is

“Maybe” The reason for this equivocal answer is

Recent Developments in Earthquake Hazards Studies 257multifold As shown by the 2005 Kashmir, Pakistan,

and 2008, Sichuan, China, earthquakes, rapid

popu-lation growth is resulting in the urbanization of

for-merly remote, seismically hazardous regions These

two earthquakes were in relatively isolated

moun-tainous regions, but both caused more than 70,000

deaths, with millions left homeless Due to the

moun-tainous topography, massive landslides and rock falls

were an additional, secondary effect, in addition to

the severe ground shaking Secondary hazards pose

a severe threat to all affected by an earthquake The

December 26, 2004 tsunami is a prime example; the

tsunami wave was a secondary hazard produced by the

Mw= 9.2 earthquake that occurred at the Sumatra

sub-duction zone Another example occurred in Chimbote,

Peru in 1971, when a Mw= 7.9 earthquake struck off

the coast of Peru, and a landslide buried the city of

Yungay and thousands of its residents These examples

should serve as reminders that one should prepare not

only for earthquake strong ground motion, but also its

secondary repercussions for example

Remote and mountainous regions are not the only

population centers at heightened risk The growth of

the world’s mega-cities also presents a great

chal-lenge There are twenty-two cities that have a

pop-ulation greater than 10 million; the list of cities at

risk from earthquakes includes: Tokyo, Lima, Los

Angeles, Mexico City, Beijing, Dhaka, Istanbul, bai, Karachi, and Tehran These ten cities comprisemore than 120 million residents Landslides and rockfalls are not the main hazard in these megacities, build-ing collapse and fires are For this reason, the miti-gation of earthquakes hazards in these cities depends

Mum-on the adoptiMum-on and enforcement of adequate buildingcodes and fire safety These new codes will require avery significant financial investment, one that in manycases may be beyond the reach of the local or nationalgovernment

Although some earthquakes have long recurrenceintervals (≥ 1,000 years), they are still extremely dan-gerous – possibly even more than those with shortrecurrence intervals because these communities mayfail to plan for earthquakes which they believe willnot happen in their lifetime Unfortunately, this wasthe case in the Sichuan province in China on May 12,

2008 The Longmen Shan fault system has a recurrenceinterval of about 2,000 years As a result, the publicwas not prepared to deal with an earthquake of suchgreat magnitude (Fig 16)

Contrary to expectations, the death toll caused by anearthquake does not always correspond with the mag-nitude Such was the case for the December 26, 2003(Mw= 6.6) Bam, Iran earthquake that killed approxi-mately 31,000 people, injured 30,000, and left 75,600

Fig 16 Ruins of an acient

temple in Hanwang-Mianzhu,

Sichuan, China The

destruction of the Wenchuan,

China, earthquake reinforces

the point that a long

recurrence interval is not

equivalent to safety This

temple withstood hundreds of

years of environmental forces

but was destroyed by the

Mw = 7.9 earthquake on May

12, 2008 Source: Sarah

Bahan, USGS

homeless Some 85% of buildings in Bam were

dam-aged or destroyed as a result of severe ground shaking,

the main cause of damage to buildings and

infrastruc-ture Furthermore, the proximity of a region to the

seis-mic source is very important The closer to the

epicen-ter, the stronger the ground shaking will be

Technological improvements hold the promise for

reducing losses due to earthquakes These

improve-ments can be divided into two types: (1) better

moni-toring and risk assessment before an earthquake

hap-pens, and (2) improved reporting and response after

the event occurs Better seismic monitoring and GPS

systems provide critical data for improved risk

assess-ments After an earthquake occurs, Shakemaps can be

used to quickly and efficiently alert the government,

media, and general public of the hazard InSAR maps,

if available quickly, provide a comprehensive picture

of the region affected Such technological advances

will make it possible to continue to improve

seis-mic hazard assessments, monitoring, mitigation and

response

References

Allen R.M and H Kanamori, 2003, The potential for earthquake

early warning in southern California Science 300, 786–789.

Bakun W, Aagaard B, Dost B, Ellsworth W, Hardbeck J,

Har-ris R, Ji C, Johnston M, Langbein J, Lienkaemper J, Michael

A, Nadeau R, Reasenburg P, Reichle M, Roeloffs E, Shakai

A, Simpson R and Waldhauser F, 2005, Implications for

pre-diction and hazard assessment from the 2004 Parkfield

earth-quake Nature V 437, p 969–974.

Barka A., 1999, The 17 August 1999 Izmit earthquake Science

285, 1858–1859.

Belardinelli M.E., Bizzarri A., Cocco M., 2000, Earthquake

triggering by static and dynamic stress changes Eos Trans.

AGU Fall Meet Suppl, 81, 48.

Beroza G.C and W.L Ellsworth, 1996, Properties of the seismic

nucleation phase Tectonophysics 261, 209–227.

Bolt B., 2006, Earthquakes Fifth edition W H Freeman and

Company, New York.

Brace W., and Kohlstedt D., 1980, Limits on lithospheric stress

imposed by laboratory measurements J Geophys Res 85,

6248–6252.

Burchfiel B.C., Z Chen, Y Liu, L.H Royden, 1995, Tectonics of

the Longmen Shan and adjacent regoins, central China Int.

Geol Rev 37, 8, edited by W.G Ernst, B.J Skinner, L.A.

Taylor.

Calais E., Han J.Y., DeMets C., d Nocquet J.M., 2006

Deforma-tion of the North American plate interior from a decade of

continuous GPS measurements J Geophys Res 111, doi:

10.1029/2005JB004253.

