T22B-01 INVITED
The Goldilocks problem of making cratons
Cratons form the cores of continents and were formed within a narrow window of time (2.5-3.2 Gy ago), the majority having remained stable ever since. Petrologic evidence suggests that the thick mantle roots underlying many cratons might have been built by lithospheric stacking, but paradoxically this requires that the building blocks of cratons are weak even though cratons must have been strong subsequent to formation. Here, we show that thickening is facilitated by thrusting of oceanic lithospheres along weak shear zones, generated in the serpentinized upper part of the oceanic lithosphere (crust + mantle) due to hydrothermal interaction with seawater. Heating of the shear zones by thermal diffusion eventually causes serpentine breakdown at 600 oC, shutting down the shear zone and culminating in craton formation. However, if shear zones are too thin, then heating is too fast and underthrusting does not occur. If they are too thick, heating is too slow and stacking is never given a chance. Shear zone thicknesses of ~18 km are found to be favorable for craton formation. Because the original depth of seawater-induced serpentinization into the lithosphere is limited by the isotherm for serpentine breakdown, the shear zone thickness should have increased with time as the Earth's heat flux decayed with time. Thus, the early Archean was too hot, the Phanerozoic too cold, and the late Archean just right for making cratons.
T22B-02 INVITED
Southern African lithospheric structures and geometries imaged by SAMTEX: Clues to Archean tectonic processes
SAMTEX – The Southern African Magnetotelluric Experiment – is a lithospheric-scale, electromagnetic imaging project being undertaken by a multi-institutional, international consortium comprising academia, government and industry. The primary objective of SAMTEX is to determine the lithospheric geometries and internal structures of the major Archean cratons and their Proterozoic bounding belts in southern Africa with a view to elucidating Archean and Proterozoic tectonic processes of formation, deformation and destruction, and comparing those processes with those deduced for other Archean cratons and with their modern counterparts. To date, in three phases of acquisition over three years, MT data have been acquired at over 550 sites on over 9000 line kilometers in a spatial area exceeding a million square kilometers, making this the largest survey of its kind ever conducted. This paper will review the data acquired and present images and models of subsurface structures. In particular, the relationship between lithospheric electrical parameters and diamondiferous kimberlite pipes will be highlighted. Comparisons will be drawn with the Slave craton in northern Canada, which is as well-known geophysically, geochemically, petrologically and tectonically as the Kaapvaal craton. http://www.dias.ie/
T22B-03
Thermal Isostasy – a new Constraint for Heat Generation Models of Continental Lithosphere
A general thermal isostasy model for continental provinces provides a new constraint for continental geotherms and therefore for the vertical distribution of heat generation throughout the lithosphere. Continental elevation results primarily from a combination of compositional and thermal buoyancy. The thermal component of elevation is isolated by using seismic velocity data and a velocity-density transform to remove crustal thickness and local crustal density effects. The remaining "adjusted" elevation of continental provinces for 73 global tectonic provinces shows an approximate 3000 m variation between hot rift zones and cold shield regions, not unlike the elevation difference between hot mid ocean ridges and cool abyssal plains in the oceanic case. Both amplitude and trends of thermal elevation are well predicted from a set of continental geotherms for which we assume a standard vertical distribution of heat production. A diagnostic plot of adjusted elevation versus heat flow allows us to identify regions of anomalous heat generation. For example the low thermal/adjusted elevation (84 m) of the Precambrian Wopmay Orogen in northern Canada is consistent with a surface heat flow of 43 mW- m-2 rather than the observed heat flow of 75 mW-m-2. The excess 32 mW-m-2 heat flow most likely arises from anomalous heat generation, the vertical distribution of which is tightly constrained because it must produce a geotherm equivalent to the standard geotherm consistent with a surface heat flow of 43 mW- m-2. Further examples are drawn from north and south Australian cratons.
