Abstract Orogens and their posthumous traces are the basic elements that can be used to understand the material circulation within the Earth. Although information preserved in the rocks on the surface ranging in age from 4.4 Ga to the present has been used to characterize orogens, it is important to understand orogens on a whole-Earth scale to evaluate global material circulation through time. In this paper, we synthesize the general concepts and characteristics of orogens and orogenic belts. The collision type and accretionary type constitute the two end-member types of orogens, both sharing similar structural features of subhorizontal disposition, bounded above and below by paired faults. Their exhumation generally occurs in two steps: first by wedge extrusion to form a sandwich structure with subhorizontal boundaries, which is followed by domal uplift of all the units. In the accretionary type, oceanic lithosphere subducts under the continental margin, and in the collision type, buoyant continents collide with each other. Of the various types of subduction and collision processes, arc–arc collision orogeny may have been widespread in the Archaean, although most of the intra-oceanic arc crust must have been destroyed and dragged down to the Archaean core–mantle boundary (CMB). Here we propose a broad two-fold classification of orogens and their subducted remnants, based on (1) their thermal history and (2) temporal constraints. Based on their thermal history, orogens are grouped into three types: cold orogens, hot orogens and ultra-hot orogens. Two extreme situations, which are anomalous and unlikely to occur on Earth, termed here super-cold and super-hot orogens, are also proposed. We discuss the characteristics of each of these subtypes. Based on temporal constraints, we group orogens into Modern and Ancient, where in both cases regional metamorphic belts occupy the orogenic core. In both groups, the overlying and underlying units of the regional metamorphic belts are weakly metamorphosed or unmetamorphosed, and are either accretionary complex in origin (Pacific type) or continental basement and cover (collision type). Major structures are subhorizontal with oceanward vergence of deformation, for both types. Orogens in the Modern Earth are grouped into four sub-categories: (1) deeply subducted orogens that are taken down to mantle depths and never return to the surface, termed here ‘ghost orogens’; (2) those that are subducted to deep crustal levels, undergo melting and are recycled back to the surface, forming resurrected and temporarily ‘arrested orogens’; (3) ‘extant orogens’, which are partly returned to the surface after deep subduction; (4) ‘concealed orogens’, which have been deeply subducted and only the traces of which are represented on the surface by mantle xenoliths carried by younger magmas. The preservation of orogens on the surface of the Earth occurred through an unusual return process from their natural course of total destruction, a phenomenon that operated more efficiently in the Phanerozoic through exhumation from ultra-deep domains against the slab-pull force of the plate, aided by fluids derived by dehydration of subducted lithosphere. Orogens at present represented on the surface of the Earth constitute only a fraction of the total volume formed in Earth history. Traces of the deeply subducted ‘lost orogens’ are sometimes returned to the surface in the form of melt or mantle xenoliths through combined processes of plume and plate tectonics. From a synthesis of the processes associated with the various categories of orogens proposed in this study, we trace the time-dependent transformations of orogens in relation to the history of the evolving Earth.

The Haiyangsuo area of the NE Sulu ultrahigh-pressure terrane, eastern China, consists of gneisses with minor granulite and amphibolite layers, metagabbros, and granitic dikes. The peak-stage assemblages of the granulites (garnet + orthopyroxene + clinopyroxene + plagioclase ± pargasite ± biotite ± quartz) formed at >750 °C and 9–11 kbar and were overprinted by amphibolite-facies phases characterized by well-developed corona layers of | garnet | amphibolite + quartz | at contacts between plagioclase and clinopyroxene or orthopyroxene, as well as by the exsolution of (orthopyroxene + ilmenite + amphibole) from clinopyroxene. These textures indicate a near-isobaric cooling history of the granulite-bearing gneiss terrane. The metagabbro preserves a relict igneous assemblage (orthopyroxene + clinopyroxene + plagioclase + pargasite ± ilmenite ± quartz) in its core, but in its margins has a metamorphic corona texture similar to the granulite that formed at ∼600–700 °C and 7–10 kbar. Sensitive high-resolution ion microprobe (SHRIMP) U-Pb dating of zircons indicates that the protolith age of the garnetbiotite gneiss is older than 2500 Ma, whereas the granulite-facies metamorphism (the first regional metamorphic event) occurred at 1846 ± 26 Ma. Gabbro intrusion took place at 1734 ± 5 Ma, and the formation of amphibolite assemblages in both metagabbro and granulite occurred at ca. 340–370 Ma. Both gneiss and metagabbro were intruded by granitic dikes, with one dated at 158 ± 3 Ma. These data, together with a lack of eclogitic assemblages, suggest that this granulite-amphibolite–facies complex is exotic relative to the Triassic Sulu high-pressure–ultrahigh-pressure terrane; juxtaposition took place in Jurassic time.

The blueschist-greenschist facies transition for a model basaltic system Na 2 O-CaO-MgO-Al2O 3 -SiO2-H 2 O is defined by a univariant reaction: 6 clinozoisite + 25 glaucophane + 7 quartz + 14 H 2 O = 6 tremolite + 9 chlorite + 50 albite; for the Fe 2 O 3 -saturated basaltic system, by a discontinuous one: 4 epidote + 5 Mg-riebeckite + chlorite + 7 quartz = 7 hematite (magnetite) + 4 tremolite + 10 albite + 7 H 2 O. These two reactions were experimentally investigated to determine the nature of the blueschist-greenschist transition. The results have located the first reaction at 350 ± 10°C, 7.8 ± 0.2 Kb and 450 ± 10°C, 8.2 ± 0.4 Kb. Reconnaissance experiments for the second reaction indicate that the minimum pressure for the occurrence of epidote + Mg-riebeckite + chlorite + quartz is about 4 Kb at 300°C for f O2 defined by the hematite-magnetite buffer. The presently determined P-T location for the blueschist-greenschist transition in the Fe-free basaltic system is about 3 Kb lower than the minimum pressure limit of glaucophane of Carman and Gilbert (1983), but is compatible with the revised stability field of jadeite + quartz determined by Holland (1980). Introduction of Fe 3+ into the model basaltic system significantly lowers the minimum pressure limit for occurrence of the buffered assemblage sodic amphibole + epidote + actinolite + chlorite + albite + quartz, and the participating phases gradually increase their Fe 3+ /Al ratio with decreasing pressure. Isopleths of sodic amphibole composition in the buffered assemblage in terms of X G1 are delineated and the effect of Fe 2+ and temperature on the isopleths are discussed. The Al 2 O 3 content of sodic amphibole coexisting with epidote + actinolite + chlorite + albite + quartz decreases systematically with decreasing pressure and hence can be used as a geobarometer. Pressure estimates for metabasites at Ward Creek of the Franciscan terrane, the Mikabu greenstones of the Sanbagawa belt, the Otago schists of Lake Wakitipu, New Zealand, and the blueschists at Ouegoa, New Caledonia, based on the proposed glaucophane geobarometry, are in agreement with those derived from sodic pyroxene geobarometry.