4

Implementing the Exoplanet Science Strategy

EXPANDING THE STATISTICAL CENSUS OF EXOPLANETS IN THE GALAXY

As described in Chapter 2, although radial velocity (RV), direct imaging, and astrometry are all, in principle, sensitive to long-period planets, the microlensing technique is uniquely sensitive to low-mass planets at large separations—in particular, very low-mass (mass greater than roughly two times the mass of the Moon) planets in orbits greater than roughly 1 AU—and analogues of the Solar System ice giants. It is therefore naturally complementary to the transit technique. However, it is not possible to achieve this full potential of the microlensing technique from the ground. Rather, to enable these capabilities, a space-based, near-infrared (NIR) mission with a relatively large field of view is required (Bennett and Rhie, 2002), for several reasons:

1. Microlensing events due to stellar lenses, which last a few days to hundreds of days, are both stochastic and rare, and thus require simultaneous monitoring of hundreds of millions of stars in order to detect a few thousand microlensing events. The only line of sight where this is possible given current technology is the galactic bulge, where the stellar surface density is approximately 20 million stars per square degree down to magnitudes of H_AB ≅ 21. Thus, microlensing surveys require monitoring a few square degrees with a resolution of <0.3 arcsecond on time scales of a few days or less.
2. The perturbations of these microlensing events last from an hour to a few days, and have probabilities (given the existence of planet) of less than a few percent to tens of percent (for planets with the mass of the Moon to the mass Jupiter). These perturbations are also stochastic, and thus require continuous monitoring of the microlensing events, at least ~15 minutes.
3. Given the crowded conditions of the galactic bulge, resolving the lens and source from unrelated background stars requires resolutions of <0.3 arcsecond. Once this resolution is achieved, it is possible to estimate the mass of the lens and the source for the majority of microlensing events (Bennett et al., 2006).
4. The ultimate limit to the mass of a planet that can be detected via microlensing is set by the angular size of the source. The smallest sources in the galactic bulge, M dwarfs, enable the detection of planets with mass as low as that of roughly 2 times the mass of the Moon. Given that M dwarfs emit the majority of their light in the NIR, given that the galactic bulge is generally heavily extincted, and given that the sky is very bright in the NIR from the ground, NIR surveys from space are optimal for microlensing surveys for exoplanets.

The WFIRST mission (Spergel et al., 2015) provides a nearly ideal architecture with the nearly ideal instru-

mentation needed to carry out the microlensing survey that enables a more complete statistical census of exoplanets in the galactic bulge. Its combination of aperture, field-of-view, and NIR (H4RG) detectors are essentially optimal for this purpose. Figure 4.1 shows the sensitivity of a WFIRST microlensing survey, highlighting its nearly perfect complementarity to the Kepler survey.

Although there are few strong constraints on the frequency, mass function, or separation distribution of planets with separations greater than roughly 1 AU, particularly for planets with the mass of Earth and smaller (indeed, this is precisely what WFIRST will measure), rough estimates of the number of bound planets WFIRST will find can be made by adopting current estimates of the planet mass distribution from ground-based microlensing surveys and modest extrapolations for planets with masses and separations outside the region of sensitivity of these surveys. The estimated yield is of order 1400 bound planets. In addition, again using relatively conservative assumptions, WFIRST should be able to detect hundreds of free-floating planets with masses as low as the mass of Mars. Assuming they originally formed in protoplanetary disks, measurements of the occurrence rate and mass function of free-floating planets provide a strong indicator of the dominant formation processes of exoplanetary systems.

The detections by WFIRST will be unambiguous and of very high significance. Figure 4.2 shows the simulated detection of a Ganymede-mass planet located at 5.2 AU from its parent star, and the detection of a potentially habitable planet. WFIRST may also be able improve upon Kepler’s estimate of η⊕, the mean number per star of rocky planets with between 1 and 1.5-2 Earth radii that reside in the habitable zone of their host star.

Finding: A microlensing survey would complement the statistical surveys of exoplanets begun by transits and radial velocities by searching for planets with separations of greater than 1 AU (including free-floating planets) and planets with masses greater than that of Earth. A wide-field, NIR, space-based mission is needed to provide a similar sample size of planets as found by Kepler.

Recommendation: NASA should launch WFIRST to conduct its microlensing survey of distant planets and to demonstrate the technique of coronagraphic spectroscopy on exoplanet targets.

The recommendation regarding the WFIRST coronagraph is explained in the section “Space-Based Studies,” later in this chapter.

Although a microlensing survey with the current incarnation of WFIRST is clearly the most capable and likely the most cost-effective mission to survey for exoplanets where the statistical census of exoplanets is most incomplete (Penny et al., 2018; Bennett and Rhie, 2002; Bennett et al., 2009), it is not strictly required to achieve the goals outlined above. The Microlensing Planet Finder, a proposed Explorer-class mission (Bennett et al., 2010), could achieve the majority of these goals, as could the 1.1 m DRM2 incarnation of the WFIRST mission proposed by Green et al. (2012).

Finding: If WFIRST is cancelled, a smaller, dedicated probe-class satellite could accomplish some of the science enabled by a space-based microlensing survey.

For the most part, a space-based microlensing survey would be self-contained, meaning that follow-up or concurrent observations would not be needed to extract the majority of the science from the survey. However, simultaneous observations from ground-based observatories such as the Large Synoptic Survey Telescope, the Korean Microlensing Telescope Network (Kim et al., 2016), or the Prime focus Infrared Microlensing Experiment

telescope that is currently being developed (Bennett et al., 2018, white paper), could yield additional information about some of the detected planets, including the masses of free-floating planets (Zhu and Gould, 2016).

As described in the updated Exoplanet Exploration Program Analysis Group Study Analysis Group (SAG)-11 report “Preparing for the WFIRST Microlensing Survey,” there are a number of activities that are needed to prepare for and optimize the WFIRST microlensing survey (Yee et al., 2014). These include a precursor NIR microlensing survey to determine the optimal location of the WFIRST microlensing fields (Shvartzvald et al., 2017), ground-based adaptive optics (AO) or Hubble Space Telescope (HST) data of past microlensing events to develop the technique that will be used by WFIRST to measure the masses of the host stars and their exoplanets, precursor simultaneous monitoring of microlensing events from Earth and from heliocentric satellites such as Spitzer or Kepler in order to measure microlensing parallaxes (Calchi Novati et al., 2015), precursor HST observations of the likely target fields to provide improved estimates of the microlensing event rates, and finally, building the (currently small) U.S. microlensing community through workshops, “hack weeks,” making more microlensing data sets publicly available, and developing open-source, easy-to-use, and well-documented microlensing modeling codes (Poleski and Yee, 2018).

Finding: A number of activities, including precursor and concurrent observations using ground- and space-based facilities, would optimize the scientific yield of the WFIRST microlensing survey.

THE CASE FOR IMAGING

A described in Chapter 2, direct imaging of exoplanets requires angularly resolving the planets from their host stars and directly detecting photons from the planets. Separating the planet’s image from its host star is fundamentally limited by the theoretical diffraction limit l/D, which is set by the observing wavelength l and telescope diameter D. The large flux ratios and small angular separations imply that directly imaging and spectroscopically characterizing planets close to their host star requires dedicated high-contrast instruments.

Two complementary methods are required to achieve a very high-contrast direct detection using the traditional technique of internal coronagraphy: a coronagraph, which helps to suppress the diffraction from the host star, and wavefront control, which helps mitigate the effect of scattering arising from the combination of atmospheric turbulence (on ground-based telescopes) and time-varying optical aberrations in the telescope and instrument. Modern stellar coronagraphs consist of a series of masks inserted in the instrument pupil or focal plane. These masks are designed to remove the diffraction pattern from the central star as efficiently as possible, while preserving the signal of an off-axis object as close to the optical axis as possible. Roughly speaking, the inner working angle (IWA) of a coronagraph quantifies the smallest angular separation from the host star beyond which a planet of a given flux ratio can be imaged, and is often expressed in l/D units. Formally, it is defined as the 50 percent off-axis throughput point. Naturally, the fundamental limit to the IWA is the theoretical diffraction limit l/D. The details of a coronagraph’s particular implementation are driven by the trade-off between IWA, contrast, and sensitivity to optical aberrations.

However, being static elements, coronagraphs do not allow for active control of scattered light induced by time-varying optical aberrations. That task is relegated to the wavefront control system, which is also known as the adaptive optics system on ground-based telescopes. Achieving a raw contrast of 10^–5 at 1 micron requires controlling wavefront aberrations at the 1 nm root-mean-square (RMS) level. To accomplish this, the wavefront control system relies on a sensor, which can be a dedicated instrument or the science camera itself, and a correcting element. Most systems use actuated deformable mirrors for the correcting element. The challenge of the wavefront controller is to sense and correct the aberrations within the time scale associated with the change of the perturbation. This requirement sets another fundamental limit on the achievable contrast—the brightness of the host star used for sensing—and its corresponding photon shot noise over the time scale of the stability of the system (telescope and instrument).