Campbell D.L., 1978 Investigation of the stress-concentration

mechanism for intraplate earthquakes, Geophys Res Lett.,

5(6), 477–479.

Chandrasekhar D.V and Mishra D.C., 2002 Some geodynamic aspects of Kutch basin and seismicity: An insight from grav-

ity studies, Curr Sci India, 83(4), 492–498.

Cloetingh S., 1982 Evolution of passive margins and initiation

of subduction zones, Ph.D thesis, Utrecht Univ., lands.

Nether-Cooper J.D., 1998 Letter to the Editor, San Francisco Daily Evening Bulletin, Nov 3, 1868 (as quoted in Lee and Espinosa-Aranda, 1998).

Cua G and T.H Heaton, 2003 An envelope-based paradigm for

seismic early warning, (abstract) Trans Am Geophys Union

84, F1094–1095 Cutcliffe C H 2000 Earthquake resistant building design codes

and safety standards: The California experience Geophys J.,

51, 259–262.

Dragert H and Hyndman R., 1995 Continuous GPS monitoring

of elastic strain in the northern Cascadia subduction zone.

Geophys Res Lett 22:755–758.

Dragert H; Wang K; James TS, 2001 A Silent Slip Event on the

Deeper Cascadia Subduction Interface Science 292(5521):

1525–1528.

Ellsworth W.L and G.C Beroza, 1995 Seismic evidence for an

earthquake nucleation phase Science 268, 851–855.

Eshghi S and Zaré M 2004 Preliminary observations on the

Bam, Iran, earthquake of December 26, 2003 EERI Special

203, SCEC Contribution #1138.

Frankel A., Mueller C., Barnhard T., Perkins D., Leyendecker E., Dickman N., Hanson S., Hopper M., 1996 National Seismic Hazard Maps: Documentation June 1996, U.S Geological Survey Open-File Report 96-532, Denver, CO, 111 pp Freed A.A., 2005, Earthquake triggering by static, dynamic, and

postseismic stress transfer, Ann Rev Earth Plant Sci., 33,

335–367.

Freed A M., Ali S T and Burgmann R 2007 Evolution of stress

in Southern California for the past 200 years from coseismic,

postseismic and interseismic stress changes Geophys J Int.,

169, 1164–1179.

Gangopadhyay A and Talwani P., 2003 Symptomatic features

of intraplate earthquakes, Seism Res Lett., 74, 863–883 Goetze C., 1978, The mechanisms of creep in olivine, Phil.

Trans Roy Soc Lond A., 288, 99–119.

Goodacre A.K and Hasegawa H.S., 1980 Gravitationally

induced stresses at structural boundaries, Can J Earth Sci.,

17, 1286–1291.

Gordon R.G., 1995 Plate motions, crustal and lithospheric

mobility, and paleomagnetism: prospective viewpoint J.

Geophys Res., v 100, p 24367–24392.

Gordon R.G and Stein S., 1992 Global tectonics and space

geodesy Science, v 256, p 333–342.

Gupta H.K., Rastogi B.K., and Narain H., 1972, Common

fea-tures of the reservoir-associated seismic activities, Bull Seis.

Soc Am., 62, 481–492.

Recent Developments in Earthquake Hazards Studies 259 Grollimund B and Zoback M.D., 2001 Did deglaciation trig-

ger intraplate seismicity in the New Madrid seismic zone?,

Geology, 29(2), 175–178.

Helmstetter A., Kagan Y.Y., and Jackson D.D., 2006

Compari-son of Short-Term and Time-Independent Earthquake

Fore-cast Models in Southern California Bull Seismol Soc Am.,

v 96, no 1, p 90–106.

Hinze W.J et al., 1988 Models for Midcontinent tectonism: an

update, Rev Geophs., 26(4), 699–717.

Hirose H., et al 1999 A slow thrust slip event following the two

1996 Hyuganada earthquakes beneath the Bungo Channel,

southwest Japan Geophs Res Lett 26(21): 3237–3240.

Irwan M., Kimata F., Hirahara K., Sagiya T., Yamagiwa A.,

2004 Measuring ground deformations with 1-Hz GPS

data: the 2003 Tokachi-oki earthquake (preliminary report).

Earth Planets Space, 56: 389–393 http://news

thomas-net.com/IMT/archives/2007/07/earthquake_japan_nuclear_

kashiwazaki-kariwa_plant_industry_automotive_production_

halt.html

Johanson I.A., Fielding E.J., Rolandone F., and Buergmann R.,

2006 Coseismic and postseismic slip of the 2004 Parkfield

earthquake from space-geodetic data Bull Seismol Soc.

Am., 96(4B):S269–S282.

Johnston A.C and Kanter L.R., 1990 Earthquakes in stable

con-tinental crust, Sci Am., 262(3), 68–75.

Johnston A C., Coppersmith K J., Kanter L.R and Cornell

C.A., 1994 The earthquakes of stable continental regions:

assessment of large earthquake potential, TR-102261, Vol.

1–5, ed Schneider J.F., Electric Power Research Institute

(EPRI), Palo Alto, CA.

Jordan T 2003 Living on an Active Earth: Perspectives on

Earthquake Science The National Academies Press:

Wash-ington, D.C.

Kanamori H., (2005) Real-time seismology and earthquake

damage mitigation Ann Rev Earth Planet Sci 33, 195–214.

