The Caribbean‐South America subduction zone is a flat subduction zone, with Laramide‐style thick‐skinned uplifts occurring in the Merida Andes, Sierra de Perija Range, and Santa Marta Massif. Geodetic measurements and historical seismicity show this region is storing strain energy and is capable of a mega‐thrust earthquake (
Phase A (Mg7Si2O8(OH)6) is one of the important dense hydrous magnesium silicates in subducting slab, because it forms after the breakdown of antigorite serpentine and could be the dominant hydrous phase in the upper‐mantle deep slab for the water transportation into deep Earth. In this study, the compressional (
- PAR ID:
- 10444586
- Publisher / Repository:
- DOI PREFIX: 10.1029
- Date Published:
- Journal Name:
- Geophysical Research Letters
- Volume:
- 48
- Issue:
- 19
- ISSN:
- 0094-8276
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
Abstract M ≥ 8.0). Previous seismic investigations of the lithosphere and upper mantle in this area are either very large scale, very local, or only peripheral to this area; therefore, details of the Caribbean plate subduction geometry beneath the Maracaibo block remain unclear. In this study, we used a new data set acquired by the Caribbean‐Merida Andes seismic experiment (CARMA), which comprised 65 temporary broadband stations and 44 permanent stations from the Colombian and Venezuelan national seismic networks. We jointly inverted ambient noise Rayleigh waveZ /H ratios, phase velocities in the 8–30 s band and ballistic Rayleigh wave phase velocities in 30–80 s band to construct a 3‐D S‐wave velocity model in the area between 75°–65°W and 5°–12°N. The 3‐D model reveals a general increase in crust thickness from the trench to the southeast. An anomalous area is the Lake Maracaibo, which is underlaid by the thinnest crystalline crust in the region. This observation may indicate that the Maracaibo block is experiencing a contortion deformation within the crust. We also identified a high velocity anomaly above the subducting Caribbean slab, likely representing a detached piece of eclogitized Caribbean large igneous province from the base of the Maracaibo block. Additionally, ourV s model clearly indicates a slab tear within the subducted Caribbean slab, approximately beneath the Oca‐Ancon Fault. -
Abstract Seismic anisotropy measurements show that upper mantle hydration at the Middle America Trench (MAT) is limited to serpentinization and/or water in fault zones, rather than distributed uniformly. Subduction of hydrated oceanic lithosphere recycles water back into the deep mantle, drives arc volcanism, and affects seismicity at subduction zones. Constraining the extent of upper mantle hydration is an important part of understanding many fundamental processes on Earth. Substantially reduced seismic velocities in tomography suggest that outer rise plate‐bending faults provide a pathway for seawater to rehydrate the slab mantle just prior to subduction. Estimates of outer‐rise hydration based on tomograms vary significantly, with some large enough to imply that, globally, subduction has consumed more than two oceans worth of water during the Phanerozoic. We found that, while the mean upper mantle wavespeed is reduced at the MAT outer rise, the amplitude and orientation of inherited anisotropy are preserved at depths >1 km below the Moho. At shallower depths, relict anisotropy is replaced by slowing in the fault‐normal direction. These observations are incompatible with pervasive hydration but consistent with models of wave propagation through serpentinized fault zones that thin to <100‐m in width at depths >1 km below Moho. Confining hydration to fault zones reduces water storage estimates for the MAT upper mantle from ∼3.5 wt% to <0.9 wt% H20. Since the intermediate thermal structure in the ∼24 Myr‐old MAT slab favors serpentinization, limited hydration suggests that fault mechanics are the limiting factor, not temperatures. Subducting mantle may be similarly dry globally.
