U42B-01 INVITED
Silicate weathering and dry vs. wet runaway greenhouse scenarios
One of the key habitability crises faced by a rocky or icy planet is the possibility of a runaway greenhouse. While the runaway greenhouse threshold is primarily governed by water vapor thermodynamics and radiative properties, the extent of irreversible loss of water can be strongly affected by the amount of CO2 that builds up in the atmosphere. In the "wet runaway" scenario, liquid water persists at the planet's surface, which means that if silicates are present in the planet's crust, there is the possibility of CO2 drawdown due to silicate weathering. I will discuss the problem of silicate weathering on a hot wet-runaway planet, and the factors governing the amount of CO2 that remains in the atmosphere. If large amounts of CO2 remain in the atmosphere, the affect on the cold trap concentration can strongly inhibit water loss. Recent planetary formation calculations suggest that rocky planets can form with a much greater water inventory than Earth, in which case it is not clear that exposed subaerial silicates would exist in sufficient quantity to permit conventional subaereal weathering. In this case, however, it is still possible that submarine weathering of ocean floor could limit the amount of CO2 remaining in the atmosphere. Geochemical calculations originally developed for study of the weathering cycle during a Snowball Earth will be used to estimate the importance of this process. Some remarks on the evolution pathways of icy moons and larger icy waterworlds, lacking significant silicates, will also be presented.
U42B-02 INVITED
Coupled Planetary Reservoirs
We can look beyond the Earth, to Venus and Mars, to find opportunities to understand interactions among crust, mantle, hydrosphere, and atmosphere reservoirs. There has obviously been coupling among some of these reservoirs on other worlds, and in some cases feedback may have been in play but that is more difficult to demonstrate. The massive CO2 atmosphere of Venus has likely fluctuated significantly over its history due to exchange with other reservoirs, with attendant greenhouse effects strongly modulating surface temperature. Additionally, release of H2O and SO2 from large-scale magmatic events may have led to significant surface temperature increases, ΔT0, and the details depend on the competition between IR radiation warming and planetary albedo increase due to cloud formation. Diffusion of Δ T0 into the shallow crust may be responsible for the rapid global formation of compressional wrinkle ridges following widespread volcanic resurfacing [Solomon et al., 1999]. Diffusion of ΔT0 into the venusian upper mantle could have increased the rate of partial melting. The accompanying increase in volatile release to the atmosphere could set up a positive feedback because of increased greenhouse warming diffusing into the planet's interior [Phillips et al., 2001, Venus]. Another outcome of deep penetration of a greenhouse-induced positive ΔT0 is the lowering of mantle viscosity and an accompanying decrease in convective stress, which could shut down an exisiting lithospheric recycling regime [Lenardic et al., 2008]. Mars offers a rich set of possibilities for coupling between reservoirs [Jakosky and Phillips, 2001]. Magmatism at the massive Tharsis volcanic complex possibly induced episodic climate changes in the latter part of the Noachian era (~3.6-4.2 Ga). This could have led to clement conditions, forming valley networks that follow a regional slope caused partly by the mass load of Tharsis itself [Phillips et al., 2001, Mars]. Earlier in the Noachian, lithospheric recycling may have ended as surface temperatures warmed due to planetary outgassing, leading to the Lenardic catastrophe. Outer core convection may have been muted as the mantle heated up, shutting down the martian dynamo [Nimmo and Stevenson, 2000]. In turn, much of the extant atmosphere could have been lost to space, no longer protected from the solar wind by a global magnetic field.
U42B-03 INVITED
Coupling the volcanic and atmospheric evolution of Earth and Venus to their long-term tectonic state
Despite their superficial similarities, Venus and Earth's atmospheric evolution have diverged significantly. Without significant CO2 sinks, ongoing volcanism has resulted in the build up of extremely high CO2 concentrations, which have contributed to the dehydration of Venus's surface, and perhaps also the cessation of plate tectonics on Venus, either due to dry faults or surface temperatures. Its degassing history is to some extant recorded in its atmospheric argon signatures. Nonradiogenic Argon-36 is ~80 times that off Earth. Given most 36-Ar is primordial, this suggests very different initial atmospheric conditions for the two planets, with Venus retaining most of its initial atmosphere due to its fortuitous impact history. On the other hand, the deficit of radiogenic Argon-40 (~24 percent escaped from the mantle, compared with ~52 percent for Earth), hints at a very different volcanic and tectonic history, particularly in its deepest past. Recent convection modelling has shown that plate tectonic regimes break down under hot mantle conditions, due to a partial decoupling of stress from the less viscous mantle to the plates, resulting in insufficient stress for plate boundary deformation. This sends the system into an "episodic overturn" regime - similar to that which has been proposed for Venus today - where long periods of stagnant lid convection are interrupted by periods of massive lid recycling and overturn. Conversely, a planet in an episodic regime may transit into a stagnant lid regime for hotter mantle conditions. We couple a model for production and degassing of radiogenic Ar-40 from the mantle, and couple it with evolutionary models for stagnant, episodic and mobile lid tectonics. Earth's deficit in radiogenic Ar-40 may in some part be due to the different degassing efficiency of episodic convection in a hot early Earth. If Venus was stagnant for a large portion of its early evolution, the cumulative degassing efficiency of Ar-40 would be much lower than Earth, providing an explanation for Venus's very low atmospheric Ar-40 concentrations.