Most direct imaging detections of exoplanets so far have relied on the combination of adaptive optics and coronagraphs on ground-based telescopes. Due to the turbulent nature of Earth’s atmosphere, the main purpose

of the wavefront controller is to correct for the large and rapidly evolving wavefront aberrations induced by the propagation of waves through the turbulent medium of the atmosphere. The most modern incarnations of adaptively corrected ground-based coronagraph instruments are currently limited to contrast ratios of about 10^–6 (Figure 4.3), which is sufficient to detect the glow of young forming giant planets that are emitting thermal infrared due to their ongoing contraction from formation.

Because they will push the boundaries of the aforementioned fundamental trade-offs to new and distinct regimes, the advent of giant adaptively corrected ground-based telescopes and (ultra)stable active space-based coronagraphs presents a game-changing opportunity for the direct imaging and characterization of exoplanets.

Ground-Based Studies

The Giant Segmented Mirror Telescope (GSMT) Opportunity

The advent of GSMTs will provide new opportunities for exoplanet imaging and characterization in the next decade (Gilmozzi and Spyromilio, 2007; Johns et al., 2012; Sanders, 2013; Wang et al., 2017). Two of these GSMTs—namely, the Giant Magellan Telescope (GMT) and the Thirty Meter Telescope (TMT)—are U.S.-led efforts, which together will cover the entire night sky. Both are actively seeking funding to complete the projects, including funding from the National Science Foundation (NSF). The 3-fold improvement in angular resolution, 10-fold improvement in light-collecting capabilities, and 80-fold improvement in sensitivity to point sources provided by 30-m class facilities will open up new vistas of exploration. From the detection and spectroscopic study of gas and ice giants in reflected light and thermal emission, to the search for biosignatures of rocky planets orbiting M-type stars, direct imaging provides complementary phase space coverage to indirect methods and is the only technique capable of spectroscopically characterizing nontransiting exoplanets (Fitzgerald et al., 2018, white paper).

Young Self-Luminous Planets

GSMTs will provide a unique opportunity to survey nearby young planetary systems within the “ice line” (a ~ 3 AU), and bridge the separation gap with Doppler RV and transit surveys. High-resolution spectroscopy with GSMTs will also enable the measurement of planet spins (Snellen et al., 2014; Bryan et al., 2018) and in some cases Doppler imaging studies, as in Crossfield et al. (2014).

Mature Giant Planets in Reflected Light with GSMTs

GSMTs will enable the direct imaging of mature gas giant and ice giant planets in both reflected light and thermal emission (Figure 4.3). RV surveys will provide promising targets for follow-up imaging and characterization studies. Moderate or high-resolution spectroscopy will probe the depths of multiple water and methane features, allowing models to recover carbon or oxygen abundances and, in turn, enable integrated studies of abundances versus planet location, mass, and host star properties (Lupu et al., 2016).

Search for Life Around M-Type Stars

The contrast ratio between a temperate Earth-size planet and a solar-type stellar host is 10^–10. This contrast ratio requires deep levels of starlight suppression that are beyond the fundamental limits of ground-based AO systems. Contrast limits are indeed set by the finite number of photons available for sensing and correcting wavefront aberrations induced by quickly evolving atmospheric turbulence (Guyon, 2005; Poyneer et al., 2007; Guyon and Males, 2017; Males and Guyon, 2018). GSMTs will, however, be able to detect starlight reflected by rocky, habitable-zone exoplanets around the nearest M dwarfs. Indeed, the low luminosities of M dwarfs means that planets need to orbit close to the star to receive Earth-like radiance levels, and so the contrast ratio is relaxed by several orders of magnitude to 10^–7–10^–8. However, while the proximity of M-dwarf habitable zones to the star currently poses a challenge for direct imaging because the habitable zone lies well inside the IWA of 8 m class telescopes,

the small IWAs enabled by 30-m class telescopes (roughly 10 mas around 1 micron) make them ideal facilities for characterizing planets in the habitable zones of M dwarfs (Figure 4.3). GSMTs will be powerful facilities not only in obtaining more complete and less biased statistics on planetary demographics through surveys that image planets orbiting low-mass stars but also in characterizing these discoveries. Together with the James Webb Space Telescope (JWST), high-contrast AO observations will be among the first opportunities to detect biosignatures in the atmospheres of other worlds. Planets around faint M-type stars are favorable targets for spectroscopic follow-up at shorter wavelengths with both GSMTs. There are abundant lines from biosignature gases, O₂, H₂O, CH₄, and CO₂, in the near infrared (roughly 1-4 μm), also where the high dispersion coronagraphy technique is expected to reach optimal performance (see below). There are roughly 20 M dwarfs within 5 pc that are observable by GSMTs, and there is at least one rocky planet per M dwarf (Dressing and Charbonneau, 2015) with one in four potentially in the habitable zone. Given the limited number of targets available and the likely different scientific emphasis of the two U.S.-initiated GSMT programs, access to the full sky through telescopes in both hemispheres is important, and investment in both GSMT projects is preferred.

Finding: The GMT and TMT will enable profound advances in imaging and spectroscopy of entire planetary systems, over a wide range of masses, semimajor axes, and wavelengths, potentially including temperate Earth-size planets orbiting M-type stars.

Thermal Infrared Studies

Thermal infrared observations with GSMTs may allow detection of warm (T = 400-600 K) rocky planets around the nearest (<5 pc) FGK stars (at 3-5 μm; Crossfield, 2013), as well as somewhat cooler Earth-size rocky exoplanets around nearby Sun-like stars (at 8-13 μm; Quanz et al., 2015). At the longest wavelengths, biomarkers such as H₂O, CH₄, O₃, and CO₂ can be identified using low-resolution spectroscopy (Des Marais et al., 2002). The spectral energy distribution can additionally be used to estimate surface temperature and cloud fraction. For T = 400 K super-Earths around K stars, GSMTs will be able to spectroscopically characterize both reflected light and 3-5 μm thermal emission for the same planets. The same will be true for some of the known, nearby, close-separation RV-detected rocky terrestrials, super-Earths, and warm giants (ice and gas) at 10 μm. Combining measurements of thermal and reflected light will make GSMTs the first instruments capable of measuring the radii and studying the energy budget and climate of other worlds (Fitzgerald et al., 2018, white paper; Meyer et al., 2018, white paper). The GMT and TMT may image and spectroscopically characterize the nearest temperate Earth-size planets orbiting G-type stars in the thermal infrared.

A Technological Roadmap for Ground-Based High-Contrast Imaging

High-contrast characterization of exoplanets from the ground, a prime science case for GSMTs, still requires substantial technology developments. The gap between second-generation AO systems on current 8-10 m class facilities and the requirement for imaging and characterization of Earth-like planets around M-type stars constitutes several orders of magnitude of improvement in contrast. A vigorous research and development (R&D) program involving specific laboratory activities and on-sky demos spread over the next decade is necessary to fill in technology gaps (Currie et al., 2018, white paper).

Breakthroughs in Extreme Adaptive Optics and Wavefront Control

The challenge in reaching the full potential of GSMTs for reflected-light spectroscopy appears twofold—namely, pushing the raw contrast ratio to more extreme values from the current state of the art of 10^–4–10^–5 to 10^–7–10^–8, and extending the effective IWA toward the diffraction limit of the telescope (l/D). Fundamentally, these stem from a single driving requirement: to adequately sense and control wavefront aberrations—in particular, low-order aberrations that dominate the raw contrast error budget at small inner working angles (see white papers from Fitzgerald et al., 2018 and Currie et al., 2018). Current wavefront control architectures are nowhere near their fundamental photon noise limits (Guyon, 2005) because they are hampered by noisy detectors, suboptimal

sensing-to-command conversion efficiencies, including control laws (simple integrators to linear predictors; see Males and Guyon, 2018), and the time lag between the wavefront measurements and the application of the correcting command. Relatively recent architectures such as Pyramid and focal plane wavefront sensors have started to surface in adaptive optics facilities (e.g., Jovanovic et al., 2015) and hold the promise of both optimal conversion efficiencies and controlling rapidly fluctuating wavefront errors including noncommon path aberrations. Developments in predictive control and sensor fusion show promise to break through the practical limitations of present-day AO systems, currently limited to 10^–4–10^–5 raw contrast, and bring us closer to the 10^–7–10^–8 requirement to image and characterize temperate Earth-size planets around M-type stars (Poyneer et al., 2007; Guyon and Males, 2017; Males and Guyon, 2018).

Considerable uncertainty remains in the maximum achievable performance of high-contrast adaptive optics systems on the GSMTs, and that uncertainty spans the range that will allow detection of potentially habitable planets. Development of better simulation tools, prototyping of concepts on laboratory and 8-10 m class facilities (see below), and maturation of conceptual instrument designs will be crucial to determining the final architecture and capabilities of a GSMT planet imaging facility. Involvement by the U.S. community and the NSF in the two GSMT projects will help to enable that.

