We present an analysis of K2 light curves (LCs) for candidate members of the young Upper Sco (USco) association (∼8 Myr) and the neighboring ρ Oph embedded cluster (∼1 Myr). We establish ∼1300 stars as probable members, ∼80% of which are periodic. The phased LCs have a variety of shapes which can be attributed to physical causes ranging from stellar pulsation and stellar rotation to disk-related phenomena. We identify and discuss a number of observed behaviors. The periods are ∼0.2-30 days with a peak near 2 days and the rapid period end nearing breakup velocity. M stars in the young USco region rotate systematically faster than GK stars, a pattern also present in K2 data for the older Pleiades and Praesepe systems. At higher masses (types FGK), the well-defined period-color relationship for slowly rotating stars seen in the Pleiades and Praesepe systems is not yet present in USco. Circumstellar disks are present predominantly among the more slowly rotating M stars in USco, with few disks in the subday rotators. However, M dwarfs with disks rotate faster on average than FGK systems with disks. For four of these disked M dwarfs, we provide direct evidence for disk locking based on the K2 LC morphologies. Our preliminary analysis shows a relatively mass-independent spin-up by a factor of ∼3.5 between USco and the Pleiades, then mass-dependent spin-down between Pleiades and Praesepe.

Theories of the formation and early evolution of planetary systems postulate that planets are born in circumstellar disks, and undergo radial migration during and after dissipation of the dust and gas disk from which they formed 1,2 . The precise ages of meteorites indicate that planetesimals -the building blocks of planets -are produced within the first million years of a star's life 3 . A prominent question is: how early can one find fully formed planets like those frequently detected on short orbital periods around mature stars? Some theories suggest the in situ formation of planets close to their host stars is unlikely and the existence of such planets is evidence for large scale migration 4,5 . Other theories posit that planet assembly at small orbital separations may be common [6][7][8] . Here we report on a newly-born, transiting planet orbiting its star every 5.4 days. The planet is 50% larger than Neptune, and its mass is less than 3.6 times Jupiter (at 99.7% confidence), with a true mass likely to be within a factor of several of Neptune's. The 5-10 million year old star has a tenuous dust disk extending in to about 2 times the Earth-Sun separation, in addition to the large planet located at less than 1/20 the Earth-Sun separation.USco 161014. 75-191909.3, hereafter K2-33, is a several million year old M-type star that was observed by NASA's Kepler Space Telescope during Campaign 2 of the K2 mission. The star was identified as one of more than 200 candidate planet hosts in a systematic search for transits in K2 data 9 . As part of our ongoing study of the pre-main sequence population of Upper Scorpius observed by K2, we independently verified and analyzed the planetary transit signal. We acquired radial velocity (RV) and high spatial resolution observations at the W. M. Keck Observatory to confirm the planet, hereafter K2-33b, and to measure its size and mass.2 Within the 77.5 day photometric time series of K2-33 (Kp = 14.3 mag), there are periodic dimmings of 0.23% lasting 4.2 hours and occurring every 5.4 d (Fig. 1). The ensemble of transits are detected at a combined signal-to-noise ratio of ≈ 32. During the K2 observations, cool, dark regions on the stellar surface (starspots) rotated in and out of view, producing semi-sinusoidal brightness variations of ∼3% peak-to-trough amplitude with a 6.3 ± 0.2 d periodicity (Extended Data Fig. 1). We removed the starspot variability prior to modeling the transit events. We fit the transit profiles using established methods 10, measuring the planet's size relative to its host star and its orbital geometry (Table 1).K2-33 is an established member of the Upper Scorpius OB association 11,12 , the nearest site , which we confirm from Keck spectra (Table 1). Furthermore, the stellar rotation rate we measure via broadening of absorption lines in the spectra and via the starspot period (Table 1), is rapid relative to field-age stars of similar mass 14 . We determined the star's systemic RV (Table 1) The inferred planet size and mass depend directly upon the host sta...