-
Hydration of the subduction zone forearc mantle wedge influences the downdip distribution of seismicity, the availability of fluids for arc magmatism, and Earth's long term water cycle. Reconstructions of present‐day subduction zone thermal structures using time‐invariant geodynamic models indicate relatively minor hydration, in contrast to many geophysical and geologic observations. We pair a dynamic, time‐evolving thermal model of subduction with phase equilibria modeling to investigate how variations in slab and forearc temperatures from subduction infancy through to maturity contribute to mantle wedge hydration. We find that thermal state during the intermediate period of subduction, as the slab freely descends through the upper mantle, promotes extensive forearc wedge hydration. In contrast, during early subduction the forearc is too hot to stabilize hydrous minerals in the mantle wedge, while during mature subduction, slab dehydration dominantly occurs beyond forearc depths. In our models, maximum wedge hydration during the intermediate phase is 60%–70% and falls to 20%–40% as quasi‐steady state conditions are approached during maturity. Comparison to global forearc H2O capacities reveals that consideration of thermal evolution leads to an order of magnitude increase in estimates for current extents of wedge hydration and provides better agreement with geophysical observations. This suggests that hydration of the forearc mantle wedge represents a potential vast reservoir of H2O, on the order of 3.4–5.9 × 1021 g globally. These results provide novel insights into the subduction zone water cycle, new constraints on the mantle wedge as a fluid reservoir and are useful to better understand geologic processes at plate margins.more » « less
-
Abstract In the present study, we use broadband seismic data recorded by 190 stations of the EarthScope program's Transportable Array to construct a 3‐D shear wave velocity model for the upper 180 km using a non‐linear Bayesian Monte‐Carlo joint inversion of receiver functions (RFs) and Rayleigh wave dispersion curves. Ambient noise and teleseismic data are used for obtaining Rayleigh wave phase velocity dispersion curves. A resonance removal filtering technique is applied to the RFs contaminated by reverberations from the thick sedimentary layers that cover most of the region. Our observations of the higher crustal shear velocities (∼3.40 km/s) beneath the Sabine Block (SB), along with the estimated relatively thicker crust (∼34.0 km) and lower crustal
V p/V sestimates (∼1.80) in comparison with the rest of the Gulf Coastal Plain (GCP) (∼3.10 km/s for crustal shear velocities, ∼29.0 km for crustal thickness, and ∼1.90 for crustalV p/V sestimates), indicating that this crustal block has different crustal properties from the surrounding coastal plain regions. The southern Ouachita Mountains have a thin crust (∼30.0 km) and low mean crustalV p/V svalue (∼1.73), suggesting that lower crustal delamination has occurred in this region. Low velocities in the upper mantle beneath most of the GCP are interpreted as a combined result of thin lithosphere, higher‐than‐normal temperatures, and possibly compositional variations. -
Abstract The hydration of tricalcium silicate (C3S)—the major phase in cement—is effectively arrested when the activity of water (
a H) decreases below the critical value of 0.70. While it is implicitly understood that the reduction ina Hsuppresses the hydration of tricalcium aluminate (C3A: the most reactive phase in cement), the dependence of kinetics of C3A hydration ona Hand the criticala Hat which hydration of C3A is arrested are not known. This study employs isothermal microcalorimetry and complementary material characterization techniques to elucidate the influence ofa Hon the hydration of C3A in [C3A + calcium sulfate (C$) + water] pastes. Reductions in water activity are achieved by partially replacing the water in the pastes with isopropanol. The results show that with decreasinga H, the kinetics of all reactions associated with C3A (eg, with C$, resulting in ettringite formation; and with ettringite, resulting in monosulfoaluminate formation) are proportionately suppressed. Whena H ≤0.45, the hydration of C3A and the precipitation of all resultant hydrates are arrested; even in liquid saturated systems. In addition to—and separate from—the experiments, a thermodynamic analysis also indicates that the hydration of C3A does not commence or advance whena H ≤0.45. On the basis of this criticala H, the solubility product of C3A (K C3A) was estimated as 10−20.65. The outcomes of this work articulate the dependency of C3A hydration and its kinetics on water activity, and establish—for the first time—significant thermodynamic parameters (ie, criticala HandK C3A) that are prerequisites for numerical modeling of C3A hydration.