High-Density Deformable Mirrors

The enabling hardware technology for GSMTs are fast (>kHz), large-stroke (>6 microns), high-order (120 × 120 actuators) deformable mirrors. Planet imager instruments on GSMTs will directly benefit from a recently concluded R&D effort initiated by European Southern Observatory for the Extremely Large Telescope (ELT). The outcome of this study and other efforts currently carried out in the industry is a clear path to fulfilling the requirements.

Coronagraphs for Segmented and Obscured Apertures

Thanks in part to the WFIRST-Coronagraph Instrument (WFIRST-CGI) technology development program and the difficulties associated with the heavily obscured WFIRST telescope aperture, as well as the Segmented Coronagraph Design and Analysis (SCDA) initiated by the Exoplanet Exploration Program (ExEP) for Large UV/Optical/IR Surveyor (LUVOIR) and Habitable Exoplanet Imaging Mission (HabEx), there are now coronagraph technologies and design tools readily applicable to GSMTs. Coronagraph technologies have significantly evolved over the past 10 years such that the challenge of designing and building a coronagraph that is insensitive to segmented/obscured apertures and low-order wavefront errors has largely been overcome.

High-Dispersion Coronagraphy

High-dispersion coronagraphy aims to optimally combine high-contrast techniques with high-resolution spectroscopy (Ruane et al., 2018, white paper). This rapidly developing technique, which leverages advances in extreme precision RV spectrographs (see the section “Radial Velocities,” later in this chapter), is still in its infancy but promises to help bridge the contrast gap mentioned above by sidestepping speckle noise, which is one of the most pervasive sources of systematics. The technique searches for specific spectral features in either thermal emission or reflected light that are unique to the planet. It multiplies the contrast ratio achieved by high-contrast imaging by the additional contrast realized by high-dispersion spectroscopy. Implemented on GSMTs, it has the potential to find and atmospherically characterize temperate terrestrial planets around the most nearby stars. Since it is inherently sensitive to Doppler effects induced by the orbital motion of the planet (10-100 km/s; see, e.g., Snellen et al., 2014; Hoeijmakers et al., 2018), it allows for an effective separation of the suite of planet molecular lines in a given absorption band from those produced in Earth’s atmosphere, enabling the search for interesting species such as H₂O, CH₄, CO₂, and O₂.

While current facilities have helped to provide a glimpse of the power of this recent method (Konopacky et al., 2013; Snellen et al., 2014; Bryan et al., 2018; Crossfield et al., 2014), no facility fully integrating high-contrast spectroscopy has yet seen the light. Several ongoing projects at the Very Large Telescope (VLT; Lovis et al., 2017),

Subaru telescope (Jovanovic et al., 2017), and Keck telescope (Mawet et al., 2017) are noted that will help assess the relevance of the technique to future GSMT facilities.

Polarimetry

Polarimetry is a useful tool for the characterization of directly imaged exoplanets (Millar-Blanchaer et al., 2018, white paper). Polarimetric observations with future telescopes have the potential to refine understanding of scattering processes in the atmospheres and on the surfaces of planets. In particular, time-series and spectropolarimetric measurements can distinguish between different cloud and surface types. Notably, polarimetry has the ability to constrain planetary albedos and may ultimately be able to reveal the presence of a liquid water surface (Williams and Gaidos, 2008). Relatively little attention has been given to polarimetry within the exoplanet community to date. This is due in part to the difficulty of obtaining high signal-to-noise ratio polarimetric measurements with current facilities, and to the lack of existing polarimetric detections. To maximize the gain from polarimetric measurements, careful attention needs to be paid to the design and implementation of future instrument designs. For example, an instrument able to suppress unpolarized speckles by a factor of 1000 (similar to current Gemini Planet Imager [GPI] and Spectro-Polarimetric High-Contrast Exoplanet Research [SPHERE] instrument levels; see, e.g., Millar-Blanchaer et al., 2016; van Holstein et al., 2017) will gain a factor of 50 in detection limits relative to the raw contrast for a 5 percent polarized planet. For reference, Earth is up to approximately 30 percent polarized at 0.5 μm (Coffeen, 1979).

High Quantum Efficiency, Low-Noise Detectors

High quantum efficiency (QE), low-noise (<1 e–) or zero-noise detectors will be needed to approach the fundamental limits of wavefront sensing and scientific measurements. The arrival and on-sky testing of new low-noise fast-readout detectors, such as microwave kinetic inductance detectors (MKIDs) and infrared (IR) avalanche photodiode arrays, have enabled much more powerful focal-plane wavefront sensing techniques (Meeker et al., 2018; Atkinson et al., 2014). These detector technologies still have a long way to go to form the backbone of future scientific instrumentation. For instance, their format is still confined to small arrays (e.g., 320 × 256), which is sufficient for wavefront control but precludes their use in conventional imagers and spectrographs.

Science and Technology Pathfinders on Existing 8 to 10 Meter Class Facilities

Ground-based telescopes have been playing a leading role in exoplanet direct imaging science and technological development for the past two decades and will continue to have an indispensable role for the next decade and beyond. Extreme AO systems will advance wavefront control, coronagraphy, and post-processing, thereby augmenting the performance of and mitigating the risk for WFIRST-CGI, while validating performance requirements and motivating improvements to atmosphere models needed to unambiguously characterize Solar System analogues with HabEx/LUVOIR (Currie et al., 2018, white paper). Current facilities are also the proving ground to develop technologies for future GSMT instruments focused on exoplanet imaging and spectroscopic characterization.

Finding: The technology roadmap to enable the full science potential of GMT and TMT in exoplanet studies is in need of investments, leveraging the existing network of U.S. centers and laboratories and current 8-10 meter class facilities.

Recommendation: The National Science Foundation should invest in both the GMT and TMT and their exoplanet instrumentation to provide all-sky access to the U.S. community.

The committee further notes that an additional finding in support of this recommendation appears at the end of the section “Opportunities to Characterize Planets Through Transits,” later in this chapter.

Space-Based Studies

Direct imaging with a space-based telescope has long been recognized as one of the most promising paths to spectroscopically characterize planets around Sun-like stars (Malbet et al., 1995; Angel and Burrows, 1995). In space, wavefront aberrations are caused only by the instrument itself, and can be stable or evolve very slowly. For a sufficiently stable telescope, deformable mirrors can provide essentially perfect correction. Since space telescopes are (necessarily) smaller than their Earth-bound counterparts, diffraction control is critical to achieving a reasonable IWA, but a wide variety of coronagraph concepts have been developed to address this requirement (Guyon et al., 2006).

In the early part of the 21st century, an architecture known as the Terrestrial Planet Finder Coronagraph (TPF-C) emerged as a mission concept (Trauger and Traub, 2006). It would have combined a 3 × 8 m elliptical monolithic mirror with a visible-light active optics coronagraph to achieve the flux ratio sensitivity of ~10^–10 at ~0.08 arcsecond IWA required to detect and characterize Earth-like planets.¹ Although understanding of coronagraphy has advanced enormously since then, and TPF-C was never selected as a mission, it remains the prototype for the planet-characterizing missions now on the drawing board. These missions remain exceptionally challenging, requiring exquisite optical stability and sophisticated components, but are feasible with current technology.

Such concepts are often referred to as “internal” coronagraphs, since the scattered starlight is controlled by components, such as coronagraphic masks and deformable mirrors, located inside the telescope spacecraft. Another concept that was originally proposed decades ago (Spitzer, 1963) but that has seen enormous recent advances is the “external” coronagraph (Cash, 2006), usually referred to as a “starshade,” which blocks the target star with a large (20-70 m) free-flying occulter. The occulter is designed to minimize diffracted light, while allowing the off-axis planet to be imaged directly. The inner working angle of a starshade system is set by the geometry of the system, roughly the ratio of size of the starshade to the distance to it, and hence is somewhat decoupled from the telescope diameter, although practical and optical considerations often limit achievable inner working angles to roughly the telescope diffraction limit for 4 m telescopes. This still improves the practical IWA by a factor of 2 to 3 over internal coronagraphs. The shape of the starshade needs to be specially designed to control diffracted starlight, thereby casting the deepest possible shadow. Starshades are well suited to small to medium-size (1-4 m) telescopes, giving them similar IWAs to larger telescopes with internal coronagraphs, but this comes at the cost of operational flexibility. In addition, current technology generally allows starshades to have larger outer working angles (OWAs) than internal coronagraphs, as the OWA of a starshade is set by the size of the detector, whereas the number of actuators in the deformable mirrors set the OWA of an internal coronagraph. The above-mentioned concepts have somewhat complementary capabilities, with internal coronagraphs being more agile and easily retargeted, while starshades have high throughput over a broad spectral bandpass, and have very deep shadows, and thus can characterize extremely faint planets; some mission concepts combine both modes to take advantage of this complementarity.