Exoplanets orbiting pre-main sequence stars are laboratories for studying planet evolution processes, including atmospheric loss, orbital migration, and radiative cooling. V1298 Tau, a young solar analog with an age of 23 ± 4 Myr, is one such laboratory. The star is already known to host a Jupiter-sized planet on a 24 day orbit. Here, we report the discovery of three additional planets -all between the size of Neptune and Saturn -based on our analysis of K2 Campaign 4 photometry. Planets c and d have sizes of 5.6 and 6.4 R ⊕ , respectively and with orbital periods of 8.25 and 12.40 days reside 0.25% outside of the nominal 3:2 mean-motion resonance. Planet e is 8.7 R ⊕ in size but only transited once in the K2 time series and thus has a period longer than 36 days, but likely shorter than 223 days. The V1298 Tau system may be a precursor to the compact multiplanet systems found to be common by the Kepler mission. However, the large planet sizes stand in sharp contrast to the vast majority of Kepler multis which have planets smaller than 3 R ⊕ . Simple dynamical arguments suggest total masses of <28 M ⊕ and <120 M ⊕ for the c-d and d-b planet pairs, respectively. The implied low masses suggest that the planets may still be radiatively cooling and contracting, and perhaps losing atmosphere. The V1298 Tau system offers rich prospects for further follow-up including atmospheric characterization by transmission or eclipse spectroscopy, dynamical characterization 2 David et al.through transit-timing variations, and measurements of planet masses and obliquities by radial velocities.

Using K2 data, we identified 23 very-low-mass members of the ρ Oph and Upper Scorpius star-forming region as having periodic photometric variability not easily explained by well-established physical mechanisms such as star spots, eclipsing binaries, or pulsation. All of these unusual stars are mid-to-late M dwarfs without evidence of active accretion, and with photometric periods generally <1 day. Often the unusual light-curve signature takes the form of narrow flux dips; when we also have rotation periods from star spots, the two periods agree, suggesting that the flux dips are due to material orbiting the star at the Keplerian co-rotation radius. We sometimes see "statechanges" in the phased light-curve morphologies where ∼25% of the waveform changes shape on timescales less than a day; often, the "state-change" takes place immediately after a strong flare. For the group of stars with these sudden light-curve morphology shifts, we attribute their flux dips as most probably arising from eclipses of warm coronal gas clouds, analagous to the slingshot prominences postulated to explain transient Hα absorption features in AB Doradus and other rapidly rotating late-type stars. For another group of stars with somewhat longer periods, we find the short-duration flux dips to be highly variable on both short and long timescales, with generally asymmetric flux-dip profiles. We believe that these flux dips are due to particulate clouds possibly associated with a close-in planet or resulting from a recent collisional event.

Age determination is undertaken for nearby early-type (BAF) stars, which constitute attractive targets for high-contrast debris disk and planet imaging surveys. Our analysis sequence consists of: acquisition of uvbyβ photometry from catalogs, correction for the effects of extinction, interpolation of the photometry onto model atmosphere grids from which atmospheric parameters are determined, and finally, comparison to the theoretical isochrones from pre-main sequence through post-main sequence stellar evolution models, accounting for the effects of stellar rotation. We calibrate and validate our methods at the atmospheric parameter stage by comparing our results to fundamentally determined T eff and log g values. We validate and test our methods at the evolutionary model stage by comparing our results on ages to the accepted ages of several benchmark open clusters (IC 2602, α Persei, Pleiades, Hyades). Finally, we apply our methods to estimate stellar ages for 3493 field stars, including several with directly imaged exoplanet candidates.

