Differentiating induced versus spontaneous subduction initiation using thermomechanical models and metamorphic soles

Hot versus cold subduction initiation

Initiation of a new subduction zone could act in two different ways, forming either a hot or cold incipient subduction channel with contrasting geological records.

Understanding subduction infancy to mature subduction in Southwest Japan via the self-consistent formation of a weak slab interface

The weak slab interface controls long-term subduction dynamics. A weak hydrous layer at the slab interface promotes mechanical decoupling between the forearc mantle and the subducting slab and converts a hot forearc mantle to a cold mantle. Often referred to as a cold nose, the cold forearc mantle, plays a key role in the transition from subduction infancy to mature subduction. This study was the first to numerically demonstrate the self-consistent formation of a weak hydrous layer with permeability anisotropy based on the Southwest Japan subduction zone case, where transition-related geological features were present. Our models showed that mechanical decoupling by spontaneous downdip growth of the weak hydrous layer created a cold nose by converting a hot forearc mantle to a cold mantle. The emergence of a cold nose explained the migration of the forearc-to-arc volcanic front, expressed as the formation of mid-Miocene forearc high-magnesium andesite and Quaternary arc adakite. Furthermore, the weak hydrous layer providing a pathway for free-water transport toward the mantle wedge tip elucidates slab/mantle-derived geochemical components in deep groundwater as well as large S-wave delay times and non-volcanic seismic tremors in the forearc.

Boron isotopes in boninites document rapid changes in slab inputs during subduction initiation

How subduction-related magmatism starts at convergent plate margins is still poorly understood. Here we show that boron isotope variations in early-formed boninites from the Izu-Bonin arc, combined with radiogenic isotopes and elemental ratios document rapid (~0.5 to 1 Myr) changes in the sources and makeup of slab inputs as subduction begins. Heterogeneous hornblende-granulite facies melts from ocean crust gabbros ± basalts fluxed early melting to generate low silica boninites. Hydrous fluids from slab sediments and basalts later fluxed the low silica boninites mantle source to produce high silica boninites. Our results suggest that initially the uppermost parts of the slab were accreted near the nascent trench, perhaps related to early low-angle subduction. The rapid changes in slab inputs recorded in the boninites entail a steepening subduction angle and cooling of the plate interface, allowing for subduction of slab sediment and basalt, and generating hydrous fluids at lower slab temperatures.

The geochemical record of subduction initiation is still not well understood, despite >50 years of study. Here, the authors use boron isotopes in Izu-Bonin boninites to document rapid changes in slab inputs to melting at the start of subduction, related to the steepening and cooling of the downgoing Pacific plate.