The Subduction Factory Initiative is a component of the MARGINS program. The MARGINS approach concentrates resources on areas targeted for intense, multidisciplinary research. In these focus areas, interaction between researchers involved in field data collection, numerical simulation and laboratory analysis promises unparalleled synergism necessary to understand complex natural systems such as the Subduction Factory. The operation of the Subduction Factory involves lithospheric deformation, mass fluxes, sedimentation, melts, and aqueous fluids. The MARGINS philosophy is realizing the goals of the Subduction Factory Initiative by sponsoring coordinated, interdisciplinary investigations in these areas.
The fundamental goal of the Subduction Factory Initiative is to understand relationships between input and output mass and energy fluxes through a subduction zone, and to use this information to address the fundamental science questions outlined above. Realizing these goals requires a theoretical framework that quantitatively reproduces the observable consequences of plate subduction. Because subduction zones are among the most complex components of the solid earth system, achieving this goal will require a coordinated effort by a wide range of earth scientists interacting as members of an interdisciplinary team. The purpose of this section is to suggest guidelines for organizing the wide range of necessary studies.
Integration of the diverse science activities at a given convergent margin should lead to a robust physico-chemical model for how the Subduction Factory operates. We see the next generation of Subduction Factory models as progressing from preliminary model development to data acquisition to construction of refined models. For several convergent margins, sufficient data exist such that work can begin to develop specific physical and chemical models. Models will reveal areas of uncertainty to be addressed by subsequent field campaigns, laboratory analyses, and experiments. These results will constrain the construction, testing, and refinement of the next generation of models. The ultimate success of the Subduction Factory Experiment will be judged by the extent to which models become increasingly comprehensive, can be tested from the observables, and are able to predict behavior at other subduction zones. This requires an interdisciplinary scientific dialogue that promises to result in a quantum leap in our understanding of forcing functions, volatile cycling and mass balance in the Subduction Factory.
A series of early Subduction Factory Theoretical Institutes should focus on how best to develop the physical and chemical models. The Institutes should result in discussions of what geochemical and geophysical data are needed to drive the field and laboratory efforts. Periodic Subduction Factory Theoretical Institutes will maximize scientific team efficiency and recruit new team members. The first of these was held in Eugene, OR in August 2000; the special volume is now in proof.
A wide range of data is needed to constrain and test physical and chemical models. Coverage for many of the geophysical methods should extend seaward of the trench to include the incoming plate. This is necessary for completeness, but more importantly as a means to evaluate possible deep hydration and serpentinization of the llithospheric mantle of the incoming plate.
Improved swath mapping techniques are providing detailed and evocative bathymetric images of the subducting plate and the leading edge of the upper plate, along with submarine portions of the arc and backarc (cross-chains and spreading centers). Such images are extremely useful for Subduction Factory research in a number of ways. They can identify fault scarps formed as the plate bends into the trench, where downdropped grabens may become sediment-filled buckets, and horsts may affect the faulting structure of the upper plate. The faults themselves may be conduits for seawater flow to basement. Bathymetric images also help to identify places where smooth sedimented sea-floor vs. rougher topography (seamounts or ridges) is being subducted. Swath mapping reveals the response of the leading edge of the upper plate to topographic perturbations on the incoming plate. The images provide clues to processes such as prism evolution, frontal accretion, development of the deformation front, and subduction erosion, all important for understanding subduction dynamics and mass transport to depth. In non-accretionary margins, swath mapping can reveal the size, distribution, and to some extent structural setting of serpentinite mud volcanoes. In either case, swath mapping is a useful tool for identifying regions likely to be structurally complex enough that 3-D seismic surveys are important.
Active-Source Seismology: Images and seismic velocities are obtainable at scales useful for probing the Subduction Factory to depths of about 25 km using newer experimental techniques and focused observational programs. Below 25 km, depth and resolution are limited by the difficulty in propagating broadband sound to great depths, and passive sources (seismicity) becomes the more powerful tool. Active source imaging can elucidate the top of the down-going slab, delineate structures within the base of the overriding plate, and define structural and velocity details within the forearc and parts of the arc-mantle wedge. This is true for accretionary margins, where the subducted material may either be underplated or carried deep into the Subduction Factory. For many non-accretionary forearcs, imaging will be difficult, but low velocities associated with serpentinite bodies should be mappable. In addition, the initial volatile losses within the upper part of the oceanic crust will be detectable given sufficient velocity resolution. Recent seismic programs to study ocean ridges are applicable to the subduction factory, including active-source tomography.
Seismic sources must be large to penetrate to the needed depths and contain a broad-band spectrum of energy to preserve resolution and allow waveform inversions of the reflections. The best way to obtain high-quality images and velocities is by using 3-D seismic reflection acquisition methods and prestack depth migration. These techniques require high quality data as well as high-performance computing capability. Multichannel seismics, particularly 3-D acquisition and processing, have been shown to provide high-quality images of the décollement and structures above and below. For instance a 3-D data set from Barbados directly mapped the location of aqueous fluids along the décollement as well as in fault ramps splaying into the overlying accretionary prism (Figure 10). Where the structures above the seismogenic zone are more complex, 3-D methods provide the essential first order corrections for the overlying structure. If the shallow structure is not properly accounted for, reflection amplitudes and waveforms of deeper events will be severely distorted.
Reflection and refraction techniques become more powerful when combined. Closely spaced ocean bottom seismographs/hydrophones (OBS/H) along modern normal incidence reflection lines have been used to extend structural imaging to depth as well as provide unique velocity data. These data also provide background velocities to combine with reflection waveform analysis. Active-source tomography has been used to map the structure of a seamounts, young oceanic crust, and rifted oceanic crust using 3-D arrays of closely-spaced OBSs with either conventional surface sources or explosive bottom sources. Similar experiments can help constrain shallow volatile losses from the subducting plate and the degree of serpentinization of non-accretionary forearcs.
