Empirical geothermobarometry for garnet peridotites and implications for the nature of the lithosphere, kimberlites and diamonds

Abstract

As a result of a study on the influence of Cr on geothermometry and geobarometry empirical garnet-orthopyroxene geobarometry has been refined. The applicability to simple and complex systems is demonstrated and the modelling procedure briefly outlined. The suitability of some geothermometers for application to garnet lherzolite xenoliths and their limits of precision are discussed. Fields of pressure-temperature estimates for garnet lherzolites from various provinces are shown. Low-temperature xenoliths (< 1050°C) give estimates in accord with a conductive geotherm for the continental lithosphere. African xenoliths (both South African and “off-craton” Namibian) record a distinctive pattern, exhibiting a “temperature discontinuity”, where high-temperature xenoliths (> 1050°C) give near-isobaric estimates for a range of temperatures. The depth of this “temperature discontinuity” is shallower for off-craton than for on-craton suites. Similar patterns are indicated for suites from U.S.A. and the U.S.S.R. The temperature discontinuity is interpreted as a geologically short-lived transient state, where the deepest parts of the lithosphere are in the process of adjustment to higher heat flux from below. At the highest temperatures recorded by those suites “wet” peridotitic solidi are intersected and thus the xenoliths record events, were P,T conditions may have allowed the formation of kimberlitic or strongly undersaturated magmas at about the depths of this discontinuity. This process is inferred to lead to thermal erosion of the base of the lithosphere. The generation and subsequent crystallization of magmas at the base of the lithosphere is inferred to be the source for diamond formation. The distinctive pattern of high- and low-temperature xenoliths in relation to the diamond stability field shows the existence of a rather small “diamond window” between about 900 and 1300°C at 40–55 kb. This window exists only underneath old, stable continents with a thick lithosphere. If recent age determinations on diamonds are correct, this implies the existence of thick continental lithosphere with similar characteristics to present-day lithosphere in the Archean. A model is presented arguing for hot, upwelling convection streams as a possible source of heat. The model implies differences in lithospheric development according to the speed of plates, carrying a continent: Continuous lithospheric thinning by thermal erosion for steady or very slow moving plates with the break-up of a continent and the formation of an ocean as one extreme, the continuous repetition of thermal erosion and growth on cooling (underplating) of the continental lithosphere as the other extreme for fast moving plates. The latter may be the cause of a number of complexities and heterogeneities both within the deep continental lithosphere as well as in the convecting mantle.