U42B-04 INVITED
Biological modulation of tectonics
Photosynthesis has had geologic consequences over the Earth's history. In addition to modifying Earth's atmosphere and ocean chemistry, it has also modulated tectonic processes through enhanced weathering and modification of the nature and composition of sedimentary rocks within fold mountain belts and convergent margins. Molecular biological studies indicate that bacterial photosynthesis evolved just once and that most bacterial clades descend from this photosynthetic common ancestor. Iron-based photosynthesis (ideally 4FeO + CO2 + H2O = 2Fe2O3 + CH2O) was the most bountiful anoxygenic niche on land. The back reaction provided energy to heterotrophic microbes and returned FeO to the photosynthetic microbes. Bacterial land colonists evolved into ecosystems that effectively weathered FeO-bearing minerals and volcanic glass. Clays, sands, and dissolved cations from the weathering process entered the ocean and formed our familiar classes sedimentary rocks: shales, sandstones, and carbonates. Marine photosynthesis caused organic carbon to accumulate in black shales. In contrast, non-photosynthetic ecosystems do not cause organic carbon to accumulate in shale. These evolutionary events occurred before 3.8 Ga as black shales are among the oldest rock types (Rosing and Frei, Earth Planet. Sci. Lett. 217, 237–244, 2004). Thick sedimentary sequences deformed into fold mountain belts. They remelted at depth to form granitic rocks (Rosing et al., Palaeoclimatol. Palaeoecol. 232, 99–11, 2006). Regions of outcropping low-FeO rocks including granites, quartzites, and some shales were a direct result. This dearth of FeO favored the evolution of oxic photosynthesis of cyanobacteria from photosynthetic soil bacteria. Black shales have an additional modulation effect on tectonics as they concentrate radioactive elements, particularly uranium (e.g. so that the surface heat flow varies by a factor of ca. 2). Thick sequences of black shales at continental rises of passive margins are trapped in collisional orogens. Radioactive heat production maintains high heat flow making these regions weak, and increasing the relative buoyancy among adjacent structural units as well as reaction progress of dehydration and decarbonation reactions. Mobile belts formed at the expense of black shales in continental interiors, similarly have high heat flow and weak lithosphere. They are preferential sites of subsequent continental break-up and thrusting.
U42B-05
Late-Archean continental emergence: consequences for the rise of atmospheric oxygen
The balance between the secular cooling of the Earth's mantle and the growth of the continental crust implies
changes in the isostatic equilibrium between continents and oceans, in the oceanic bathymetry, and in the
area of emerged continental crust. The evolution of the latter is of fundamental importance to the
geochemical coupling between the continental crust, the atmosphere and the oceans. The area of emerged
land can be estimated from models that depend on mantle temperature, continental area and continental
hypsometry.
In the Archean, the mantle was probably 150-200°C hotter than present and the continental area
could have increased from 20% of present at ~~3.5Ga to 80% of present by ~~2.5Ga. Using
these values, and comparing different thermal evolution models for the Earth, we calculate that the area of
emerged continental crust would be reduced to 1-12% of the Earth's area during the Archean (compared to
27.5% for present-day Earth). As for the continental hypsometry, a greater radiogenic crustal heat
production and a greater mantle heat flow would have reduced the strength of the continental lithosphere in
the Archean, thus limiting the crustal thickening due to mountain building processes and the maximum
elevation in the Earth's topography [Rey and Coltice, Geology 36, 635-638 (2008)]. Taking this into account,
we show that the continents were mostly flooded until the end of the Archean and that less than 3% of the
Earth's area (which is roughly the superficy of South America) consisted of emerged continental crust by
~~2.5~Ga. These results are consistent with widespread Archean submarine continental flood basalts,
and with the emergence of a sialic geochemical reservoir recorded from ~~2.5~Ga in (a) the
composition of shales, (b) the isotopic ratio 87Sr/86Sr of marine carbonates and (c) the
δ18O signature of igneous zircons.
The progressive emergence of the continents as shown by our models from the late-Archean onward had
major implications for the Earth's environment and for the evolution of early life. It contributed to the
oxygenation of the atmosphere as it would result in (a) the reduction of the proportion of submarine LIPs, a
major sink of oxygen [Kump and Barley, Nature 448 (2007)], (b) an increase in the weathering of emerged
mafic material, a major sink of carbon dioxide, and (c) an increase of the release of nutrients such as iron
and phosphorus into the oceans, increasing the activity of oxygen-producing cyanobacteria.
U42B-06
Continental fragmentation and the strontium isotopic evolution of seawater.
