ESS 8 Earthquakes week 2

1
ESS 8 - Earthquakes (week 2)
Prof. Didier Sornette sornette_at_moho.ess.ucla.edu
Sara Cina saracina_at_ucla.edu Jelen
a Tomic tomic_at_moho.ess.ucla.edu
Northridge earthquake
http//www.ess.ucla.edu/academics/courses/web/fall
_2005/ess_8/index.asp
2
on the Way to a Geodynamo in a Computer
Geophysicists have been worrying out loud lately
that a 10 weakening of the magnetic field during
the past 170 years may presage an overdue
flipping of Earth's magnetic poles, a so-called
magnetic reversal. It's been 780,000 years since
the last reversal, although millennia-long
reversals have occurred on average every 100,000
years or so. Aside from millennia of confused
magnetic navigation, a reversal would greatly
weaken Earth's magnetic shield that fends off
cosmic radiation.
Science, Vol 309, Issue 5733, 364-365 , 15 July
2005
3
The cracks were detected before but researchers
now know they can remain open for long periods,
rather than opening and closing for just very
brief intervals. This new discovery about how the
Earth's magnetic shield is breached is expected
to help space physicists give better estimates of
the effects of severe space weather. 
Immense cracks in the Earth's magnetic field
remain open for hours, allowing the solar wind to
gush through and power stormy space weather,
according to new observations from the IMAGE and
Cluster satellites.
4
Left A simulation of Earth's magnetic field
structure. Right An image of what Earth's
magnetic field might look like during a reversal,
something humans may have to worry about
thousands of years from now. Images courtesy Gary
Glatzmaier
5
(No Transcript)
6
A very large earthquake triggered landslides,
toppled an apartment building and flattened
villages of mud-brick homes Saturday, killing
more than 3,000 people across a mountainous swath
touching Pakistan, India and Afghanistan.
Saturday, 8 October 2005 The 7.6-magnitude quake
with the epicentre 80km (50 miles) north-east of
Islamabad wiped out several villages.
The casualty toll from the 7.6-magnitude tremor
was rising early Sunday as rescuers struggled to
dig people from the wreckage, their work made
more difficult as rain and hail turned dirt and
debris into sticky muck. The worst damage was in
Pakistan, where the dead included 250 girls
crushed at a school and 200 soldiers on duty
in the Himalayas. For hours, aftershocks rattled
an area stretching from Afghanistan across
northern Pakistan into India's portion of the
disputed Himalayan region of Kashmir.
Hospitals moved quake victims onto lawns, fearing
tremors could cause more damage. The earthquake,
which struck just before 9 a.m., caused buildings
to sway for about a minute in the capitals of
Afghanistan, Pakistan and India, an area some 625
miles across. Panicked people ran from homes and
offices, and communications were cut to many
areas.
India reported at least 250 people killed and 800
injured when the quake collapsed 2,700 houses and
other buildings in Jammu-Kashmir state. Most of
the deaths occurred in the border towns of Uri,
Tangdar and Punch and in the city of
Srinagar, said B.B. Vyas, the state's divisional
commissioner.
Landslides and rain hindered rescue efforts.
7
(No Transcript)
8
Earthquakes and active faults in northern
Pakistan and adjacent parts of India and
Afghanistan are the direct result of the Indian
subcontinent moving northward at a rate of about
40 mm/yr (1.6 inches/yr) and colliding with the
Eurasian continent. This collision is causing
uplift that produces the highest mountain peaks
in the world including the Himalayan, the
Karakoram, the Pamir and the Hindu Kush ranges.
As the Indian plate moves northward, it is being
subducted or pushed beneath the Eurasian plate.
Much of the compressional motion between these
two colliding plates has been and continues to be
accommodated by slip on a suite of major thrust
faults that are at the Earths surface in the
foothills of the mountains and dip northward
beneath the ranges. These include the Main
Frontal thrust, the Main Central thrust, the Main
boundary thrust, and the Main Mantle thrust.
These thrust faults have a sinuous trace as they
arc across the foothills in northern India and
into northern Pakistan. In detail, the modern
active faults are actually a system of faults
comprised of a number of individual fault traces.
In the rugged mountainous terrain, it is
difficult to identify and map all of the
individual thrust faults, but the overall
tectonic style of the modern deformation is clear
in the area of the earthquake north- and
northeast-directed compression is producing
thrust faulting. Near the town of Muzaffarabad,
about 10 km southwest of the earthquake
epicenter, active thrust faults that strike
northwest-southeast have deformed and warped
Pleistocene alluvial-fan surfaces into anticlinal
ridges. The strike and dip direction of these
thrust faults is compatible with the style of
faulting indicated by the focal mechanism from
the nearby M 7.6 earthquake.
9
(No Transcript)
10
(No Transcript)
11
Summary