T22B-04 INVITED
A Reference Heat Generation Model for the Continental Lithosphere Constrained by Heat Flow and Elevation
We propose a new general heat generation model within the continental lithosphere. Heat generation within the upper crust (~16~km) is parameterized as a percentage of the surface heat flow to account for the general decrease in heat flow and heat generation with crustal age. We explore constant, linear and exponential heat production relations as viable models. The best-fitting thermal isostatic model suggests ~30% of the surface heat flow is due to heat generation within this layer. Our model assumes a lower crustal heat generation of 0.4~μW-m-3, consistent with observations from exposed granulite terranes (0.15--2.08~μW- m-3, median 0.4~μW-m-3) but somewhat higher than mafic lower crustal xenoliths (0.13~μW-m-3). We estimate mantle heat generation as 0.02~μW-m-3, consistent with estimates of non-kimberlite hosted cratonic xenoliths (0.019~μW-m-3) and previous modeling studies (0 - 0.03~μW-m-3). Tests of elevation sensitivity to heat generation variations show that much of the scatter about the best-fitting thermal isostatic relationship may be due to deviations from the reference heat generation. Continental elevation provides a new constraint to improve lithospheric heat generation estimates, thus allowing more accurate determination of lithospheric and boundary temperatures and heat flow, all of which have important geodynamic implications.
T22B-05 INVITED
Heat flow, crustal heat production, and crustal evolution in the Canadian Shield
More than 200 heat flow values are available for the entire Canadian Shield, in provinces that range in age between >3Ga and <1Ga. Heat production has been determined at nearly all the heat flow sites. In addition, systematic heat production studies have been conducted that sample through the most important lithologies and all structural levels. At the largest scale, we find that the mean heat flow is the same 41mW~m-2 in all provinces regardless of age, with the exception of the Archean Slave Province where the limited data set suggests that surface heat flow is higher (50mW~m-2). The Slave Province is also characterized by high surface heat production. On the other hand, geothermobarometry on mantle xenoliths indicate lower mantle temperature in the Slave than in the Superior Province. Heat flow and xenolith data can be reconciled if variations in heat flow are accounted for by crustal heat production. Variations are found at the smaller scale of individual subprovinces with different lithologies. The southern part of the Superior Province consists of E-W trending belts with ages decreasing southwards. This age trend seems to be reflected in a progressive southward increase in heat flow in the greenstone belts. Within a single belt, variations in heat flow reflect changes in the average crustal composition, essentially the proportion of mafic to felsic rocks. In the PaleoProterozoic (1.8Ga) Trans Hudson Orogen, juvenile crust is less radiogenic and has lower heat flow (37 mW~m-2) than the reworked Archean crust of the Thompson Belt (53mW~m-2). Several examples of high amplitude-short wavelength variations in heat flow have been found suggesting that these variations are due to crustal sources. Different methods have been used to estimate the crustal and mantle contribution in different regions of the Shield. The range of mantle heat flow values is very narrow 11-15mW~m-2 and within the error limits of the estimate (± 2 mW~m-2), implying that most of the heat flow variations are accounted by crustal heat production. With such narrow range of Moho heat flow variations, we can define a differentiation index, as the ratio of the surface to the mean crustal heat production. This index is positively correlated with the surface heat flow at the time of crustal stabilization, which is consistent with high temperature in the lower crust resulting in a very differentiated crust. Consequently, temperature differences are small at the base of the crust and in the lithospheric mantle.
T22B-06 INVITED
A Mechanism to Explain the Exponential Model for Heat Production
Evidence for chemical interaction between plutons and meteoric ground water and analyses of the micro-scale and mega-scale distribution of uranium and thorium suggests that thermal convection in batholith complexes mobilizes uranium and may partially explain the exponential model for the vertical distribution of crustal heat production. Investigations of the distribution of uranium, thorium, and potassium within and around eight shallow plutons show evidence for mobilization and redistribution of uranium that was not bound in resistate-refractory minerals. Induced fission track radiography of thin sections was used to reveal the mineral associations of uranium within the rocks. Shallow plutons showing evidence of significant chemical interaction with meteoric groundwater (delta 18O) contain significant amounts of uranium in secondary hydrous minerals and secondary oxides. Plutons from deeper levels in the crust that show no evidence of interaction with meteoric ground water show uranium occurs predominately in resistate minerals such as zircon, apatite, allanite, sphene, and monazite. Thorium may behave similarly. These observations suggest the hypothesis that an exponential decrease in fracture permeability with depth likely controls the volume of hydrothermal fluid activity within the crust. Soils, sediments, volcanics, metamorphic, igneous, and sedimentary rocks contain uranium and thorium in both resistate minerals and as mobile ions. A significant amount of uranium, and possibly thorium, may be assimilated from hydrothermal fluids that interact with cooling plutons in the upper crust. Metasomatism may similarly affect the distribution of uranium and thorium in metamorphic environments.