Kayal J.R., Zhao D., Mishra O.P., De R., Singh O.P., 2002, The

2001 Bhuj earthquake: Tomographic evidence for fluids at

the hypocenter and its implications for rupture nucleation,

Geophys Res Lett., 29, 2152–2155.

Kenner S.J and Segall P., 2000 A mechanical model for

intraplate earthquakes; application to the New Madrid

seis-mic zone, Science, 289(5488), 2329–2332.

King G.C.P., Stein R.S and Lin J., 1994, Static stress changes

and the triggering of earthquakes, Bull Seismol Soc Am.,

84, 935–953.

Kostoglodov V., et al., 2003 A large silent earthquake in

the Guerrero seismic gap, Mexico Geophs Res Lett Col.

32(15): 1807, doi:10.1029/2003GL017219.

Lawson A.C (Ed.), 1908, The California earthquake of April 18,

1906 Report of the State Earthquake Investigation

Commis-sion (reprinted in 1969 by the Carnegie Institution of

Wash-ington, D.C.).

Lee W H K and Espinosa-Aranda J M., 1998 Earthquake

Early Warning Systems: Current Status and Perspectives

in Early Warning Systems for Natural Disaster Reduction.

J Zschau and A N Kuppers, Eds Springer, New York:

834 pp.

Lee W.H.K and Espinosa-Aranda J.M., 2003 Earthquake early

warning systems: Current status and perspectives: in ‘Early

Warning Systems for Natural Disaster Reduction’, J Zschau

and A N Kuppers, Eds Springer, Berlin: pp 409–423.

Liu L and Zoback M.D., 1997 Lithospheric strength and

intraplate seismicity in the New Madrid seismic zone,

Tec-tonics, 16(4), 585–595.

Long L.T., 1988 A model for major intraplate continental

earth-quakes, Seism Res Lett., 59(4), 273–278.

Lowry AR; Larson KM; Kostoglodov V; Bilham R Transient

fault slip in Guerrero, southern Mexico Geophys Res Lett.,

Murray J and Langbein J., 2006 Slip on the San Andreas Fault

at Parkfield, California, over two earthquake cycles, and the

implications for seismic hazard Bull Seismol Soc Am.,

subduction zone Nature 378: 371–374.

Quinlan G., 1984 Postglacial rebound and the focal

mecha-nisms of eastern Canadian earthquakes, Can J Earth Sci.,

21, 1018–1023.

Raphael A., and Bodin P., 2002, Relocating aftershocks of the

26 January 2001 Bhuj earthquake in western India Seis Res.

Lett 73, 417–418

Reid H.F., 1910, The Mechanics of the Earthquake, The nia Earthquake of April 18, 1906, Report of the State Investi- gation Commission, Vol.2, Carnegie Institution of Washing- ton, Washington, D.C.

Califor-Roeloffs E, May 2006, Evidence for Aseismic

Deforma-tion Rate Changes Prior to Earthquakes Annual Review

of Earth and Planetary Sciences Vol 34: 591–627.

(doi:10.1146/annurev.earth.34.031405.124947) Rogers G., and Dragert H., 2003 Episodic Tremor and Slip on the Cascadia Subduction Zone: The Chatter of Silent Slip.

Science 300(5627): 1942–1943.

¸Saroglu F., ˝ O Emre, and I Ku¸sçu, Active Fault Map of Turkey, General Directorate of Mineral Research and Exploration (MTA), Eski¸sehir Yolu, 06520, Ankara, Turkey, 1992 Satake K and Atwater B.F., 2007 Long-Term Perspectives on

Giant Earthquakes and Tsunamis at Subduction Zones Annu.

Rev Earth Planet Sci 35:349–374.

Savage J.C., Lisowski M., Prescott W.H., 1981 Geodetic

strain meansurements in Washington J Geophys Res 86:

4929–4940.

Savage J.C., Lisowski M., Svarc J., 1994, Postseismic

deforma-tion following the 1989 ( M = 7.1) Loma Prieta, California,

earthquake, J Geophys Res., 99, 13757–13765.

Sbar M.L and Sykes L.R., 1973 Contemporary compressive

stress and seismicity in eastern North America: An example

of intra-plate tectonics, Geol Soc Am Bull., 84, 1861–1882.

Schulte S.M and Mooney W.D., 2005 An updated global

earth-quake catalogue for stable continental regions:

reassess-ing the correlation with ancient rifts Geophys J Int 161,

707–721.

Segall P and Davis J.L., 1997, GPS applications for

geodynam-ics and earthquake studies, Ann Rev Earth Planet Sci., 25,

301–336.

Shishikura M., 2003, Cycle of interpolate earthquakes along the

Sagami Trough deduced from tectonic geomorphology, Bull.

Earthquake Res Inst Univ Tokyo 78, 245–254.

Simpson D.W., 1986, Triggered earthquakes, Ann Rev Earth

Plant Sci., 14, 21–42.

Sokoutis D et al Insights from scaled analogue modeling into

the seismotectonics of the Iranian region Tectonophysics

Volume 376, Issues 3–4, 4 December 2003 pp 137–149

Stein R.S., Barka A.A., and Dieterich J.H., 1997, Progressive

failure on the North Anatolian fault since 1939 by earthquake

stress triggering, Geophys J Int., 128, pp 594–604.

Stein R.S., 1999, The role of stress transfer in earthquake

occur-rences, Nature402, 605–609.

Stein R.S., Toda S., Parsons T And Grunewald E., 2006, A

new probabilitstic seismic hazard assessment for greater

Tokyo, Phil Trans R Soc A, 1965–1988, doi:10.1098/rsta.

2006.1808.