• - Overview of the inside of the Earth
• - Crust, mantle, core
• - Historical development of Plate Tectonics
• - Continental Drift
• -Matching fossils coastlines
• - Mantle convection
• - Rocks flow over long timescales
• - Heat drives them
• - Magnetic stripes on the ocean floor
• - Record sea-floor spreading and
• reversals of the Earths magnetic field
• - How the plates have moved over time
• - Pangea (250 My), Gondwanaland, Rodinia

12
Most seismicity occursat plate boundaries
but not all
13
Exception 1Blurring of the Plate Boundary

• More common on continental boundaries

Continents are weaker
1977-1997, USGS
14
Exception 2Intraplate Earthquakes

• Earthquakes in the middle of continents
• Ex New Madrid, Missouri 1811-1812

1977-1997, USGS
15
Exception 3Hotspots

• Why is Hawaii in the middle of the ring of fire?

16
Mantle Plumes
17
With all these exceptions, why do we believe
plate tectonics?

• Provides unified explanation for
• Matching geology fossils
• Magnetic stripes
• Locations of most earthquakes
• Locations of most volcanoes
• What does it predict?
• Plate velocities (From geological record plate
  reconstructions)

18
Testing the Predictions

• Modern techology (GPS) lets us measure plate
  velocities (cm/year) and compare to the
  geological predictions

Nuvel-1a is 3 Mya average GPS is current motion
http//www.earth.northwestern.edu/ research/snapp.
html
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
From large scale motion to earthquakes

• Smooth large scale plate motions
• Abrupt crust rupture during earthquakes

23
Earthquake is sliding/rupture on a fault
24
Faults

• Discontinuity (break) in the Earths crust along
  which motion takes place
• Can be active (still has
• motion) or inactive
• (moved in the past
• Can be seismic (makes
• earthquakes) or aseismic
• (creeps smoothly)

San Gabriel Fault Photo Matt DAllessio
25
Faulting at plate boundaries

• Three main types, like three types of boundaries
• Normal faults - divergence
• Thrust faults - convergence
• Strike-slip faults - sideways

26
Ocean ridges

• It is where new crust is being created
• Pacific, Atlantic, and Indian Oceans each have
  mid-ocean ridges
• Called ridges because youngest, hottest crust is
  lighter than old crust, and floats higher on the
  mantle, forming ridges
• Normal faults

27
Normal faults common at spreading centers.
Ocean crust is forming here.
28
Normal fault - divergence
Press, 18-12a,b
29
Convergent zones

• Collision zones
• India-Asia
• Subduction zones
• Around Pacific Rim
• Earthquakes, volcanoes, tsunamis, landslides
• Ring of Fire
• Also by Indonesia

Thrust faults
Press, 18-12b
30
To distinguish normalfrom thrust faults

• Imagine a vertical line through the fault. The
  crust above the intersection of the line with the
  fault is called the hanging wall, the crust below
  the intersection is called the footwall.
• If the hanging wall is moving up, the fault is a
  thrust (reverse) fault.
• If the hanging wall is moving down, the fault is
  a normal fault.