A high-contrast imaging mission would combine starlight control (coronagraph or a starshade) with a suite of scientific instruments. Notionally, these would include an imaging camera (optimized for rapid detection of planets, measurement of their orbital motions, and broad photometric characterization); an imaging spectrograph such as an integral field unit (for obtaining low-resolution spectra of the vicinity of target stars, and to identify molecular species in planetary atmospheres); and perhaps a polarimetric capability (see the preceding section, “Ground-Based Studies”) for further characterizing planetary clouds or circumstellar dust. Imaging spectroscopy is broadly similar to transit spectroscopy (see the section “Atmospheric Characterization Through Transit Spectroscopy,” later in this chapter), but because it does not require the “slant geometry” of transit spectroscopy, it is less sensitive to high-atmosphere clouds and hazes. Thus, directly imaged spectra potentially allow for the detection of molecular species in cloudy planets (Robinson et al., 2016). The fundamental and practical limitations of coronagraphs and starshades likely limit missions to the UV/Optical/NIR region.

___________________

¹ A parallel interferometer-based concept, TPF-I, was designed for mid-IR operations; see NASA Exoplant Program, “Documents,” https://exoplanets.nasa.gov/exep/resources/documents/.

Science Case for Space-Based High-Contrast Imaging

Space-based direct imaging is capable of a wide variety of scientific programs in the study of extrasolar planets. A well-designed direct imaging space mission would address a range of the key questions identified in this report.

Characterization of Mature Exoplanets in Reflected Light

Space-based direct imaging will be able to access mature giant planets, Saturn-size or smaller, from approximately 0.5 to 15 AU. Space-based coronagraph imaging is a powerful complement spectroscopic characterization with JWST, which will characterize giant planets close to their stars. Spectroscopy can constrain atmospheric properties such as methane abundance (Figure 4.4). With the combination of both missions, it will be possible to study planets whose different locations imply different formation histories, and whose different temperatures access different chemistries at different atmospheric depths.

Since giant planets are relatively bright, missions capable of discovering Earth-equivalent planets can easily characterize a large sample of giant planets. Most of them will already be known through RV or astrometric surveys (see the section “Exoplanet Masses,” later in this chapter), making this approach a straightforward, high-reward science program early in any direct imaging mission. Such observations will also be capable of characterizing “super-Earth” planets over a narrower range of separations (approximately 0.5-3 AU), again complementing the sample of super-Earths close to low-mass stars. For example, detection of atmospheric constituents could help determine whether these planets have primordial hydrogen-dominated atmospheres or secondary atmospheres.

Search for Earthlike Planets Around Solar-Type Stars

Detecting an Earth-size planet in the habitable zone of a Sun-like star at a distance of 10 pc in reflected light requires detecting a planet with a flux ratio of roughly 10^–10 (equivalent to a V = 30 magnitude star) less than 0.1 arcsecond away from a V = 5 magnitude star. Daunting as this is, high-fidelity simulations indicate it is within the reach of a properly optimized space telescope. Such measurements are essentially impossible with any other technique; it is extremely unlikely that such a planet will be found to transit any nearby star, and even if it did, accumulating enough transits to characterize the atmosphere of the planet (which would produce a signal of roughly 0.1 percent of the transit depth, which itself is 0.01 percent), would obviously take many years. Ground-based extremely large telescopes, although capable of imaging at very small IWAs, will be unable to reach these contrast levels (see the section “Ground-Based Studies,” earlier in this chapter). A large space-based direct imaging mission, using either the coronagraph or starshade starlight suppression techniques, is the best path to finding an “Earth twin” in the next two decades.

Finding: A coronagraphic or starshade-based direct imaging mission is the only path currently identified to characterize Earth-size planets in the habitable zones of a large sample of nearby Sun-like stars in reflected light.

Finding: Recently acquired knowledge of the frequency of occurrence of small planets, and advances in the technologies needed to directly image them, have significantly reduced uncertainties associated with a large direct imaging mission.²

As with giant planets, determining the nature of an Earth-size planet will require spectroscopic characterization. Similar techniques can be applied—for example, simulated retrievals indicate that high signal-to-noise ratio (SNR) moderate-resolution (R > 70) visible-light spectra can determine atmospheric abundances of O₂ and H₂O (Feng et al., 2016). However, determining whether a world is truly life bearing may require additional study beyond the identification of molecules. Multiple observations will allow determination of the orbit of a planet and (when combined with Doppler measurements) the mass of a planet. More capable coronagraphic or starshade missions can also access NIR or UV wavelengths, allowing species such as CH₄ to be measured. It is important to emphasize, however, that unless observers are very fortunate, no single mission is likely to be able to identify life, or even a habitable world, with high confidence. Once these first discoveries are made, they will, instead, guide future observations and the design of more capable future missions that will shape the understanding of these new worlds.

Architecture of Mature Exoplanetary Systems

Imaging a planetary system can provide (almost by definition) a portrait of multiple members of a planetary family. Such a portrait could show many of the planets within a given system. Follow-up observations will determine their orbits (co-planarity, stability, etc.). In addition, circumstellar dust from smaller bodies—analogous to zodiacal light near Earth—will likely be detectable, providing information on the small-body population and mass constraints on both seen and unseen planets. For example, the masses of the planets seen in the directly imaged systems HR8799 and Beta Pictoris are constrained by dynamical interactions between multiple planets or dust disks (Fabrycky and Murray-Clay, 2010). In the youngest systems, these disks are thousands of times brighter than the Solar System’s and easily detected, particularly in the outer portions of stellar systems. Future coronagraphic missions will approach the level of sensitivity needed to see mature disks; simulations predict that the WFIRST coronagraph could detect roughly 10 times the solar level of scattered light (Mennesson et al., white paper).

Near-Term Imaging with JWST

Although it will carry a suite of coronagraph masks for different modes and wavelengths, the JWST has not been optimized for high-contrast imaging. It will have similar capabilities to ground-based AO instruments for studying young giant planets. JWST will be able to detect lower mass (Saturn-like) or low-entropy planets (Fortney et al., 2008) by observing at longer wavelengths where they are brightest, but at slightly larger IWA than the most advanced ground-based systems (Beichman et al., 2018, white paper). Discovering a significant number of Saturn- to Jupiter-mass planets will require a large-scale systematic survey and could be a key JWST giant-planet result. Combining JWST mid-infrared (MIR) observations with ground-based spectroscopy will allow precise determination of planetary luminosities and temperature.

Sensitivity Scaling with Telescope Diameter

The capability of a large space mission to discover any given planet type depends on its inner working angle and contrast, and also on its total sensitivity. Planets are extremely faint; even with starlight perfectly blocked,

___________________

² See later in this section for further discussion.

adequate spectroscopic characterization requires a large photon collection rate and, thus, a large aperture or high throughput.

The discovery process is quite complex—a single observation may fail to find a planet in an edge-on orbit if it is passing almost in front of or behind its parent star, for example. Target selection, exposure time selection, timing of observations to confirm a candidate, and so on, are a complex optimization problem that has been extensively studied (Stark et al., 2015; Garrett et al., 2017; Brown, 2004). Nonetheless, with the increasing confidence in planetary occurrence rates and increasingly sophisticated strategies, researchers can now forecast the likely yield of Earth-size habitable-zone (HZ) planets within a factor of roughly 2. Adjustments to optimization strategies and criteria may vary planet yields, particularly the number of planets for which a high-SNR spectrum can be obtained, but the overall relative capabilities of missions of different scales are unlikely to vary significantly from these estimates.

Figure 4.5 shows the general trend of the number of habitable zones surveyed for potentially Earth-like planets (discovery and spectroscopic characterization) by notional coronagraph missions versus telescope-inscribed diameter and architecture, based on modeling by Christopher C. Stark. In addition to telescope diameter, sensitivity is a function of telescope architecture. Off-axis telescopes, without a secondary mirror obscuring part of the pupil, can incorporate coronagraphs that can operate at small inner working angles with high throughput. Small obscurations such as gaps between hexagonal telescope segments do not significantly degrade the yield. Larger features such as a secondary mirror, particularly when combined with realistic wavefront jitter such as vibration-induced tip/tilt, restrict the telescope to lower performance coronagraphs such as apodized pupil Lyot coronagraphs (APLCs). These limitations may not be fundamental but have not yet been overcome in spite of extensive exploration of coronagraph concepts. In Figure 4.5, the two lines correspond to high-performance coronagraphs suitable for telescopes

without secondary mirrors, and lower performance coronagraphs for very large conventional telescopes. The figure shows that missions now under consideration have a high probability of yielding spectra of many Earth-analogue planets in the habitable zones of intermediate-mass stars.

Roadmap for Direct Imaging

The overall path to direct imaging of Earth-like planets is relatively clear through currently active programs and missions to be proposed to the 2020 Decadal Survey.

First, the WFIRST mission will carry a technology demonstration coronagraph to advance readiness for future missions. The NASA Technology Demonstration for Exoplanet Missions program continues to support laboratory development of technology for future missions, including advanced coronagraphs suitable for the telescope designs of the next major mission, validation of starshade concepts and technology, and key components such as low-noise detectors.

During the decade of the 2020s a probe-class starshade could be launched to rendezvous with WFIRST and carry out a strong science program, characterizing a diverse set of planetary systems while potentially searching for Earth analogues around the nearest stars. If the WFIRST mission loses the coronagraph or starshade capabilities, a stand-alone probe-class coronagraph launched late in the decade could fill the gap between current capabilities and the large strategic mission, laying scientific groundwork and validating technical approaches.

