Lecture Outlines
Natural Disasters, 7th edition
Patrick L. Abbott
Some Earthquakes
in Western North America
Natural Disasters, 7th edition, Chapter 6
Plate tectonic history of Western North America
• Opening of Atlantic
Ocean basin pushes
North and South
America west
• Farallon plate
subducts beneath North
America
• By 28 million years
ago, enough of the
Farallon plate has
subducted under North
America for the Pacific
spreading center to
make contact with
North America
Figure 6.1
Plate tectonic history of Western North America
• Remaining spreading
centers to the north and
south are connected by
a transform fault 
ancestor of San
Andreas fault
• In last 5.5 million years,
Gulf of California has
opened 300 km, tearing
peninsular California
(Baja California and
California west of the
San Andreas fault)
away from the North
American plate and on
the Pacific plate
Figure 6.1
Plate tectonic history of Western North America
• Eventually peninsular
California will move
north to Alaska (can
not sink into Pacific
Ocean or be subducted)
• Consider size of Pacific
and North American
plates as they move past
one another  stress is
not just along San
Andreas fault but over
broad area of western
North America
Figure 6.1
Subduction Zone Earthquakes
Subduction Zone Earthquakes
• Subduction of Pacific plate under Alaska creates truly
large earthquakes (not San Andreas fault)
• The Good Friday Earthquake, Alaska, 1964
– Major subduction movement created magnitude 9.2
earthquake, as Pacific slab shoved under Alaska in seven
lurches
– Seven minutes of shaking induced avalanches, landslides,
ground settling and tsunami, killing 131 people
– Relatively low loss of life because area is sparsely inhabited,
few people downtown, low tide and warm weather
Subduction Zone Earthquakes
Mexico City, 1985
• Subduction of Cocos plate under North America created magnitude
8.1 earthquake, with magnitude 7.5 and 7.3 aftershocks
• Earthquake was 350 km from Mexico City, but intense damage
• 8,000 people were killed by building collapses
• Earthquake was
expected 
Michoacan
seismic gap
• Guerrero seismic
gap still exists
Figure 6.3
Subduction Zone Earthquakes
Mexico City, 1985
• Earthquakes Don’t Kill, Buildings Do
• 5,700 buildings in Mexico City were damaged, while small towns
closer to the epicenter had much less damage
Figure 6.2
Figure 6.4
Subduction Zone Earthquakes
Mexico City, 1985
• Earthquakes Don’t Kill, Buildings Do
• Resonance between seismic waves, soft lake-sediment foundations
and improperly designed buildings, all vibrating most intensely at 1
to 2-second periods, destroyed many buildings
Figure 6.6
Figure 6.5
Subduction Zone Earthquakes
Pacific Northwest, The
Upcoming Earthquake
• Cascadia subduction zone –
small plates subducting under
North America
• Why no major earthquakes
in last 200 years in Pacific
Northwest?
– Has subduction stopped?
• No, because volcanism still occurs
Figure 6.7
– Are plates subducting so smoothly that earthquakes don’t occur?
• Probably not, since subducting lithosphere is so young
Subduction Zone Earthquakes
Pacific Northwest, The Upcoming Earthquake
• Similar subduction setting to Japan and Chile
– Japan’s subduction zone relieved stress with two magnitude 8.1
earthquakes in 1944 and 1946
– Chilean subduction zone generated world’s largest measured
earthquake in 1960, magnitude 9.2
– 1960 Chilean events would be comparable to earthquake series
that shifts entire Cascadian subduction zone
• Could Cascadia have a magnitude 9 earthquake?
