Middle America Subduction Experiment (MASE)

A cooperative project involving

Tectonics Observatory (TO), Caltech
Institutes of Geophysics and Geology, UNAM
Center for Embedded Network Sensors (CENS), UCLA

TO UNAM CENS

Scientific Objectives

We seek to combine theory and observation into a dynamic model of the evolution of a subduction system over tens of millions of years, including the thermal and compositional state of the mantle wedge and melt production, the shape of the slab, coupling between the overriding and downgoing plates, thermal history of the upper plate, and forces operating on the entire system. We choose the Central America subduction zone as a good location to study this problem, because significant variations in behavior and characteristics along strike are present there. These variations do not appear to correlate with slab age or plate convergence rate. However, they may be controlled by thermal or compositional variations in the mantle wedge itself, by features that were riding on the downgoing plate and subsequently were subducted, or by characteristics of the subduction interface itself. The geochemistry of the magnatic rocks in this region has the potential to record the temperature, water content, source region, and the type and extent of melt in the mantle wedge.

We intend to construct a dynamical (numerical) model of the subduction process that matches the variety of subduction scenarios present in the Central America subduction zone. The parameters for the model will be determined from data that already exist in the region and from surveys we will conduct ourselves.

Data Availability     Bibliography     Links     Papers & Abstracts     Image of the Acapulco-Tampico Transect

Proposed Activities

1) Seismic

We propose three passive seismic profiles to sample the subduction system at different dip/age/rate/volcanic-effusion-rate scenarios. The seismic will provide the upper plate and subducted slab geometry and an estimate of the viscosity (via attenuation) in the mantle wedge. The profiles are planned for central Mexico (Guerrero), southern Mexico (Oaxaca), and Nicaragua. The main estimates of structure will come from receiver function analysis and tomography. The estimates of attenuation will come from the change in frequency content of body waves as a function of distance of the receiver from the trench. Surface waves will also be used to determine structure and attenuation. Measuements of anisotropy will be used to estimate flow direction in the mantle and wedge. We will also look for phases traveling along the slab that will be sensitive to melt and other properties.


Acapulco to Tampico Line

Preliminary results based on a small array of stations that are either part of the UNAM/SSN seismic array or are stations installed as part of the TO project is is shown here, superimposed on Valdez's model.

2) Geodetic
We propose detailed GPS studies looking in particular at the uplift along the seismic lines to estimate the coupling between the upper plate and subducted slab.
3) Plate Tectonics
The subduction rates, plate age and initial slab thermal state for the model will be provided by the plate tectonic studies. We will also examine the other side of the East Pacific Rise system to look for evidence of unusual features that might have been subducted in the Middle America Trench, but whose conjugate features might still be preserved on the Pacific Plate.
4) Geochemistry
We will construct a database on the geochemistry of Mexican and Central American volcanic rocks by way of four tasks: (1) Compilation of published and unpublished but publicly distributed data (including theses and internet databases) for Tertiary volcanic rocks from the trans-Mexican volcanic belt, southern Mexico, the Central American arc between Guatemala and central Costa Rica, and putative adakites from southern Costa Rica and Panama. (2) Collection of a new suite of Tertiary volcanic rocks from the northern and southern edges of the trans-Mexican volcanic belt, from suspected volcanic centers in southern Mexico and northern Guatemala, from Miocene parts of the Central American arc, and from Panama. (3) Determination of the mineralogy, mineral chemistry, and major- and trace-element geochemistry of collected samples (including thin-sectioning, electron-probe analyses, and contract measurements from outside labs), and ages of select samples (by Ar-Ar, contracted at an outside lab). We anticipate studying between 100 and 200 samples by some or all of these methods. (4) Study of the isotopic compositions and melt-inclusion properties from select samples using a variety of in-house techniques (oxygen isotopes, FTIR) and by sending group members to use outside labs (ion probe, TIMS and MC-ICP-MS). We anticipate studying between 50 and 100 samples by one or more of these methods. The collective database will be used to infer the compositions of primitive melts and their degrees of differentiation and hybridization with the crust, and the compositions of those putative primitive melts will be used in turn to infer the temperature and water content of the sub-arc mantle and the phase identity and specific source of slab-derived metasomatic agents added to that mantle.
5) Numerical Modeling
A suite of dynamic models are anticipated that will allow integration of geophysical and geochemical observations in time and space. Instead of creating just one master model of southern Mexico and Central America, a variety of models will be used to predict specific classes of observations, test specific hypotheses on mechanical and other interior properties, and guide the deployment of instrumentation and sample collection. We will formulate instantaneous models which fully integrate all structural constraints inferred from seismology and geochemistry. Time dependent models (2-D & 3-D) which will fully incorporate plate tectonic history since the early Miocene will be formulated. Models will be tested against observed topography, gravity, GPS derived surface strain field, surface uplift & subsidence on geological time-scales, structural & stratigraphic constraints on fault growth and slip, and geochemically inferred temperatures and melt transport. During the course of the study, few restrictions will be placed on the classes and geometry of numerical models. For example, the crust can be treated as a cohesive solid, the lithosphere as a visco-elastic solid, and the mantle as a non-linear creep fluid. Percolation of magma through the solid system will be accounted for and faults will either be imposed apriori or allowed to nucleate and grow through time. Local and regional models will not be influenced by artificial boundary conditions as all models will be incorporated into large scale formulations with realistic time-dependent tectonic boundary conditions imposed from global plate/mantle circulation models. We will use all models and data within a framework allowing code coupling and data assimilation in both GIS and plate tectonic reference frames.

