Jupiter and Saturn formed in a few million years (ref. 1) from a gas-dominated protoplanetary disk, and were susceptible to gas-driven migration of their orbits on timescales of only ∼100,000 years (ref. 2). Hydrodynamic simulations show that these giant planets can undergo a two-stage, inward-then-outward, migration. The terrestrial planets finished accreting much later, and their characteristics, including Mars' small mass, are best reproduced by starting from a planetesimal disk with an outer edge at about one astronomical unit from the Sun (1 au is the Earth-Sun distance). Here we report simulations of the early Solar System that show how the inward migration of Jupiter to 1.5 au, and its subsequent outward migration, lead to a planetesimal disk truncated at 1 au; the terrestrial planets then form from this disk over the next 30-50 million years, with an Earth/Mars mass ratio consistent with observations. Scattering by Jupiter initially empties but then repopulates the asteroid belt, with inner-belt bodies originating between 1 and 3 au and outer-belt bodies originating between and beyond the giant planets. This explains the significant compositional differences across the asteroid belt. The key aspect missing from previous models of terrestrial planet formation is the substantial radial migration of the giant planets, which suggests that their behaviour is more similar to that inferred for extrasolar planets than previously thought.

Exoplanetary transmission spectroscopy in the near-infrared using Hubble/NICMOS is currently ambiguous because different observational groups claim different results from the same data, depending on their analysis methodologies. Spatial scanning with Hubble/WFC3 provides an opportunity to resolve this ambiguity. We here report WFC3 spectroscopy of the giant planets HD 209458b and XO-1b in transit, using spatial scanning mode for maximum photon-collecting efficiency. We introduce an analysis technique that derives the exoplanetary transmission spectrum without the necessity of explicitly decorrelating instrumental effects, and achieves nearly photon-limited precision even at the high flux levels collected in spatial scan mode. Our errors are within 6% (XO-1) and 26% (HD 209458b) of the photon-limit at a resolving power of λ/δλ ∼ 70, and are better than 0.01% per spectral channel. Both planets exhibit water absorption of approximately 200 ppm at the water peak near 1.38 µm. Our result for XO-1b contradicts the much larger absorption derived from NICMOS spectroscopy. The weak water absorption we measure for HD 209458b is reminiscent of the weakness of sodium absorption in the first transmission spectroscopy of an exoplanet atmosphere by Charbonneau et al. (2002). Model atmospheres having uniformly-distributed extra opacity of 0.012 cm 2 g −1 account approximately for both our water measurement and the sodium absorption. Our results for HD 209458b support the picture advocated by Pont et al. (2013) in which weak molecular absorptions are superposed on a transmission spectrum that is dominated by continuous opacity due to haze and/or dust. However, the extra opacity needed for HD 209458b is grayer than for HD 189733b, with a weaker Rayleigh component.

Living systems produce more than 90% of Earth's atmospheric methane; the balance is of geochemical origin. On Mars, methane could be a signature of either origin. Using high-dispersion infrared spectrometers at three ground-based telescopes, we measured methane and water vapor simultaneously on Mars over several longitude intervals in northern early and late summer in 2003 and near the vernal equinox in 2006. When present, methane occurred in extended plumes, and the maxima of latitudinal profiles imply that the methane was released from discrete regions. In northern midsummer, the principal plume contained approximately 19,000 metric tons of methane, and the estimated source strength (>/=0.6 kilogram per second) was comparable to that of the massive hydrocarbon seep at Coal Oil Point in Santa Barbara, California.