– Yes, it has and will again
Subduction Zone Earthquakes
Earthquake in 1700
• Last major earthquake in Pacific Northwest was about
magnitude 9, in 1700
– Trees in drowned forests died by salt-water flooding when
ground dropped in 1700
– Tsunami hit Japan in 1700 – might have been caused by
magnitude 9 earthquake along Washington-Oregon coast
• Effects of magnitude 9
earthquake:
Figure 6.8
– Three to five minutes of
intense shaking
– Tsunami 10 m high, 15 to 40
minutes after earthquake
Spreading-Center Earthquakes
• Spreading center underlies Gulf of
California, extends north into U.S.,
including Salton Sea, Coachella and
Imperial Valleys
• Salton Trough:
– High heat flow, glassy volcanic domes,
boiling mud pots, geothermal reservoirs,
earthquake swarms from moving magma
– One of most earthquake-active U.S. areas
– Brawley seismic zone – often hundreds
of small earthquakes over few days
(swarms), continually releasing stress
rather than building for big earthquake
– Other nearby faults generate larger
earthquakes, up to magnitude 7
Figure 6.9
Transform Fault Earthquakes in California
San Francisco, 1906
• Home to about 400,000 people
• Intense ground shaking for about one minute in early
morning hours
• Damage much worse in areas constructed on artificial
fill rather than rock or consolidated sediment
• Many fires broke out and water lines were destroyed,
making fire-fighting impossible – ten times more
damage than shaking
Transform Fault Earthquakes in California
San Andreas Fault
Earthquakes
• In 1906 earthquake, 430 km of
San Andreas fault between
Cape Mendocino and San Juan
Bautista shifted up to 6 m
• This section of the fault has
experienced no major
earthquakes since 1906:
‘locked’ section of fault
– Stress is stored until large
rupture releases energy
Figure 6.10
Transform Fault Earthquakes in California
San Andreas Fault Earthquakes
Figure 6.12
Figure 6.13
Transform Fault Earthquakes in California
San Andreas Fault Earthquakes
• Different sections of the fault behave differently:
– South of 1906 section, between San Juan Bautista and
Cholame:
• Frequent small to moderate earthquakes and creep
(millimeters per year of ongoing offset)
– South of creeping section, between Cholame and San
Bernardino:
• Locked section with recent deficit of earthquake activity
• Most recent large rupture in 1857 Fort Tejon earthquake of
magnitude 7.8
– South of locked Fort Tejon section:
• Complex zone with locked sections
In Greater Depth:
Neotectonics and Paleoseismology
• Compressive bends in strike-slip cause land to uplift;
pull-apart bends cause land to drop down
• Down-dropped areas
accumulate sediment
from
– Sand washing in from
heavy rains
– Clays settling in
ponded water
– Vegetation buried by
clay and sand
Figure 6.15
In Greater Depth:
Neotectonics and Paleoseismology
• Fault movements can result disturbances and offsets in
sediment deposition in down-dropped basins or faultdammed areas
• Determine size of earthquakes by amount of offset
• Dates of prehistoric earthquakes from radioactive carbon
Figure 6.16
Figure 6.17
In Greater Depth:
Neotectonics and Paleoseismology
Earthquake Prediction – Intermediate and Long
Term
• Rockwell classifies fault-movement timing into:
– Quasi-periodic timing, detectable with
paleoseismology
– Clustered movements, such as North Anatolian fault
in Turkey
– Random movements, inherently unpredictable, such
as San Andreas fault in California
Transform Fault Earthquakes in California
• December 1988: Using paleoseismology, assessed 30% probability
of magnitude 6.5 earthquake on Loma Prieta segment of San
Andreas within 30 years
– Loma Prieta earthquake occurred 10 months later
• In 2003: Assessed 62% probability of magnitude 6.7 earthquake in
San Francisco Bay region before 2032, and 85% probability of
magnitude 7 or higher earthquake in southern California before
2024
Figure 6.19
Transform Fault Earthquakes in California
World Series (Loma Prieta) Earthquake, 1989
• Just prior to scheduled start of Game 3 of the World
Series, at Candlestick Park of San Francisco
• Rupture of southernmost 42 km of 1906 section
• Fault rupture was deep, did
not offset ground surface
• Fault movement was
vertical and horizontal,
not just horizontal
• Rupture lasted only 7
seconds (short for
Figure 6.21
magnitude 7.0 earthquake)
Transform Fault Earthquakes in California
World Series (Loma Prieta)
Earthquake, 1989
• Occurred at bend in San
Andreas fault near Santa Cruz
Mountains’ highest peak, Loma
Prieta
– Bend in fault uplifted Santa Cruz
Mountains
• At bend, fault plane is inclined
at 70 degrees (not vertical), so
fault movement shifted Pacific
plate 1.9 m to north, and uplifted
Santa Cruz Mountains 36 cm
Figure 6.20
Transform Fault Earthquakes in California
World Series (Loma Prieta) Earthquake, 1989
• Mainshock was magnitude 7.0, numerous aftershocks
• Loma Prieta region was seismic gap before earthquake
– Mainshock and aftershock filled in seismic gap
• Destruction was much less than could have been, as
shaking lasted only 11 seconds
– 67 people killed, 3,757 people injured, 12,000 people homeless
– $6 billion in damages
Figure 6.22
Transform Fault Earthquakes in California