Initially, models will be used to plan optimal seismic and GPS network configurations. With existing structural and geochemical constraints on slab and mantle wedge melting, we will be able to estimate acceptable models with locally weak mantle wedges (mantle wedges will be weakened through a combination of partial melting and water induced low viscosities). Such weak zones will be mapping into zones of high seismic attenuation. We will then use such attenuation models in the computation of the full seismic wavefield from which we will determine the optimal configuration of broad band seismic networks. As our modeling ability improves and our data base enlarges, we will refine our dynamic picture of how slabs influence the mantle wedge through melting and other transport mechanisms, and in turn how these properties influence surface deformation.

With time-dependent models of the downgoing oceanic lithosphere, we would attempt to match the plastic failure of the oceanic lithosphere through the development and growth of normal faults. Such models would be tied to existing bathymetric and gravity observations within the trench, OBS surveys of seismicity within the trench & forebulge, and potentially MCS constraints on the faults at depth within the oceanic crust.

6) Geology and morphology
Geological and geomorphological studies of the active deformation in the upper plate near the trench will provide an additional constraint on the details of the plate coupling, and long term deformation. Geologic analysis will also be important for determining the history of the subducted plate.
7) Magneto-telluric
Long-range magneto-tellurics can provide a measurement of the conductivity in the subsurface. It can be used to directly detect fluids in the mantle wedge.

Participating Scientists


Jorge Arzate, UNAM-Queretaro, magnetotellurics
Jean-Philippe Avouac Caltech, tectonics
Ilya Bindeman, Caltech, geochemistry
Robert Clayton, Caltech, seismology
Paul Davis, UCLA-CENS, seismology
John Eiler Caltech, geochemistry
Ken Farley, Caltech, geochemistry
Luca Ferrari, UNAM-Queretaro, tectonics
Michael Gurnis, Caltech, geodynamics
Juan Martin Gomez, UNAM-Queretaro, seismology
Don Helmberger, Caltech, seismology
Arturo Iglesias, UNAM, seismology
Hiroo Kanamori Caltech, seismology
Vladimir Kostoglodov, UNAM, GPS,tectonics
Carlos Mortera, UNAM, seismology
Javier Pacheco, UNAM, seismology
Xyoli Perez-Campos, F1-UNAM, seismology
Osvaldo Sanchez, UNAM, GPS
Peter Schaaf, UNAM, geochemistry
Shri Krishna Singh, UNAM, seismology
Joann Stock, Caltech, geology, GPS, tectonics
Yuri Taran, UNAM,
Arturo Gomez Tuena, UNAM, geochemistry
Carlos Valdes, UNAM, seismology
Raul Valenzuela Wong, IFG-UNAM, seismology
Dante Moran Zenteno, UNAM, tectonics,geochemistry
Progress Report 2006