PowerPoint Presentation EarthScope Scientific Goals

1
Bolivia Chile University of Arizona Carnegie
Institute Lawrence Livermore Nat. Lab. San
Calixto Observatory, Bolivia University of Chile,
Santiago
Lithospheric Thickening and Foundering Mountain
and Capacity Building in the Andes S. Beck, P.
Alvarado1, H. Gilbert, G. Zandt Department of
Geosciences University of Arizona Tucson, Arizona
USA 1Also at the Universidad Nacional de San
Juan, San Juan, Argentina

Argentina Chile University of Arizona Univ.
Nacional de San Juan INPRES University of Chile,
Santiago
IRIS PASSCAL projects Funded by the Geophysics
Program at NSF
2

Earthquake Tsunami Hazards in South America
Local seismic networks focus on earthquake
tsunami hazards. Both intra- and inter-plate
destructive earthquakes tsunamis. Chile 1960
- largest earthquake ever recorded. Numerous
deadly earthquakes Peru 1970 - 65,000 Chile 1939
- 30,000

Gap?
Gap
3

IRIS PASSCAL Seismic Deployments in South America

Proposed Research Mountain building
processes Plate boundary processes Lithospheric
removal Flat slab subduction Ridge
subduction Volcanic arc magmatic processes Seism
ogenic zone Lithospheric structure Continental
formation accretion
4

• Central Andean Plateau
• Normal dipping slab
• Active arc
• Fold thrust belt with 300-450 km of total
  shortening
• Thick weak crust
• Heterogeneous
• mantle wedge
• South-central Andes
• Accreted terranes
• Flat slab geometry, Arc shut off 8 Ma
• Active basement
• cored uplifts
• Fold thrust belt
• Cold dry mantle wedge
• Active arc South of 33S

Altiplano
Western Cordillera, Bolivia Chile Border Active
Arc
High Cordillera above flat slab
5

• Questions
• What is the relative importance in mountain
  building of lithospheric shortening versus
  lithospheric removal?
• How important is subduction erosion and/or
  continental subduction in modifying the upper
  plate and/or mantle wedge?
• Are continental arcs and back-arcs sites of
  significant continental destruction rather than
  addition?

• Observations
• Central Andes 300-450 km of crustal shortening
  (McQuarrie, 2002 Kley and Monaldi, 1998) suggest
  a significant amount of lithosphere has been
  removed as the Brazilian craton subducted.
• Paleoelevation indicate the Altiplano region has
  uplifted 2.5-3.5 km in the last 10 Ma (Garzione
  et al., 2006) suggesting rapid large scale
  lithospheric removal.
• Seismic studies in the central Altiplano show
  heterogeneous upper mantle suggestive of
  piecemeal, rather than wholesale, lithospheric
  removal.
• Seismic studies show variable composition lower
  crust (felsic quartz-rich lower crust versus
  partially ecologitized lower crust).
• Seismic tomography studies show regions of low
  Vp/Vs mantle material that may indicate a
  contribution of subducted continental crustal
  material.

6
Central Andes

• 300-450 km of total crustal shortening
  (McQuarrie, 2002).
• Paleoelevation studies suggest rapid uplift
  between 10 and 6.8 Ma (Garzione et al., 2006).
• Much of the shortening was prior to 10 Ma - how
  do you keep the eleva-tion lt 1 km until 10 Ma?

Paleoelevation estimates from oxygen isotopes in
carbonates (shaded gray) Garzione et al., 2006.
Paleoleaf physiognomy (red bar), from
Gregory-Wodzicki et al., 1998).
7
Crust

• Northern central Altiplano has thick, weak,
  felsic quartz rich crust. No evidence of high
  velocity lower crust.
• Is the lack of high velocity lower crust due to
  modification during mountain building or
  inherited?
• Southern Altiplano shows hints of higher velocity
  lower crust.

8
Comparison of average crustal S wavespeeds
9
E-W Cross-section of Stacked Receiver Functions
ALVZ
Moho
Common Conversion Point (CCP) stacked receiver
functions for the E-W BANJO transect. Positive
arrivals are shown as red (increase in velocity)
and negative arrivals (decrease in velocity) are
shown as blue.
10
N-S Cross-section of Stacked Receiver Functions
APMB
ALVZ
Moho
Vp/Vs 1.75
Vp/Vs 1.8?
Change in crustal character at 20S
11
Central Andes P-wave velocities from travel time
tomography

• Upper mantle is heterogeneous and consistent with
  piecemeal removal.
• Paleoelavation studies (Garzione et al, 2006
  Gregory-Wodzicki et al., 1998) suggest 2.5-3.5 km
  of uplift in the last 10Ma.

Modified from Beck and Zandt 2002
12
Regional Seismic Tomography for Vp/Vs
From Myers et al, 1998

• High Vp/Vs (1.82-1.86, red/pink) under the arc
  and part of the Altiplano.
• Low Vp/Vs (1.72-1.76, blue/green).
• Low Vp/Vs may indicate subduction of continental
  material (Brazilian craton).
• 300-400 km of Brazilian lithosphere has been
  removed.

