The postimpact Earth model

Modelling shows that the giant impact had a major effect on the evolution of the planet. It is expected that the distribution of the impact energy was highly heterogeneous in space and time. The sizes of the target and the impactor imply that both already had iron cores (Section 12.5). The energy released in this process heated the impactor core up to many thousands of degrees, causing it to melt and partially evaporate (the latter process is the most heat consuming) and the temperature of the Earth's mantle could have risen to more than 3000 K, which is above the melting temperature of silicate rocks even at high pressures. Models of this complicated process and their results have been discussed in the scientific literature for more than 20 years (Benz et al., 1986, 1987, 1989; Stevenson, 1987; Cameron, 2000, 2001b; Canup and Asphaug, 2001; Canup, 2004).

Furthermore, the modelling tells us that a huge crater, almost covering the facing overheated hemisphere, would be formed and rocky material thrown out to form a disk around the proto-Earth (Melosh et al., 1993). Part of the ejected material, mainly the silicate mantle of the trailing hemisphere of the impactor, would have escaped beyond the Roche limit mostly as partially molten clumps. The Moon would have been formed from these bodies soon afterwards, probably within ~ 1000 years (Pritchard and Stevenson, 2000).

Vaporization, causing the severe loss of volatile elements, occurred at the ejection stage, when a layer of hot, mainly silicate, vapour flowed around the proto-Earth and covered it. This dense, hot, thick atmosphere also received energy from the re-accretion of ejected material. The temperature increased with height, thus preventing convective cooling and extending the lifetime of the silicate atmosphere. Before the silicate component of this atmosphere condensed and fell back to Earth, the high temperatures in its upper reaches would have promoted the thermal escape of even the heavy elements (Benz and Cameron, 1990; Cameron and Benz, 1991).

The fate of the initial atmosphere during the giant impact is not clear (Melosh, 2003). Some models envisage its almost total loss by escape of the impact plume, antipodal shock or thermal (and hydrodynamic) escape (Cameron, 1983; Cameron and Benz, 1991; Ahrens, 1993; Chen and Ahrens, 1997). Genda and Abe (2003) revised the antipodal-shock escape model of Chen and Ahrens and concluded that this is not an efficient mechanism for escape. Thermal and hydrodynamic outflow during the hot stage of the atmosphere remains likely, however. Below are several observations pointing to extensive loss of the early terrestrial atmosphere, some of which are discussed in more detail in Chapter 20.

(1) The noble-gas abundances of the Venusian atmosphere exceed the terrestrial abundances by a factor ~ 100 (Fig. 20.1). The deficit of non-radiogenic noble gases (e.g.36 Ar) in the Earth is recorded by the 40Ar/36Ar ratio, which equals 295.5 in the Earth's atmosphere and only ~ 1 in the Venusian atmosphere (Istomin et al., 1983). Even though the K abundance in Venus is considered to be somewhat lower than that in Earth (by a factor ~ 2) and degassing models for Venus predict better retention of 40 Ar (again by a factor ~ 2; e.g. Kaula, 1999), the low 40 Ar/36Ar ratio still indicates a roughly one-hundred-fold deficit of terrestrial36 Ar. The simplest interpretation of this discrepancy is that in the course of accretion the twin planets acquired similar amounts of noble gases. However, the Earth lost its initial atmosphere mainly during the giant impact and (partially) retained a post-impact supply only, whereas Venus, which most probably did not suffer a giant collision (as deduced from e.g. its slow rotation and the absence of a satellite, Fig. 14.1), has better preserved its noble gases. Following this interpretation, the similarity of the abundances of major volatile species, e.g. N2 or CO2, in the Earth and Venus can be explained by late (post-impact) delivery of a major part of the atmophile species to both planets (Section 20.4).

(2) The inventory of radiogenic xenon isotopes, e.g. 129xe(I), definitely shows that more than 90% (probably ~ 99%) of the total amount generated in the Earth via the decay of extinct 129I has been lost from the Earth-atmosphere system. xenon loss from the atmosphere is also recorded by the244 Pu-136Xe(Pu) systematics. The xenon loss could have been partially caused by the impact. Correspondingly, the loss of all lighter gases from the atmosphere appears to have been inevitable (Chapter 20).

(3) After the giant impact, it is thought that the lower mantle solidified on the short time scale of ~ 103 yr, but a transient ~ 500-km-thick magma ocean could have survived for a longer time. This depends crucially on the rate of heat loss from the Earth. A thick early atmosphere would retard heat loss and slow down convection (Section 17.5). It has been argued that this negative feedback could extend the lifetime of a magma ocean up to ~ 108 yr and even longer without any heat supply (Abe, 1993; Solomatov, 2000). An irreversible fractionation of the magma ocean and the formation of a thick buoyant crust would appear to be inevitable consequences of such a development. If the giant impact caused the loss of the early atmosphere, then vigorous convection (Section 17.5) could have prevented mantle fractionation and formation of the crust and would also have led to fast mantle cooling and magma-ocean solidification within ~ 106 yr or less (Tonks and Melosh, 1990). The Earth's mantle and crustal compositions suggest that this is a much more likely scenario (Chapter 17).

Post-giant-collision accretion proceeded via interactions of the Earth and small bodies, which served to reduce the planet's eccentricity and spin rate (Agnor et al., 1999). Metal-silicate fractionation in transient magma seas, induced by these impacts, could have caused the final stages of core formation, characterized by low metal-silicate ratios in the mantle. Therefore it is expected that moderately siderophile elements would have been left in the silicate fraction (Azbel et al., 1993; Kramers, 1998). Even though such late impacts contributed additional energy they also stripped the outer boundary layer, allowing melts to reach the planetary surface and thus enhancing heat dissipation and enabling cooling of the planetary interior (Pritchard and Stevenson, 2000).

0 0

Post a comment