Thermal structure of the mantle wedge beneath northeastern Japan: Magmatism in an island arc from the combined data of seismic anelasticity and velocity and heat flow

Abstract

The thermal structure of the mantle wedge beneath northeastern Japan is investigated from seismic anelasticity ( Q −1). Laboratory measurements in dry peridotite at high pressures show a strong temperature-dependence of anelasticity; this may be used to determine the detailed thermal structure of the subduction zone. Since, however, water or hydrous minerals may exist in the mantle wedge, the laboratory Q results for dry peridotite may not be readily applicable to the region. We here compare temperatures derived from the dry- peridotite Q with those from heat flow. The comparison yields a correction factor that is used to correct for the unknown water effect on temperatures derived from the seismic Q. The results show that no temperature corrections are necessary in the back-arc side of the volcanoes, whereas corrections of as much as 250–660°C are necessary in the fore-arc side. We then determine the corrected temperatures in the mantle wedge, and study the role ofthe wedge and the descending slab in island arc magmatism. In the uppermost mantle wedge, the temperature decreases from ∼ 1000°C beneath the Japan Sea coast to ∼ 200–300°C in the apex of the mantle wedge, in good agreement with the location of the volcanoes and the cold slab. Beneath the volcanic front and to the Japan Sea, temperatures lie between the dry and wet solidus. Thus, partial melting should occur only if water is present in the mantle wedge. The seismic structure and the temperature profile indicate that melting as well as volatiles may exist only locally (in regions less than 25 km in width), possibly forming a plume (or a diapir) beneath volcano. Therefore, an uprising of hot material from deep mantle appears to produce magmatism and high heat flow in the back-arc region. The low velocity and low Q in this region is generally not due to the occurrence of localized melt, but due to the grain boundary weakening of mantle rocks by high subsolidus temperatures. On the other hand, the derived temperatures are well below the wet solidus, and no melting should occur even under water-saturated conditions in the fore-arc region. The temperature profile is thus consistent with the site of the volcanic front. Some low-velocity regions observed in the fore-arc mantle wedge are due to the presence of hydrous minerals or free-water. Seismic structure also shows some low velocity zones in the mantle wedge adjacent to (or near) the slab. This could be again produced by volatiles derived from the subducted oceanic crust. The high velocity and high Q of the slab indicate a fairly low temperature of the region. Therefore, we infer that the slab itself will not melt, but instead provides water to the overlying mantle wedge for generating island arc magmas.