Stein S., Sleep N., Geller R.J., Wang S.C and Kroeger G.C.,

1979 Earthquakes along the passive margin of eastern

Canada, Geophys Res Lett., 6(7), 537–540.

Sykes L.R., 1978 Intraplate seismicity, reactivation of

preexist-ing zones of weakness, alkaline magmatism, and other

tec-tonism postdating continental fragmentation, Rev Geophys.,

16(4), 621–688.

Tagare G V 2002 Earthquakes: Some ancient speculations.

Speech given at the Institute for Oriental Study, Thane, India.

http://www.orientalthane.com/speeches/gvtagare/1.html

Talwani P., 1988 The intersection model for intraplate

earth-quakes, Seism Res Lett., 59(4), 305–310.

Talwani P., 1999 Fault geometry and earthquakes in continental

interiors, Tectonophysics, 305, 371–379.

Talwani P and Rajendran K., 1991 Some seismological and

geometric features of intraplate earthquakes, Tectonophysics,

186, 19–41.

Tuttle M.P., Schweig III, E.S., Campbell J., Thomas P.M., Sims

J.D., Jafferty III, R.H., 2005, Evidence for New Madrid

earthquakes in

Tse S., and Rice J., 1986 Crustal Earthquake Instability in

Rela-tion to the Depth VariaRela-tion of FricRela-tional Slip Properties, J.

Geophys Res., 91(B9), 9452–9472.

United Nations Population Division World population prospects: the 2002 Revision.http://www.un.org/popin/data html

USGS Earthquake Hazard Summary, Magnitude 8.0 – NEAR THE COAST OF CENTRAL PERU, 2007 http://earthquake.usgs.gov/eqcenter/eqinthenews/2007/us2007 gbcv/#summary

USGS earthquake summary for the July 16, 2007 earthquake

in Japan http://earthquake.usgs.gov/eqcenter/eqinthenews /2007/us2007ewac/

Vinnik L.P., 1989 The origin of strong intraplate

earth-quakes, translated from O prirode sil’nykh vnutrilplitovykh

zemletyaseniy, Doklady Akademii Nauk SSSR, 309(4), 824–827.

Wald D., Wald L., Worden B., and Goltz J., 2003 ShakeMap –

A tool for earthquake response, US Geological Survey Fact Sheet FS-087-03 (available at: http://pubs.usgs.gov/fs/fs- 087-03/)

West M., Sanches J.J., McNutt S.R., 2005 Periodically Triggered Seismicity at Mount Wrangell, Alaska, After

the Sumatra Earthquake Science Vol 308, No 5725,

Xu X., Yu G., Klinger Y., Tapponnier P & Van Der Woerd J.,

2006 Reevaluation of surface ruion of the 2001 shan earthquake (M w 7.8), northern Tibetan Plateau, China,

Kunlun-J Geophys Res., 111, B05316, doi:10.1029/2004JB00

3488.

Zoback M.L and Zoback M.D., 1980, State of stress in

the conterminous United States, J Geophys Res., 85,

6113–6156.

Zoback M.D., 1983 Intraplate earthquakes, crustal deformation

and in-situ stress, in A workshop on ‘The 1886 Charleston,

South Carolina, earthquake and its implications for today’,

Open-File Report, No 83–843, pp 169–178, eds Hays, W.W., Gori, P.L and Kitzmiller, C., US Geological Survey, Reston, VA.

Zoback M.D and Townend J., 2001, Implications of static pore pressure and high crustal strength for the

hydro-deformation of intraplate lithosphere, Tectonophysics 336,

19–30.

Zoback M.L., 1992, First- and second-order patterns of stress

in the lithosphere: the world stress map project J Geophys.

Passive Seismic Monitoring of Natural and Induced

Earthquakes: Case Studies, Future Directions

and Socio-Economic Relevance

Marco Bohnhoff, Georg Dresen, William L Ellsworth, and Hisao Ito

Abstract An important discovery in crustal

mechan-ics has been that the Earth’s crust is commonly stressed

close to failure, even in tectonically quiet areas As

a result, small natural or man-made perturbations to

the local stress field may trigger earthquakes To

understand these processes, Passive Seismic

Moni-toring (PSM) with seismometer arrays is a widely

used technique that has been successfully applied to

study seismicity at different magnitude levels ranging

from acoustic emissions generated in the laboratory

under controlled conditions, to seismicity induced by

hydraulic stimulations in geological reservoirs, and up

to great earthquakes occurring along plate boundaries

In all these environments the appropriate deployment

of seismic sensors, i.e., directly on the rock sample,

at the earth’s surface or in boreholes close to the

seis-mic sources allows for the detection and location of

brittle failure processes at sufficiently low

magnitude-detection threshold and with adequate spatial

reso-lution for further analysis One principal aim is to

develop an improved understanding of the physical

processes occurring at the seismic source and their

relationship to the host geologic environment In this

paper we review selected case studies and future

direc-tions of PSM efforts across a wide range of scales and

environments These include induced failure within

small rock samples, hydrocarbon reservoirs, and

natu-ral seismicity at convergent and transform plate

bound-aries Each example represents a milestone with regard

to bridging the gap between laboratory-scale

large-Keywords Earthquakes · Passive Seismic ing · Borehole Seismology · Crustal mechanics ·Physics of Faulting

monitor-Introduction

Global monitoring of seismicity detects the occurrence

of earthquakes down to about M = 4 The resultingpattern of their distribution traces the plate bound-aries and highlights the most active intraplate seismiczones In many parts of the globe the detection thresh-old is lower because of the presence of regional andlocal seismic networks Within regions such as the westcoast of North America, Japan and Western Europeregional thresholds on the order of M = 1–2 havebeen achieved Such well-designed local seismic net-works not only record earthquake activity at low mag-nitude detection thresholds but also resolve the focaldepths and focal mechanisms of earthquakes The tools

of modern Passive Seismic Monitoring, referred to asPSM in the following, allows the refinement of earlier

