The role of pargasitic amphibole in the formation of major geophysical discontinuities in the shallow upper mantle

Abstract Several explanations have been proposed for variation in geophysical properties and depths for the lithosphere–asthenosphere boundary (LAB) and mid-lithospheric discontinuities (MLD). Here, we investigate the proposal that the dehydration solidus of pargasitic amphibole-bearing upper mantle...
Ausführliche Beschreibung

Abstract Several explanations have been proposed for variation in geophysical properties and depths for the lithosphere–asthenosphere boundary (LAB) and mid-lithospheric discontinuities (MLD). Here, we investigate the proposal that the dehydration solidus of pargasitic amphibole-bearing upper mantle with very low bulk water (hundreds ppm) may be one of the main reasons for the observed geophysical anomalies. The dehydration solidus may be associated with a very small degree of partial melting in the upper mantle at temperatures and pressures in excess of 1050 °C (for geochemically more depleted) or 1100 °C (for geochemically less depleted upper mantle) and from 1 to 3 GPa (~30 to 90 km) respectively. This small amount of partial melt may be responsible for changes in geophysical properties (e.g. lower seismic velocity, higher attenuation of seismic waves, higher electrical conductivity) in association with the LAB and MLD. This simple petrologic model is tested on the abundant geophysical data of the Carpathian–Pannonian region (CPR), central Europe. The high resolution heat flow data available in the CPR allows us to estimate the depths to intersection of area specific depth-temperature curves with the dehydration solidus temperatures (1050 and 1100 °C isotherms). There is relatively small mismatch (<5 km) between the position of these intersections and the geophysically determined LAB in the central area of the CPR. These observations lend support for the proposition that the dehydration solidus may be largely responsible for depth variation of the LAB in young continental rift areas. Towards the margins of the CPR, however, where the heat flow is lower (≲70 mW/$ m^{2} $), the predictive capability of the dehydration solidus model deteriorates. This is because, for lower geothermal gradients, pargasitic amphibole breaks down at ~90 km (or ~3 GPa) before temperature exceeds the dehydration solidus temperatures. Consequently, at ~90 km depth we expect no changes in geophysical properties indicative of hydrous silicate melt, in areas where surface heat flow is lower (i.e. Precambrian cratonic shields, Phanerozoic continental lithospheres or, possibly older oceanic plates). Alternatively, in these areas, the intersection of the geotherm with pargasitic amphibole breakdown may cause small changes in properties which correlate with the MLD rather than the LAB, which is at deeper levels. Ausführliche Beschreibung