Passive-Source Seismology: The wave trains of seismic signals sampling slabs can provide unique information about the structure and composition of subduction zones, at depths relevant to magmagenesis and tectonic driving forces. The development of high-fidelity broadband recording has made it possible to gain far more information from earthquake recordings than previously, particularly from the late parts of seismic signals.
At regional event-receiver distances, subduction zones produce several unusual seismic phenomena. Large P-to-S and S-to-P conversions are commonly observed, which reveal boundaries of the subducted crust and the associated material contrasts. These conversions map boundary locations at accuracies of a few km, and provide a good picture of the relationship between the earthquakes usually used to define subducting slabs and the actual material boundaries at depth. Body waves traveling along slabs are also severely distorted or dispersed, a phenomenon that is used to constrain otherwise inaccessible properties of the subducted plate such as seismic velocities within subducted crust and the thickness of that layer. These measures can provide in situ constraints on the extent to which the basaltic crust metamorphoses to eclogite, and the depths to which it persists at blueschist facies. Finally, large temperature variations in the subducting slab and mantle wedge generate strong changes in seismic attenuation, which is now observable over a wide range of frequencies. Attenuation studies can provide constraints on temperature variations beneath arc volcanoes independent of those provided by velocity tomography.
Teleseismic wave trains reveal strong converted-wave signals from the subducting slab at overlying stations, such as P-to-S and S-to-P conversions. These signals, usually analyzed as receiver functions, provide direct information on the location of the slab and other discontinuities, and on their impedance contrasts. They also provide sensitive measures of Poisson’s ratio—useful because serpentinite has a anomalously high Poisson’s ratio. Another kind of observation is provided by shear-wave splitting measurements, which constrain the flow-induced fabric of the mantle to the extent that olivine crystals align and produce bulk anisotropy. These observations provide tests of dynamical models in the mantle wedge and elsewhere. Both of these measurements are now being made routinely from portable PASSCAL-type deployments, and are rapidly expanding our understanding of the Earth’s deep interior.
Requirements for Seismic Imaging of the Mantle Wedge: A good distribution of earthquakes and locations for seismic stations is essential for passive imaging of upper mantle structure associated with the Subduction Factory. Ideally, an arc would show a high level of seismic activity throughout the upper mantle to depths of ~600 km, and a broad land region for operating a land seismic network. In practice, the sub-arc magma production region can be well imaged as long as earthquake activity extends beneath the volcanic front to depths of 150-200 km. In addition, ocean bottom seismographs (OBSs) may be used in lieu of land seismic stations for arcs with limited land exposure, if seismicity rates are high. Some OBS deployment in the forearc and backarc are generally necessary in most arcs to image a wide region. Table 1 lists the seismicity rates at various depths for the arcs under consideration.
Tonga shows the highest seismicity rate, with seismicity distributed throughout the upper mantle, but would require an OBS deployment. The Mariana and Izu-Bonin slabs also show seismicity throughout the upper mantle, but have lower seismicity rates, and would also require OBSs. Japan has good intermediate depth seismicity and an exceptionally dense seismic network, allowing the best detail in tomographic images (Figure 7), although offshore areas are not well imaged. The Aleutians and Central America lack deep seismicity, but have adequate intermediate depth seismicity, and good local networks in place at some locations along strike. Cascadia lacks seismicity deeper than 60 km, and thus does not permit detailed imaging, although the dense land network permits some tomography with teleseismic rays.
From magnetotelluric (MT) studies at Cascadia, we know that the uppermost part of the subducting plate is about ten times as electrically conductive as normal mantle. Enhanced electrical conductivity at subsolidus temperatures is principally caused by the presence of water; the addition of 0.1 wt% water to dry olivine enhances conductivity by nearly two orders of magnitude. Consequently, elucidating conductive pathways serves as a geophysical tracer of the flow of water into the mantle. Numerical 2D modeling shows that moderately good MT data has the potential to distinguish between hydration of the upper slab, hydration of the adjacent mantle wedge, and localized enhanced conductivity in the thin central part of the wedge (Figure 11).
High-quality heat flow data provide critical information on the thermal structure of a subduction zone, which forms the basis of seismological and petrological models. Surface heat flow data need to be collected on the incoming plate and in the forearc, arc, and backarc regions. For example, offshore and onshore heat flow data have been used to demonstrate that frictional heating is negligible in some subduction zones. In North-East Japan, comparison of seismic velocities derived from tomographic inversion with the sublithosphere temperatures derived from heat flow suggest that the forearc mantle is hydrated in this mature arc. Heat flow data can also constrain the geometry and magnitude of fluid flow through the forearc.
ODP-IODP: The ODP-IODP Program has been, and will continue to be, essential for studies of the subduction factory. Much of what we know about the alteration of the incoming plate and the composition of its sedimentary veneer comes from ODP drilling. Recovery of the sedimentary section outboard of the trench will continue to be important. In addition, scientific and technical progress has changed the way in which the ODP-IODP Program can be used. Casing techniques can provide better hole stability for deeper penetration and core recovery in the compressive regime of the accretionary prism. Logging-while-drilling techniques provide high quality logs for density and porosity in fore-arc sites. Pore fluid sampling and analysis both outboard and inboard of the trench provide a very sensitive look at the diagenetic, hydrological and chemical changes in the earliest stages of subduction. Deep drilling with the ODP-IODP Program would allow penetration into the altered oceanic crust in the deep waters near trenches. In any study of the Subduction Factory, it is essential that a reference site be drilled outboard of the deformation front.
Riser Drilling: Aspects of the Subduction Factory Initiative that were discussed at the CONference on Cooperative Ocean Riser Drilling (CONCORD) include subduction zone earthquakes, the initiation of subduction, formation of juvenile arc crust, and mass fluxes at convergent margins. Riser drilling would ultimately allow deeper penetration, improved hole stability and better recovery under difficult drilling conditions typical of convergent margins. Drilling through the seismogenic zone will provide samples of aqueous fluids and of accreted and subducting sediments, necessary for understanding shallow subduction processes, and their effect on the slab delivered to depth. Riser drilling could provide improved access to deeper fore-arc serpentinites and associated pore-fluids. It could also provide a longer record of arc evolution through deep drilling in well-chosen locations in the arc edifice or in subsiding basins that receive volcanic sediments or ash.