Marco Bohnhoff now at Helmholt - Centre Potsdam GFZ (bohnhoff@gfz-potsdam.de)

261

S Cloetingh, J Negendank (eds.), New Frontiers in Integrated Solid Earth Sciences, International Year of Planet

Earth, DOI 10.1007/978-90-481-2737-5_7, © Springer Science+Business Media B.V 2010

seismotectonic models pioneered in the 1970s through

the application of modern methods for determining

crustal structure, locating earthquakes and

determin-ing focal mechanisms These include 3-D mappdetermin-ing of

active faults and fault systems, routine moment tensor

determination of source processes, analysis of

earth-quake interaction, high-resolution characterization of

active faults within hydrocarbon and geothermal

reser-voirs, and investigation of the systematics of the

earth-quake cycle for large magnitude earthearth-quakes along

plate-bounding faults

The concepts behind PSM date back to the 1930s

and 1940s and were accompanied by the quantification

of the earthquake phenomenon The introduction of the

earthquake magnitude scale for regional events in

Cal-ifornia by C.F Richter (1935) and the discovery of the

earthquake frequency-magnitude relation by B

Guten-berg and C.F Richter (1941) are certainly the most

prominent and relevant examples At about the same

time, the idea of developing earthquake studies from

regional and local seismic networks was introduced to

study aftershock sequences in Japan (Imamura et al.,

1932)

Based on technical developments in the 1960s that

permitted the low-power operation and recording of

many seismic stations on a common time base, new

seismic networks and processing methods were

devel-oped that permitted the routine analysis of earthquakes

of M < 2, commonly referred to as

“microseismic-ity” A comprehensive review of principles and

appli-cations of microearthquake networks of this period

was given by Lee and Stewart (1981) Using a similar

approach extensive research using local arrays of

seis-mic stations was undertaken to implement a nuclear

test ban treaty Over the last two decades, progress in

the field of PSM has occurred primarily as a result

of advances in seismic instrumentation and

computa-tion facilities to store and serve large data sets The

transition to digital high-frequency full waveform data

acquisition systems with increased dynamic range also

stimulated the development of more sophisticated

anal-ysis schemes allowing refinement of existing models at

local, regional and plate-boundary scale

A key objective of modern PSM is to collect data

that can be used to resolve earthquake source

pro-cesses in space and time during the rupture process

Seismic waves observed at a receiver carry

informa-tion about the source process, but are also modified

by propagation through the earth When earth structure

is sufficiently well known, it is possible to correct the

observed waveforms for many propagation effects Asthe waves propagate in the heterogeneous and inelasticearth, information about the source process is lost due

to scattering and attenuation Once lost, this tion cannot be recovered, and so the solution is to placethe receiver “close enough” to the source Because thelosses are greatest at the highest frequencies, to record

informa-a signinforma-al with informa-a ~1 m winforma-avelength the sensor must bewithin less than 1 km of the source, assuming anelas-tic and scattering losses corresponding to a dampingfactor of Q ~500 Clearly, if one intents to understandearthquake processes on a specific scale, one needs torecord close to the source

In this paper we summarize the principal objectivesinvolved in PSM and review selected examples andfuture directions from key-locations representing var-ious environments such as hydrocarbon and geother-mal reservoirs, seismically quiet intra-plate regions,and large-scale transform faults and subduction zones

We also consider rock-deformation experiments inthe laboratory thus giving examples that cover rup-ture length scales from millimeter to hundreds of km.Table 1 gives an overview on the different magni-tude ranges and the relevant scales of rupture dimen-sion, displacement, dominant frequency and seismicmoment for the different environments discussed inthis paper We highlight that a comprehensive under-standing of the physical processes responsible for brit-tle failure requires investigations that span this largespatial and frequency range The effort to monitorthese processes using adequately designed receivergeometries is important, especially with regard tosocio-economic implications as shown in the cases ofsubduction megathrusts and large earthquakes alongplate-bounding transform faults

Quantifying the Earthquake Process

Earthquakes are the vibratory motion of the earth ated by the sudden release of energy within the solidrock mass of the planet Most earthquakes are caused

cre-by slip on faults, and as a consequence the term quake” is commonly used to refer to the earthquakesource process rather than the seismic waves it causes.Because the waves travel to great distances throughthe earth even for small underground disturbances theyprovide a powerful observational basis for studying thelocation, strength and fundamental nature of the earth-

“earth-Passive Seismic Monitoring of Natural and Induced Earthquakes 263

Table 1 Overview on different earthquake magnitude ranges

and relevant scales for rupture length, displacement, dominant

frequency and seismic moment Length and displacement scales

are approximate and appropriate for crustal earthquakes with

stress drops of 3 MPa Note that ranges given may overlap between earthquake class depending on source-receiver dis- tances and type of wave recorded

Magnitude range Class Length scale

Displacement

Seismic moment ∗

∗1 Aki (Ak) is defined as 1018 Nm The unit is named after Keiiti Aki, who pioneered the use of seismic moment in theory and practice The International Association of Seismology and Physics of the Earth’s Interior recommended in 2007 the adoption of the Aki as the standard unit of earthquake size.

