On the decompression melting structure at volcanic arcs and back‐arc spreading centers

Abstract

Mantle dynamics can strongly affect melting processes beneath spreading centers and volcanic arcs. A 2‐D numerical model of the Tonga subduction zone, with the slab viscously coupled to the mantle beneath the brittle‐ductile transition but faulted above, shows that induced corner flow may cause asymmetric melting at the Lau back‐arc spreading center, 400 km away. The down‐going slab also entrains the high‐viscosity base of the overlying lithosphere, drawing hot, low‐viscosity asthenosphere upwards into the gap, triggering decompression melting in the wedge. Because the slab is decoupled from the brittle overlying plate, a cold upper corner develops, inhibiting melting where the slab is shallow. The cold corner is consistent with seismic attenuation and heat flow at arcs. Decompression melting may be a substantial fraction of magma production at some arcs, but less at others. Possibly more important, the shallow decompression melting structure may govern the pathways of melt extraction beneath volcanic arcs.

Thermal structure due to solid-state flow in the mantle wedge beneath arcs

“…The recent models of van Keken et al [2002] (see their Figure 3) and Conder et al [2002] also show this thinning, although in their models the amount of thinning appears to be limited by the presence of a rigid "lithospheric" layer of prescribed thickness. [Furukawa, 1993a;Furukawa, 1993b] also prescribed a fixed lithospheric thickness.…”

“…These models differ in many respects, but most agree that subduction of oceanic crust that is more than 20 million years old at down-dip rates greater than 20 km/Myr will not produce temperatures at the top of the subducting plate that are high enough to allow fluid-saturated melting of sediment or basalt. The sole exceptions are recent models by Conder et al [2002] which, like ours, incorporate only temperature-dependent viscosity, and a model by van Keken et al which incorporates both temperature-and stress-dependent viscosity.…”

Shear wave splitting from local events beneath the Ryukyu arc: Trench-parallel anisotropy in the mantle wedge

“…It has been demonstrated that both lithospheric and asthenospheric contributions to anisotropy are important in many regions (e.g., Fouch et al, 2000;Simons et al, 2002;Simons and van der Hilst, 2003;Fischer et al, 2005;Waite et al, 2005), although splitting measurements in subduction zone settings are nearly always interpreted in terms of flow in the asthenosphere (e.g., Fischer et al, 1998Fischer et al, , 2000Smith et al, 2001;Anderson et al, 2004). The lithosphere beneath the Ryukyu arc stations is likely to be thin, due to thermal erosion associated with mantle wedge flow (e.g., Conder et al, 2002), and it is unlikely that the lithosphere is thick enough to explain the large split times. A second line of argument against primarily lithospheric anisotropy comes from the data itself.…”

“…Because we observe trench-parallel fast directions at stations located approximately 100 km from the trench in the interior of the arc (most noticeably station KGM; see Table 1), a mechanism such as flow around the edges of the slab is probably insufficient. Conder et al (2002) proposed a model for flow associated with decompression melting in the mantle wedge and suggested that in some arc systems, corner flow may act in concert with trench-parallel flow. It has been suggested (James Conder, Washington University, personal communication) that elevated temperatures in the interior of the mantle wedge lead to a dramatic reduction in viscosity that may form hot, low-viscosity channels whose flow regime is decoupled from the rest of the wedge and the downgoing slab.…”

A community benchmark for subduction zone modeling

“…This focusing increases the temperature at the top of the slab, results in narrow thermal and rheological boundary layers, and enhances the effects of decompression melting and importance of buoyancy (or secondary convection) in the mantle wedge (Furukawa, 1993;Billen and Gurnis, 2001;Van Keken et al, 2002;Conder et al, 2002;Kelemen et al, 2003;Gerya and Yuen, 2003;Cagnioncle et al, 2007;Morgan et al, 2007). Since the targeted accuracy for use of these models in petrological studies is on the order of tens of degrees there is a strong need for methods that can solve the governing equations accurately.…”