31
Strike-slip fault - transform
Left-lateral
Press, 18-12a,d
32
Left- and Right-lateral
Right-lateral
Left-lateral
33
Other ways to deform rock
Press, 10-6
34
From large-scale motion to earthquakes

• Ductile - Smooth motion in space and time
• Large-scale plate motions are smooth
• Due to flow in ductile mantle underneath
• Brittle - Abrupt rupture when stressed
• During earthquakes
• Due to brittle nature of crust

35
Liquids versus Solids (ex Silly-Putty)

• Liquids flow viscosity h resists.
• Solids deform elasticity G resists.
• Maxwell characteristic time t h/G
• h is viscosity coefficient and G is elastic
  modulus
• t 10-12 seconds for water
• t 106 years for earth crust
• Time scale of deformation lt t gt solid
• Time scale of deformation gt t gt liquid
• See experiment with silly-putty

36
Steady flow and unsteady faults
Crust Mantle
Slip
Motion on earthquake faults (Seismic faults)
Mantle flow or creeping (aseismic) faults
Time
Total slip has to add up to the same over a long
time
37
Some faults creep slowly

• Motion at the surface matches flow at depth

USGS
38
No Large Historic Earthquakes on the Creeping
Section
1906 San Francisco Earthquake to the North
1857 Fort Tejon Earthquake to the South
39
Definition Stress

• Stress is force per area
• A force is something that pushes or pulls
• Ex gravity creates a force (mass x acceleration
  g)
• Walking on ice
• Better to spread your weight out to avoid falling
  through. Spreading out the weight reduces the
  stress with the same force.

40
Definition Strain

• Strain is the fraction of size that a body is
  deformed
• ex Squeezing a block

1.5 m
Strain -0.5m/1.5m -1/3 -0.33
1 m
41
Qualitatively

• Strain is how much something changes shape

Which one is more strained?
42
Elasticity

• Definition A material is elastic if stress is
  proportional to strain
• Properties of an elastic body
• The more your stretch (or squeeze) it the more it
  deforms
• If you stop stretching, then it returns to its
  original shape
• It is not ductile, i.e., it does not flow.

43
Elastic Rebound

• A fault remains locked (by friction) while stress
  slowly accumulates, gradually twisting the rock
• Then it suddenly ruptures in an earthquake,
  releasing the stored-up stress.
• Energy is released in the form of heat and
  seismic waves.

Fault
44
Elastic Rebound

• Deformation during the earthquake cycle

time 0 yrs
width of deformed zone set by thickness of crust
time 100 yrs
after earthquake!
45
Elastic Rebound

• After 100 years of accumulating strain

old road
new road
46
Elastic Rebound

• After earthquake

old road
new road
47
Consequences of Elastic Rebound

• Faults store energy slowly
• over decades to centuries to millenia
• But release the energy rapidly (10s of seconds
  or less)
• in an earthquake!!!

stress
time
48
How often should we have large earthquakes?

• Assume
• Slip in earthquakes rate x time between
  earthquakes
• The biggest recorded earthquake (Chile 1960)
  slipped 25 m.
• If the plate convergence rate is 77 mm/yr (0.077
  m/yr)
• Solve for time between earthquakes slip/rate
  25/0.077 yr 325 yrs

49
A few things missing in this calculation

• Little earthquakes take up some of the slip
• Some aseismic creep happens
• Earthquakes trigger each other to happen before
  they are due
• The last slide gives a very rough estimate of how
  active a boundary is.
• It is most useful as a warning
• Ex Cascadia

50
Thrust faulting cyclelike strike-slip, but
vertical strain
Strain build-up
Earthquake
51
Strain accumulation

• Steady strain rate over many years
• Distributed across zone about 100 km wide
• Only top 20 km build strain in California
• Deeper rocks seem to flow due to higher temp.
• We see strain accumulate with satellite
  measurements
• Ex Global Positioning System (GPS) and InSAR
• If build-up of strain is steady and featureless,
  there may be no clues of coming quakes

52
Relationship betweenEarthquakes and Faults

• Chicken-and-egg problem
• Earthquakes occur on faults
• Faults grow by earthquakes
• Answer Faults
• begin on
• weaknesses in
• Rocks self-organization process

53
What does a fault look like?
Punchbowl Fault Photos by Fred Chester, Texas AM
54
Nojima Fault Core (1995 Kobe Earthquake)
Mori et al., Kyoto Univ.
55
Internal structure of faults

