The Upsilon Andromedae system was the first multi-planet system discovered orbiting a main-sequence star. We describe the detection of water vapor in the atmosphere of the innermost non-transiting gas giant upsAndb by treating the star-planet system as a spectroscopic binary with high-resolution, ground-based spectroscopy. We resolve the signal of the planet's motion and break the mass-inclination degeneracy for this non-transiting planet via deep combined flux observations of the star and the planet. In total, seven epochs of Keck NIRSPEC L band observations, three epochs of Keck NIRSPEC short-wavelength K band observations, and three epochs of Keck NIRSPEC long wavelength K band observations of the upsAndsystem were obtained. We perform a multi-epoch cross-correlation of the full data set with an atmospheric model. We measure the radial projection of the Keplerian velocity (K P =55 ± 9kms M J ), and orbital inclination (i b 24°± 4°), and determine that the planet's opacity structure is dominated by water vapor at the probed wavelengths. Dynamical simulations of the planets in the upsAndsystem with these orbital elements for upsAndb show that stable, long-term (100Myr) orbital configurations exist. These measurements will inform future studies of the stability and evolution of the upsAndsystem, as well as the atmospheric structure and composition of the hot Jupiter.

We target the thermal emission spectrum of the non-transiting gas giant HD 88133 b with high-resolution nearinfrared spectroscopy, by treating the planet and its host star as a spectroscopic binary. For sufficiently deep summed flux observations of the star and planet across multiple epochs, it is possible to resolve the signal of the hot gas giant's atmosphere compared to the brighter stellar spectrum, at a level consistent with the aggregate shot noise of the full data set. To do this, we first perform a principal component analysis to remove the contribution of the Earth's atmosphere to the observed spectra. Then, we use a cross-correlation analysis to tease out the spectra of the host star and HD 88133 b to determine its orbit and identify key sources of atmospheric opacity. In total, six epochs of Keck NIRSPEC L-band observations and three epochs of Keck NIRSPEC K-band observations of the HD 88133 system were obtained. Based on an analysis of the maximum likelihood curves calculated from the multi-epoch cross-correlation of the full data set with two atmospheric models, we report the direct detection of the emission spectrum of the non-transiting exoplanet HD 88133 b and measure a radial projection of the Keplerian orbital velocity of 40±15 km s , and an atmosphere opacity structure at high dispersion dominated by water vapor. This, combined with 11 years of radial velocity measurements of the system, provides the most up-to-date ephemeris for HD 88133.

We describe the detection of water vapor in the atmosphere of the transiting hot Jupiter KELT-2Ab by treating the star-planet system as a spectroscopic binary with high-resolution, ground-based spectroscopy. We resolve the signal of the planet's motion with deep combined flux observations of the star and the planet. In total, six epochs of Keck NIRSPEC L-band observations were obtained, and the full data set was subjected to a cross correlation analysis with a grid of self-consistent atmospheric models. We measure a radial projection of the Keplerian velocity, K P , of 148 ± 7 km s −1 , consistent with transit measurements, and detect water vapor at 3.8σ. We combine NIRSPEC L-band data with Spitzer IRAC secondary eclipse data to further probe the metallicity and carbon-to-oxygen ratio of KELT-2Ab's atmosphere. While the NIRSPEC analysis provides few extra constraints on the Spitzer data, it does provide roughly the same constraints on metallicity and carbon-to-oxygen ratio. This bodes well for future investigations of the atmospheres of non-transiting hot Jupiters.

We report the 6.5σ detection of water from the hot Jupiter HD187123b with a Keplerian orbital velocity K p of 53±13 km s −1. This high-confidence detection is made using a multi-epoch, high-resolution, crosscorrelation technique, and corresponds to a planetary mass of-+ 1.4 0.3 0.5 M J and an orbital inclination of 21°±5°. The technique works by treating the planet/star system as a spectroscopic binary and obtaining high signal-tonoise, high-resolution observations at multiple points across the planet's orbit to constrain the system's binary dynamical motion. All together, seven epochs of Keck/NIRSPEC L-band observations were obtained, with five before the instrument upgrade and two after. Using high-resolution SCARLET planetary and PHOENIX stellar spectral models, we were able to drastically increase the confidence of the detection by running simulations that could reproduce, and thus remove, the nonrandom structured noise in the final likelihood space well. The ability to predict multi-epoch results will be extremely useful for furthering the technique. Here, we use these simulations to compare three different approaches to combining the cross correlations of high-resolution spectra and find that the Zucker log(L) approach is least affected by unwanted planet/star correlation for our HD187123 data set. Furthermore, we find that the same total signal-to-noise ratio (S/N) spread across an orbit in many, lower S/N epochs rather than fewer, higher S/N epochs could provide a more efficient detection. This work provides a necessary validation of multi-epoch simulations, which can be used to guide future observations and will be key to studying the atmospheres of farther separated, non-transiting exoplanets.

