Seismic waves generated by large earthquakes arrive at different times at seismographs arranged in a world-wide network. When they arrive depends on the relative positions of epicentres and receivers, but most importantly on variations in physical properties within the Earth that affect the speed at which they travel. Given enough high-quality seismic records and powerful computing, such data allow geophysicists to map how wave speeds change with depth in the mantle and produce 3-D models. In other words, seismic energy can produce geophysical homologues of medical CAT scans. The second important means of visualizing the unseeable comes from the geochemistry of basaltic lavas formed by partial melting of the mantle in different tectonic settings. Results from such studies reveal that the composition of the mantle is not homogeneous. Combining information from both sources, in the light of motions of the lithosphere, provides a powerful input to modelling how the Earth behaves as a whole (see Earth Pages, July 2000, Geodynamics).
Seismic tomography’s most important derivative stems from the manner in which wave speed depends on variations in the mechanical properties of the mantle. For P-waves, speed varies with the mantle’s differing resistance to compression, and S-wave speed is directly proportional to the rigidity of the mantle. Unusually high mantle temperatures cause decreases in compression resistance and rigidity, and therefore drops in the speeds of both kinds of body wave. The cooler the temperature, the higher both speeds. So, velocity variations in seismic tomographs are proxies for changing mantle temperature, and in turn for regions of different density – the hotter a material is, the lower is its density. The implications are quite simple; high-speed anomalies signify cool, potentially sinking regions in the mantle, whereas low speeds suggest that matter is able to rise. In practice, modelling the fundamental dynamics of the Earth’s mantle using seismic tomography is computationally difficult, often ambiguous and blurred because of the lack of suitable data.
Seismic tomography gave the first clues to the idea that subducted slabs penetrate all the way down to the core mantle boundary, and that at least some of the plumes suspected to underpin hot spots have their source at such depths. Together, these findings support whole-mantle convection. As well as improving the amount of high-quality seismic data and the software to analyse them, combining physical parameters with sketchy knowledge of variations in mantle chemistry and mineralogy is the next step in “sharpening” the focus of mantle models. That seems to have been taken by Alessandro Forte and Jerry Mitrovica of the Universities of Western Ontario and Toronto (Forte, A.M. and Mitrovica, J.X. 2001. Deep-mantle high-viscosity flow and thermochemical structure inferred from seismic and geodynamic data. Nature, v. 410, p. 1049-1056). Their work confirms the concept of whole-mantle convection resulting from thermal anomalies, but has an added bite. They show evidence for vary large variations in deep-mantle composition – to megaplumes they have added “mega-blobs”. Although the results of their analyses are limited by data availability and reliability, and by simplifying assumptions, they imply that such blobs can respond to temperature changes by rising and sinking periodically. That is, the mantle may move as vast domes and downwellings as well as in the more tightly constrained plumes and sinking slabs. One intriguing possibility is that such blobs may be primitive and retain high concentrations of elements that evolution of other parts of the mantle has transferred to the continental crust. Such primitive signatures are passed on to the geochemistry of basalts forming from plumes beneath ocean islands. However, there is a long way to go before a blob-plume-ocean island connection can be made. If it proves to be plausible, then such ancient blobs would have to be very viscous to have resisted mixing over time with more evolved mantle. Another possibility is that the blobs are themselves highly evolved, through the progressive accumulation of subducted slab material.
(See also: Manga, M. 2001. Shaken, not stirred. Nature, v. 410, p. 1041-1042)
Atmospheric oxygen: yet more
Following last month’s Earth Pages briefing (Mantle overturn and oxygenation of the atmosphere) Nature (19 April 2001) ran a news feature on the competing theories for when oxygen began to accumulate in Earth’s atmosphere (Copley, J. 2001. The story of O. Nature, v. 410, p. 862-864). The paradox between evidence for oxygen production by photosynthetic cyanobacteria since 3.5 Ga and that supporting the first major influence of oxygen in redbeds at 2.2 Ga may be resolved by the ideas of Hiroshi Ohmoto of Pennsylvania State University.
Redbeds – terrestrial sediments containing abundant ferric hydroxides – form when iron enters its Fe-3 state, and are insoluble. That results in weathering processes being unable to leach soils of their iron content, unless the waters involved have been rendered reducing by bacterial activity. The most dramatic expression of this is laterite that blankets ancient erosion surfaces of most of the Gondwanan continents, much of which formed in Palaeocene times. Palaeosols older than 2.2 Ga do not show the characteristic laterite ferricrete cap, implying that iron existed consistently in its soluble Fe-2 form and could be leached away. Most geochemists regard that as evidence for a reducing atmosphere, lacking oxygen except as a trace. Ohmoto suggests that organic acids formed by terrestrial cyanobacteria might also create the reducing conditions necessary for iron leaching.. He sees such “blue-greens” as having had a dual role, fixing iron in soils through oxidation and then releasing it to solution by formation of organic acids. Ohmoto and Antonio Lasaga are developing a geochemical model for the iron, oxygen, carbon and sulphur cycles during the Archaean. Early runs suggest that only 30 Ma after the appearance of cyanobacteria at 3.5 Ga their release of oxygen would have built up high levels in the early atmosphere.
That bucks the evidence for low oxygen provided by detrital sulphides and uranium oxide grains in Archaean high-energy sediments, such as the conglomerates of the Witwatersrand basin in South Africa – in the presence of oxygen, both should break down quickly in water. Archaean banded iron formations, thought to form by reaction between Fe-2 ions in ocean water and oxygen produced locally by shallow-water cyanobacteria, have a dual significance – abundant oceanic Fe-2 suggests global lack of oxygen, and BIF deposition of ferric oxide would have formed a sink for any oxygen in the environment. Ohmoto cites the re-appearance of BIFs at several times in the Proterozoic Eon as a sign that BIF formation was possible when atmospheric oxygen was abundant.
The debate seems destined to run, for two reasons. Studies of sulphur isotopes – Ohmoto’s speciality – give evidence for fractionation through the influence of ultraviolet radiation. Once oxygen rose in the air, its formation of ozone gas would have blocked UV and ended this kind of selective take-up of sulphur isotopes. James Farquhar of the University of California in San Diego has found its effects common in Archaean rocks, but no sign in later rocks. That favours an oxygen-poor early atmosphere. Ohmoto counters with abundant evidence in the Archaean for the activity of bacteria that reduce sulphate ions to sulphide – in an oxygen-poor world, sulphate formation would have been suppressed.
Oxygen build-up demands complementary burial of organic matter formed by photosynthesis before it oxidized. The influence of organic carbon burial is to take with it 12C that biological processes favour over heavier 13C, so that carbon-rich rocks show higher 12C than carbonates precipitated from the seawater that was left. Such enrichment in 12C shows up most clearly after 2.7 Ga ago, when carbon burial must have been stoked up somehow. That points to a late build-up of oxygen in the air. But why? James Kasting, also of Pennsylvania State University, suggests a change in the Earth’s mantle from reducing to oxidizing conditions. Before that time volcanic gases would have been dominated by reduced gases that could mop up any free oxygen. Afterwards, oxidized volcanic gases could have co-existed with free oxygen.