Borehole Observatories: Increasingly, boreholes are used for hydrological or seismic experiments while the drill ship is on station, or as long-term observatories. CORK (Circulation Obviation Retrofit Kit) technology allows boreholes to be sealed and isolated from seawater. Perforated casing or screened intervals allow pore fluids from formation levels of interest to percolate into the borehole. The pressure and temperature is monitored, and borehole fluids and gases are sampled osmotically and stored for later chemical analysis. Tracer injection studies allow estimates of fluid flow rates. Filtered water samples can also be used to investigate the microbiology of the site. New developments are leading to a second generation CORK that will allow multiple levels in the borehole to be isolated from each other and from seawater, so that the hydrology, chemistry and microbiology can be investigated at different levels in the borehole. Such observatories would be particularly useful in the fore-arc of margins selected for focused study, with isolation of intervals above and below the décollement and in the basement of the subducting plate. In the often unstable hole conditions of convergent margins, it will be necessary to develop packer technology that will make effective seals against sometimes unstable formations.
To date, CORKs have only been used with ODP drill hole equipment. A mechanism whereby smaller CORKS (mini-CORKs) can be used in conjunction with gravity core or piston core equipment is currently under development. The corers and mini-CORKS can be used to study sites of fluid flow along fault zones or conduits in convergent margin settings. They can be deployed cheaply and in large numbers to effect “arrays” of seafloor monitoring sites. The mini-CORKs and down-hole monitoring devices are designed to be compatible with ODP dedicated borehole observatories and with JAMSTEC designed long-term monitoring devices. Thus, these mini-observatories could be linked with borehole observatories as arrays for the investigation of 3-D aspects of various phenomena associated with convergent margin processes, such as hydrologic processes, heat flow, seismicity, regional strain, geochemical variability, and biologic processes.
ROVs: One aspect of 3-D mini-CORK observatories in convergent margins is that deployment and servicing of these devices may be in water depths exceeding 6500 m. Thus some applications will require deep water ROVs (Remotely Operated Vehicles), control devices or AUVs (Automated Underwater Vehicles). In addition to servicing seafloor observatories, next generation ROV capability will allow investigating otherwise inaccessible parts of the submarine subduction zone. Ideally, new ROVs will combine the robustness necessary for operation in deep waters, maneuverability, video for high quality mapping and sample recovery (aqueous fluids, sediments, rocks).
GPS is currently the premier method for determining 3D displacements in a global reference frame. For the Subduction Factory, GPS will be important for several reasons. It will allow precise measurements of contemporary convergence rates, and how they vary along the margin. It can constrain intra-arc strain, deformation and crustal shortening in response to subduction of bathymetric features such as seamounts and volcanic ridges. GPS studies and associated modeling can also be used to investigate modes of back-arc spreading and rifting, constraining the role of actively driven (magmatic) rifting.
Arc Magma Production Rates: It is essential to know not only what the Subduction Factory makes, but how fast it makes it. Magma production rates are necessary to assess the influence of the forcing functions, calculate volatile and other elemental fluxes, and constrain the rate of continental growth. Assessing magma production rates in arcs is more complicated than for mid-oceanic ridges, where crustal production is simply the product of crustal thickness and the full spreading rate. In contrast, arc crust may include pre-existing material and growth may be non-steady-state. Further complications arise because arcs grow vertically, arc magmas are fractionated, and the most productive arc volcanoes are explosive and subaerial. The arc environment is also conducive to mass losses through crustal delamination.
One method for estimating convergent margin crustal growth takes the total crustal volume for the magmatic arc and divides it by the age of the arc system. This results in an estimate around 1 km3/yr globally. While this is probably sufficient for a global average (with a factor of 2 uncertainty), more precise regional rates are needed to address the central scientific issues.
Another approach for calculating magma production rates uses the arc eruption rate and the ratio of intrusives to extrusives. A ratio of 2:1 has been inferred for the Aleutian Arc, but this is poorly constrained. Further petrologic studies in addition to study of deep crustal exposures of plutonic arc sections should be pursued to aggressively address intrusive/extrusive partitioning. In addition to the petrological approaches, refraction studies of intra-oceanic arc systems, such as shown in Figure 9, may provide an independent means to estimate plutonic layer thicknesses, assuming that the velocity structure of arc crust can be interpreted as due to either intrusive or extrusive igneous rocks. This is another reason why deep geophysical sounding of intra-oceanic arc crust is a top priority of the MARGINS program.
Nor will it be easy to quantify arc eruption rates. The largest arc volcanoes generally have the best resolved chronologies, and conical stratovolcanoes are the simplest geometries for estimating eruptive volumes, both prerequisites for reliable eruption rate estimates. But a significant fraction of the volume of volcanoes has been lost as violently ejected ash or washed away by glaciers and streams. An alternative strategy is to estimate eruptive volumes over submarine arc volcanoes, where erosion is negligible and violent dispersal is minimized. This approach will require detailed marine reflection studies that can be tied to drill cores in order to reliably estimate volumes and establish chronologies.
The eruption rates and intrusive/extrusive ratio at different arcs will naturally vary due to different crustal structures and stress regimes. For example, low-density continental crust will retard rising of mafic magmas so that the ratios will be higher than for arcs built on high density oceanic crust. An important site selection consideration for at least one of the arcs to be studied is that it should be a good place to determine both intrusive to extrusive ratios and eruption rate.
The Importance of Primitive Arc Melts: In order to understand how the Subduction Factory operates, we must know how one of its most important product—magmas—are produced. To do this, we first must know the composition of unfractionated, primitive melts. This knowledge is essential for calculating mass fluxes, which is itself a paramount goal of the Subduction Factory initiative, but many other benefits accrue. For example, if we know the composition of primitive arc melts, we can reproduce these experimentally to constrain temperatures, pressures, and volatile contents in the mantle at the point of melt generation, providing constraints for theoretical models of the mantle wedge that can be obtained no other way. Furthermore, this information will allow us to move from speculation to quantification of otherwise intractable problems such as formation of the lower crust and lower crustal delamination.