The time evolution of the strontium isotopic composition of seawater over the last 600 million years has the form of an asymmetric trough. The values are highest in the Cambrian and recent and lowest in the Jurassic. Superimposed on this trend are a number of smaller oscillations. The mechanisms responsible for these global isotopic fluctuations are subject to much debates. In order to get a quantitative picture of the changing paleogeography, we have characterized land-ocean distributions over Late Proterozoic to Phanerozoic times from measurement of perimeters and areas of continental fragments, based on paleomagnetic reconstructions. These measurements served to calculate geophysically constrainted breakup and scatter indexes of continental land masses from 0 to 1100 Ma (Cogne and Humler, 2008). Both parameters (strontium isotopic ratios of seawater and continental fragmentation indexes) are obviously highly correlated during the last 600 Ma. Low continental dispersion (that is large continental land masses) are associated with low seawater strontium isotopic ratios (that is when the continental inputs to oceans are minimum) and high continental dispersion (that is relatively small and widely distributed continents) with high seawater strontium isotopic ratios (that is when the continental input to ocean is maximum). Importantly, this first order evolution appears to conflict with the common idea of mountains erosion as a source for radiogenic strontium to oceans because high strontium isotopic ratios in seawater correspond to period of maximum dispersion of continents and not with period of general collisions. At first glance, it would seem that continental erosion increases with the degree of continental dispersion. Models showing that continental precipitation increases when continental masses are smaller and more widely dispersed and/or the length of continental margins available for rivers to carry continental material to oceans are thus favoured in order to resolve the paradox.
U42B-07
Feedback Between Volcanism and Milankovitch Cycles
Deglaciation is known to induce volcanism in many regions, notably Iceland. Since volcanism contributes CO2 to the atmosphere, we have investigated the global extent of glacially induced magmatism, and whether such volcanism may contribute to the co-variation between atmospheric CO2 and glacial cycles over the course of the late Pleistocene. Investigation of two combined global data sets on dated eruptions shows that global frequency of subaerial volcanic events increases substantially between 12Ka and 7Ka. An important aspect of the data is the temporal bias. While the record extends to 40,000 years, 80% of dated eruptions occur in the last 1000 years. The observation appears robust despite the temporal bias because it is apparent when comparing the 12-7Ka data with both older and younger time intervals. Application of a correction for the temporal bias indicates that global volcanic activity increases by a factor of 3-5 during this time interval. Increased volcanism can be confidently linked to deglaciation both in terms of location and mechanism. All of the increase occurs in regions thought to have experienced significant deglaciation, that is, predominantly high latitude or high elevation regions with significant precipitation. We show that two possible mechanisms may be important. As previously known from Iceland, deglaciation leads to decompression melting of the underlying mantle, yielding more magmatic input to volcanoes beneath ablating ice. In addition, pacing of volcanic eruptions reflects a balance between the forces generated by melt production and degassing, and the confining pressure and integrity of the surrounding rocks. A simple pacing model based on an eruption threshold and observed power law behavior of eruption frequency with number of volcanoes shows that ice volumes changes (estimated from oceanic δ18O) could also cluster the timing of volcanic eruptions near times of deglaciation. The influence of deglaciation upon volcanism is thus physically viable. To assess whether volcanism feeds back upon deglaciation through atmospheric CO2 we use data from arc volcanism to estimate primary CO2 contents of arc magmas, and a Monte Carlo simulation of the resulting atmospheric CO2 concentrations, accounting for the uncertainties in the eruption frequency analysis and model parameters for CO2 uptake by the ocean. The increased volcanism leads to a 30-80ppm increase in atmospheric CO2, and can account for about half of the rise in CO2 associated with the deglaciation. Effects from oceanic mechanisms remain important, particularly for the initial CO2 rise from 17-12Ka. The volcanic influence thus appears to be a positive feedback that may contribute to the sawtooth pattern of Milankovitch cycles.
U42B-08
Andean growth and the deceleration of South American subduction: Time evolution of a coupled orogen-subduction system
Present-day orography at the Andean margin is a result of isostasy, tectonic accretion, and erosional processes. The resulting excess mass of the Andes gives rise to frictional stresses on the seismogenic plate interface that resist the sinking of the subducting slab into the upper mantle. Thus, subduction rates should be sensitive to the time-dependent dynamics of a back-arc orogen, as well as erosional or accretional processes that affect orogen growth. Here we develop a two-dimensional coupled orogen-slab model that allows for the prediction of orogen size and plate motion in response to both tectonic and erosional forcing. We find that the frictional force exerted by the orogen on the subducting slab grows quadratically with orogen width and that the frictional resistance typically balances 10-50% of the slab pull force. The time evolution of the coupled orogen-subduction zone system is largely controlled by the rate of orogen growth, which is controlled by the rate of convergence and the erosivity of the climate state. In the case of the Andean margin, our models show that Miocene aridification leads to reduced erosion, increased orogen growth, greater frictional resistance to subduction, and, ultimately, to a ~50 reduction in the convergence rate between the Nazca and South American plates.