In the same decade design would begin on a direct imaging mission for launch in the 2030s, based around the large strategic mission concepts of HabEx or LUVOIR. It would be capable of detecting and characterizing a significant sample of Earth-size HZ planets. Characterization would include high-SNR spectroscopy to determine abundances of O₂, O₃, H₂O, and, if possible, CH₄ for a subset of the closest targets. The mission would also enable a balanced exoplanet science portfolio through studies of planetary system architectures (with sensitivity to true Solar System analogues), metallicity and atmospheric properties of giant planets, atmospheric diversity of super-Earth planets inside and outside the habitable zone, density and structure of circumstellar dust, and the properties of young planet-forming systems. Combined with characterization such as planetary mass measurements from other facilities, this mission could provide a comprehensive picture of the evolution of planetary systems and may provide the first evidence of a habitable Earth twin—an epochal moment in human history.

Recommendation: NASA should lead a large strategic direct imaging mission capable of measuring the reflected-light spectra of temperate terrestrial planets orbiting Sun-like stars.

WFIRST Opportunity

The WFIRST mission (Spergel et al., 2015) has gone through a series of evolutions of its direct imaging capability. Originally conceived by the 2010 Decadal Survey as a wide-field imaging mission on a 1-1.5 m telescope (ASTRO2010), with exoplanet science focused on microlensing, it was redesigned based on the addition of a newly available 2.4 m telescope assembly. In addition to enhancing the core science capabilities, this telescope was large enough to potentially carry a coronagraph. Such a coronagraph would both satisfy the 2010 Decadal Survey mandate for advancing coronagraph technology and could carry out a significant science mission characterizing giant planets and zodiacal dust disks. The telescope was not ideal for coronagraphy, with a large and complex secondary mirror structure and other complications such as large reaction wheels for survey modes, but three coronagraph concepts were identified that could give sufficient performance. After study (Spergel et al., 2015), the newly configured mission was approved by the mid-decadal survey (NRC, 2016), contingent on a $3.2 billion cost. The coronagraph instrument (CGI) was included in the mission baseline, with a primary role as a technology demonstrator but also with a significant and capable science program.

In Phase A, the mission experienced significant cost pressure (not primarily driven by the coronagraph), and both NASA cost estimates and an independent review, the WFIRST Independent External Technical/Management/Cost Review, identified that it was likely to exceed the $3.2 billion cap, and also highlighted the risk that the coronagraph was underresourced for a full scientific instrument (NASA, 2017). In response, NASA implemented

several cost control measures on the primary mission and down-scoped the coronagraph to a pure technology demonstration, reducing the filter set to the bare minimum and eliminating any funded science program or institutional observer support. Some science may still be carried out under a “participating scientist” model as long as there is no significant additional cost.

WFIRST-CGI will provide significant risk reduction for future missions. Already, the detailed systems engineering and modeling needed to produce a flight-ready CGI design has led to rapid maturation of coronagraph concepts and identification of technology readiness issues such as deformable mirror stability and radiation damage to ultra-low-noise charge coupled devices (CCDs). Interactions between the telescope and coronagraph, such as polarization or jitter effects, are challenging and uncertain in ground simulations, and performance predictions for future missions carry modeling uncertainty terms that lead to challenging future-mission telescope requirements. Most importantly, operating an advanced coronagraph on exoplanetary targets in flight will lead to the development and quantification of the capabilities of existing analysis techniques. For example, techniques developed to remove correlated point spread function (PSF) noise and extract planetary spectra for ground-based coronagraphs will operate differently on space mission data, and quantifying the final noise properties of extracted spectra as a function of collected flux is crucial to determining the scope of future missions. Algorithms that combine wavefront control and PSF analysis hold great promise for extracting faint planetary signals while relaxing telescope stability requirements, but are best tested with a realistic target and in a flight environment. Given these opportunities, WFIRST-CGI’s benefits are greatest when it allows flexibility for testing new approaches and retains some ability to carry out meaningful quantitative exoplanet spectroscopy, even on a very small number of targets.

Finding: Flying a capable coronagraph on WFIRST will provide significant risk reduction and technological advancement for future coronagraph missions. The greatest value compared to ground testing will come from observations and analysis of actual exoplanets, and in a flexible architecture that will allow testing of newly developed algorithms and methods.

Even after the cost reductions in the coronagraph, it is still highly sensitive for imaging circumstellar dust. Simulations show that it could detect dust disks at the level of roughly 10 times solar zodiacal dust densities around nearby (<10 pc) Sun-like stars, which represents a significant subset of the targets for future large strategic missions. This sensitivity level would significantly surpass that of the Large Binocular Telescope Interferometer (LBTI) Hunt for Observable Signatures of Terrestrial Systems campaign (Ertel et al., 2018), also directly measuring visible-wavelength scattered light, reducing the uncertainty of extrasolar zodiacal light impacting a future planet-imaging mission.

Finding: The WFIRST-CGI at current capabilities will carry out important measurements of extrasolar zodiacal dust around nearby stars at greater sensitivity than any other current or near-term facility.

WFIRST Starshade

Although the WFIRST telescope aperture and spacecraft design make it a challenging mission for an internal coronagraph, a starshade occulter could provide it with a capable exoplanet imaging capability. Occulters have high throughput (even with an obscured aperture) and can achieve a relatively small IWA (limited by occulter size and cost rather than telescope size) with good flux ratio sensitivity. The long retargeting time scales of a starshade work well with a general astrophysics mission such as WFIRST, and the WFIRST starshade mission has an order of magnitude higher effective throughput and hence higher science reach than the WFIRST-CGI.

A starshade “rendezvous” mission would combine the existing WFIRST instrument package with a separately launched starshade spacecraft. This was studied as one starshade mode by a NASA-led science definition team and the mission concept is being updated in preparation for the decadal survey (NASA, 2015). This would be a probe-class mission, for which no NASA Astrophysics Division funding line currently exists. It would therefore have to be selected as either a strategic mission by the 2020 Decadal Survey, or selected as part of a competed probe-class mission line established by the 2020 Decadal Survey. Either way, if it were to proceed, it would almost

certainly launch in the latter half of the baseline WFIRST mission. The probe study found that WFIRST with a starshade would be capable of spectroscopically characterizing a large sample of known giant extrasolar planets at high SNR, mapping extrasolar zodiacal dust, and exploring nearby stars for smaller planets. It has some potential to detect (but not characterize) Earth-size habitable-zone planets, as well, but only around a small number of the very nearest bright stars. A final evaluation of its scientific capability will be available at the end of the mission study in early 2019.

Large Strategic Coronagraph Missions

In preparation for the 2020 Decadal Survey, NASA is funding two design studies for large strategic missions with significant capability to directly image exoplanets in reflected light. The HabEx observatory is designed to prioritize such exoplanet science, achieving Earth-analogue sensitivity through careful optimization of the telescope and instrument, while the LUVOIR telescope is a general-purpose astrophysics mission with similar capabilities in a larger and more flexible architecture. Both designs are currently being optimized, and both studies will likely include both a larger and a smaller design. They can be thought of as points on a continuum of missions, with the ultimate trade between cost, complexity, exoplanet characterization capability, and other astrophysical capabilities to be made by the 2020 Decadal Survey. The scale of such a mission is ultimately a question for the next decadal survey, which will balance the cost and capabilities against other opportunities. A 4 m class mission probing ~50 effective habitable zones would yield ~10 habitable-zone terrestrial planets and begin to explore the diversity of such worlds in atmospheric properties composition and provide the first meaningful measurement or upper limit on the occurrence rate of potentially habitable worlds. Even if planet occurrence rates are lower than expected, it would have a high probability of at least one characterizable planet. A 10-15 m extremely large mission would of course produce a correspondingly larger sample: with ~50 worlds, rare classes of planets could be seen, and statistical trends in planet properties with the other properties of their host systems (such as stellar age or the presence of giant planets) would begin to emerge. The notional HabEx and LUVOIR represent points in a continuum of exoplanet science. All large, strategic, direct imaging mission architectures are capable of transformative science in the integrated study of planetary systems.

HabEx

The HabEx concept prioritizes exoplanet detection science above all other science. To maximize coronagraph performance, its baseline architecture uses a 4 m monolithic telescope with an off-axis secondary mirror. It is capable of operations from 120 to 1800 nm, although exoplanet observations for typical targets would be in the 300-1000 nm range. In addition to UV/Vis general astrophysics instruments, there would be an exoplanet imager and integral field spectrograph. An internal coronagraph would allow an inner working angle of ~60 mas at 500 nm and be used primarily for discovery and orbit determination; a starshade would allow a similar IWA at all wavelengths shorter than 1200 nm and be used for high-sensitivity spectroscopic follow-up of planetary systems. If equipped with a starshade, HabEx could also have a large OWA, potentially enabling a more complete assay of the outer parts of planetary systems than an internal coronagraph alone.

