1. The Evolution of Earth’s Atmosphere and Surface James F. Kasting Department of Geosciences Penn State University

2. Brief review of theories of early atmospheric composition 1.Oparin (1924, 1938), Haldane (1929) Atmosphere was a strongly reduced mixture of CH 4, NH 3, H 2, and H 2 O Good from an origin of life standpoint Belief in this model was greatly strengthened by the success of the Miller-Urey experiments (early 1950’s): spark discharge in such a mixture created many organic compounds, including amino acids Image from Wikkipedia

3. The giant planets are not good analogs for the early terrestrial atmosphere, as Urey had argued CH 4 and (especially) NH 3 can be photolyzed and converted into other gases Modern (surface) volcanic gases are relatively oxidized, as nicely documented in books and papers by Dick Holland. The major constituents, in decreasing order of abundance, are H 2 O, CO 2, CO, H 2, and sometimes SO 2 and H 2 S. Little or no CH 4 is released from surface volcanoes. –And there is now good evidence that the redox state of the upper mantle has not changed with time Problems with the old model

4. Brief review of theories of early atmospheric composition 2.Rubey (1955), Walker (1977), Kasting (1993) The atmosphere was a weakly reduced mixture of N 2, CO 2, and smaller amounts of H 2 and CO The H 2 concentration was determined by the balance between outgassing from volcanoes and escape to space Hydrogen was assumed to escape at the diffusion- limited rate J. F. Kasting, Science (1993)

5. Diffusion-limited escape Hydrogen escape can be limited either at the homopause (~100 km altitude) or at the exobase (~500 km altitude) On Earth today, H escape is limited by diffusion through the homopause This may not be true, however, in hydrogen- rich primitive atmospheres 100 km Homopause 500 km Exobase

6. Slow hydrodynamic escape? If hydrogen escape was limited by energy rather than by diffusion, then the escape rate could have been 10-100 times slower than the maximum rate This could have helped the atmosphere remain highly reduced for a long time But, the problem is complex because the young Sun should have been much more active, leading to greater upper atmosphere heating F. Tian et al., Science (2005) Hydrogen escape rate (cm -2 s -1 ) Homopause hydrogen mixing ratio Diffusion limit

7. Solar EUV flux vs. time  Studies of nearby young solar- type stars give estimates for the time history of short- wavelength solar emissions (CoRoT press release drawing of young solar-analog star) I. Ribas et al., Ap. J. (2005) Young stars rotate faster, are more magnetically active, and hence have greatly enhanced EUV and XUV emissions 

8. Estimated exospheric temperatures on the early Earth Helmut Lammer’s group has worked on this problem. They calculate high exospheric temperatures on the early Earth (Kulikov et al., Space Sci. Rev., 2006) –Solar XUV flux was 100 times greater at 4.5 Ga, 10 times greater at 3.8 Ga –Based on observations of young, solar-type stars –Earth’s atmosphere would be swept away by the solar wind if not for the presence of a magnetic field –But, these atmospheres were hydrostatic… Bauer and Lammer, Planetary Aeronomy (2004)

9. Earth’s present thermosphere at different XUV fluxes: hydrodynamic model Work in progress with Feng Tian Physics packages from the NCAR TIEGCM Atmosphere is dynamic, so as it expands into space, it cools adiabatically Still needs work to accurately model H 2 -rich atmospheres F. Tian et al., JGR Planets (2008)  EUV --Enhanced solar EUV flux --Present Earth atmosphere composition

10. An aside: Thermal escape of heavy gases (C and O) from early Mars High solar EUV flux from the young Sun, combined with low gravity, leads to fast thermal escape of C and O If most CO 2 is lost by photodissociation and escape, as opposed to formation of carbonates, then the atmosphere can become oxidized Helps explain early Mars’ climate --NO 2 and O 3 provide additional warming (by lowering the albedo) Could provide another “false positive” for life

11. Back to Earth… Impacts may also have helped shape the early environment Artist’s rendition of the K-T impact (http://commons.wikimedia.org/wiki/ File:KT-impact.gif)

12. The late heavy bombardment The Moon, Mars, and Mercury are all covered with impact craters These impacts must have postdated the oldest Moon rock, which has an age of ~4.44 Ga Many of the Apollo moon rocks have ages of 3.8-3.9 Ga 1200 km Mare Imbrium

13. Mare Orientale Located just over the western limb (on the lunar farside) Crater is 930 km in diameter Requires a 100-km diameter impactor The Earth must have been getting hit with objects at least this big Photograph from NASA Lunar Orbiter 4 (1967)

14. $64,000 question: Was there a “pulse” of bombardment at ~3.9 Ga, or was there a 700-m.y. exponential tail to accretion? Probable answer: Yes, i.e., both of these are true

15. The Nice model A new theory of outer Solar System evolution can explain the clustering of lunar age dates near 3.9 Ga Current orbital periods: Jupiter – 11.8 yrs Saturn – 29.4 yrs In the Nice model, Saturn starts closer in, and its orbital period is less than twice that of Jupiter http://images.spaceref.com/news/2006/ 1196_web.l.jpg

