|A map of seafloor MT/CSEM instruments deployed during the SERPENT experiment and the paths of the CSEM transmitter completed. ()|
SERPENT: Serpentinite, Extension, and Regional Porosity Experiment across the Nicaraguan Trench
Dan Lizarralde, WHOI Kerry Key, Steve Constable, Scripps Institution of Oceanography http://marineemlab.ucsd.edu/Projects/SERPENT
Water plays an important role in many of the processes occurring at convergent margins, as the release of water from the downgoing slab affects the rheology of the mantle, impacts seismicity, allows melting to occur more readily by lowering the solidus temperature, and alters the chemistry of arc-lavas. Yet, the amount of water entering the subduction system remains poorly constrained. One of the major uncertainties in terms of fluid inputs into the subduction factory, concerns the extent of serpentinization of the oceanic upper mantle and the volumes of water that are being carried into the subduction system through this route.
In 2010 we completed a large marine CSEM and magnetotelluric deployment off the coast of Nicaragua. Our project is the largest combined controlled-source electromagnetic (CSEM) and magnetotelluric (MT) data set ever collected on an active subduction zone. During the single 28 day research cruise aboard the R/V Melville we collected 54 stations of broadband marine magnetotelluric (MT) data and deep-towed nearly 800 km of controlled-source electromagnetic (CSEM) data. Robust multiple-station array processing of the MT data yields high quality MT responses from 10 to 20,000 s period. Two circular CSEM tows of 30 km radius were measured by special long-wire EM (LEM) sensors on the abyssal plain and the outer rise. Conventional CSEM data recorded at a broad suite of transmission frequencies along the 300 km long profile and a 50 km along strike profile provide constraints on crustal conductivity variations.
The CSEM data indicate that crustal conductivity increases by a factor of 2-3 with the onset of the trench normal faults, providing the first evidence that seawater penetrates through the outer rise faults in the subduction zone.
CSEM data from the continental slope indicate an anomalously high crustal conductivity beneath a band of observed seafloor seeps, indicating that deep faulting on the margin slope allows for fluids generated from the dewatering of subducted sediments to rise up through the margin slope. However, the CSEM data is only sensitive to the upper 6 km, and hence doesn't no image the deeper dewatering region.
Anisotropic lithosphere conductivity obtained by circular tows of long-offset CSEM data show a slightly anisotropic fabric beneath the abyssal plain that becomes highly anisotropic beneath the outer-rise faults, likely due to a combination of parallel conductive faults in the crust and a conductive serpentinized uppermost mantle.
The magnetotelluric data show a predominantly 1D conductivity beneath abyssal plain that becomes progressively 2D near the trench. Sites located at the base of the margin slope are highly 3D, perhaps indicative of a localized electric current leakage path along the plate boundary. A discrete high conductivity anomaly occurs on the margin slope at 10-20km depth, right beneath the seafloor seeps, and exactly where dewatering of the subducted sediments is expected. This correlates with the high conductivity seen in the CSEM data, but the MT data have imaged the deeper dewatering fluid source region at the plate interface.
The analysis of these data is ongoing and will provide a comprehensive picture of the electrical conductivity structure from the seafloor to the upper mantle, representing the entire input into this part of the Central American subduction system. Since conductivity is highly dependent on thermal structure, crack porosity and the presence of serpentinite, our experiment will provide constraints on the depth of active fluid circulation within the oceanic crust and mantle, the variation of fluid circulation with distance from the trench and hence with the degree of plate bending and faulting, and the extent of deawatering of the slab in the shallow portions of the subduction system.