U44A-01
Re-evaluation of a 2.7-Ga Mantle Event from U/Pb Zircon Age Distributions in Orogenic Granitic Rocks
During the late Archean, seven major episodes of orogenic granitoid magmatism occur at 2750, 2700, 2675, 2650, 2630, 2600 and 2550 Ma as recorded by U/Pb isotopic ages of single zircons from orogenic granitic rocks (largely TTG). Age peaks are defined by igneous zircon ages with < 5% discordancy and < 20 My uncertainty. Each episode is represented on four or more cratons and has an uncertainty generally less than 10 My. The 2750, 2700, 2675, and 2600 Ma peaks are also represented in the detrital zircon population. Only the 2700-Ma peak appears to be global in extent, and an average uncertainty of 5 My, it occurs on nine cratons (Baltica, North China, India, Yilgarn, Superior, Nain, Slave, Amazonia, Rae/Hearne). The 2600 and 2550-Ma peaks are widespread only in India, North China and Africa. The age distribution of ENd(T) values greater than zero falling within one epsilon unit of the mantle growth curve indicates that a large volume of juvenile continental crust was added at 2700 Ma with possible significant amounts also at 2750 and 2550 Ma. Although a mantle plume event at 2700 Ma is a viable model for the 2700-Ma granite-forming event, a single plume event does not seem capable of explaining the other six granitic episodes in the late Archean. A peak in relative komatiite abundance at 2700 Ma correlates well with the global 2700-Ma granitoid event. Other phenomena in the late Archean, which are less precisely dated, yet may correlate with the 2700-Ma granitoid event include: a negative excursion in δ13C of the organic reservoir (2750-2650 Ma),δ18O change in detrital zircons and in the composition of seawater (2700-2200 Ma); peak in Re/Os mantle depletion age (2900-2700 Ma); La/Yb peak in TTG (2750-2650 Ma); flood basalt eruption (Fortescue at 2740, Nunavut at 2710; and Ventersdorp at 2720); ENd(T) and Nb/Th changes in the mantle (2700-2100 Ma); and Ce and Eu anomalies in BIF/chert (2700 Ma).
U44A-02
Plume Magmatism, Continental Tectonics, Increased Subaerial Volcanism and the Rise of Atmospheric Oxygen
The remarkable coincidence of the rise of atmospheric oxygen with the transition from the Archean to the Proterozoic Eons suggests that a likely cause-and-effect relationship with the defining tectonic change (the stabilization of continental cratons). A shift in the locus of volcanism from largely submarine to a mix of submarine and subaerial is argued to be the primary cause. This shift would have been accompanied by a significant reduction in the sink for oxygen provided by reduced volcanic gases: Archean submarine volcanoes emitted molecular hydrogen in excess, allowing hydrogen to accumulate in the atmosphere and prevent the build up of oxygen, whereas Proterozoic subaerial/submarine volcanoes did not, so oxygen came to dominate atmospheric composition. Evidence for a possible early rise to low oxygen levels between 2.94 and 2.74 Ga coincides with stabilization of the Pilbara and Kaapvaal Cratons and subsequent continental subaerial flood basalt volcanism. In contrast, the most intense episode of dominantly submarine mantle plume activity and growth and stabilisation of continental crust recorded in Earth history between 2.74 and 2.66 Ga resulted in a return to anoxic conditions. An abrupt switch close to the Archean-Proterozoic transition from dominantly submarine volcanic eruptions to a mixture of submarine and subaerial eruptions similar to that of the Phanerozoic during the 2.45 Ga plume break out and the final stages of amalgamation of Archean cratons into Earth's first large stable continents or supercontinent was coincident with strong evidence for the permanent establishment of atmospheric oxygen. Either cyanobacterial evolution was somehow tied to the 2.45 Ga event, or more likely, cyanobacterial oxygen photosynthesis evolved earlier, but its expression in the atmosphere was delayed by an overwhelming volcanic sink.