The mass and age of substellar objects are degenerate parameters leaving the evolutionary state of brown dwarfs ambiguous without additional information. Theoretical models are normally used to help distinguish between old, massive brown dwarfs and young, low-mass brown dwarfs but these models have yet to be properly calibrated. We have carried out an infrared high-contrast imaging program with the goal of detecting substellar objects as companions to nearby stars to help break degeneracies in inferred physical properties such as mass, age, and composition. Rather than using imaging observations alone, our targets are pre-selected based on the existence of dynamical accelerations informed from years of stellar radial velocity (RV) measurements. In this paper, we present the discovery of a rare benchmark brown dwarf orbiting the nearby (d = 18.69 ± 0.19 pc), solar-type (G9V) star HD4747 ([Fe/H]=−0.22 ± 0.04) with a projected separation of only ρ=11.3±0.2 au (θ = 0 6). Precise Doppler measurements taken over 18 years reveal the companion's orbit and allow us to place strong constraints on its mass using dynamics ( =  m i M sin 55.3 1.9 Jup ). Relative photometry (ΔK s = 9.05 ± 0.14, =  M 13.00 0.14 2.3 Gyr based on gyrochronology. Combining astrometric measurements with RV data, we calculate the companion dynamical mass ( =  m M 60.2 3.3 Jup ) and orbit (e = 0.740 ± 0.002) directly. As a new mass, age, and metallicity benchmark, HD4747B will serve as a laboratory for precision astrophysics to test theoretical models that describe the emergent radiation of brown dwarfs.

Surveys of nearby field stars indicate that stellar binaries are common, yet little is known about the effects that these companions may have on planet formation and evolution. The Friends of Hot Jupiters project uses three complementary techniques to search for stellar companions to known planet-hosting stars: radial velocity monitoring, adaptive optics imaging, and near-infrared spectroscopy. In this paper, we examine high-resolution K band infrared spectra of fifty stars hosting gas giant planets on short-period orbits. We use spectral fitting to search for blended lines due to the presence of cool stellar companions in the spectra of our target stars, where we are sensitive to companions with temperatures between 3500 and 5000 K and projected separations less than 100 AU in most systems. We identify eight systems with candidate low-mass companions, including one companion that was independently detected in our AO imaging survey. For systems with radial velocity accelerations, a spectroscopic non-detection rules out scenarios involving a stellar companion in a high inclination orbit. We use these data to place an upper limit on the stellar binary fraction at small projected separations, and show that the observed population of candidate companions is consistent with that of field stars and also with the population of wide-separation companions detected in our previous AO survey. We find no evidence that spectroscopic stellar companions are preferentially located in systems with short-period gas giant planets on eccentric and/or misaligned orbits.

No abstract

While high-resolution cross-correlation spectroscopy (HRCCS) techniques have proven effective at characterizing the atmospheres of transiting and nontransiting hot Jupiters, the limitations of these techniques are not well understood. We present a series of simulations of one HRCCS technique, which combines the cross-correlation functions from multiple epochs, to place temperature and contrast limits on the accessible exoplanet population for the first time. We find that planets approximately Saturn-sized and larger within ∼0.2 au of a Sun-like star are likely to be detectable with current instrumentation in the L band, a significant expansion compared with the previously studied population. Cooler (T eq 1000 K) exoplanets are more detectable than suggested by their photometric contrast alone as a result of chemical changes that increase spectroscopic contrast. The L-band CH 4 spectrum of cooler exoplanets enables robust constraints on the atmospheric C/O ratio at T eq ∼ 900 K, which have proven difficult to obtain for hot Jupiters. These results suggest that the multi-epoch approach to HRCCS can detect and characterize exoplanet atmospheres throughout the inner regions of Sun-like systems with existing highresolution spectrographs. We find that many epochs of modest signal-to-noise ratio (S/N epoch ∼ 1500) yield the clearest detections and constraints on C/O, emphasizing the need for high-precision near-infrared telluric correction with short integration times.