Deducing the composition of primitive arc melts is not simple, because although scientists agree that most arc magmas originate by melting of subduction-modified mantle peridotite, we rarely find the aphyric and unfractionated lavas that record this equilibrium. In contrast to basalts from other tectonic environments, arc lavas have lost nearly all of the volatiles bestowed at the time of melt generation. This is true for lavas erupted from submarine as well as subaerial arc volcanoes, and it is likely that degassing and melt fractionation are closely linked. Just adding water lowers mantle melting temperatures by several hundred degrees, and crystals will form rapidly as decompressing melts approach the surface. An important part of the Subduction Factory initiative must be learning to interpret magmatic evolution from degassed, porphyritic arc lavas.
Geochemical and Microbeam Approaches: One approach to reconstructing primitive arc melts is to use long-established geochemical and isotopic techniques. Some studies, principally isotopic investigations (Sr, Nd, Hf, Pb; Rare gases; U-Th disequilibrium; 10Be), which provide essential information such as mantle or subducted slab isotopic signatures or melt generation and ascent timescales, may still be carried out to good effect on fractionated or accumulative lavas.
Tremendous opportunities to surmount problems posed by porphyritic lavas are provided by recent technological advances in microsampling and microanalysis. Small melt samples (< 100 microns) are captured in phenocrysts and frozen as glass. These glass inclusions are extremely valuable because they are more easily reconstructed to pure melt compositions, sometimes more primitive, and much less degassed than erupted lavas. In fact, glass inclusions have provided the only direct means to determine magma volatile concentrations. Several established or developing microanalytical techniques have opened-up the study of melt inclusions: electron microprobe for major elements, ion probe for trace elements, Fourier transform-infrared spectroscopy for H2O and CO2 contents, and laser ablation-multiple-collector ICP-MS for isotopes. These microanalytical techniques can also permit study of individual crystals in mantle xenoliths and stratigraphically controlled tephra glasses.
It is essential that petrologic-geochemical-isotopic studies of the arc suites selected for study be coordinated among the various laboratories where this work can be carried out. Because of the wide range of lava types that can be encountered at a single arc and the many directions that studies of these rocks can take, it will be important to form a team committed to the complete range of studies on a sample suite.
Temporal Evolution of Arcs and Approach to Steady-State: We need to understand how the Subduction Factory has evolved through its life. This is needed to assess the extent to which the present operation of the factory reflects its past. Is the system in steady state? If so, how long did it take to attain this condition after subduction began? There are three ways to do this, each of which samples arc history differently. All three require drilling, but at different distances from the volcanic apex. Because sedimentation rates decrease over several orders of magnitude as distance from arc volcanoes increases, we can recover much longer histories more efficiently by drilling farther away from the arc. Drilling through the volcanic carapace yields a detailed history through the lifespan of a single edifice. Studying samples recovered by drilling through volcaniclastic (mass flow) deposits at some distance (10's of km) from the arc integrates the histories of several arc volcanoes through the life of the sedimentary basin. Studying samples drilled downwind 100's of km from the arc or in forearc basins reveals the history of subaerial, explosive volcanism in the arc system, provided contributions from other volcanoes can be resolved or neglected. The microanalytical techniques discussed above are increasingly important in moving through these three scales of dispersal, not only for the reasons outlined previously but also because the far-traveled tephra in particular is so fine that it cannot be analyzed any other way. Application of this technology to arc history has already contributed tremendously to our understanding of arc magmatic history of the Mariana-Izu Arc system.
Sedimentation and Arcs: Finally, the evolution of sedimentary basins built on the roof of the Subduction Factory—both forearc basin and back-arc basin - provide an easily accessible record of the past activity of the factory. In addition to chemical evolution of the arc preserved in these volcaniclastic sediments, the subsidence history of these basins and the diagenetic history of these sediments constrains the thermal evolution of the arc lithosphere. Assuming these deposits are submarine in the arc system being studied, it is essential that studies of forearc and back-arc basin sedimentary sequences be based on MCS and heatflow surveys leading to ODP drilling.
Rapid Response Plans: We need to be able to quickly reach places where earthquakes have just occurred or volcanic eruptions are in progress or about to happen. A framework for rapid responses to these or other phenomenon must be developed and implemented.
Seismic Calibration: Interpretation of the P and S-wave velocity requires calibration with measurements at relevant temperatures and pressures under laboratory conditions. These are difficult measurements, particularly with hydrous materials, but are critical for interpreting the seismic data.
Experimental Petrology: The following experimental goals are essential for constraining models of how matter is transferred from the subducting plate to the overriding plate, how different elements equilibrate with and migrate through the mantle wedge, how melts are generated in the mantle wedge, how they rise to the surface, and how they fractionate before they erupt:
1) In order to know at what temperature and pressure critical dehydration reactions occur, we need experimental studies of the subsolidus transformations of water and CO2-bearing phases and rocks in the subducting slab. Emphasis will be placed on understanding the dehydration behavior of natural mixtures or analogs appropriate to the focus margins.
2) In order to reconstruct metamorphic reactions in the subducted slab, as well as melting in the slab and mantle, we need to understand the partitioning of specific tracers between melts, solids and aqueous fluids. Tracers chosen for focused study will reveal fundamental processes in the Subduction Factory, constrain development or testing of models, or are essential for mass balance. A list of such high priority tracers include species that illuminate sources in the subducted slab, (i.e., 10Be, B, Li, Cl, Ar), elements that reveal transport timescales (i.e., the U-series nuclides: U, Th, Ra, Pa), species mostly derived from the mantle wedge (e.g., 3He, Nb, Yb), elements that reveal mineralogy where melting occurs (REE, Sc, Y), radiogenic isotopes that reveal source histories (i.e., Sr, Nd, Hf, and Pb), and species comprising the fluid itself (H2O and CO2). There is also virtually no partitioning information for key minerals in the slab, including lawsonite, phengite, serpentine, and aragonite. Determining partition coefficients for these phases will allow us to estimate the composition of aqueous fluids or melts at the point where these leave the slab and enter the mantle wedge.
