We explore whether increases in thermal conductivity and viscosity could create a stagnant layer in the lower mantle that acts as a reservoir for material feeding deep-rooted plumes with distinct geochemical signatures from mid-ocean ridge basalts. Changes in the spin state of iron in the lowermost mantle may increase (Sherman 1991, Badro et al. 2003,2004) or decrease (Lin et al. 2005) the radiative heat transfer and should increase the viscosity (Badro et al. 2003). We use a finite-element model of mantle convection with variable viscosity and thermal conductivity to simulate aspects of this transition. The models include passive tracer particles to track the stability of material originating in lower mantle. We examine the stability of different candidate reservoirs by calculating the configurational entropy of the passive tracer particles through time using the method of Goltz and Bose (2002). As stirring takes place, the configurational entropy increases with time, until it reaches a peak and levels off. Models with a lower overall mixing rate require more time before the peak in configurational entropy occurs. The rate of increase of configurational entropy, and the time required to reach the configurational entropy peak, can serve as a quantitative measure of the mixing between different mantle reservoirs. We also model noble gas compositions of different regions in order to determine whether elevated viscosity and thermal conductivity can produce the distinct noble gas isotopic compositions observed in OIB and MORB. Previous calculations show that large increases in viscosity and thermal conductivity at the mantle mid-point fail to isolate reservoirs in the lower mantle. Material crosses the interface and mixing takes place between regions with varying viscosity and thermal conductivity. These models therefore do not produce the distinct, long-lived, isolated reservoirs in the mantle required by neon and xenon studies. To isolate parts of the mantle for most of the Earth's history requires intrinsic density contrasts. Models containing intrinsic density contrasts in the D'' region along with thermal conductivity and viscosity increases near 2000 km are also examined. Increasing viscosity and thermal conductivity near 2000 km depth noticeably increases the size of upwellings in the lower mantle, but fails to significantly retard flow passing through 2000 km depth. These preliminary results suggest that increases in viscosity and thermal conductivity near 2000 km depth likely do not stagnate lower mantle flow enough to create the stable, long-term reservoirs suggested by neon and xenon studies. The variations in viscosity and thermal conductivity at 2000 km depth, however, significantly affect the average temperature of the system as well as the basal and surface heat fluxes.

[ return to abstract list ]