World Series (Loma Prieta) Earthquake, 1989
• Earthquakes don’t kill, buildings do
• Near epicenter, damage was mostly to older and
poorly designed buildings
• Seismic waves were amplified by soft muds and
artificial-fill foundations around San Francisco Bay
• Ground motion on soft-sediment sites was 10 times
stronger than at nearby rock sites
• Farther from epicenter, considerable damage to
poorly sited or poorly designed structures
– Marina District
– Interstate 880
Transform Fault Earthquakes in California
World Series (Loma Prieta)
Earthquake, 1989
• Marina District
Figure 6.23
Transform Fault Earthquakes in California
World Series (Loma Prieta) Earthquake, 1989
• Marina District
– Built on fill from debris from 1906 earthquake
• Amplified shaking
• Deformation and liquefaction of artificial-fill
foundations
• Soft first-stories – ground floors of buildings cleared
out to make room for parking, leaving building
without adequate support so rest of building pancakes
onto first floor
Transform Fault Earthquakes in California
World Series (Loma Prieta) Earthquake, 1989
• Interstate 880
– 42 people killed when upper deck collapsed onto lower deck
Figure 6.24
Transform Fault Earthquakes in California
World Series (Loma Prieta) Earthquake, 1989
• Interstate 880
– Section where collapse occurred was built on soft muds (other
sections on sand and gravel survived)
– Freeway’s natural resonance was amplified by shaking of muds
– Could have been predicted by collapses of freeways in 1971 San
Fernando earthquake
Figure 6.25
Transform Fault Earthquakes in California
Bay Area Earthquakes – Past and Future
• Historic records for last 225 years (good records only last
150 years)
– Major earthquake recurrence
interval is more than 150 years
• 19th century earthquake history
quite different than 20th century
– 19th century: Seven greater than
6 magnitude earthquakes before
1906 earthquake
– 20th century: No large
earthquakes until 1970s
Figure 6.27
Transform Fault Earthquakes in California
Bay Area Earthquakes – Past and Future
Pattern 1: common large earthquakes vs. rare giant earthquakes
• Plate tectonic movement is satisfied by
either one magnitude 8 earthquake in a
century or one magnitude 6-7 earthquake
every decade
Pattern 2: East-west pairings of
earthquakes
• In 1836, Monterey Bay area had
earthquake about magnitude 6.5; in 1838,
San Francisco peninsula had earthquake
of magnitude 6.8
• In 1865, Santa Cruz had a large
earthquake; in 1868, Hayward had a
Figure 6.27
magnitude 6.9 earthquake
Transform Fault Earthquakes in California
Bay Area Earthquakes – Past and Future
Pattern 3: northward progression of earthquakes
• Earthquakes of 1865 and 1868 on the Hayward fault were
preceded by five moderate earthquakes that moved
northward up the Calaveras fault
– From 1974 to 1988 four moderate earthquakes
occurred from south to north on the Calaveras fault
– Area of Hayward fault today is populated by 2 million
people, with schools, hospitals, city halls, houses and
the University of California built on the fault itself
Transform Fault Earthquakes in California
Kobe, Japan, 1995 Versus
Oakland, California, 20??
• Most expensive earthquake in
history
• Magnitude 6.9 earthquake with
50 km long rupture of Nojima
fault and 100 seconds of shaking
• Tile roofs and little lateral
support caused collapse of
buildings causing 6,308 fatalities
• More than 140 fires resulted
Figure 6.29
Transform Fault Earthquakes in California
Kobe, Japan, 1995 Versus
Oakland, California, 20??
• Hayward fault has 27% probability of
causing magnitude 6.7 or greater
earthquake before 2032
• San Francisco Bay region, 62%
probability of 6.7 magnitude
earthquake on a fault before 2032
• Hayward fault and Nojima fault:
– About 50 km long, densely
populated areas, with large areas
of weak ground materials
– Both capable of generating
magnitude 7 earthquakes
Figure 6.31
How Faults Work
How Faults Work: Old View
Elastic-rebound theory
• Movement takes place on both
sides of a fault but not on the
fault itself, until buildup of
stress on the fault overcomes
friction on the fault and the
rocks on both sides of the fault
move quickly in an earthquake
• After earthquake, all stress on
the fault has been removed and
buildup begins again
Figure 6.32
How Faults Work
How Faults Work: Newer View
Windows of opportunity
• Fault movement begins at hypocenter and radiates
outward
• How much of stored stress is relieved depends on
how long the fault moves
– Opening gate for line of people: how long the gate is
open determines how many people will get through
– Moving a carpet by making a ripple and propagating
ripple across floor: ripple may be small (small offset 
little relieved stress) but may travel great distance (large
area of fault slips  large earthquake)
How Faults Work
Landers, California,
1992
• Sequence began in 1992
with magnitude 6.1
Joshua Tree earthquake,
followed two months later
by magnitude 7.3
Landers earthquake,
then followed few hours
later by magnitude 6.3
Big Bear earthquake
• Followed in 1999 by
magnitude 7.1 Hector
Mine earthquake
Figure 6.33
How Faults Work
Landers, California, 1992
• What was learned about
how faults move?:
– Unlike most fault
movements, Landers
rupture front slowed when
it reached right-step but
did not stop
– Rupture velocity was
initially 3.6 km/sec,
slowed almost to stop then
continued northward at
varying speeds
Figure 6.34
How Faults Work
Landers, California, 1992
• What was learned about how faults move?:
– Amount of slip varied from centimeters to 6.3 m
– Amount of fault movement at the ground surface differs
from that at depth
– Fault movement and shaking at the hypocenter were
much less than later, farther from hypocenter
Figure 6.35
How Faults Work
Landers, California, 1992
• What was learned about how faults move?:
– Only small portion of fault slipped at any one time
• Fault movement lasted 23 seconds but no spot on the fault moved
for more than 4 seconds
– When the rupture front slowed at the fault step, stress
built up behind the front until it moved through the step
– Triggered other earthquakes in areas to the north
(directivity), fortunately for Los Angeles to the south
– 1992 Landers series, faults moved from south to north;
1999 Hector Mine event, fault moved from north to south
– Patches of the faults without much movement became
origin for later earthquakes
How Faults Work
Southern San Andreas Fault
• Coachella Valley segment was assigned 40% probability of
magnitude 7.5 earthquake before 2018
– Neotectonic analyses indicate earthquakes about 250
years apart, with last one 315 years ago
– Recent large earthquakes at both ends of the Coachella
Valley segment of the San Andreas fault (like before 1906
earthquake on northern San Andreas)
• Might be not just Coachella segment (6 m unreleased stress)
– Could continue on San Bernardino Mountains segment
(4.3 m unreleased stress) creating magnitude 7.8
earthquake, or even on Mojave segment (4.7 m
unreleased stress), creating magnitude 8 earthquake
In Greater Depth:
Earthquake Prediction – Short Term
• Plate tectonics: can know why and where earthquakes
occur
• Neotectonics: can know how big and how often
earthquakes occur on a given fault
• Short-term prediction of earthquakes?
– Detailed behavior of faults may be too unpredictable to
ever allow short-term prediction of earthquakes
– False lead: Parkfield, California, with magnitude 6
earthquakes every 22 years, should have allowed prediction
of next earthquake to occur in 1988, +/- five years (finally
occurred in 2004)
Thrust Fault Earthquakes in
Southern California
• Big Bend: left step in San Andreas fault in southern
California creates thrust (reverse) faults
– Thrust faults uplift mountains in Transverse Ranges
– Blind thrust faults do not reach surface, difficult to detect
• Los Angeles region is
shortening by 10-15
mm/yr
– Detectable by global
positioning system (GPS)
– Could generate magnitude
6 earthquake every six
years
Figure 6.36
Thrust Fault Earthquakes in
Southern California
Northridge,
California, 1994
• Pico blind thrust
ruptured in magnitude
6.7 earthquake under
Northridge (shaking
fortunately directed
away from Los Angeles)
• 61 people killed, 9,000
people injured, $20
billion in damages
Figure 6.37
Thrust Fault Earthquakes in
Southern California
Northridge,
California, 1994
• Similar to 1971 San
Fernando Valley
earthquake except
shaking directed away
from Los Angeles in
1994
Figure 6.39
Thrust Fault Earthquakes in
Southern California
Northridge, California, 1994
• Los Angeles region seems to be experiencing fewer
earthquakes than it could
– GPS shows deformation that should result in twice as many
magnitude 6 earthquakes and three times as many
magnitude 7 earthquakes
– Radar interferometry indicates that ground is warping
without large earthquakes being generated
The “Big One”
• What is meant?
– Comparable earthquake in Alaska, 2002 on Denali fault
• Over 140 seconds, ruptured 340 km in magnitude 7.9
earthquake with offsets up to 8.8 m
• Significant directivity – triggered earthquake swarms up to
3,660 km to the southeast as far as Utah
• Unpopulated area, minimal effects on people
– 1857 Fort Tejon earthquake on San Andreas fault
• Over 130 seconds, ruptured 360 km in magnitude 7.8
earthquake with offsets up to 9.5 m
• Relatively unpopulated area at that time
– San Bernardino section of San Andreas fault now houses 3
million people
The “Big One”
Earthquake hazard studies in southern California evaluate:
• San Andreas fault, along eastern and northern side of
Los Angeles metropolitan area, last movements in 1812,
1857
• Thrust faults (caused 1971 San Fernando, 1994
Northridge earthquakes)
“Big One” would be comparable to 1857, 1971 and 1994
southern California events occurring simultaneously
The “Big One”
Worst-case scenario for southern California involves both:
• 1957 Mongolian earthquake sequence
– Major strike-slip fault moved in magnitude 8 earthquake at
same time as thrust faults moved in magnitude 7 earthquakes
Figure 6.41
The “Big One”
• Potential future costs can be predicted as annualized
earthquake losses by software HAZUS
– Uses data on population, buildings and shaking potential
• U.S.: $4.4 billion projected annual earthquake losses
– Almost half in southern California
– Los Angeles County $1 billion in losses each year
End of Chapter 6