3) In order to understand how dense aqueous fluids and hydrous melts move through the mantle and interact, we need thermodynamic modeling and experimental investigations of the wetting and transport properties of dense aqueous fluids in slab and mantle lithologies. These constraints will help refine chromatographic and fluid migration models for the mantle wedge.
4) In order to understand how magmas are generated in the mantle wedge, we need experimental analogs for hydrous flux melting of peridotite, amphibole peridotite melting, and decompression melting of hydrous and amphibole-bearing peridotite, over a pressure range of 2-4 GPa. Experimental analogs are also needed for mafic melt crystallization during volatile loss in order to understand how arc magmas fractionate. Experimental constraints on the generation of felsic magmas are necessary to understand how continental crust forms.
As outlined above, developing models for the Subduction Factory is integral to the strategy of the Initiative. A testable model must be able to describe the observable geochemical consequences of slab and mantle processes. Thus geodynamic, physical models for subduction and media flow must eventually incorporate chemical partitioning such that the chemistry of fluid and melt outputs can be used to constrain the models. The following aspects should be a part of any modeling effort:
1) Subduction of lithosphere and mantle convection beneath the arc and back-arc, if present.
2) Metamorphism, dehydration, and partial melting of subducted crust and sediments
3) Thermal structure in the convecting mantle wedge and the subducting plate.
4) Flow of aqueous fluids and melts from the subducting plate to the site of initial melt generation in the mantle wedge, including porous and channelized flow, the effect of the convecting mantle on the migration path and composition of aqueous fluids and melts, and fore-arc sites of aqueous fluid egress.
5) Ascent of melts through the mantle to the base of the crust, including diapiric ascent, porous flow, channelized flow, and melt-rock interactions.
6) Storage of melts in the crust and their subsequent fractionation, degassing and eruption
A focus for initial work should be the internal workings of the factory, the engine that transfers down-going material to up-going material. This includes wedge convection and melting. Melting in the mantle wedge probably results from adiabatic decompression as well as hydrous fluxing. To assess the significance of decompression melting requires tracking the movement of the solid, including the residue of melting. Existing kinematic models do not allow the thermal component of mantle buoyancy to be considered rigorously; hence dynamic models of subduction will be required. Observations that are needed to constrain dynamic models include slab shape, rollback rate, surface heat-flow, seismic imaging (velocity, attenuation and anisotropy), gravity field, surface stress state and surface subsidence history. To constrain the inputs to these models will require a better understanding of the density and rheology of hydrated, molten, fertile and residual mantle. Such dynamic thermal models will form the template on which models of melt and fluid migration and chemical interactions can be developed and tested with magma geochemistry.
Interdisciplinary studies and international cooperation require free and easy access to a wide array of data. Despite its importance to the broadest community of scientists, development of databases is a neglected effort. Different communities within the Earth Sciences have organized their data to varying degrees, from well-managed databanks (like seismic data through IRIS) to data that is scattered among individual scientists. Because of the multidisciplinary team approach of all MARGINS initiatives, it is essential to develop open databanks. For the same reason, samples collected under MARGINS aegis must be properly curated.
We envision two types of databases: one for each of the focus areas, and one for global data. The first database developed should be for the sites chosen for focused, interdisciplinary study. Much data is already organized through such organizations as ODP and IRIS, and MARGINS should not duplicate these efforts. Instead, emphasis will be on data not routinely curated, such as bathymetric and geologic data, seafloor imagery, seismic reflection data, measurements of potential fields and heatflow, and geochemical analyses. We envision an RFP for web-based or GIS-based database that takes advantage of experience gained from development of RIDGE databases. A link should be made to GERM (the Geochemical Earth Reference Model) which has a program already underway to develop geochemical databases for volcanic and sedimentary rocks.
Another vital aspect of the Subduction Factory Initiative is to develop global databases, such that comparative studies can be made using results coming out of the focus areas. An effort identified for initial work is development of geochemical databases for input and output products. The geochemical data for subduction-related igneous rocks is poorly organized and not currently available to a wide community of investigators. A modest investment of resources and effort could lead to potentially great gains in identifying relationships to geophysical measurements, or providing constraints on theoretical models. Also essential for SubFac-sponsored projects is sample curation and distribution. Although marine samples are generally well curated at the major marine institutions and ODP, with clear labeling and distribution protocols, the same is not always true of terrestrial samples. Samples collected from subaerial volcanoes are dispersed throughout the geology departments of the country, with non-standard numbering schemes, vulnerable to separation from their geographic location. At the minimum, organized and open sample collections should be developed for each of the focus sites.
To sample the products of the Subduction Factory and to image its internal workings requires integrated interdisciplinary experiments that are focused on a small number of convergent margins. Subsets of the above studies used in different margins reveal the power of the approaches. However, the complexity of convergent margins, such as variations in slab temperature, water flux and slab and mantle chemistry, make it very difficult to understand the underlying processes except in the context of a focused experiment. The magnitude of the investment needed for a focused experiment requires guidance to margins where focused study promises scientific breakthroughs.
Criteria for Selection of Focus Sites: The following are a refined set of guidelines developed at the various workshops for focusing discussion on the optimal margins for an integrated experiment:
• Must have an active volcanic arc. Sampling the volcanic output of the factory is clearly essential. Nankai, endorsed by SEIZE, does not have an active arc.
• Weather, local infrastructure and government regulations must not impede study of margin.
• Enough background information must be available to formulate a focused experiment.
• The subducting input must be accessible to sampling by existing drilling technology. Output (gases, aqueous fluids, volcanic sediments, volcanic rocks) must be available as needed to address the major questions.
• Seismic illumination of the subducting slab and overlying mantle wedge between 0 and 200 km depth must be feasible with on-land stations or in a realistic time frame for OBS deployment. Illumination to > 200 km is desirable. Both active and passive source methods are essential.