∗∗The term “microearthquake” traditionally refers to earthquakes M < 3 The earthquake class names used here are a compromisebetween the SI naming conventions, which would require that a microearthquake had a magnitude between M = 2 and M = 4, and traditional practice.

quake source Seismology is the science of the

anal-ysis of these waves, and over the past century it has

become a deep and sophisticated branch of

mathemati-cal physics (e.g., Aki and Richards, 2002) From the

waves it is possible, in theory, to extract a detailed

description of the earthquake source process in space

and time The study of the earthquake source,

how-ever, is of necessity an empirical science, as we have

little control over when and where earthquakes occur

Aside from analog experiments performed in the

lab-oratory or earthquakes induced by industrial

modifica-tion of underground condimodifica-tions, the seismologist must

be prepared at all times to capture the earthquake when

it happens

Modern seismological instruments are designed to

record the wide range of frequencies and amplitudes

contained in the seismic waves of particular interest

Successful PSM also requires a geographic distribution

of instruments that encircle the source Only by

record-ing waves from a range of azimuths and distances is

it possible to accurately determine the location of the

earthquake source (hypocenter) and determine its basic

source properties (moment tensor, focal mechanism,

etc.) For most natural earthquakes recorded by

sur-face stations, this requires a network of instruments

with at least one station within a focal depth of each

earthquake When these conditions are met, the initial

point of rupture in an earthquake can be determined

to a precision of a few hundred meters Substantially

higher precision locations can be obtained using theseismic wave field (Rubin et al., 1999; Waldhauser andEllsworth, 2000)

Obtaining an accurate geographic description ofwhere earthquakes occur is among the most basicsteps toward developing a tectonic understanding of

a region Other physical measures of the complexmechanical event producing the earthquake take manyforms, including the dimensions of the faulted region,the direction and amount of slip in both space and time,

as well as traditional measures based on the tudes of the radiated elastic waves To relate the char-acteristics of one event to another, the observed quan-tities must generally be summarized through the use

ampli-of either an empirical relation, such as magnitude, or aquantity derived from a physical model, such as seis-mic moment Based on recordings from the SouthernCalifornia Seismic Network that initially consisted of

~10 stations at 100 km spacing, Richter (1935) oped the local magnitude scale as a first approach toquantify the earthquake size in a physical sense on aninstrumental basis He defined

devel-M1= log10A− log10A0()

where Ml is the event magnitude, A is the mum amplitude recorded by the Wood-Anderson seis-mograph, and log10A0 is the reference term used

maxi-to account for amplitude attenuation with epicentral

distance () This concept was further developed and

generalized by Gutenberg (1945) then including also

teleseismic events

Based on Richter’s results, Ishimoto and Ida (1939)

as well as Gutenberg and Richter (1941) discovered

a systematic relation between the magnitude of

earth-quakes and their frequency resulting in

log N= a − b∗Mwhere N is the number of events and a and b are con-

stants representing the overall level of seismic

activ-ity and the ratio between small and large earthquakes,

respectively To date, numerous studies of earthquake

statistics extending over as much as 34 magnitude units

from phonon emission in crystals to devastating

earth-quakes have confirmed this relation and determined

the value b to be close to 1 on average (Davidson

et al., 2007) Consequently, the number of earthquakes

increases by about a factor of 10 for each decrease

of one magnitude unit This simple relation

exempli-fies that the number of seismic events detected by a

seismic network entirely depends on its

magnitude-detection threshold or magnitude of completeness (Mc)

for which the network detects all events in the

tar-get area Mc is directly linked to the average

source-receiver distance of a seismic network and station

den-sity The value of the resulting earthquake catalog for

scaling-related studies clearly increases with

decreas-ing Mc

Of similar importance as the quantitative

descrip-tion of earthquake magnitude are physical descripdescrip-tions

of the source process The lowest-order

approxima-tion of an earthquake is a double-couple point source

Describing the source by force equivalents leads to

ele-gant mathematical approaches for propagation of

elas-tic waves (Aki and Richards, 2002) Further refinement

of the source processes replaces the point model by a

spatially extensive description, such as a circular crack,

distribution of point sources, or more complex

geome-tries Digital waveforms with sufficient dynamic range

allow routine measurement of the properties of a

seis-mic source such as the seisseis-mic moment tensor, rupture

duration and stress drop

It is beyond the scope of this paper to provide an

introduction into the theory of faulting and the

propa-gation of elastic waves Rather, we briefly describe the

role of stress as a key-parameter for the generation of

earthquakes Following the great 1906 San Francisco

earthquake, H.F Reid (1910) developed the theory ofelastic rebound that has been largely confirmed as thebasis of earthquakes by modern GPS measurements.Reid concluded that the slip during the earthquakeresulted from the release of previously accumulatedelastic strain energy stored in the crust astride the SanAndreas Fault Once the applied stress exceeded therock strength the stored energy was released by rapidslip on the fault We now know that the elastic strainenergy accumulated over centuries due to plate motion,and the forces are building up today for another greatearthquake, as this seismic cycle of accumulation andrelease continues The fault is a plane of weakness

in the earth that will be re-activated when the appliedshear stress (τ) exceeds a critical value This value

is primarily a function of the effective normal stressrepresenting the difference between normal stress (σn)and fluid pressure (PF) across the fracture plane and theroughness of the fracture surfaces Assuming a con-stant normal stress, the behavior remains elastic untilthe peak-strength is reached and slip occurs Amongstvarious formulations for the peak shear strength (T),the simplest one is the linear Mohr-Coulomb criterion