T22B-07
Birch's Crustal Heat Production-Heat Flow Law: Key to Quantifying Mantle Heat Flow as a function of time
Birch (1968) first showed the linear correlation of surface heat flow and radioactive heat production (Qs = Qo + bAs ) in granites in New England, USA and discussed implications to the vertical scale of radioactive heat generation in the crust. Subsequently similar relationships have been found worldwide and numerous papers written describing more details and expanding the implications of Birch's Law. The results are a powerful contribution from heat flow research to the understanding of the lithosphere and its evolution. Models are both well constrained experimentally and simple in implications. However, there still exist thermal models of the crust and lithosphere that do not have the same firm foundation and involve unnecessary ad hoc assumptions. A main point of confusion has been that the several of the original relationships were so low in error as to be considered by some to be "fortuitous". Interestingly a "similar" relationship has been proposed based on regional scale averaging of Qs -As data. A second point of confusion is that one admissible crustal radioactivity distribution model (the constant heat generation to depth b) has been criticized as unrealistic for a number of reasons, including the effect of erosion. However, it is appropriate to refer to the Qs -As relationship as a law because in fact the relationship holds as long as the vertical distribution is "geologically realistic." as will be demonstrated in this paper. All geologic and geophysical models of the continental crust imply decreasing heat production as a function of depth (i.e. the seismic layering for example) except in very special cases. This general decrease with depth is the only condition required for the existence of a "linear" Qs -As relationship. A comparison of all the Qs -As relationships proposed for terrains not affected by thermal events over the last 150 to 200 Ma shows a remarkably uniformity in slope (10 ± 3 km) and intercept value (30 ± 5 mWm-2 ). Therefore these parameters of Birch's Law equation represent the starting place for discussions of lithospheric thermal regime and evolution. The stability of the values of intercept Qo for areas with thermal ages of Paleozoic and older prove that the lithosphere heat flow does not vary significantly with age as is demonstrated in the companion paper. The minimum mantle heat flow for preMesozoic thermal terrains is 20 - 25 mWm-2. This value is consistent with the lack of indication from xenolith data that lithosphere thickness changes with age and with theoretical models of mantle convection.
T22B-08
Radiogenic Heat Generation and the Evolution of Continental Crust During the First Half of Earth History
Reports during the past few years that many early Proterozoic terrains are reworked Archean crust support the interpretation by Morgan (1985, JGR, v 90 Suppl., C561-C570) that "low" heat flow in modern Archean cratons is a manifestation of selective survival of Archean continental crust with low intrinsic radiogenic heat production. The importance of intrinsic crustal radiogenic heat production in the reworking of early continental crust has been extended to explain the lack of preservation of crust from the first 550 Ma of Earth history (the Hadean: Morgan, 2006, Eos, Trans. AGU, 87 (52), Fall Meet. Suppl., Abstract VC11D-0630). Low crustal heat production implies thicker lithosphere relative to higher crustal heat production and does not preclude the recording of chemical evolution of the mantle lithosphere through time, as indicated by xenolith data. However, the thick lithosphere models do not explain the generally low measured heat productions values in Archean cratons, nor the statistically chemically distinct composition of Archean upper crust, including low U, Th, and K (for this discussion the Archean-Proterozoic boundary is taken as a transition from ~2.2 to ~2.8 Ga based on general crustal characteristics rather than a globally synchronous chronologically boundary at 2.5 Ga). Based on radiogenic heat production arguments alone, global heat loss must have been higher in the Archean than at present. Although the partitioning between continental and oceanic heat flow in the Archean cannot be assumed, any mechanism that assumes more efficient oceanic heat loss in the Archean requires higher asthenospheric temperatures, which are likely to have affected continental as well as oceanic lithosphere. Assuming that radiogenic heat producing elements have remained essentially immobile in Archean cratonic crust since the stabilization of these cratons, with the exponential decay of radioactive decay with time, ancient geotherms should have been hotter than modern geotherms. This conclusion suggests that xenoliths do not record ancient geotherms, nor do they constrain the thickness of ancient continental lithosphere.