Planets with sizes between those of Earth and Neptune divide into two populations: purely rocky bodies whose atmospheres contribute negligibly to their sizes, and larger gas-enveloped planets possessing voluminous and optically thick atmospheres. We
show that whether a planet forms rocky or gas-enveloped depends on the solid surface density of its parent disk. Assembly times for rocky cores are sensitive to disk solid surface density. Lower surface densities spawn smaller planetary embryos; to assemble a core of given mass, smaller embryos require more mergers between bodies farther apart and therefore exponentially longer formation times. Gas accretion simulations yield a rule of thumb that a rocky core must be at least 2$M_oplus$ before it can acquire a volumetrically significant atmosphere from its parent nebula. In disks of low solid surface density, cores of such mass appear only after the gas disk has dissipated, and so remain purely rocky. Higher surface density disks breed massive cores more quickly, within the gas disk lifetime, and so produce gas-enveloped planets. We test model predictions against observations, using planet radius as an observational proxy for gas-to-rock content and host star metallicity as a proxy for disk solid surface density. Theory can explain the observation that metal-rich stars host predominantly gas-enveloped planets.
Numerous telescopes and techniques have been used to find and study extrasolar planets, but none has been more successful than NASAs Kepler Space Telescope. Kepler has discovered the majority of known exoplanets, the smallest planets to orbit normal
stars, and the worlds most likely to be similar to our home planet. Most importantly, Kepler has provided our first look at typical characteristics of planets and planetary systems for planets with sizes as small as and orbits as large as those of the Earth.
We establish the three-dimensional architecture of the Kepler-419 (previously KOI-1474) system to be eccentric yet with a low mutual inclination. Kepler-419b is a warm Jupiter at semi-major axis a = 0.370 +0.007/-0.006 AU with a large eccentricity e=
0.85 +0.08/-0.07 measured via the photoeccentric effect. It exhibits transit timing variations induced by the non-transiting Kepler-419c, which we uniquely constrain to be a moderately eccentric (e=0.184 +/- 0.002), hierarchically-separated (a=1.68 +/- 0.03 AU) giant planet (7.3 +/- 0.4 MJup). We combine sixteen quarters of Kepler photometry, radial-velocity (RV) measurements from the HIgh Resolution Echelle Spectrometer (HIRES) on Keck, and improved stellar parameters that we derive from spectroscopy and asteroseismology. From the RVs, we measure the mass of inner planet to be 2.5+/-0.3MJup and confirm its photometrically-measured eccentricity, refining the value to e=0.83+/-0.01. The RV acceleration is consistent with the properties of the outer planet derived from TTVs. We find that, despite their sizable eccentricities, the planets are coplanar to within 9 +8/-6 degrees, and therefore the inner planets large eccentricity and close-in orbit are unlikely to be the result of Kozai migration. Moreover, even over many secular cycles, the inner planets periapse is most likely never small enough for tidal circularization. Finally, we present and measure a transit time and impact parameter from four simultaneous ground-based light curves from 1m-class telescopes, demonstrating the feasibility of ground-based follow-up of Kepler giant planets exhibiting large TTVs.
Gas giants orbiting interior to the ice line are thought to have been displaced from their formation locations by processes that remain debated. Here we uncover several new metallicity trends, which together may indicate that two competing mechanisms
deliver close-in giant planets: gentle disk migration, operating in environments with a range of metallicities, and violent planet-planet gravitational interactions, primarily triggered in metal-rich systems in which multiple giant planets can form. First, we show with 99.1% confidence that giant planets with semi-major axes between 0.1 and 1 AU orbiting metal-poor stars ([Fe/H]<0) are confined to lower eccentricities than those orbiting metal-rich stars. Second, we show with 93.3% confidence that eccentric proto-hot Jupiters undergoing tidal circularization primarily orbit metal-rich stars. Finally, we show that only metal-rich stars host a pile-up of hot Jupiters, helping account for the lack of such a pile-up in the overall Kepler sample. Migration caused by stellar perturbers (e.g. stellar Kozai) is unlikely to account for the trends. These trends further motivate follow-up theoretical work addressing which hot Jupiter migration theories can also produce the observed population of eccentric giant planets between 0.1 and 1 AU.
The exoplanets known as hot Jupiters---Jupiter-sized planets with periods less than 10 days---likely are relics of dynamical processes that shape all planetary system architectures. Socrates et al. (2012) argued that high eccentricity migration (HEM)
mechanisms proposed for situating these close-in planets should produce an observable population of highly eccentric proto-hot Jupiters that have not yet tidally circularized. HEM should also create failed-hot Jupiters, with periapses just beyond the influence of fast circularization. Using the technique we previously presented for measuring eccentricities from photometry (the photoeccentric effect), we are distilling a collection of eccentric proto- and failed-hot Jupiters from the Kepler Objects of Interest (KOI). Here we present the first, KOI-1474.01, which has a long orbital period (69.7340 days) and a large eccentricity e = 0.81+0.10/-0.07, skirting the proto-hot Jupiter boundary. Combining Kepler photometry, ground-based spectroscopy, and stellar evolution models, we characterize host KOI-1474 as a rapidly-rotating F-star. Statistical arguments reveal that the transiting candidate has a low false-positive probability of 3.1%. KOI-1474.01 also exhibits transit timing variations of order an hour. We explore characteristics of the third-body perturber, which is possibly the smoking-gun cause of KOI-1474.01s large eccentricity. Using the host-stars rotation period, radius, and projected rotational velocity, we find KOI-1474.01s orbit is marginally consistent with aligned with the stellar spin axis, although a reanalysis is warranted with future additional data. Finally, we discuss how the number and existence of proto-hot Jupiters will not only demonstrate that hot Jupiters migrate via HEM, but also shed light on the typical timescale for the mechanism.
We have detected transits of the innermost planet e orbiting 55 Cnc (V=6.0), based on two weeks of nearly continuous photometric monitoring with the MOST space telescope. The transits occur with the period (0.74 d) and phase that had been predicted b
y Dawson & Fabrycky, and with the expected duration and depth for the crossing of a Sun-like star by a hot super-Earth. Assuming the stars mass and radius to be 0.963_{-0.029}^{+0.051} M_sun and 0.943 +/- 0.010 R_sun, the planets mass, radius, and mean density are 8.63 +/- 0.35 Mearth, 2.00 +/- 0.14 Rearth, and 5.9_{-1.1}^{+1.5} g/cm^3. The mean density is comparable to that of Earth, despite the greater mass and consequently greater compression of the interior of 55 Cnc e. This suggests a rock-iron composition supplemented by a significant mass of water, gas, or other light elements. Outside of transits, we detected a sinusoidal signal resembling the expected signal due to the changing illuminated phase of the planet, but with a full range (168 +/- 70 ppm) too large to be reflected light or thermal emission. This signal has no straightforward interpretation and should be checked with further observations. The host star of 55 Cnc e is brighter than that of any other known transiting planet, which will facilitate future investigations.