U44A-03
Age distribution of passive margins through earth history and tectonic implications
The ages and lifespans of all existing passive (Atlantic-type) margins plus 59 ancient ones were compiled. Passive margins have existed on Earth almost continually since at least 2685 Ma. Their abundance has fluctuated dramatically, and most of the fluctuations have clear tectonic causes. For the past 2200 Ma, the compiled age distribution of passive margins appears to be robust and not an artifact of an incomplete rock record or flaws in the compilation. It closely tracks all the first-order highs and lows of the seawater 87Sr/86Sr curve, which has been derived from utterly independent data. The main features of the age distribution are as follows: (1) A present-day maximum in number and aggregate length of passive margins corresponds to a time of continental dispersal following breakup of Pangea. (2) A 250- 350-Ma minimum corresponds to Pangea's greatest extent. (3) A 500-600-Ma maximum represents a time of continental dispersal following staged breakup from 600 to 1000 Ma of one or more larger continents (Rodinia in most models). (4) Passive margins are rare between 1000 and 1650 Ma, and none are known at all between 1650 and 1750 Ma. This 750-m.y. low in the passive margin record coincides with the heyday of massif anorthosites. Whereas the Mesoproterozoic may have seen few modern-style Wilson Cycles involving the opening and closing of Atlantic-type oceans, it nonetheless was a time of plate tectonics involving subduction and collision (e.g., Grenville orogeny). (5) Passive margins were abundant between 1750 to 2250 Ma. The close of this interval at 1750-1800 Ma was marked by the collisional assembly of Laurentia, Baltica, and other cratons, which may have been part of a supercontinent (Columbia in some models). A maximum at 1850-2050 Ma corresponds to a time of dispersed small continents. (6) The record of passive margins before 2250 Ma is patchy. It definitely extends back to 2685 Ma (Kaapvaal craton, W margin), and possibly to about 3000 Ma. For all but a few margins with an unusual tectonic history such as re-rifting, the lifespan was calculated from the time between the rift-drift transition and the passive margin to foredeep transition. The present-day passive margins have a mean age of 104 m.y. and a maximum age of 180 m.y.; these are partial lifespans. Fifty-nine ancient margins have a mean lifespan of 187 m.y. and a maximum lifespan of 550 m.y. Subdivided into natural age groupings, mean lifespans are 182 m.y. for the Archean to Paleoproterozoic, 211 m.y. for the Neoproterozoic, 145 m.y. for the Cambrian to Carboniferous, and 142 m.y for the Permian to present. Of the 59 ancient margins, 20 had lifespans that were longer than that of the oldest modern margins, and all 20 are either wholly or partly Precambrian in age. Several margins in the Mesoproterozoic and Neoproterozoic are very long-lived by modern standards, though in each case, the tectonic interpretation is debatable and (or) age control is poor. The tentative world-record holder, at 550 m.y., is the Mesoproterozoic eastern margin of Siberia. The number of long-lived Precambrian margins is inconsistent with the widely held notion that the tempo of Wilson Cycles was faster in the Precambrian than at present, as has been be predicted from the long-term decline in Earth's radiogenic heat production. Thus, the duration of Wilson Cycles involving passive margins is not a good proxy for rates of plate motion. Greater heat production in the Precambrian might still be linked to faster rates of plate motion if the effect was confined to oceanic (Pacific-type) plates.
U44A-04
Episodic Precambrian Subduction
Peaks in the preserved crustal record at 1.1, 1.9-2.1, 2.7 and 3.5Ga coincide with periods of anomalous tectonism and volcanism. These events do not conform with the supposition of continuous, smooth plate tectonics, and so previous workers have postulated mantle avalanches, superplumes, or exceptional plume activity as their ultimate cause. Here we show in numerical simulations that plate tectonics breaks down under hotter mantle conditions of the Precambrian, and is replaced by an episodic regime characterized by rapid pulses of subduction followed by long periods of relative quiescence. We also reanalyze the Precambrian paleomagnetic record, and demonstrate the existence of large anomalies in apparent polar wander record, coinciding with peaks in crustal production, and consistent with rapid plate motion during these events. These results shed light on the nature and style of Precambrian subduction, and the origin of the anomalous events which dominate the Precambrian geological record.
U44A-05
Growth rate of the lithospheric mantle: variations in time and space
Two global databases for the continents, (a) for tectono-thermal ages and (b) for lithospheric thermal thickness
(Artemieva, Tectonophysics, 2006 and available for download at the web-site), are used to calculate (i) the volume
of the preserved continental lithosphere of different ages within the individual cratons, (ii) the lithospheric growth
rate for different continents over the past 3.6 Ga, (iii) a global model of lithosphere growth rate since the Archean.
