This is a combined active and passive seismic experiment along the Eastern Lau Spreading Center to test the following hypotheses. 1. Circulation in the mantle wedge is dominated by slab driven flow. 2. Interaction of the arc and backarc magma production controls the character of the ridge by influencing melt flux, petrology, and geochemistry. 3. Variations in the mantle melt supply control ridge crest features such as morphology, thermal structure, and hydrothermal venting. The passive experiment consists of 55 broadband ocean bottom seismographs and five land seismographs deployed for 10 months to image the larger-scale structure of the melt production region and the mantle flow pattern. The active source experiment consists of 100 ocean bottom seismographs deployed along a 250 km section of the spreading center extending from the inflated Vala Fa region to the magma-starved northern Eastern Lau Spreading Center where the axial melt lens is absent.
In broader terms this experiment addresses a first order problem in ridge dynamics in a unique subduction zone setting and is essential for the stated goals of R2K program. Graduate students will be supported by all four PIs.
The project was focused on improving our understanding of mantle melting, mantle flow, crustal formation, and the effect of subducted water at the Eastern Lau Spreading Center (ELSC), a back-arc spreading center near the Tonga arc and trench (figure 1). The Lau backarc provides an ideal location for such a study because different spreading center segments are located at different distances from the subducting slab, and thus have very different amounts of subducted water in the melt source region. Through techniques of passive seismic imaging and geodynamic modeling we have obtained a much more detailed picture of how melt is produced and transported to the surface, and how water affects this process. The rapid transition between a "low water" and "high water" geochemistry midway along the ELSC has been difficult to understand because the distance from the slab, and thus the source of water, varies slowly. Seismic images obtained in this project show that the pattern of mantle velocity anomalies changes suddenly at this location, with the northern "low water" spreading center fed by upwelling mantle from the west, away from the water source, and the southern "high water" spreading center sampling only wet mantle near the subducting slab (Wei et al, Nature, accepted) (figure 2). Melt extraction from the mantle is far more efficient at the water-influenced portions of the ridge, as demonstrated by the reduced seismic anomalies suggesting lower melt contents to the south. We interpret this as resulting from lower melt viscosity or larger matrix grain size due to the effect of water (Wei et al.), causing the melt to move easily and quickly through the matrix to the surface, with less melt present at any particular time. Another outstanding puzzle has been why the northern ELSC (non-subduction influenced) has thinner crust than the Central Lau spreading center (also non-subduction influenced) despite being closer to the arc. The geodynamic portion of this work shows the importance of the 3D nature of the system in controlling the crustal production. Because the southern subduction-influenced portion of the ridge has a lower viscosity mantle, fertile, upwelling mantle from the northern ELSC is pulled towards the south, adding to the melt production (and crustal thickness) there, while decreasing it in the north (Tarlow and Conder, AGU fall meeting, 2014) (figure 3).