16. Orbital evolution of the giant planets Simulated evolution of outer planets interacting with a ‘hot’ disk of planetesimals Uranus and Neptune flip positions as Jupiter and Saturn pass through the 2:1 mean motion resonance! Tsiganis et al., Nature, 2005, Fig. 1 Neptune Uranus Saturn Jupiter

17. Implications of the Nice model Asteroids would be continually kicked into new orbits (some of them Earth-crossing) as Jupiter’s mean motion resonances swept through the asteroid belt –Helps to explain asteroid belt clearing –This would give rise to more-or-less continuous bombardment in the inner Solar System Comets would be strongly perturbed when Uranus and Neptune make their move –This may have created a pulse of impacts near 3.8-3.9 Ga Impacts involving both types of objects would have influenced atmospheric composition (creating CO and CH 4 ) and may have frustrated the early origin of life

18. What was Earth’s surface like prior to 4.0 Ga? –Atmosphere was probably weakly reduced, assuming that the impact flux was not too large –No evidence for atmospheric O 2 until about 2.4 Ga (despite continued claims to the contrary) –Climate is a big question: Was it hot, or were temperatures relatively cool? –Information about early climates can, in principle be derived from O isotopes in various types of sedimentary rocks

19. after Bassinot et al. 1994 O isotopes—the last 900 k.y. O isotopes are measured in carbonate minerals in deep sea drill cores Low  18 O values correspond to warm temperatures

20. Marine carbonate  18 O vs. time Shields & Veizer, G 3, 2002 Warm Cold On longer time scales, the O isotope composition of carbonates exhibits a steep trend Taken at face value, a decrease of 10‰ in  18 O of carbonates corresponds to a temperature increase of ~54 o C, putting the Archean Earth at roughly 70 o C! Time 

21.  18 O of modern and ancient cherts (SiO 2 ) P. Knauth, Paleo 3 219, 53 (2005) Warm Cold Cherts, which are better preserved, tend to show the same trend, i.e., they become increasingly depleted in 18 O as they get older (SMOW)  Time

22. Chert data: Mean surface temperature was 70  15 o C at 3.3 Ga –Ref.: Knauth and Lowe, GSA Bull., 2003 Carbonate data: Surface temperatures remain significantly elevated until as recently as the early Devonian (~400 Ma) These arguments have been supported by evidence from silicon isotopes (Robert & Chaussidon, Nature, 2006) and from molecular biology (Gaucher et al., Nature, 2008)

23. I don’t believe this, however, because there were glaciations at ~2.4 Ga and 2.9 Ga

24. Geologic time Warm (?) Rise of atmospheric O 2 First shelly fossils (Cambrian explosion) Snowball Earth ice ages Warm Ice age Origin of life (Ice age) Loss of methane greenhouse

25. The faint young Sun problem Kasting et al., Scientific American (1988) Furthermore, the Sun was less bright back at that time, making it even more difficult to make the early Earth hot

26. CH 4 /CO 2 /C 2 H 6 greenhouse with haze One can compensate for low solar luminosity by having higher concentrations of greenhouse gases, specifically, CO 2 and CH 4 Warming by CH 4 is limited, however, by the formation of organic haze Hence, one needs very high CO 2 in order to make the early Earth hot  Late Archean Earth? Water freezes Paleosols J. Haqq-Misra et al., Astrobiology (2008)

27. Surface temperature vs. pCO 2 and fCH 4 Kasting and Howard, Phil. Trans. Roy. Soc. B (2006) 3.3 Ga S/S o = 0.77 Getting surface temperatures of ~70 o C at 3.3 Ga would require of the order of 3 bars of CO 2 (10,000 times present) pH of rainwater would drop from 5.7 to ~3.7 Would this be consistent with evidence of paleoweathering? Probably not…

28. Possible explanations for the O (and Si) isotope data The oxygen isotope composition of seawater may have changed with time –Kasting et al., EPSL, 2006 The cherts may have been formed by mixing of warm, hydrothermal fluids with cold seawater on the ocean floor –Hofmann et al., Precambrian Res., 2005; van den Boorn et al., Geology, 2007 Seawater isotopic composition is thought to be controlled by water- rock interactions within the midocean ridge hydrothermal circulation systems (Muehlenbachs and Clayton, 1976)

29. Conclusions The Earth’s atmosphere prior to 4.0 Ga was probably weakly reduced (mostly CO 2 + N 2 ). Some CH 4, however, was almost certainly present, especially after the origin of life –O 2 was not present until about 2.4 Ga The early surface environment of the Earth, along with climate and atmospheric composition, depends critically on the impact history of the early Solar System –The Nice model suggests that the heavy bombardment included a pulse at 3.8-3.9 Ga, along with an extended tail to accretion –Going back to the Moon with people may allow us to determine if this is really true Despite the apparent congruity between isotopic and biological data that support the hot early Earth hypothesis, it seems unlikely that this was actually the case