• Zones of deformation

56
Appearance of fault trace

• Fault trace - where fault plane intersects the
  Earths surface
• Surface fault geology results from the
  accumulation of offsets in many repetitions of
  earthquake slip, combined with erosion
• Fault scarp steep slope formed by fault motion

57
1992 Landers Earthquake Scarp
58
But

• Many faults are not visible at the surface
• Blind thrust faults
• Subduction megathrust faults
• visible only deep under the ocean

59
Strike-slip fault - transform
Left-lateral
Press, 18-12a,d
60
A stream channel offset by the San Andreas fault,
Carrizo Plain, central California
61
San Andreas
Transform margins such as California have
strike-slip earthquakes. Stream is offset here
by 4000 years of motion. Is this left or
right-lateral?
62
CarrizoPlainSanAndreasFault
Note offset stream with right-lateral motion
Yeats et al., 8-23
For scale, dots are trees
63
Anothernearbyplace
64
Famous1906example
San Andreas Fault
Note right-lateral strike-slip motion
Colorized!
NOAA web site
65
1979 Imperial Valley Strike-slip
motion Right-lateral or left-lateral?
NOAA web site
66
Imperial Valley
NOAA web site
67
Imperial Valley orchard
68
(No Transcript)
69
Transform boundaries (continued)
The Dead sea is on a transform fault. The
transform fault has an offset which causes a
pull-apart basin (with normal faulting) Transform
s have shallow, strike slip earthquakes. The same
plate boundary has both strike-slip and normal
faulting
70
Guatemala
1976 1 m offset here
NOAA web site
71
Guatemala 1976
Left-lateral
NOAA web site
72
Sharp fault -split tree
An occasional way to exactly date earthquakes
Guatemala again
NOAA web site
73
Thrust fault - convergence
Press, 18-12a,c
74
Thrust fault trace

• Topographically irregular scarp
• In contrast, strike-slip has straight trace
• Low dip-angle fault plane
• Perched terraces
• Formed during previous interseismic intervals
• Deeply incised canyons
• From rapid uplift

75
Quake 1
Erosion
Thrust fault development
Quake 2
Erosion
76
Thrust fault in Alps
77
Terraces at Wairarapa coast (New Zealand) from
quakes in 1855, 1460, ...
78
Thrust fault scarp, 1980, El Asnam,Algeria
NOAA web site
79
A difficult track - Taiwan 1999
80
A closer look
81
Normal fault - divergence
Press, 18-12a,b
82
Normal fault trace diagram

• Typically steep dip-angle fault plane
• Perched terraces
• Like near thrust fault traces
• Alluvial fans
• Can be very large, as in Death Valley
• Often large, clean fault surfaces

83
Normal fault trace
84
Normal fault - Sierra Nevada
Note big alluvial fan
Mountains up
Valley down
Keller, 2-18
85
Edgecumbe fault, New Zealand,1987
86
Normal fault, Hebgen Lake, 1959
NOAA web site
87
Slickensides flat surfaces made as two sides of
fault slide past each other during fault motion
Corona Heights, SF, CA
Slickenside showing polish structures and
striations.
88
Normal fault face
Slickenlines (large-scale) or grooves in normal
fault, Coyote Mountains, Salton Basin,
California. Photo by Ed Beutner
89
Where are quakes?

• Mostly near plate boundaries Interplate
• Greatest number at subduction zones
• But also plenty at ridges and transform zones
• A few near hot spots Intraplate
• Well discuss with volcanoes
• Some anomalies in US, also intraplate
• Failed rift - New Madrid
• Unloading after ice age? East coast

90
Earthquakes Mgt5, 1963-1988
Keller, 1-5
91
How deep are quakes?

• All types of boundaries have shallow quakes
• 0 to 30 km depth
• Subduction zones also have deeper events
• As deep as 650 km
• Subduction is dragging cold material down
• Cold material is more brittle
• Deeper events breakage of subducting slab
• Mostly from the pull of the weight of the sinking
  slabs
• Some are also caused by bending of sinking slab
• Not from rubbing together of plates