The Upper Scorpius OB association is the nearest region of recent massive star formation and thus an important benchmark for investigations concerning stellar evolution and planet formation timescales. We present nine EBs in Upper Scorpius, three of which are newly reported here and all of which were discovered from K2 photometry. Joint fitting of the eclipse photometry and radial velocities from newly acquired Keck-I/HIRES spectra yields precise masses and radii for those systems that are spectroscopically double-lined. The binary orbital periods in our sample range from 0.6-100 days, with total masses ranging from 0.2-8 M . At least 33% of the EBs reside in hierarchical multiples, including two triples and one quadruple. We use these EBs to develop an empirical mass-radius relation for pre-main-sequence stars, and to evaluate the predictions of widely-used stellar evolutionary models. We report evidence for an age of 5-7 Myr which is self-consistent in the mass range of 0.3-5 M and based on the fundamentally-determined masses and radii of eclipsing binaries (EBs). Evolutionary models including the effects of magnetic fields imply an age of 9-10 Myr. Our results are consistent with previous studies that indicate many models systematically underestimate the masses of low-mass stars by 20-60% based on H-R diagram analyses. We also consider the dynamical states of several binaries and compare with expectations from tidal dissipation theories. Finally, we identify RIK 72 b as a long-period transiting brown dwarf (M = 59.2 ± 6.8 M Jup , R = 3.10 ± 0.31 R Jup , P ≈ 97.8 days) and an ideal benchmark for brown dwarf cooling models at 5-10 Myr.Meanwhile, theoretical models have evolved from simple hydrostatic contraction and basic nuclear reaction networks to including the effects of deuterium burning, proto-stellar and circumstellar disk accretion, realistic surface boundary conditions, convection, and magnetic fields or starspots (see D'Antona 2017, for a review). Our best method of evaluating such models is through detailed characterization of benchmark PMS stars allowing dynamical mass determinations from, e.g., circumstellar disk rotation curves or binary orbit determi-

We present the discovery and characterization of four low-mass ( < M 0.6  M ) eclipsing binary (EB) systems in the sub-Gyr old Praesepe open cluster using Kepler/K2 time series photometry and Keck/HIRES spectroscopy. We present a new Gaussian process EB model, GP-EBOP, as well as a method of simultaneously determining effective temperatures and distances for EBs. Three of the reported systems (AD 3814, AD 2615 and AD 1508) are detached and double-lined, and precise solutions are presented for the first two. We determine masses and radii to 1%-3% precision for AD 3814 and to 5%-6% for AD 2615. Together with effective temperatures determined to ∼50 K precision, we test the PARSEC v1.2 and BHAC15 stellar evolution models. Our EB parameters are more consistent with the PARSEC models, primarily because the BHAC15 temperature scale is hotter than our data over the mid-M-dwarf mass range probed. Both ADs 3814 and 2615, which have orbital periods of 6.0 and 11.6 days, are circularized but not synchronized. This suggests that either synchronization proceeds more slowly in fully convective stars than the theory of equilibrium tides predicts, or magnetic braking is currently playing a more important role than tidal forces in the spin evolution of these binaries. The fourth system (AD 3116) comprises a brown dwarf transiting a mid-M-dwarf, which is the first such system discovered in a sub-Gyr open cluster. Finally, these new discoveries increase the number of characterized EBs in sub-Gyr open clusters by 20% (40%) below < M 1.5 M e ( < M 0.6 M e ).

We report the discovery of three low-mass double-lined eclipsing binaries in the pre-main sequence Upper Scorpius association, revealed by K2 photometric monitoring of the region over ∼78 days. The orbital periods of all three systems are <5 days. We use the K2 photometry plus multiple Keck/HIRES radial velocities (RVs) and spectroscopic flux ratios to determine fundamental stellar parameters for both the primary and secondary components of each system, along with the orbital parameters. We present tentative evidence that EPIC 203868608 is a hierarchical triple system comprised of an eclipsing pair of ∼25 M Jup brown dwarfs with a wide M-type companion. If confirmed, it would constitute only the second double-lined eclipsing brown dwarf binary system discovered to date. The double-lined system EPIC 203710387 is composed of nearly identical M4.5-M5 stars with fundamentally determined masses and radii measured to better than 3% precision (M M 0.1183 0.0028from combination of the light curve and RV time series. These stars have the lowest masses of any stellar mass double-lined eclipsing binary to date. Comparing our derived stellar parameters with evolutionary models, we suggest an age of ∼10-11 Myr for this system, in contrast to the canonical age of 3-5 Myr for the association. Finally, EPIC 203476597 is a compact single-lined system with a G8-K0 primary and a likely mid-K secondary whose lines are revealed in spectral ratios. Continued measurement of RVs and spectroscopic flux ratios will better constrain fundamental parameters and should elevate the objects to benchmark status. We also present revised parameters for the double-lined eclipsing binary UScoCTIO 5 (M M 0.3336 0.0022, which are suggestive of a system age younger than previously reported. We discuss the implications of our results on these ∼0.1-1.5 M  stars for pre-main-sequence evolutionary models.