
Thelower mantle, historically also called themesosphere, occupies about 56% of the total volume ofEarth, and is the region from 660 to 2,890 km (410 to 1,800 mi) belowEarth's surface; between thetransition zone and theouter core.[1] Thepreliminary reference Earth model (PREM) separates the lower mantle into three sections, the uppermost (660–770 km (410–480 mi)), mid-lower mantle (770–2,700 km (480–1,680 mi)), and the D layer (2,700–2,890 km (1,680–1,800 mi)).[2]
Pressure and temperature in the lower mantle range from 24–127 GPa (3,500,000–18,400,000 psi)[2] and 1,900–2,600 K (1,630–2,330 °C; 2,960–4,220 °F).[3] It has been proposed that the composition of the lower mantle ispyrolitic,[4] containing three major phases ofbridgmanite,ferropericlase, and calcium-silicate perovskite. The high pressure in the lower mantle has been shown to induce a spin transition of iron-bearing bridgmanite and ferropericlase,[5] which may affect bothmantle plume dynamics[6][7] and lower mantle chemistry.[5] The mantle moves at about 1 cm (0.39 in) per year.[8]
The upper boundary is defined by the sharp increase inseismic wave velocities anddensity at a depth of 660 km (410 mi).[9] At a depth of 660 km (410 mi),ringwoodite γ-((Mg,Fe)2SiO4) decomposes intoMg-Si perovskite andmagnesiowüstite.[9] This reaction marks the boundary between theupper mantle and lower mantle. This measurement is estimated from seismic data and high-pressure laboratory experiments. The base of the mesosphere includes theD″ zone which lies just above themantle–core boundary at around 2,700–2,890 km (1,678–1,796 mi).[9]
The lower mantle was initially labelled as the D-layer in Bullen's spherically symmetric model of the Earth.[10] The PREM seismic model of the Earth's interior separated the D-layer into three distinctive layers defined by the discontinuity inseismic wave velocities:[2]
The temperature of the lower mantle ranges from 1,960 K (1,690 °C; 3,070 °F) at the topmost layer to 2,630 K (2,360 °C; 4,270 °F) at a depth of 2,700 kilometres (1,700 mi).[3] Models of the temperature of the lower mantle approximateconvection as the primary heat transport contribution, while conduction and radiative heat transfer are considered negligible. As a result, the lower mantle's temperature gradient as a function of depth is approximately adiabatic.[1] Calculation of the geothermal gradient observed a decrease from 0.47 kelvins per kilometre (0.47 °C/km; 1.4 °F/mi) at the uppermost lower mantle to 0.24 kelvins per kilometre (0.24 °C/km; 0.70 °F/mi) at 2,600 kilometres (1,600 mi).[3]
The lower mantle is mainly composed of three components, bridgmanite, ferropericlase, and calcium-silicate perovskite (CaSiO3-perovskite). The proportion of each component has been a subject of discussion historically where the bulk composition is suggested to be,
Laboratory multi-anvil compression experiments ofpyrolite simulated conditions of the adiabaticgeotherm and measured the density usingin situX-ray diffraction. It was shown that the density profile along the geotherm is in agreement with thePREM model.[11] The first principle calculation of the density and velocity profile across the lower mantle geotherm of varying bridgmanite and ferropericlase proportion observed a match to the PREM model at an 8:2 proportion. This proportion is consistent with the pyrolitic bulk composition at the lower mantle.[12] Furthermore, shear wave velocity calculations of pyrolitic lower mantle compositions considering minor elements resulted in a match with the PREM shear velocity profile within 1%.[13] On the other hand,Brillouin spectroscopic studies at relevant pressures and temperatures revealed that a lower mantle composed of greater than 93% bridgmanite phase has corresponding shear-wave velocities to measured seismic velocities. The suggested composition is consistent with a chondritic lower mantle.[14] Thus, the bulk composition of the lower mantle is currently a subject of discussion.
The electronic environment of two iron-bearing minerals in the lower mantle (bridgmanite, ferropericlase) transitions from a high-spin (HS) to a low-spin (LS) state.[5] Fe2+ in ferropericlase undergoes the transition between 50–90 GPa. Bridgmanite contains both Fe3+ and Fe2+ in the structure, the Fe2+ occupy the A-site and transition to a LS state at 120 GPa. While Fe3+ occupies both A- and B-sites, the B-site Fe3+ undergoes HS to LS transition at 30–70 GPa while the A-site Fe3+ exchanges with the B-site Al3+ cation and becomes LS.[15] This spin transition of the iron cation results in the increase inpartition coefficient between ferropericlase and bridgmanite to 10–14 depleting bridgmanite and enriching ferropericlase of Fe2+.[5] The HS to LS transition are reported to affect the physical properties of the iron bearing minerals. For example, the density and incompressibility was reported to increase from HS to LS state in ferropericlase.[16] The effects of the spin transition on the transport properties andrheology of the lower mantle is currently being investigated and discussed using numerical simulations.
Mesosphere (not to be confused withmesosphere, a layer of theatmosphere) is derived from "mesospheric shell", coined byReginald Aldworth Daly, aHarvard Universitygeology professor. In the pre-plate tectonics era, Daly (1940) inferred that the outer Earth consisted of threespherical layers:lithosphere (including thecrust), asthenosphere, and mesospheric shell.[17] Daly's hypothetical depths to the lithosphere-asthenosphere boundary ranged from 80 to 100 km (50 to 62 mi), and the top of the mesospheric shell (base of the asthenosphere) were from 200 to 480 km (124 to 298 mi). Thus, Daly's asthenosphere was inferred to be 120 to 400 km (75 to 249 mi) thick. According to Daly, the base of the solid Earth mesosphere could extend to the base of the mantle (and, thus, to the top of thecore).
A derivative term,mesoplates, was introduced as aheuristic, based on a combination of "mesosphere" and "plate", for postulated reference frames in which mantlehotspots exist.[18]