Intellectual Merit. There is strong evidence from experimental petrology and olivine-hosted melt inclusions that H2O is important in the formation of supra-subduction zone mafic magmas. However, despite the increasing sophistication of geochemical and geodynamic models based in part upon these data, there is still a major gap in our knowledge of the P-T-XH2O conditions of magma generation in mantle wedges above subducting slabs. This project will test some fundamental ideas about arc magma generation by addressing the following questions: 1. Do primitive magmas from the TMVB saturate with a lherzolite, harzburgite or pyroxenite assemblage at mantle pressures under hydrous conditions? 2. Are there differences in the pressures at which H2O-rich magmas (probably formed by flux melting) and H2O-poor magmas (probably formed by decompression melting) last equilibrated with a mantle assemblage? Such differences could relate to transport and storage of subduction derived components within the wedge (e.g. do H2O-rich melts derive from closer to the slab). 3. Are mantle equilibration temperatures and pressures for the H2O-poor primitive magmas consistent with advection and upwelling of hotter mantle from behind the arc? 4. How do equilibration temperatures and pressures compare with predictions for the thermal structure of the mantle wedge based on 2D and 3D geodynamic models? This project will provide important constraints for evaluating the relative roles of fluid-flux and decompression melting in arcs and for testing geodynamic models of subduction systems by providing data on temperatures and H2O contents at various depths within the mantle wedge. Liquidus phase relations will be determined for five primitive melt compositions from the Trans-Mexican Volcanic Belt (TMVB) as a function of H2O content at mantle wedge pressures. Starting compositions span much of the global range of K2O and H2O for primitive arc magmas. Compositional effects of deep crustal fractional crystallization such parental basalts will be evaluated to test whether more evolved compositions can be "back-corrected" to primary melt compositions.
Broader Impacts. This project will integrate research and education through the involvement of a Ph.D. student at the University of Oregon. The student will participate in all aspects of the experimental research and will learn a spectrum of modern analytical and imaging techniques (FTIR, electron probe, SEM). An undergraduate student will also be involved in a senior thesis project. The interpretation of results will involve collaboration with a young scientist in Mexico, Dr. Vlad Manea (UNAM, Juriquilla, Mexico and Caltech), who is working on geodynamic modeling of subduction zones. John Donovan, a highly regarded electron beam instrument operator, will be involved in developing robust methods for analyzing hydrous run products.
Subduction zones are places where oceanic crust gets recycled back into the deep parts of the Earth. They are typically associated with a deep sea trench and a chain of volcanoes that lie on a separate tectonic plate that is positioned above the sinking oceanic plate. The Cascade volcanic chain in the Pacific Northwest region of the United States is an example of a subduction zone. The goal of this project was to better understand how magma is produced above the sinking plate, and to investigate this, we did laboratory experiments at high temperatures and pressures to simulate the melting conditions in the Earth. The rock samples that we used for this study mostly came from volcanoes in central Mexico, though we also used one sample from the Aleutian Islands. Both of these regions are also subduction zones. A key focus of our study was to investigate the role of water in causing rocks deep inside the earth to melt. When subducted oceanic crust heats up, it releases water that is locked inside minerals and this water escapes upwards from the crust. Water has a profound effect in lowering the melting temperature of rocks, so the rising water comes into contact with hotter rocks above the subducted oceanic crust in a region known as the mantle and causes the rock to melt. We knew from independent evidence how much dissolved water was present in the magmas that formed the rocks used in our studies, so we used this information in our experiments so we could simulate the exact melting conditions in the Earth that produced different types of magma. Although our study focused on how magma forms inside the Earth, the results also help us understand why magmas in subduction zone volcanoes are water rich, which has the effect of making these volcanoes erupt explosively, creating a hazard to human communities. There were three main projects that were supported by this grant. The work on the projects was done by two graduate students at the University of Oregon. One received a Ph.D. for her work and the second received a Masters Degree. In the first project, we compared melting beneath the Aleutians, an example of an island arc, where one oceanic plate subducts beneath another, and central Mexico, an example of a continental arc, where an oceanic plate subducts beneath a continental plate. We found that in both arcs the magmas formed at about 45 km depth beneath the Earth’s surface. The mantle rocks beneath Mexico, however, appear to be at a lower temperature (~1200 °C) compared to the Aleutians (~1300 °C). Interestingly, our experiments showed that the mantle rocks beneath Mexico have a different composition than beneath the Aleutians. The difference probably arises because the rocks beneath Mexico (which is a long-lived subduction zone) have been melted before, whereas the rocks beneath the Aleutians have not been melted before and are thus more ‘fertile’. Our results give us a basis for looking at the chemical composition of volcanic rocks in arcs around the world and using them to infer the complicated history of the mantle beneath the arc that melts to form magma. In the second project, we focused on volcanic rocks from the Mexico City region because the continental crust in this part of the arc is very thick compared to just about every other arc worldwide, and the volcanic rocks have an unusual composition (high-magnesium andesite rather than basalt). Our results again showed that melting had occurred about 45-50 km depth, just beneath the base of the continental crust. In this case we found that the unusual composition was caused by even more previous melting of the mantle than we had found in the first part of the project summarized above. Our work showed that melting of this extremely refractory rock in the mantle produces a type of magma that crystallizes unusually magnesium-rich olivine. So this provides another clue that can be used in looking at rocks from other arcs around the world to infer parts of the mantle that have been melted repeatedly and thus are very refractory. In the third part of the project, we did experiments on some of the more uncommon volcanic rock types in Mexico. The reason for choosing these is that they seem to represent cases where more water from the subducted oceanic plate may have been involved in melting than is normally the case in subduction zone magmas. The results, when combined with the other experiments mentioned above, showed that there is a range of different mantle rock types beneath the arc. Some of these rocks have received a lot of fluid from the subducted plate, resulting in veins that are chemically enriched in many of the elements that are distilled from the plate during subduction.
