Chapter 13 Making an Earth Thus God knows the world, because melts, and peridotites are thought to be He conceived it in His mind, as if from residues, some mixture of these should approximate the composition of the upper the outside, before it was created, mantle. Peridotites are the main mantle and we do not know its rule, because reservoirs for elements such as magne­ we live inside it, having found it sium, chromium, cobalt, nickel, osmium already made. and iridium. The continental crust is an important reservoir of the crustal elements; Brother William of Baskerville potassium, rubidium, barium, lanthanum, uranium and thorium. Enriched magmas, Overview such as kimberlites, are rare but they are so enriched in the crustal elements that they cannot be ignored. Thus, each of these Attempts to estimate mantle composition fall components plays an essential role in deter­ into two broad categories. mining the overall chemistry of the prim­ Cosmochemical approaches take meteorites as itive mantle. An alternate approach is to the basic building blocks; these materials search for the most 'primitive' ultramafic are processed or mixed in order to satisfY rock or component and attribute its compo­ such constraints as core size, heat flow and sition to the whole mantle or the original crustal ratios of certain elements. These mantle, or to take the most depleted MORB models constrain the bulk chemistry of the and attribute its source region to the whole Earth rather than that of the mantle alone. upper mantle. The only primitive meteorites that match satisfactorily both the stable isotope and The earliest, and simplest, petrological nwdels redox characteristics of the Earth are the tended to view the mantle or the upper mantle as enstatite chondrites. The actual material homogenous, and capable of providing primitive forming the Earth is unlikely to be repre­ basalts by partial melting. When the petrological sented by a single meteorite class. It may be data are combined with isotopic and geophysical a mixture ofvarious kinds of meteorites and data, and with considerations from accretional the composition may have changed with calculations, a more complex multi-stage evolu­ time. The oxidation state of accreting mate­ tion is required. Similarly, simple evolutionary rial may also have changed with time. models have been constructed from isotope data Petrological models take the present continen­ alone that conflict with the broader database. tal crust, basalts and peridotites as the basic There is no conflict between cosmochemical, geo­ building blocks. Since basalts represent chemical, petrological and geophysical data, but 154 MAKING AN EARTH a large conflict between models that have been provides information about physical properties constructed by specialists in these separate fields. and boundaries in the mantle. The mantle is heterogenous on all scales and Recycling of crust into the upper mantle is there are vast regions of the Earth that are an important process. It is possible to estimate unsampled today because they were isolated by the composition of the f e r til e u pper man­ gravitational stratification during accretion of t l e by combining known components of the the planet. upper mantle- basalts, peridotites, recycled crust The composition of the crust and the upper and so on - in such a way as to satisfY cosmic mantle are the results of a series of melting ratios of the lithophile refractory elements. Other and fractionation events, including the high­ elements are not necessarily concentrated into temperature accretion of the planet. Attempts the crust and upper mantle. The MORB source is to estimate upper-mantle chemistry usually start just part of the upper mantle and it is not the from the assumption that it initially was the only LIL depleted part of the mantle. It is not same as bulk s ilicate Ea rth (ESE) and dif­ necessarily convectively homogenized. fers from it only by the extraction of the Attempts to establish an average composi­ crust, or that the most depleted midocean-ridge tion for the upper mantle have focused on basalts (MORB) plus their refractory residues MORB because of the assumption the whole constitute the entire upper mantle. Tradition­ upper mantle is the MORE-reservoir. This pro­ ally, geochemists have assumed that the lower cedure involves major assumptions about melt mantle is still undifferentiated ESE . On the other generation, melt transport and differentiation hand, large-scale melting and differentiation processes that have affected these melts, and upon accretion probably pre-enriched the upper the sources of non-MORB melts. The d e ple ted mantle with incompatible elements, including upper ma n t l e , that part of the mantle that the radioactive elements; the crust and the vari­ is assumed to provide MORB by partial melt­ ous enriched and depleted components sampled ing is variously called DUM, DM, DMM and the by current melting events were probably