USING XENOLITHS TO EXPLORE VARIATIONS IN UPPER MANTLE COMPOSITION AND THE RELATION OF COMPOSITION TO SEISMIC VELOCITY STRUCTURE BENEATH THE COLORADO PLATEAU

The Colorado Plateau (CP) is a low-relief, high elevation region that has remained relatively undeformed and amagmatic within the actively-deforming western United States. Upper mantle seismic velocities beneath the central and northern CP are about 3 to 4% higher than those in the surrounding Rio Grande Rift and Basin and Range extensional provinces. These differences in seismic velocities may be due to several factors including temperature, composition, incipient melting, and the effects of mantle metasomatism. Previous workers have argued that xenoliths show that the mantle beneath the CP is depleted in basaltic components (especially iron) relative to the mantle from surrounding provinces. Garnet peridotite xenoliths are found only in the central part of the plateau, coinciding with a region of higher Vp, Vs and lower heat flow, which collectively suggest lateral variations in mantle composition and temperature beneath the plateau. We explore the contribution of compositional heterogeneity to seismic velocity variations in the mantle beneath the plateau and test the hypothesis that mantle compositional heterogeneity, as represented by variations in Mg-number (Mg#) and/or metasomatism, may be correlative with seismic velocity variations. The chemical and mineralogical compositions of mantle xenoliths from the Southwest are used to calculate density, Mg#, Vp, and Vs and these results are compared to published Vp and Vs velocity models for western North America. Depths of equilibration for xenoliths are estimated from existing thermobarometry. Preliminary results indicate that some calculated seismic velocities from xenoliths are consistent with observed velocities for depths of 55 to 95 km. Calculations of seismic velocities also show that the presence of a few percent of metasomatic minerals (e.g. chlorite, carbonate, amphibole) has <1% effect on Vp or Vs. It is possible that the decrease in Vp from the center of the plateau outwards is related to an increase in the ratio of spinel to garnet peridotite, which could reflect a steepening of the lithospheric mantle geotherm, or to variations in lithosphere thickness. An initial comparison of the empirically observed scaling between xenolith Mg#s and seismic velocity structure suggests that published scaling relations between Vs and Mg# need to be revised.