13
The Central Andes were built by underthrusting of
the foreland lithosphere, accretion and
shortening of felsic crust into the thickened
orogenic wedge, and piecemeal removal of the
subducted foreland mantle lithosphere.
100s of kms of mantle lithosphere have been
recycled into the mantle.
14
South Central Andes Flat Slab (30S) Juan
Fernandez Ridge - hot spot track Subduction
Erosion of continental margin Frontal Cordillera
- high Cordillera, volcanic arc shut off 6-8
Ma Precordillera - fold thrust belt, 65-75
shortening Sierras Pampeanas - basement cored
uplifts, 10-15 shortening, lt5Ma.
Terraines Chilenia terrane Cuyania composite
terrane Pampia terrane Rio del la Plata craton
Active arc (36S) -Andesitic arc volcanism
Quaternary basaltic back arc volcanism

JFR
Volcano CHARGE Stations U. Chile/INPRES
15
Inter-station Pn Analysis
We used 9 earthquakes that occurred offshore to
analyze inter-station Pn arrivals along the
northern CHARGE transect. We find crust
thicker (50-55 km) than expected from the
elevations in the western Sierras Pampeanas
(Cuyania terrain). How do we explain the low
elevations in the WSP? High density lower crust?
From Fromm et al., 2004
16
W-E Cross-section of Stacked Receiver Functions
at 30.5S
W
E
Moho
100 Km
Cross section of Common Conversion Point (CCP)
stacked receiver functions for the northern
CHARGE transect. Positive arrivals are shown as
red (increase in velocity) and negative arrivals
(decrease in velocity) are shown as blue. From
Gilbert et al., (2006).
17
High velocity lower crust from regional waveform
modeling
Cuyania terrane has a higher velocity lower crust
consistent with the receiver function
studies. Consistent with partially ecologitized
lower crust?
Alvarado et al., 2006
18
South-central Andes regional travel time
tomography

Low Vp/Vs seismic anomaly above the flat slab
has a trend similar to the JFR. Is it related to
subduction erosion transporting continental
material from the trench and trapping it above
the flat slab?
JFR
Vp and Vs Red/yellow slow blue/purple fast
Red/yellow high Vp/Vs blue/purple low Vp/Vs
Wagner et al., 2005 Wagner et al., in preparation
19
Combining tomography and RF results at 30.5S

• Summary
• Changes in crustal properties in the Sierra
  Pampeanas correlate with terranes.
• Cuyania terrane has 50-55 km thick crust with
  Vp/Vs1.8 and partial ecologite (high velocity)
  lower crust.
• The flat slab prevents lithospheric removal.
• Subduction erosion of the continental margin
  resulted in a 50 km eastward migration of the
  margin since 7 Ma and could contribute to the low
  Vp/Vs anomaly.

From Gilbert et al, 2006
20

• Summary
• The central and south-central Andes show
  different stages of crustal shortening,
  lithospheric removal, and crustal modification.
• The central Andes backarc
• Central Altiplano has felsic quart-rich crust.
  In contrast, southern Altiplano Puna may have
  higher velocity lower crust.
• Heterogeneous upper mantle suggest piecemeal
  lithospheric removal and modification of the
  crust.
• Low Vp/Vs material in the mantle may indicate
  subduction of the Brazilian craton.
• Tectonic shortening may drive lithospheric
  removal.
• The south-central Andes (30S) backarc
• Cuyania terrane has a weak Moho conversion
  amplitude and high velocity lower crust
  consistent with partial eclogitization. The flat
  slab geometry prevents lithospheric removal.
• Large amounts of subduction erosion of the
  continental margin (due to the subduction of the
  JFR) may contribute to the low Vp/Vs anomaly
  above the flat slab.

21
Capacity Building Difficulties

• South American Universities/Local Seismic
  Networks
• Different goals.
• Local seismic networks need to locate earthquakes
  in near real time - so cant wait for PASSCAL
  data that might not show up for months.
• Often difficult to go back and relocate
  earthquakes due to lack of time and resources.
• Some local collaborators are not be set up to use
  broadband data - some local networks use
  triggered not continuous recording.
• Difficulties in integrating PASSCAL and local
  network data.

• Educational Issues
• Training is often sporadic as PIs come and go
  quickly (especially with declining budgets and
  increased costs).
• In country computers are often not adequate for
  large PASSCAL data sets/software packages.
• Difficult for in country students to get data
  from the DMC as they are not familiar with data
  retrieval methods.

22
Capacity Building Opportunities

• Educational Science Opportunities
• Work to help with archiving local data to make a
  useful combined data set for everyone.
• Improved lithospheric-scale seismic velocity
  models that can be easily used in earthquake
  location programs.
• Help in-country seismologists obtain funding to
  build their seismology program.
• Include projects on local seismicity, historic
  earthquakes tsunamis when possible.
• Build student/faculty exchange programs (weeks to
  months).
• Recruit new graduate students that might go back
  to help build seismology programs in Latin
  America.
• Joint classes and/or workshops in-country.
• Provide training at the PASSCAL center on
  broadband instrumentation.
• Help promote collaboration between Latin America
  countries.

23
SIG Capacity Building in Latin America

• How can we better integrate PASSCAL deployments
  with local seismic networks?
• How can we enhance both the science and
  educational collaborations of PASSCAL projects in
  Latin America?
• How can we get more permanent and temporary
  broadband seismic station coverage in Latin
  America and make the data available?
• How can IRIS and individual PIs contribute to
  improved earthquake tsunami hazards in Latin
  America ?
• Should we consider a more coordinated
  international approach to collaborations in Latin
  America and if so what can IRIS do to facilitate
  this?

24
Seismic Crustal Properties Correlate with
Terranes Regional Waveform Modeling
High Cordillera (flat slab) th 60 km Vp 6.6
km/s, Vp/Vs 1.85 Active Arc th 45 km Vp
5.8 km/s Vp/Vs 1.80
Cuyania th 50 km Vp 6.4 km/s , Vp/Vs
1.80 Pampia th 35 km Vp 5.9 km/s, Vp/Vs
1.70
Alvarado et al, (2006)
Slab contours from Anderson et al., 2006