already con vecting upper mantle. Simplified mass in the upper mantle shortly after accretion and balance calculations suggested to early workers solidification. that this depleted mantle constituted ~ 30 % of Midocean-ridge basalts represent large the mantle; the 650-670-km discontinuity was degrees of melting of a large source volume, and adopted as the boundary between DUM and ' the blending of magmas having different melting primitive u n d e pleted undegassed l ower histories. The central limit theorem explains mant l e '. The starting condition for the upper many of the differences between MORB and other mantle (UM) was taken as identical to primi­ kinds of melts that sample smaller volumes of tive ma ntle (PM) and the present lower man­ the heterogenous mantle. Observed isotopic tle (LM). There are many estimates of PM and arrays and mixing curves of basalts, including ESE based on various cosmological and petrologi­ ocean-island basalts (OIB), can be generated by cal considerations. The primitive upper mantle - various stages of melting, mixing, melt extrac­ crust plus DUM - is labeled PUM. It was further tion, depletion and enrichment and do not assumed that the upper mantle was vigorously require the involvement of unfractionated, prim­ convecting, well-stirred and chemically homoge­ itive or lower, mantle components. However, the nous, and extended from the base of the plate to first stage in building an Earth - the accretional 650-km depth. Thus, this part of the mantle was stage - does involve large degrees of melting that also called the convecting mantle. The non-MORB essentially imparted an unfractionated - but basalts that occur at the initiation of spreading enriched - chondritic REE pattern to the upper and at various locations along the global spread­ mantle. Small-degree melts from this then serve ing system were attributed to plumes from the to fractionate LIL. Mass-balance and box-model lower mantle. calculations can go just so far in constrain­ Most or all of the mantle needs to be depleted ing the chemistry of the mantle. Geophysics and degassed to form the crust and upper mantle PETROLOGICAL BUILDING BLOCKS ISS and the 40 Ar in the atmosphere (e.g. the first more reasonable that the Earth accreted from edition of Theor y o f t h e Earth, or TOE). material with major-element compositions that Depletion of the upper mantle alone cannot were distinct from primitive meteorite types than explain the continental crust (CC); the MORB to accept chemical stratification and petrological reservoir and the CC are not exactly comple­ differentiation of the mantle. mentary. There must be other components and 'Primitive mantle' or PM is the silicate frac­ processes beyond single-stage small-degree melt tion of the Earth, prior to differentiation and removal from part of the primordial mantle to removal of the crust and any other parts of the form CC. There are other enriched components present mantle that are the result of differen­ in the mantle, probably in the shallow man­ tiation, or separation, processes. This is called tle. Other depleted components or reservoirs, in bulk silicate Earth (ESE). In geochemical models, addition to the MORB-source, are required by which were popular until very recently, it was mass-balance calculations. TI1e upper mantle can­ assumed that large parts of the Earth escaped not be treated as if its composition can be uni­ partial melting, or melt removal, and are there­ quely determined from the properties of depleted fore still 'primitive'. Some petrological models MORB - NMORB or DMORB - and depleted peri­ assumed that melts being delivered to the Earth's dotites, continental crust, and an undifferential surface are samples from previously unprocessed starting condition. material. It is difficult to believe that any part Both the cosmochemical and petrological of the Earth could have escaped processing dur­ approaches utilize terrestrial and meteoritic ing the high-temperature accretional process. data. The common theme is that the Earth 'Primitive mantle', as used here, is a hypothetical should have an unfractionated chondri tic pattern material that is the sum of the present crust and of the refractory elements. This can be used as a mantle. Some petrological models assume that formal a priori constraint in geochemical model­ it is a mixture of the MORB-source and conti­ ing of the composition of the Earth. This mass nental crust. 'Primitive magma' is a hypotheti­ balance approach is consistent with the idea cal magma, the parent of other magmas, which that most of the radioactive elements, and other formed by a single-stage melting process of a par­ crustal elements, are in the crust and upper man­ ent rock and has not been affected by loss of tle. Some investigators, however, decouple their material (crystal fractionation) prior to sampling. models of the Earth from meteorite composi­ It is much more likely that magmas are the result tions.