T= Co+ (σn− PF)∗ tanφwhereφ is the friction angle and Cois cohesion In thecase of a natural earthquake, shear failure results fromshear stress accumulation through tectonically drivenloading In the case of induced seismicity the effectivenormal stress and the frictional resistance to slidingcan be lowered by increasing the pore fluid pressurethrough fluid injection at depth as well as contraction

in geothermal fields and depleted oil reservoirs Ifthe fracture is subject to shear stress greater thanthe product of the effective normal stress and thecoefficient of friction, the rocks will slip and generate

an earthquake (e.g., Raleigh et al., 1972) Earthquakesgenerally may be classified into natural and inducedevents where induced microseismicity is meant to

be associated with a wide range of engineeringactivities These include deep well activities related tohydrocarbon production, hydro-fracturing treatments,

or dam impounding A review on different types oftriggered earthquakes was given by McGarr et al.(2002)

In the following we review selected examples andcase studies representing the present state-of-the-artfor the respective environments

Passive Seismic Monitoring of Natural and Induced Earthquakes 265

Case Studies

Monitoring the Failure Process: Acoustic

Emission Activity and Fracturing

in the Laboratory

The lowermost end of magnitude scales systematically

observed using PSM approaches is represented by

rock-deformation experiments in the laboratory under

controlled conditions Non-destructive testing methods

using advanced analysis of acoustic emissions (AE)

are being used more routinely to investigate processes

of brittle fracturing and frictional sliding in situ in the

laboratory The technique exploits the radiation of

elas-tic waves from fractures that may propagate as shear

or tensile cracks in a stressed rock sample AE analysis

is an indispensible tool to analyze fracture nucleation

and propagation on the sample scale However, the

technique is also used to analyze very high frequency

microseismic activity associated with excavation

dam-age and rockburst in underground mining (e.g., Young

and Collins, 2001; Plenkers et al., 2008) Typically,

the AEs are recorded by piezoceramic transducers

that are attached to the rock surface converting the

recorded elastic waves into a voltage signal (Fig 1a)

Suitably designed P- and S-wave sensors have

res-onant frequencies that range from a few hundred

kHz to several MHz Modern multichannel transient

recording systems allow storage of pre-amplified full

waveform signals with broad bandwidth (up to 16

bit) and high sampling rates (>10 MHz) Figure 1b

shows typical waveforms of AE signals recorded

during a rock-deformation experiment The numbers

of events that are being recorded largely depend on

the material and the characteristics of the recording

system They may range between a few hundred to

several thousand AEs for a single test Consequently,

signal processing, such as location of AE hypocenters

and source type analysis have to build on automated

procedures (Zang et al., 1996, 2000; Stanchits et al.,

2006) The recorded AEs typically represent <5% of

the experimentally-induced cracks (Lockner, 1993;

Zang et al., 2000) but have been found to provide very

accurate spatial images of brittle deformation

struc-tures such as shear and compaction bands (Fortin et al.,

2006)

Shear fracture nucleation was observed in triaxial

compression tests performed on low-porosity igneous

rocks and porous sandstones at confining pressures

<100 MPa (Lockner et al., 1991; Lockner and Byerlee,

1991, 1992; Lei et al., 2000, Stanchits et al., 2006)and in experiments performed using an indenter toinduce a shear fracture (Zang et al., 2000) In general,location of AE hypocenters in these experimentspermits precise monitoring of the formation ofmicrocrack clusters that precede unstable growth of amacrofracture (Lockner et al., 1991) During loading

of the specimens randomly distributed AEs developinto planar nucleation spots with increasing stress(Lockner and Byerlee, 1991; Stanchits and Dresen,2003) The nucleation patches consist of spatiallycorrelated events contained in a volume of about

a cubic centimeter from which future shear bandsdevelop In most experiments, fracture nucleationhas been found to be associated with a burst in AEactivity, a decrease in ultrasonic velocities along raypaths crossing the future fault trace (Stanchits et al.,2006) and a significant drop in b-values that develop

at 60–80% peak load prior to sample failure (Scholz,1968; Stanchits and Dresen, 2003)

Propagation of macroscopic fractures in rock fromnucleation spots involves evolution of a crack damagezone of finite width surrounding the fracture tip In thebrittle deformation regime the dominant dissipativemechanisms operating in the fracture process zoneinclude cracking and concomitant formation andclosure of pore space In experiments, the distribution

of AE hypocenters typically shows a close spatialcorrespondence with the macroscopic fracture trace(Fig 1c) with crack damage increasing stronglytowards the fault (Zang et al., 2000) However, thewidth of the process zone in cross section as defined bythe distribution of AEs or microcracks depends on theresolution of the method and may vary considerably(Janssen et al., 2001)

Dense spatial coverage of transducers surroundingthe sample allows determination of focal mechanismsfor larger amplitude AEs with a dominantly doublecouple radiation pattern Although AE hypocenterscommonly align with the fault trace, their focal mech-anisms often reveal a complex orientation pattern ofnodal planes This likely reflects the corrugated path

of the macroscopic fracture on the scale of ual grains (Zang et al., 1998) AE first motion polar-ities and full waveform analysis of larger AEs allowseparating the events in dominantly tensile, shear andimplosive (pore collapse) sources (Lei et al., 1992;Zang et al., 1998; Graham et al., 2009) With an

individ-a) b)

c)

Fig 1 a) Experimental setup to monitor acoustic emissions

(AE) during rock-deformation experiments in the laboratory.

