The mantle isotopic printer: Basic mantle plume geochemistry for seismologists and geodynamicists

High-temperature geochemistry combined with igneous petrology is an essential tool to infer the conditions of magma generation and evolution in the Earth's interior. During the past thirty years, a large number of geochemical models of the Earth, essentially inferred from the isotopic composition of basaltic rocks, have been proposed. These geochemical models have paid little attention to basic physics concepts, broadband seismology, or geological evidence, with the effect of producing results that are constrained more by assumptions than by data or first principles. This may not be evident to seismologists and geodynamicists. A common view in igneous petrology, seismology, and mantle modeling is that isotope geochemistry (e.g., the Rb-Sr, Sm-Nd, U-Th-Pb, U-Th-He, Re-Os, Lu-Hf, and other less commonly used systems) has the power to identify physical regions in the mantle, their depths, their rheological behavior, and the thermal conditions of magma generation. We demonstrate the fallacy of this approach and the model-dependent conclusions that emerge from unconstrained or poorly constrained geochemical models that do not consider physics, seismology (other than teleseismic travel-time tomography and particularly compelling colored mantle cross sections), and geology. Our view may be compared with computer printers. These can reproduce the entire range of colors using a limited number of basic colors (black, magenta, yellow, and cyan). Similarly, the isotopic composition of oceanic basalts and nearly all their primitive continental counterparts can be expressed in terms of a few mantle end members. The four most important (actually “most extreme”, because some are extraordinarily rare) mantle end members identified by isotope geochemists are DMM or DUM (depleted MORB [mid-ocean-ridge basalt] mantle or depleted upper mantle), HIMU (high mu, where mu = μ = ^(238)U/^(204)Pb), EMI, and EMII (enriched mantle type I and type II). Other mantle end members, or components, have been proposed in the geochemical literature (e.g., PHeM, FOZO, LVC, PreMa, EMIII, CMR, LOMU, and C), but these can be considered to be less extreme components or mixtures in the geochemical mantle zoo. Assuming the existence of these extreme "colors" in the mantle isotopic printer, the only matter for debate is their location in the Earth's interior. At least three of them need long-term insulation from convection-driven homogenization or mixing processes. In other words, where these extreme isotopic end members are located needs to be defined. In our view, no geochemical, geological, geophysical, or physical arguments require the derivation of any magma from deep mantle sources. Arguments to the contrary are assumption based. The HIMU, EMI, and EMII end members can be entirely located in the shallow non-convecting volume of the mantle, while the fourth, which is by far the more abundant volumetrically (DMM or DUM), can reside in the transition zone. This view is inverted compared with current canonical geochemical views of the Earth's mantle, where the shallowest portions are assumed to be DMM like (ambient mantle) and the EMI-EMII-HIMU end members are assumed to be isolated, located in the deep mantle, and associated with thermal anomalies. We argue that the ancient, depleted signatures of DMM imply long-term isolation from recycling and crustal contamination while the enriched components are not free of contamination by shallow materials and can therefore be shallow