The submerged areas with continental crust are excluded from the analysis.
On the scale of a craton, significant differences in the rates of lithosphere growth are observed between the
individual cratons. These data are compared with independent estimates of growth rate of juvenile crust on
different continents as constrained by sedimentary record, geological and isotope data.
On the global scale, the results show a general agreement between the global cumulative growth rate of the
continental lithosphere and juvenile crust (Condie, 1998). The most pronounced peak in lithosphere growth
occurred at 2.1-1.7 Ga, when the volume of lithospheric mantle was increasing with the rate of ca. 10-20 (km3 per
year). Contrary to growth models of juvenile crust, the peaks in growth rate of the lithospheric mantle at ca. 2.7-
2.6 Ga and 1.3-1.1 Ga are weak, ca. 5-8 (km3 per year). The differences between the growth rates of the
lithosphere and juvenile crust are interpreted as indicator of the preservation rate of the cratonic lithosphere since
the Archean.
http:www.lithosphere.info
U44A-06
Can we derive a continental growth curve from global S seismic tomography?
The origin of continents is a standard question in geology; the worldwide application of geochronology lead to propose several models for the rate of formation of continental surfaces: see Hurley and Rand (1969), Amstrong (1981), Allègre (1982), McLennan and Taylor (1982), McCulloch and Bennett(1994), Condie (1998) ... In a diagram plotting the percentage of continental surface as a function of surface rock age, each model is characterized by a different curve - see Rino et al (2004) for example. In seismology, the structure of the continental lithosphere has been mapped in detail on all continents during the last 20 years. Large scale studies suggest a close relationship between the average P and S vertical travel times across the lithosphere inside continents and the age of surface rocks. The vertical travel time is directly related to the velocity and to the thickness of the lithosphere. It is correlated to the geotherm which is also known to be nearly dependent on the age of the continents. Using such a vertical traveltime - age relationship, it is straightforward to build a seismological curve which would be the seismic equivalent of the crustal growth curve familiar to geochronologists. This curve relates the percentage of continental surface to the average S vertical travel time and indirectly to age. We present various attempts to build such a curve. Our basic input is the S-tomography model obtained from global surface wave measurements by Shapiro and Ritzwoller (2002). P-delays derived from ISC residuals could in principle be used but they are a local measurement. S-tomography models have the advantage that they are surface measurements. We finally compare our growth curve with standard crustal growth curves and find that it is an "intermediate" curve. The S-tomography derived growth curve does not exhibit fast growth episodes as in McCulloch and Bennett (1994) and Condie (1998). A regular rate of formation of continents seems the most probable hypothesis at a worldwide scale from the S-tomography data set. This may reflect the smoothing inherent in computing S-models. On the contrary, P-station delay histograms are compatible with a few faster growth episodes but unfortunately, their spatial coverage is not good enough to conclude if fast growth episodes have been the rule or not.
U44A-07
Potential Dynamical Mechanisms Behind Global Mantle Events
By numerical models we have investigated three potential mechanisms behind global mantle events. Plumes, originating in the thermal boundary layers of the mantle convection system can exhibit a significant episodicity, once a strong temperature-dependence of the viscosity of the mantle material is taken into account. An increase of the viscosity with pressure, as sometimes believed to suppress plumes, acts in fact to focus buoyancy into a few strong upwellings, which are potentially able to generate events on global scale. Plumes originating self- consistently from a thermal boundary layer, transport mostly material from their source region, while they entrain only little material during ascent. Compositionally dense material at the Core-mantle boundary has been proposed to explain seismological observed anomalies. The stability of such heterogeneities against entrainment by the overlying mantle-flow is determined by a complex set of properties, rather than by the density difference alone. Model calculations, taking into account a combined dependence of viscosity on temperature, pressure and , as mostly neglected; on composition, demonstrate, that under such conditions the D", can function as an isolated reservoir form some time, that however the destruction of the compositionally distinct layer, shielding the Earth'core can take place rapidly., with a profound effect also on the surface heat flow.. Finally we observe that episodic mobilization events of the surface are dynamically plausible for appropriate rheologies. A combination of temperature- and stress-dependent viscosity leads to an intermittent type of temporal behavior, where periods showing no surface motion (stagnant lid) are interrupted by phases with strong plate motions at the top. It seems at least possible that plate motion is not a continuously operating process.