LUVOIR

The LUVOIR design study is exploring an extremely large and capable general-purpose observatory, a true successor to HST and JWST. The current study covers two architectures, with an 8 m and 15 m diameter, with the 15 m architecture being on axis, whereas the 8 m architecture is off axis. The full suite of instruments operates from 100 to 2400 nm (although performance beyond 1800 nm is limited by the warm telescope), with coronagraphic operations primarily from 200 to 1800 nm. The obscured aperture of the 15 m architecture somewhat degrades coronagraph performance (see Figure 4.5), but this is offset by the larger telescope diameter, and the mission is capable of detecting and characterizing Earth-analogue planets to considerable distances.

Origins Space Telescope

The Origins Space Telescope’s (OST’s) greatest exoplanet science capabilities are in the area of protoplanetary disk characterization and transit spectroscopy, which are discussed in other subsections of this chapter. The OST study is also considering a secondary coronagraphic mode, but this would be a relatively low performance mid-infrared coronagraph, improved from JWST but still capable primarily of studying young giant planets, because the IWA of a MIR traditional monolithic (noninterferometric) telescope such as OST would be larger by a factor of roughly l/D, or a factor of roughly 10 for observations at 10 microns versus observations at 1 micron for the same diameter telescope.

Probe-Class Missions

The scaling of inner working angle with telescope diameter sets a minimum practical size for a coronagraph mission capable of studying habitable-zone Earth-size planets, while the sheer faintness of Earth analogues (which have apparent magnitudes of roughly V = 30 for systems at 10 pc) renders spectroscopic characterization challenging with moderate-size telescopes even with a starshade (which generally has higher throughput). However, absolute flux ratio floors are somewhat independent of telescope diameter, so small telescopes can still detect large planets outside the habitable zone or around very nearby stars, and characterize the brighter (giant) planets. Coronagraph missions with diameters of 1-2 m can have significant science reach for studying giant planets in wider (1-10 AU) orbits, as well as some ability to study super-Earth- or mini-Neptune-size planets, which are known to be very common (see Chapter 2).

NASA funded two studies of probe-class missions, Exoplanet Direct Imaging: Coronagraph (Exo-C; a 1.4 m telescope with an internal coronagraph) and Exoplanet Direct Imaging: Starshade (Exo-S; a low-cost 1.1 m telescope paired with a starshade, or a larger starshade to operate with WFIRST). Both showed significant science reach, capable of spectroscopically characterizing 10-20 known giant planets at high signal-to-noise ratio, discovering 1-4 R_E planets around a significant sample of stars, mapping and detecting circumstellar dust around young and mature stars, and (potentially) finding Earth-size planets around the very nearest stars. Both are significantly more capable than the current WFIRST-CGI implementation. These studies had a $1 billion cost cap; expanding slightly beyond that cost cap would further enhance their science capabilities. If WFIRST does not fly with a coronagraph or starshade, such missions would also provide risk reduction for a future large mission, while developing a scientific community and analysis techniques that will lay the groundwork for the scientific operation of a large coronagraph.

Finding: A probe-class coronagraph or stand-alone starshade mission has significant scientific capability for studying giant planets and would provide risk reduction for a future large mission.

Technology Gaps for Space-Based Imaging Missions

For the past few years, the NASA Exoplanet Exploration Program (ExEP) has been maintaining a list of technology gaps pertaining to possible exoplanet missions, working with the community to identify, track, and prioritize these gaps and, ultimately, close them via investment in technology development projects. The technology gaps are summarized in ExEP’s annually updated Technology List and captured in detail in their Technology Plan Appendix.³ A possible roadmap to advance these technologies is described in Crill and Siegler (2017) and Crill et al. (2018, white paper).

___________________

³ See NASA, “Technology Needs and Gap Lists,” https://exoplanets.nasa.gov/exep/technology/gap-lists/, and NASA, “Technology Overview,” https://exoplanets.nasa.gov/exep/technology/technology-overview.

Internal Coronagraphs

Internal coronagraph technologies include coronagraphic masks, deformable mirrors, low-order wavefront sensors, and wavefront control algorithms. The requirement for imaging an Earth-like planet around a Sun-like star is 1-2 orders of magnitude more demanding than WFIRST’s expected performance. The latter is limited by a combination of suboptimal aperture shape geometry, which is heavily obscured, and pointing specifications. Internal coronagraph performance demonstrations with clear off-axis apertures (no central obscuration) are close to achieving the 10^–10 contrast goal in the laboratory. A raw contrast of 10^–10 is the baseline requirement for HabEx architecture A (off-axis monolith).

Some of the alternative architectures for HabEx, on the other hand, are composed of segmented apertures, which also encompass all the architectures that are being considered by LUVOIR. The latter may also be on axis and so will have a central obscuration. Several efforts are under way to address the challenge of high-contrast coronagraphy on segmented and obscured telescopes, leveraging the technology and tools developed for the WFIRST-CGI. In particular, the ExEP chartered the SCDA in 2016. The study, led by the Jet Propulsion Laboratory (JPL), has gathered experts in all available coronagraph technologies.

To address the overall maturity of coronagraphs for future missions, ExEP has established the Decadal Survey Testbed (DST), intended to validate the performance of advanced coronagraphs suitable for HabEx and LUVOIR. The most promising coronagraph designs will be fabricated and tested in air and then in the DST in 2019. By 2020 the ExEP should know how well each coronagraph design would work with all architectures considered for future exoplanet missions.

Coronagraph and wavefront control technologies are critical to future direct imaging missions. However, due to the extremely low rate of photons detected from distant exoplanets, performing spectroscopy at a sufficient SNR will require the contrast to be maintained for integration periods lasting hundreds of hours. In the case of coronagraphy, this is expected to translate to sensing and controlling wavefront errors typically between 10 and 100 pm RMS for a telescope and instrument system (Nemati et al., 2017). While instrument-level laboratory demonstrations to date are within a factor of a few of this requirement, this is one to two orders of magnitude more demanding than the performance of current and upcoming space telescopes.

This level of extreme wavefront stability needs to be maintained, as the space observatory and its coronagraph experience typical environmental disturbances during operation, such as dynamic jitter and thermal drifts. Large mirrors, both monolithic and segmented, will be challenged by the need to achieve a stable back-structure, and segmented mirrors will need to maintain a large number of individual segments as a single paraboloid. Due to these tight stability requirements, coronagraphs can no longer be designed as separate payload instruments and, instead, need to be designed along with the observatory as a single system. Ongoing analyses by the HabEx and LUVOIR study design teams are determining the best approach to these challenges for space-based telescopes of a range of sizes. Their work, and the assessment thereof, will determine the likelihood that these telescope systems can meet these very demanding wavefront error stability requirements.

Starshades

Starshade technology is currently being advanced under a single ExEP technology development activity whose objective is to advance five key technologies to technology readiness level (TRL) 5. WFIRST is being used as a reference mission for the design and engineering work (a starshade, however, is not baselined for WFIRST). While a starshade’s optical performance can never be demonstrated at full scale on the ground, a preliminary assessment (Seager et al., 2015) has developed design models with error budgets predicting better than 10^–10 contrast.⁴ A subscale validation demonstration has already achieved 4.6 × 10^–8 starlight suppression at flight Fresnel numbers, and is expected to demonstrate the 10^–9 starlight suppression goal in 2018. However, to test at these regimes and operate within a practically sized testbed, the demonstration is being conducted with only a 25 mm starshade. (Note that the testbed is already 77 m long; testing larger size starshades requires very long testbeds, as the required separation between the starshade and “telescope” increases by the square of the starshade radius.)

___________________

⁴ See NASA, “Probe Scale Science and Technology Definition Teams,” https://exoplanets.nasa.gov/exep/studies/probe-scale-stdt/.

Hence, the confidence with which these subscale starshade demonstrations represent full-scale performance will depend on their ability to validate their performance models. Although scalar diffraction theory predicts optical performance to be independent of scale, wavelength-scale features in small starshades could introduce additional effects, and their final performance remains uncertain. Additional suppression demonstrations are planned to be completed by CY20 at different wavelengths, starshade sizes, and a range of key perturbations to demonstrate the reliability of the models.

Another key starshade technology that is currently being advanced is the reduction of the scattering of sunlight off the starshade’s petal edges. Materials that are sufficiently thin, have low reflectivity, and are suitable for stowage are being investigated as “optical edges.” Amorphous metals are a promising candidate and are currently being tested. Unlike other large structural deployments, the starshade requires precise and stable positioning of a 30 m structure (or larger) to better than 1 mm. A half-scale or larger prototype is currently being planned, in order to demonstrate that it meets deployment tolerances.

In the case of a starshade-only mission, telescope stability requirements are significantly looser and do not exceed the state of the art. Solutions for sensing and alignment control between the two spacecrafts have been developed, and subscale demonstrations are being conducted in the laboratory. Thus, the technology development for missions that utilize starshades falls primarily on the starshade itself, and not on the optical telescope assembly.

