A.R. Calderwood
Dept. of Physics, University of Nevada-Las Vegas, Las Vegas, USA
An updated model of the lower mantle thermal conductivity profile ($k(T, P)$) is presented and used to calculate the present day Net Core Heat Flux ($Q_{CMB}$) for 4 types of boundary layers in the $D''$ layer. I re-evaluate the temperature and pressured dependent thermal conductivity ($k(T, P)$) profile of the lower mantle using the method of Brown (1986), but with high temperature and pressure input parameters for Mg-perovskite and magnesiowustite in a simplified pyrolite mineralogy that ignores Ca-perovskite. The magnesiowustite profile is normalized with the experimentally derived high temperature and pressure value at the top of the lower mantle of Katsura (1997), while the Mg-perovskite profile is normalized to the high temperature, but ambient pressure, theoretical estimate of Hofmeister (1999). The resulting lower mantle $k(T, P)$ profile increases across the lower mantle from $\approx$10 W/m K at 660 km to values of $\approx$19 W/m K at $\approx$2,500 km. The present day Net Core Heat Flux across the $D''$ thermal boundary layer is calculated following the approach of Leitch (1995) but adopting 260 km for the global mean thickness of the $D''$ layer from seismological studies (Kendall and Shearer, 1994) and $\Delta$$T_{CMB}$ = 1,550 K for the temperature jump across the thermal boundary layer. $\Delta$$T_{CMB}$ is determined from the temperature offset at the core-mantle boundary between a mantle adiabatic geotherm for a chemically uniform pyrolite composition and the latest core geotherm. $Q_{CMB}$ is sensitive to the assumed type of $D''$ thermal boundary layer. Leitch (1995)derived the equations for 3 different types of boundary layers which might exist in the $D''$ layer and these are used to determine $Q_{CMB}$ for a stagnant, ablating, convecting, and composite thermal boundary layers. For a purely stagnant conducting $D"$ layer, the net core heat flux is $\approx$19 TW. If the entire $D''$ layer is solely an ablating plume source region then $Q_{CMB}$ is $\approx$29 TW, while for a convecting type, $Q_{CMB}$ is $\approx$12 TW. Lastly, for a binary composite boundary layer composed of 5$\%$ by area thinned, ablating plume catchment regions and 95$\%$ ponded slabs, the resulting $Q_{CMB}$ is $\approx$24.4 TW. This range of $Q_{CMB}$ values is significantly greater than the often cited canonical value for the net core heat flux (2.5 TW) that is estimated from the present day surface plume flux which demonstrates that the plumes are only a minor mechanism for removing heat from the core. Although nonunique, the $Q_{CMB}$ for the composite boundary layer is in good agreement with an independent estimate of the net core heat flux that is determined simply by subtracting the crust, lithosphere, average mantle, and recycled altered ocean crust radiogenic heat production (8.45 TW, 0.37 TW, 2.0 TW, and 0.54 TW) together with the mantle secular cooling heat flux (6.6 TW) from the global surface heat flux (42.5 TW). The radiogenic heat production values are determined from a self-consistent mass balance model together with an independent estimate for the mass of recycled ocean crust in the $D''$ layer calculated from parameterized cooling models. A net core heat flux of $\approx$24 TW indicates that the present day mantle is largely heated from below and not from within. This refutes conventional wisdom in the isotopic community that holds that most of the lower mantle is isotopically enriched relative to the shallow, depleted upper mantle. This study indicates that to first order, the whole mantle must be depleted and chemically uniform, with only a relatively insignificant mass of altered ocean crust ponded on the CMB.