• An historical perspective is necessary to evaluate the question of whether steady-state conditions are an acceptable approximation for the subduction time of the margin (i.e., the time for trench input to be processed through the factory). For margins with relatively fast convergence rates, this characteristic time is about 2 Ma, and lengthens as the rate slows. Where the subduction time is greater than about 2.6 Ma, changes in sedimentation patterns, particularly at high latitudes, must be factored in.
• Margin(s) selected must show variation in forcing functions, either through variation along-strike in a single margin or through contrasting values between margins. Key parameters include convergence vectors, slab temperature, sediment transport to depth, and upper plate structure.
• At least one margin endorsed for focused study should allow a cross-arc perspective. Factory outputs (aqueous fluids, gases, magmas, volcanic sediments, hydrothermal deposits) should be recoverable across a wide swath from fore-arc to back-arc, in order to integrate the sum of factory processes.
• Continental contamination of ascending magmas must either be minimal or decipherable.
• The margin must be an optimal candidate for addressing one or more of the three science objectives highlighted previously: 1) Subduction parameters as forcing functions, 2) Volatile cycle, 3) Towards mass balance and crustal growth.
Assessment of Candidate Margins: Many candidate convergent margins fail to meet some of the above criteria. For example, weather conditions preclude extended access to the Scotia arc. Drillship access to Indonesian waters has been limited. Infrastructure in Kamchatka makes field work difficult and expensive. The slab beneath Cascadia is hot enough that little seismic energy is released, and seismic imaging would be difficult except by teleseismic methods. Crustal contamination of many lavas in the Lesser Antilles, Andes and New Zealand makes it difficult to invert magma composition for processes operating deeper in the factory. The tectonic complexity of the Philippine collision zone, Bismarck and Vanuatu makes it unlikely that these systems are in steady state. The splendid seismic imaging and groundbreaking petrologic work done in Japan makes it a natural candidate for further study; the large body of work currently underway, however, indicates that a focused experiment is, in fact, already being done. Unfortunately, the presence of continental crust and widespread occurrence of fractionated and contaminated magmas means that this is not the optimal place to study the subduction factory.
Central America as a High Priority Focus Area: Central America has emerged as an optimal margin for focused study for several reasons (Figure 12). Changing subduction dynamics result in sharply varying differences in the apparent sediment transport to depth. Seismic and geochemical imaging suggest that all incoming sediments are subducted to depth beneath Nicaragua, while much of the upper hemipelagic sediments are underplated off Costa Rica, leaving a largely carbonate section to subduct to depth. The relatively large proportion of carbonate subducted here sets the stage to begin investigating the carbon cycle through a subduction zone, a unique part of the volatile cycle. Melt inclusion studies of Nicaragua volcanics have revealed among the highest water contents in any basaltic liquid on the planet (up to 6 wt% H2O).
Central American volcanoes are extremely active; several are erupting now. A modern eruption in Nicaragua, equivalent to the 2500 yr. bp Masaya eruption, will obliterate the capital, Managua, and completely disrupt the country. Most volcanoes erupt basalts free from obvious upper plate contamination.
Central America has geochemical characteristics like an island arc but has the continental advantage of easy access to all the volcanoes and on-land sites for seismic stations. Volcanoes in Nicaragua record the global maximum in recycled sediment signals, such as 10Be and Ba/La. The uplifted Cordillera de Talamanca provides exposures of the deeper crustal section, allowing investigation of the plutonic arc crust. Due to arc migration, a long-term record of arc volcanism through the Tertiary is exposed for study. Serpentinites in the Guatemala forearc may provide samples of hydrated fore-arc mantle and intermediate aqueous fluids. Changes in forcing functions along-strike allow some parameters to be investigated while others are held constant. Convergence rates increase slightly southward from Nicaragua to Costa Rica (from 70 to 90 mm/yr), while slab dip shallows from 75° to 65° at relatively constant plate age (22- 23 Ma). Dramatic along-strike variations in sediment tracers in the volcanoes attest to dramatic changes in the sediment subducted to depth, despite a relatively constant thickness of pelagic sediments (400-500 m of hemipelagic ooze and carbonate). Crustal thickness increases from Nicaragua to Costa Rica (30-40 km), along with an apparent decrease in the extent of melting in the mantle.
In short, Central America provides the opportunity to investigate all three of the major themes highlighted earlier. Forcing functions vary smoothly but lead to dramatic regional variations in the volcanic output. Carbonate subduction, and actively venting water-rich volcanics all show promise for study of the volatile cycle. Lower crustal exposures and high-fidelity tracer studies will help to pave the way to mass balance. Many of the objectives link very naturally with those of the SEIZE science plan in Central America. Further value-added for Central America comes from the excellent marine geology and seismology work underway at German institutions. Other considerations are the strong field effort already underway and opportunities for determining gas flux.
Izu-Bonin Mariana as a High Priority Area in the Western Pacific: A second margin for focused study should ideally contrast Central America in terms of forcing functions. The slab subducting beneath Central America is relatively young and the margin is towards the warmer end of the arc spectrum. Central America has few back-arc volcanoes and hence offers a weak cross-arc perspective. Parts of the Central American fore-arc are sedimented. Natural counterpoints to Central America exist in the western Pacific arcs characterized by the subduction of old, cold slabs, back-arc spreading and sediment-barren forearcs: Tonga, Izu-Bonin, and Marianas.
It was difficult to choose one of the three Western Pacific margins for focused study: Tonga, Izu-Bonin, and Marianas. This is partly because each margin offers different opportunities and limitations. For example, Tonga has the fastest convergence rate in the world and is a natural end-member for investigating convergence rate as the forcing function that drives the factory (Figure 13). Confusingly, however, volcanic activity here is apparently rather low. Tonga also has a very depleted mantle and thus the slab and mantle signatures may be distinguished easily. Seismicity is deep enough and abundant enough to allow good seismic imaging with OBS deployments.