Low-Noise Detectors for Space-Based Applications

The low flux from small exoplanets requires a detector with read noise and spurious photon count rate as close to zero as possible, which also maintains adequate performance for many years in the space environment (Rauscher et al., 2016). The state of the art is dependent on the wavelength band, but detectors need to perform at or near the photon counting limit from the UV through the NIR for current mission concepts. Across this wavelength range, the state-of-the-art detectors are semiconductor-based devices. WFIRST’s electron multiplying charge coupled device (EMCCD) detectors have achieved adequate noise performance in the visible band, although longer lifetime in the space radiation environment and higher quantum efficiency at long wavelengths are needed. Similar EMCCD devices, with delta doping, may already have adequate performance in the near UV. HAWAII4 HgCdTe detectors with multiple-readout noise levels of a few electrons are the state of the art in the NIR, but photon-counting NIR detectors may also be required. Avalanche photodiode arrays are a promising approach but currently at low TRL. Energy-resolving superconducting sensors, such as MKIDs, are also a potential option, but at low TRLs for spaceflight. JWST/Mid-Infrared Instrument (MIRI)’s detectors are expected to establish the state of the art in MIR detection sensitivity, and future MIR direct imaging is likely to require detectors that exceed it. It is likely that the detection sensitivity gap can be closed in the next decade, as a range of choices are close to meeting the requirements.

Development Beyond the New Worlds Vision

Practically, the vision laid out above for space-based missions focuses on UV to near-infrared characterization of planets at low spectral resolution, primarily enabling the measurement of (for terrestrial planets) Rayleigh scattering, oxygen, ozone, water vapor, and potentially methane and carbon dioxide. Significant additional information about planetary atmospheres (see Chapter 3) is available at longer wavelengths or higher spectral resolutions. However, both are technically challenging; in simple l/D scaling, a thermal-IR (10 micron) telescope would have to be 10-20 times larger than a visible-light optimized telescope to achieve the same diffraction limit. In practical terms, this would likely have to be implemented as either a sparse aperture or a multispacecraft interferometer such as the TPF-I or Darwin concepts. Moderate- to high-resolution spectroscopy of all but the nearest and brightest planets similarly requires larger collecting areas. Ultimately, the details of such future characterization will be informed by the discoveries that will be made over the next decade. While the core path flows to the large strategic exoplanet imaging mission discussed in earlier sections, it is important to retain the flexibility in planning, implementing, and operating the missions of the next two decades to allow them to respond to the evolution of the scientific and technological landscape. It is also important to develop the technology that will be needed for the successor missions 20 years from now.

Finding: Technology development support in the next decade for future characterization concepts such as MIR interferometers or very large/sparse apertures will be needed to enable strategic exoplanet missions beyond 2040.

OPPORTUNITIES TO CHARACTERIZE PLANETS THROUGH TRANSITS

The transit technique has now eclipsed RVs in the number and dynamic range of detected exoplanets. NASA’s Kepler mission was a resounding success, and has revealed the demographics of planets on close-in orbits as well as discovering numerous exoplanets orbiting in the liquid water habitable zones of their host stars. Building on these successes, the next decade of transiting exoplanet science involves new missions to provide a more complete census of transiting planets orbiting bright stars, and a push toward a statistical census of Earth-size habitable-zone planets orbiting Sun-like stars. These missions will provide many of the targets for atmospheric characterization efforts, which rely on high SNR to detect the minute signature of atmospheric absorption, emission, or scattering in a differential sense against the bright background of a much larger host star. Below, the committee outlines first the upcoming transit survey missions and how they will alter the landscape of the known transiting exoplanet population toward fulfilling the goals of the Exoplanet Science Strategy. It then presents the next decade and beyond of observational efforts to characterize the atmospheres of transiting exoplanets.

Planet Discoveries Through Transits

Transiting Exoplanet Survey Satellite (TESS)

NASA’s TESS mission launched successfully in April 2018 and will shortly begin its near-all-sky survey for transiting planets (Ricker et al., 2015). TESS will mainly be sensitive to close-in planets and will find approximately 2000 planets with radii less than Neptune (see Figure 4.6; Barclay et al., 2018). This will give the opportunity to explore the statistics of the mass-radius diagram (in concert with the radial velocity technique; see the section “Exoplanet Masses,” later in this chapter) and provide essential targets for large atmospheric studies in the near term with JWST, and the longer term with the Atmospheric Remote-Sensing Exoplanet Large-Survey (ARIEL; see the section “Atmospheric Characterization Through Transit Spectroscopy,” earlier in this chapter, and Tinetti et al., 2017) and the GSMTs. TESS will also find terrestrial planets around M dwarfs that will be key targets for studies of temperate terrestrial planets, with JWST and the GSMTs playing a key role. An extended TESS mission could do important work maintaining precise ephemerides for follow-up observations, discovering longer period planets, or covering the portions of the sky omitted in the original survey (Huang et al., 2018).

Ground-Based Photometry

In the era of large space-based transit surveys, ground-based photometric observations of transiting planets and their host stars have a continued role to play in several senses, as follows:

• Ground-based surveys such as MEarth and Transiting Planets and Planetesimals Small Telescope (TRAPPIST)/Search for Habitable Planets Eclipsing Ultra-Cool Stars are designed for planet detection in regimes of parameter space that provide unique sensitivity from the ground and can detect habitable-zone planets orbiting the smallest stars, which are typically too faint for TESS.
• Ground-based follow-up to TESS (and Planetary Transits and Oscillations of Stars [PLATO]; see the next section) candidates will be necessary for vetting and confirming candidate exoplanets by ruling out systems of blended stars. This would include high-resolution imaging to spatially resolve multistar systems (or chance alignments), which are expected to be prevalent in TESS data due to the large pixel size; and multiband transit photometry to search for color dependencies in transit depth that are indicative of stellar blends.
• Long-term monitoring of transiting systems, possible only from the ground, will enable the search for additional planets in systems with one or more known close-in planets. Such monitoring can also provide ephemeris maintenance that is necessary to accommodate future atmospheric characterization follow-up studies.

The latter two points in this list would benefit from a coordinated network of ground-based telescopes with broad sky coverage purposed with monitoring transiting exoplanet hosts and following up on exoplanet candidates. Existing examples of such networks include the Las Cumbres Observatory Global Telescope⁵ and the Kilodegree Extremely Little Telescope Follow-Up Network (Collins et al., 2018).

Planetary Transits and Oscillations of Stars (PLATO) Mission

The European PLATO mission is a dedicated transit survey scheduled for launch in 2026 to an L2 halo orbit.⁶ The mission design consists of a 0.59 m telescope with 24 cameras, with a total sky coverage of up to 50 percent, and a nominal mission duration of 4 years (with consumables lasting up to 8 years). Its observing strategy is being designed to back a science goal of detecting Earth-size habitable-zone planets orbiting Sun-like stars. All primary science targets with V < 11 would also be characterized asteroseismologically, thus providing precise radii and ages for the host stars and their planet candidates. An accompanying ground-based RV follow-up campaign is also envisaged to measure planetary masses. The exact observing strategy is still being developed, with trade-offs being considered between sky coverage and duration of observations on individual fields.

___________________

⁵ For additional information, see the Las Cumbres Observatory Global Telescope website at https://lco.global/.

⁶ For additional information on PLATO, see ESA (2017).

The unique transiting planet science accomplished by PLATO will include improving knowledge of the occurrence rate of Earth-like planets orbiting Sun-like stars over that provided by Kepler; delivering a sample of habitable-zone terrestrial planets orbiting bright G and K stars for future atmospheric characterization with transit or direct imaging techniques; and mapping out the evolution of planet occurrence rates and properties as a function of age, by providing precise host star dating via asteroseismology.

Atmospheric Characterization Through Transit Spectroscopy

To date, the characterization of exoplanet atmospheres has occurred primarily through transit techniques (transit and secondary eclipse spectroscopy, and phase curves, which can capture both the transit and secondary eclipse). Atmospheric characterization efforts have therefore been focused on the population of close-in transiting exoplanets with large planet-star radius ratios. While large space-based direct imaging missions are currently being conceived of that would obtain spectra of terrestrial and gas giant planets at wider separations, transit spectroscopy techniques will remain the primary mode for atmospheric characterization over the next decade and beyond.

James Webb Space Telescope (JWST) Mission

JWST provides photometric and spectroscopic wavelength coverage from 0.6 to 28 microns (see Figure 4.7). JWST will provide high-precision spectroscopic characterization of transiting exoplanets across the near- and mid-IR, which can be used to investigate many intriguing science questions that could only previously be addressed in a limited manner by ground-based instruments, HST, and Spitzer. For example, JWST’s wavelength coverage allows it to probe many spectroscopically interesting molecules including major oxygen-, carbon-, and nitrogen-bearing species. This capability will allow for accurate measurements of atmospheric metallicities and elemental abundance ratios (e.g., C/O) as a tracer of planet formation and evolution, and will provide a window into nonequilibrium chemical processes in exoplanet atmospheres. JWST will also yield access to smaller and cooler planets than can be targeted with HST and Spitzer due to its larger aperture and longer wavelength coverage.