The cylindrical sample is jacketed in rubber sleeve clamped to

steel endcaps to protect specimen from oil confining medium.

Sensors are piezoceramic transducers in brass housings and

glued directly to the sample surface b) Typical waveforms

of acoustic emission (AE) signals recorded by a piezoceramic

transducer during a rock-deformation experiment c) Left: Shear

fracture in low porosity sandstone Length of sample is 100 mm Specimen was recovered after the test, impregnated with blue epoxy and cut in half along its axis The shear fracture initiated

at an indenter placed at the top of the specimen and propagated

towards the bottom Right: Black dots represent located AE

hypocenters Note the close spatial correspondence of ter cloud and fracture trace

hypocen-increase in loading and specimen damage a shift in

dominant AE source types is commonly observed In

dense brittle rock like granite, the relative proportion

of shear and collapse events increases at the expense of

tensile sources as the sample approaches peak stress

In porous rocks like sandstone deformed at elevated

confining pressures cracking is dominated by shear andimplosive source types

Detailed analysis of AE activity in rocks tributes significantly to unravel the microphysics ofnucleation, shear zone formation and brittle failure

con-in laboratory experiments In addition it is currently

Passive Seismic Monitoring of Natural and Induced Earthquakes 267used successfully to study laboratory-scale compaction

bands in porous rocks and the mechanisms involved in

the formation of borehole breakouts

Tracking the Hydro-Frac: Passive Seismic

Monitoring in Hydrocarbon Reservoirs

At the scale of hydrocarbon reservoirs, the PSM

of microseismicity induced by stress changes during

hydraulic fracturing through massive fluid-injection is

a key method to directly image fracture growth and

permeability enhancements in hydrocarbon or

geother-mal reservoirs Microseismic data are also crucial in

other rock engineering applications, such as mining

activity, excavation stability in nuclear waste

reposito-ries or geotechnical stability studies Moreover,

sim-ilar techniques are currently developed and tested

to enhance underground storage of carbon dioxide

into different target formations such as depleted oil

and gas reservoirs, saline aquifers and coal-bearing

mines

Systematic fluid-injection was pioneered at the

Rangely Oil Field, Colorado, confirming the

hypoth-esis that earthquakes may be triggered by the increase

of fluid-pressure (Raleigh et al., 1972) The injection

was seismically monitored by a temporary local

seis-mic array at the surface and at regional distance

dur-ing a 7-year period detectdur-ing almost 1,000 events

Changes in the number of earthquakes were

corre-lated with changes in the fluid-injection rates over the

years (Gibbs et al., 1973) confirming earlier findings

by Evans (1966) and Healy et al (1968) who

stud-ied the Denver earthquakes in the context of

nearby-injection of chemical waste In the Geothermal

Indus-try, field efforts began with the pioneering work at

Los Alamos National Laboratory in the early 1970s at

the Fenton Hill site then referred to as the “Hot Dry

Rock” (HDR) project Later, this term was replaced by

“Enhanced Geothermal Systems” (EGS) to more

cor-rectly reflect the quality of resource

PSM of hydraulic fracturing experiments aims at

following the spatiotemporal growth of the created

hydro-fractures by imaging the associated small-scale

seismicity (e.g., Rutledge et al., 1994) Since the

Rangely experiment, an increasing number of PSM

field campaigns have been conducted in the petroleum

industry confirming the direct correlation between

injection flow rate and pressure and the rate of inducedseismicity (e.g., Kovach, 1974; Oppenheimer and Iyer,1980; Albright and Pearson, 1982; McGarr, 1991;Philips et al., 1998; 2002) An inherent need for alow magnitude detection threshold monitoring close

to the seismic source are additional monitoring wells

in direct vicinity to the injection well The toring wells are usually equipped with geophones todetect elastic waves and to determine the hypocen-ter location and source parameters The quality of thehydraulic fracture characterization is directly linked tothe source-receiver geometry and the quality of thevelocity model for the source area In that respect, thestudy by Rutledge et al (2004) exemplifies that an ade-quate data set of downhole seismic recordings formsthe base for further evaluation such as interpretation offaulting mechanisms (Fig 2)

moni-However, using adequately spaced sensors at voir depth is not a simple task due to the enor-mous costs involved in deep drilling and the fact thatthe downhole sensors are operated somewhat remote-controlled with a single cable of several kilometerlengths being the only connection between sensor andsurface equipment In that respect, an illustrative exem-plification for downhole PSM may be using high-techequipment in outer space where also no direct accessexists to the actual sensor and its periphery In bothcases a single dysfunction within the whole monitor-ing system is crucial to shut-down the entire operation

reser-A general observation in monitoring of injection induced microseismicity was the alignment

fluid-of the seismic cloud along the direction fluid-of the mum principal stress Furthermore, most events clus-ter into well-defined small-scale geometrical patterns(e.g., Phillips et al., 2002) In recent years the enhanceduse of such data sets has significantly moved forwardapplying collapsing methods (Jones and Stewart, 1997)and relative relocation techniques (Waldhauser andEllsworth, 2000) During hydraulic fracturing experi-ments the seismicity follows the generation and growth

maxi-of newly created fractures as discussed above natively, induced slip on pre-existing fault planes mayreflect linear and planar features indicating networks ofintersecting fractures or fracture containment betweenstratigraphic layers of differing mechanical properties

Alter-or states of stress This is further discussed and orated in the following chapter Today, PSM is a stan-dard tool applied in the hydrocarbon (and geothermal)industry to image permeability enhancements at depth