The Marianas offer a great opportunity to investigate the volatile cycle and its consequences across the entire factory from trench to back-arc (Figure 14). Serpentinite diapirs in the fore-arc actively vent aqueous fluids from the slab and transport metamorphic rocks (blueschists) from inside the factory to the surface. Ore-forming hydrothermal fluids at the arc and back-arc have slab signatures. Chemical variation along strike in the Marianas is pronounced and reflects either variation in the subducted input or in the mantle. Low seismicity, however, means that long OBS deployments would be necessary and resolution may be rather coarse.
Seismic imaging of the Izu-Bonin margin reveals the presence of the Vp=6 layer of middle and lower arc crust, with a few submarine locations where tonalite is exposed, making this an excellent candidate for investigating initial crust formation in a juvenile intra-oceanic arc. The Izu-Bonin margin is similar to Tonga in that the mantle here is depleted, making the slab signature easy to read. Serpentine diapirs are present in the fore-arc although no active venting has been reported.
After extensive discussion, IBM was selected as the Western Pacific Focus Site. A one-day MARGINs sponsored workshop at the December 1998 AGU meeting was convened to select focus sites.
In addition to the focus sites, allied studies at selected margins and paleosubduction zones are necessary to make global comparisons to models that will emerge from the focus areas and to provide valuable further insight into subduction factory processes. In some cases these may occur after initial studies in the focus areas.
Aleutians: The Aleutians show pronounced variation along-strike in plate age, convergence rate and obliquity, sediment thickness and composition, and upper plate thickness and structure (Figure 15). In addition, this margin subducts sediment that is unusually rich in silica due to high-latitude diatom productivity and thus provides a silica-rich end-member for forcing function considerations. With high magma production rates and volcanic hazards to US citizens and planes flying in US airspace, the Aleutians are a strategic region for focused study. At this stage, however, too little is known to frame such a study. Recommend studies in the Aleutians include: ODP-type drilling of the incoming plate, swath-mapping, MCS surveys, and sampling and analysis of volcanic and plutonic products of the lesser known parts of the arc.
Cascadia: Another US margin, the Cascades, is at the hottest end of the arc spectrum. Indeed, the slab beneath Cascadia is hot enough that little seismic energy is released and seismic imaging is made difficult. As a consequence of the higher slab temperature, however, many elements apparently leave the slab at shallower depths than elsewhere, resulting in a smaller slab signature in the arc, and possibly a lesser supply of water to depth. Recent studies of intrinsic water contents of primitive lavas from the Cascades show that some are relatively water rich while others are apparently dry. While Cascadia might not be the best place to study slab inputs (the incoming sediment section is very thick and complexly partitioned in the shallow part of the margin), it is a good place to study other inputs to the factory—the mantle wedge and upper plate lithosphere. Selected studies in Cascadia would also be useful in examining the relative roles of water fluxing, decompression and mantle temperature in mantle melting.
Paleo-Subduction Zones: The chemical processing and P-T conditions of the slab between about 40 km and 100 km depth can be studied directly only in metamorphic assemblages from paleo-subduction zones. When exhumed and exposed subaerially, subduction assemblages such as the Catalina, Pelona, and Kodiak record the prograde metamorphism of the subduction zone. Petrological and chemical studies can illuminate the behavior of volatiles during metamorphism, the localized presence of melting, and the changing composition of the slab as dehydration and metamorphism distill elements out of the slab at increasing pressure and temperature. Allied studies in paleo-subduction zones will be an important part of understanding the subduction factory in the intermediate-depth interval.
An important component to the Subduction Factory Initiative is the periodic convening of theme institutes and results workshops. Such gatherings are necessary to educate, exchange ideas, and pose problems across the disciplines.
Subduction Factory Workshop participants in 1998 recognized the immediate value of convening a Theoretical and Experimental Institute to address the internal workings of the subduction factory. Many of the fundamental processes—melt generation, crustal recycling, slab-mantle interactions—occur within the most inaccessible reaches of the factory. How do forcing functions such as convergence rate, dip and slab temperature affect flow and temperature in the mantle wedge? Where does the slab dehydrate, and how does this release of fluid relate to the melting process in the mantle? Where does melting occur in the mantle wedge? These fundamental questions are still with us after 30 years of subduction studies. Quantum progress can be made only by combining seismic imaging, laboratory experiments, geodynamic modeling of solid and fluid flow, and petrological and geochemical constraints provided by input and output products. Successful institutes combining these elements have been held by RIDGE and have led to vigorous exchanges between modelers and petrologists and the recent MELT experiment. By comparison to ridges, the models for mantle flow and melting at subduction zones are crude, and the models are lagging behind the observations. A theoretical and experimental institute on the “Inside of the Subduction Factory” held in 2000 brought together geodynamicists,petrologists and seismologists to develop the models for the internal workings of the subduction factory and the ways to test them. Other institutes for future years will be proposed by the community.
A ten-year program is necessary for an integrated study of the Subduction Factory. The timeline in Table 2 follows the implementation strategy outlined in Section 3.
The first three years have focused on developing the geophysical and geological background (seafloor mapping, MCS, geodetics, dredging) to guide later large-scale efforts (drilling and seismic imaging). Other critical activities in these first years include developing databases for rapid dissemination of information, and establishing seismic stations for long-term monitoring of earthquakes to image the mantle wedge and slab. On-land mapping programs have begun, in order to provide samples for geochemical analysis, which will also help to focus future drilling and imaging programs. Modeling in the early stages of the program will help to guide data acquisition. Theoretical institutes have been and will be held in order to galvanize the diverse community and to provide models to be tested with the field experiments. In addition to the on-going modeling, geochemical analysis, and earthquake monitoring, included ODP drilling studies of submarine fluxes in the subduction factory: incoming materials and forearc output. Borehole monitoring began immediately following drilling. Also beginning in this time period are seismic refraction and magnetotelluric studies of upper plate structure, in order to guide future arc drilling. Results workshops will occur throughout the intermediate stages of the programs, to integrate results from the different disciplines and experiments.
The final observational phases of the subduction factory studies will include riser (or land) drilling in the arc, to test predictions from the refraction studies of arc structure and evolution, as well as riser drilling of holes in the fore-arc and back-arc. Modeling and laboratory experiments will be critical to interpreting results from the various observational phases.
