A matter of time: The importance of the duration of UHT metamorphism

The duration of granulite (G) and ultra-high temperature (UHT) conditions in regional metamorphism is critical to arguments regarding the tectonic settings of granulites and their relationships to Supercontinent assembly. Analysis of zircon geochronology integrated with trace element (REE) constraints on the timing of zircon growth or modification and evidence for metamorphic temperatures from Ti-in-zircon reveals that zircon can form episodically at or near the metamorphic peak in long-lived UHT granulites. This reflects the segregation, transfer and stagnation or trapping of melts which leads to local melt-rock interactions that promotes zircon crystallisation. In contrast, although rutile can retain UHT information in short duration ('fast') G-UHT terrains it is afflicted by Zr loss in long duration ('slow') terrains and yields temperatures significantly lower than those preserved by zircons formed in the same G-UHT events. An assessment of the age-duration evidence from several well-documented G-UHT terrains reveals that most are 'slow', having durations of UHT, Delta t(900), in the range 30-100 Myr. Some UHT terrains previously considered to be short-lived (Delta t(900) < 10 Myr) have longer durations of UHT in the light of recent geochronology and hence are also classed as 'slow'. Of the models proposed to account for UHT metamorphism, the 'large hot orogen' (LHO) model for collisional orogeny provides the best setting for the formation of such 'slow' UHT granulites. LHO models can account not only for the P-T paths, which can range from UHT with near-isothermal decompression (UHT-ITD) through to decompression-cooling and UHT followed by near-isobaric cooling (UHT-IBC), but also for residence times under UHT conditions. The Napier Complex, the Earth's premier UHT terrain, probably formed as trapped deep crust in the hot underbelly of a late-Archaean LHO. Shorter duration UHT and near-UHT granulites also exist, mostly with Delta t(900) less than 10 Myr. A number of these are likely to have formed as a consequence of severe lithospheric thinning and crustal extension accompanied by voluminous magmatism, which could occur in arc and back-arc settings affected by subduction roll-back or, as advocated by previous workers, continental arcs undergoing extension. However, attaining long-lived UHT conditions in these settings is unlikely unless the crust has inherently high radioactive heat production in addition to the transient heat added during extension and magmatism.