Prior models of lunar accretion from a disk of solid particles found that the Moon accretes in less than 1 yr (Ida et al. 1997; Kokubo et al. 2000). Such rapid formation precludes compositional equilibration of the Earth and protolunar disk, a possible explanation to the Earth and Moon compositional similarities, which requires ~100 yr (Pahlevan & Stevenson 2007). Also the Moon would form completely molten (Pritchard & Stevenson 2000), at odds with multiple observations suggesting that the Lunar Magma Ocean did not extend throughout the Moon’s interior, e.g. recent GRAIL observations believed to show a system of subsurface dikes formed as an early cool interior was warmed from above (Andrews-Hanna et al. 2012).
The protolunar disk, however, is actually a mixture of melt and vapor, and portions of it evolve much more slowly. In particular, interior to the Roche limit, Earth’s tidal forces prevent accretion, and gravitational instabilities in the melt layer instead produce a high viscosity, causing the disk to spread radially and dissipating enough energy to vaporize it. As melt vaporizes, the dissipation rate drops, so material re-cools and condenses. In this way the Roche-interior disk maintains a vapor-melt state in which the dissipation rate in the melt nearly balances the cooling rate of the vapor atmosphere (Thompson & Stevenson 1988; Ward 2012). This results in a ~100 yr timescale for the inner disk to spread and deliver material outward to the growing Moon, which may be long enough to yield both a partially molten Moon and equilibration. Thus, whether the Giant Impact model can be consistent with these constraints will both depend sensitively on the lunar accretion rate, which in turn depends critically on an accurate model of the Roche-interior disk.
We have developed a new lunar accretion model that includes a more accurate description of the inner protolunar disk, taking into account thermal processes that limit the disk’s evolution rate (Salmon and Canup 2012). Our model predicts a 3-phase accretion of the Moon, in which material initially orbiting in the outer disk accumulates rapidly, followed by a much slower evolution of the vapor/fluid inner disk and a final phase of accretion of inner disk material onto the Moon. The Moon’s total accretion then occurs over about 100 years, which could be compatible with equilibration and partial cooling of protolunar material.
We have coupled our accretion code to a simple model for the Moon’s initial thermal state. Using a 3D spherical grid to model the forming Moon, we estimate heating due to each accretionary event using the impactor properties predicted by our accretion model, and include cooling from the Moon’s surface. We use this model to estimate the Moon’s initial thermal state as a function of the temperature of accreting material and specific accretionary histories.