Thus, throughout the ten-year initiative, the different major off-shore and on-land programs are staged in a natural progression of events, with early experiments paving the way for later ones. This timeline, however, is a generic one, and the details of the activities and the exact sequence of events will be dictated by the compelling proposals written. It may be that the full barrage of activities may not be necessary at the chosen focus areas.
A comprehensive study of the subduction factory will cost 15 to 20 million dollars, exclusive of ship time and drilling costs. The probability of extensive international cooperation will distribute these costs over the scientific funding agencies of a number of nations. Less ambitious programs will provide valuable information on the selected subduction zones; however, the synergism of a comprehensive study should yield a greater benefit per dollar invested than the more limited approach.
A Subduction Factory web site is maintained as part of the MARGINS Office (currently at Washington University, St. Louis). The SubFac web site will provide the following: (1) information concerning upcoming field expeditions and experiments, (2) access to databases and data recently acquired by SubFac, and (3) a news bulletin board to foster communication across the different disciplines in the SubFac community.
By making information concerning upcoming cruises and field experiments available in a timely fashion, other researchers can capitalize on these opportunities and secure funds to participate in these projects or design piggy-back projects. In addition, recently acquired SubFac data and/or pathways to access the data will be available on the web site soon after acquisition. This time frame will vary for different data types.
Existing data acquired in the SubFac natural laboratory will also be compiled, catalogued, and entered into the data base. Rapid dissemination of data and new ideas will help focus the community, which, in turn will lead to a more interdisciplinary approach toward studying the subduction factory. For example, preliminary tomographic and seismic reflection images will be available on the web site soon after they are developed; geochemical data and phase diagrams will be posted on the web site so that geodynamic models can be tested. Data availability, together with the news bulletin board, will improve communication between observationalists, experimentalists and theoreticians. This enhanced communication will allow rapid determination of the critical observations and experiments necessary for constraining models of the subduction factory. In this way, first order model predictions can be immediately tested. Such an iterative approach between modeling and data analysis is the necessary first step towards developing realistic quantitative models of the subduction factory.
In addition to the web site, international meetings and publications will promptly communicate the results of SubFac. Workshops on the main scientific themes (volatile cycling, forcing functions and crustal growth) will also be an important component to focusing efforts within the broad community. These workshops will bring together the diverse community of researchers needed to cross-fertilize ideas and develop multidisciplinary approaches to make progress on the SubFac scientific objectives.
Finally, an important component of SubFac science communication will be sharing SubFac science objectives and results with professional educators, whose primary concern is the development of curriculum, teacher training, and textbook preparation for K-12 and undergraduate students.
A number of current and planned efforts in different countries address aspects of convergent margins complementary to those of the US Subduction Factory Initiative.
For example, the Japanese subduction factory efforts build on many decades of groundbreaking research in convergent margins and will focus on subduction initiation and birth of the continents. Methods will include field work and analytical studies, superb seismic imaging, and ultimately riser drilling through OD 21. A Japanese workshop on the Subduction Factory held in the fall of 1998, and organized by Gaku Kimura, led to a science plan for subduction factory research in and around Japan. MARGINS strongly endorse international cooperation to further Japanese objectives in studies of the Subduction Factory.
German activities along convergent margins are global but include much effort in the Middle America area. The R/V Sonne has geophysically mapped and sampled from the Cocos Ridge to central Nicaragua and south to Costa Rica on multiple cruises since 1992. The TICOSECT project has just been completed and included crustal transects from the Pacific to the Caribbean oceanic plates across Costa Rica and Nicaragua, studies of volcanoes, on-land geologic mapping and the geochemistry of on- and offshore samples. A project from the GEOMAR Research Institute for a long term research study of fluid in the subduction zone and cycling of subducted chemical components (e.g., methane) is now well established.
A French Margin Initiative—the Group De Recherche MARGES or GDR-MARGES, http://gdrmarges.lgs.jussieu.fr —has been established. Their convergent margin program builds on an extensive body of work studying great earthquakes and the seismogenic zone (Eastern Nankai, Northern Andes), aqueous fluids in accretionary wedges (Mediterranean Ridge, Barbados), accretion versus tectonic erosion (Peru, Chile, Hikurangi margin), oblique subduction, collision, and incipient subduction (South Philippines, South Ryuku and Taiwan, New-Hebrides, Puysegur); and back-arc opening (North Fiji Basin, Manus Basin, Okinawa Trough).
New Zealand efforts focus on a North Island transect, which is a natural follow on from the highly successful South Island Geophysical Transect (SIGHT). Funding from the New Zealand Science foundation has been approved and assigned to groups at the Institute of Geological and Nuclear Sciences (IGNS) and the School of Earth Sciences, Victoria University of Wellington. International collaborations are underway. Some of the principal issues to be explored include the origin of andesites, the link between Quaternary uplift and upper mantle processes, and the initiation of subduction.
A critical aspect of international collaboration will be interaction with scientists from those nations hosting the target margins, both onshore and offshore. This is particularly important in Central America, where volcanological and seismological teams in Costa Rica and Nicaragua have made great progress in research, cooperation and infrastructure: 1) planning workshops held in Costa Rica have facilitated close scientific and logistical ties, and; 2) collaborative work with Japanese scientists is well underway in the IBM system.
This range of studies provide cooperative ways to maximize scientific return. They also emphasize the need for continuing rapid development of the international aspects of the MARGINS.
We realized that the Subduction Factory will not be block funded but progress through a series of peer reviewed proposals/grants. This funding mechanism provides continual evaluation that functions every time a proposal is submitted and reviewed. In addition, the MARGINS Office will issue progress reports, and the SubFac initiative will be periodically reviewed. Such reviews should not only demonstrate the scientific progress that is resulting from individual proposals but should include evaluations of the effectiveness of interdisciplinary and international activities. After 5 and 10 years MARGINS should hold a progress meeting/workshop to evaluate the initiative and to refine/redefine directions. NSF program managers and our peers will also evaluate the initiative.