92
Quake cross-section
Press, 1-16
93
Why dont quakes extend deeper?

• Deeper in the Earth, it is hotter
• There is also more pressure, variations in
  composition, and changes in crystal structure,
  but these dont matter as much
• If material is within a few hundred degrees of
  its melting temperature, it quietly flows rather
  than suddenly cracks in an earthquake

94
Earthquake ideas
Mongolians believed Earth rode on the back of a
giant frog
Brumbaugh 1-1
95
Ancient Japan
Japanese folklore, quakes were caused by giant
catfish flailing about under ground
Brumbaugh 1-2
96
Greece
Vibrations caused by Poseidon striking the ground
with his trident were thought to cause quakes.
Brumbaugh 1-3
97
More quake ideas

• Siberia - mighty dog Kozei shaking snow off his
  coat
• China - Earth rode on the back of an ox
• Two Indian ideas -
• Earth held up by seven serpents, or
• Held up by four elephants, who stood on a turtle,
  which stood on a cobra

98
Still more

• St. Matthew 2751
• (at the time of Christs death) the Earth did
  quake and the rocks rent
• 17th century Belgian chemist
• God signals an angel, who strikes a bell
• 1752, British Royal Society
• Earthquakes can only occur where people need
  chastening

99
Scientists misconceptions

• Anaxogoras (500-428 B.C.)
• Explosion of gases underground
• Archelaus
• Compressed air escaping from underground
• Aristotle
• Volcanic gases moving toward the surface
• John Mitchell, 1760
• Subterranean fires
• Finally got it right, 1890-1910 in US Japan

100
Tectonics of western N. Am.

• Pacific and North America are big plates
• Juan de Fuca, Cocos are smaller plates
• Mix of transform, ridge, and subduction
  boundaries
• Location of boundaries has evolved over past 30 My

101
(No Transcript)
102
(No Transcript)
103
(No Transcript)
104
Changes in west coast

• Basically, a mid-ocean ridge subducted
• Before that, just subduction on coast
• First hit near LA 25 Mya
• San Andreas fault system started then
• As ridge is subducting, two triple junctions are
  moving apart on coast
• Mendocino Triple Junction moving north
• NA-Pacific-Juan de Fuca junction
• NA-Pacific-Rivera junction off Baja California

105
Cartoon ofWesternN. Americatectonics
Vidale
106
US and Mexico coast

• Three little plates subducting offshore Oregon,
  Washington, and B. Columbia
• Juan de Fuca Plate
• Gorda Plate
• Explorer Plate
• Spreading ridge splitting Gulf of California
• Oblique because ridges are combined with
  transform faults
• Cocos Plate subducting to the south

107
N.Am.map
USGS Prof. Paper, 1-2
108
Sierra Nevada Mts.

• From 80 to 30 My Farallon plate subducted under
  west coast.
• Juan de Fuca Cocos plates are remnants of
  Farallon plate
• Produced great range of volcanoes, like
  present-day Andes Mt.
• Sierras are the cooled, solidified, uplifted
  magma chambers of the volcanic arc (Yosemite
  granite)

109
US earthquakes
110
North American tectonics

• Notice that all the action is in the west
• Few earthquakes in the East
• Most dramatic topography in the west
• Yosemite, Cascade Mts., Big Sur coast, etc.
• Some faults as far east as Yellowstone
• No action to west in Pacific plate either
• Hawaii is special case - Hot Spot

111
Oregon-Washington subduction

• Two Triple Junctions
• Like a little convection cell
• Up at the ridge, down at subduction zone
• Notice line of volcanoes inland
• From hard-to-sink stuff floating off subducting
  slab, includes Mt. St. Helens
• Best guess - M9 quakes
• Break entire subduction zone
• Come every 1000 years, last in 1700 AD

112
Oregon-Wash.Close-up
Note Volcanoes San Andreas Triple Junctions
Yeats, 5-2
113
Oregon-Washington X-sect
Cascadia cross-section
Cascade Mts
Olympic Mts
Ridge
Keller, 7-5