Ideal targets for atmospheric characterization with JWST are planets orbiting bright host stars. Hot giant planets are straightforward atmospheric targets for JWST, and in many cases a single transit or eclipse will return a high-SNR exoplanet spectrum (Figure 4.8). Warm Neptune- and sub-Neptune-size planets orbiting smaller stars are also high-quality JWST targets. Archetype objects in this class include GJ 436b and GJ 1214b, and it is expected that TESS will discover on order of 100 additional similar-size planets from which additional JWST targets will be drawn (Kempton et al., 2018). Temperate terrestrial planets remain a significant challenge for JWST; for Sun-like stars such planets will remain out of reach, and for M dwarfs studies are envisioned but will require many transits or eclipses to build up a spectrum at moderate SNR (Morley et al., 2017). Nevertheless, JWST should make significant progress toward determining if terrestrial planets orbiting late-type stars can retain atmospheres.

The exoplanet community has the ambition to characterize the atmospheres of a large number of transiting planets to reveal trends with planet mass, size, level of irradiation, and host star properties, which should inform planetary formation and evolution, and atmospheric physics and chemistry (Chapter 3). There are several challenges to accomplishing this goal. First, with a limited lifetime shared-resource facility there is a trade-off between characterizing many easy-to-observe hot giant planets versus using substantial observing time to perform detailed characterization of a much smaller number of terrestrial exoplanets. Second, there is the issue of the time allocation process itself. It is unknown what fraction of JWST time will ultimately go to exoplanet investigations, but if the history with HST provides some guidance, that number might fall around 20 percent.

The JWST launch delay means that most TESS-discovered planets will be in hand prior to the commencement of regular JWST science operations. With hundreds of high-quality atmospheric characterization targets to choose from, multiple choices of observing modes and wavelength coverage, and many competing research groups that have spent years eagerly awaiting the launch of JWST, one might expect an onslaught of observing proposals in the early cycles of the JWST mission with no clear overarching science vision. This leads to a third challenge, which is one of community organization. It would have a powerful impact if the transiting exoplanet community could come together behind a shared strategic vision of atmospheric characterization science with JWST. The resound-

ing success of the Early Release Science (ERS) program with JWST has shown that such an effort is possible among exoplanet researchers (Bean et al., 2018). The “Transiting Exoplanet Community Early Release Science Program” was a highly ranked proposal in the competitive ERS selection process, and it garnered the largest time allocation of any of the selected proposals. The proposal team consisted of 61 investigators and 43 collaborators, representing multiple individual (and competing) research groups.

The Exoplanet Science Strategy outlined in this report provides some of the framework for a shared strategic vision in atmospheric characterization with JWST. For the first time JWST will bring exoplanet atmospheric characterization efforts from a regime of limited observations to one of high-fidelity spectroscopic investigations of a comparative sample (Cowan et al., 2015). To take advantage of this opportunity, JWST should undertake a strategic and systematic survey of exoplanet atmospheres that will benefit the entire research community and guide future observing strategies for years, if not decades (Cowan et al., 2015). While a mechanism is not currently in place for key science programs with JWST, it would behoove the exoplanet community to perform such a survey as early in the mission as possible, to inform future science strategies both in later JWST cycles and in the development of next-generation facilities.

Recommendation: NASA should create a mechanism for community-driven legacy surveys of exoplanet atmospheres early in the JWST mission.

Future Large Missions for Transit Spectroscopy Beyond JWST

Despite JWST’s substantial power for transiting planet observations, the committee concludes that some significant science questions will remain in this topic after the mission is completed, particularly because JWST will not have sufficient collecting area to probe a large number of habitable planets and also because it will not cover all the wavelengths of interest. The OST, LUVOIR, HabEx, and Lynx mission concepts would all probe exoplanet atmospheres in new ways given their unique wavelength coverage compared to JWST, while an optical or infrared mission with substantially larger collecting area (e.g., LUVOIR or a large version of OST) would expand the number of potentially habitable planets that can be studied.

The characterization of a large number of potentially habitable planets will likely be out of reach for JWST (Louie et al., 2018; Batalha et al., 2018). The expected outcome is that JWST will at best be able to detect molecules for only a handful of especially favorable potentially habitable planets (i.e., planets around mid to late M dwarfs within 10 pc). Among the current mission concepts being studied in preparation for the decadal survey, OST and LUVOIR could make further progress on the topic of terrestrial planet habitability using transit techniques.

OST is being designed specifically to extend JWST transmission and emission spectroscopy science with observations at wavelengths between 5 and 25 microns in a single pointing. The spectral coverage and sensitivity of OST will enable the detection of molecules that govern planetary climate, such as carbon dioxide and water, as well as the exciting biosignature combination of ozone and methane. Although JWST will have spectroscopic capability for wavelengths longer than 12 microns (see Figure 4.7), it is currently anticipated that these will not perform well for transit spectroscopy, because the Medium Resolution Spectroscopy modes of MIRI are fed with an integral field unit that has relatively small entrance apertures. Thus, it is expected that classical transit spectroscopy will not be possible with these modes due to uncorrectable slit losses (but see Snellen et al., 2017). OST is expected to be capable of transit spectroscopy at longer wavelengths (l > 12 microns) compared to JWST. This opens up the possibility of seeing past aerosol scattering and determining aerosol compositions through the detection of vibrational transitions. Nevertheless, the committee notes that Earth’s thermal emission peaks at 10 microns and the spectra of temperate terrestrial planets drop precipitously beyond 18 microns due to the decline of thermal blackbody emission and strong water absorption. JWST does have photometric capabilities that extend beyond the MIRI Low Resolution Spectrometer cutoff, and since observations of potentially habitable planets at these wavelengths will likely be limited by photon counting noise, OST spectroscopy would offer only a modest improvement over JWST in this area.

Unlike JWST, OST is designed from conception to have low levels of systematic noise for time-series observations. However, the noise floor of JWST is unknown and the success of the community in using HST and Spitzer, which are much less optimized for transit observations than JWST, leads the committee to be optimistic about its expected performance. Ultimately, the limiting factor for transit spectroscopy of potentially habitable planets will be the number of stellar photons that can be collected, and the committee finds that JWST-size and smaller telescopes do not have sufficient light-gathering capabilities to study a large number of such objects. Therefore, the committee concludes that OST would need a much larger aperture to yield a transformative advance over JWST unless it is found that JWST’s reach is severely limited by time-series systematics.

LUVOIR would also advance transiting planet science in the post-JWST era. LUVOIR would reopen the UV and optical wavelength range (0.1 < l < 0.6 microns) for transiting planet observations that will close once HST is no longer operational, and it would maintain optical to near-infrared (0.6 < l < 2.5 microns) capability once the JWST mission ends. The short wavelength range is important for exploring planetary mass loss and characterizing aerosol scattering, and a substantial number of gas-phase absorbers for potentially habitable planets will also be accessible in the optical and near infrared, including carbon dioxide, water, and the biosignature combination of molecular oxygen, ozone, and methane. The addition of a high-resolution optical or infrared spectrograph on LUVOIR would yield a substantial advance for transit observations using the cross-correlation technique because of the absence of contaminating telluric lines. Similar to OST, LUVOIR’s exact advantages compared to JWST beyond its unique bandpass and potential for high spectral resolution will depend on the differences in collecting areas of the telescopes and JWST’s time-series systematics, the latter of which will not be ascertained until the mission becomes operational.

Finding: The combination of transiting planet detection with TESS, mass measurements with radial velocities, and atmospheric characterization with JWST will be transformative for understanding the nature and origins of close-in planets. Future space missions with broader wavelength coverage, a larger collecting area, or reduced instrumental noise compared to JWST would have greater reach to potentially habitable planets.

Atmospheric Remote-Sensing Exoplanet Large-Survey (ARIEL) Mission

The European ARIEL⁷ mission is a dedicated transit spectroscopy experiment planned for launch in 2028. ARIEL will comprise a 1 m-equivalent telescope equipped with an infrared spectrograph, giving spectral coverage from 2.0 to 7.8 microns in a single pointing. ARIEL will carry out a nominal 4-year mission at Lagrange 2 point (L2), with a possible extension to 6 years.

ARIEL’s goals are to characterize the atmospheres of approximately 1000 planets in order to search for trends that can be revealed only statistically. By comparison, JWST is expected to characterize the atmospheres of approximately 100 exoplanets, but at significantly higher precision (Cowan et al., 2015). The trends that ARIEL will characterize include the relationship between atmospheric metallicity and planet mass, the distribution of atmospheric carbon-to-oxygen ratios, and the variations of energy transport and thermal structure with irradiation level. Ultimately, ARIEL provides a wide and shallow survey that complements JWST’s more narrow and deep observations. Furthermore, ARIEL’s capability to observe a large number of sub-Neptune to giant planets would allow JWST to focus on the smaller and cooler planets that it is uniquely capable of studying.

NASA is currently studying the possibility of contributing to the ARIEL mission by providing fine guidance sensors (FGSs) for the required precise pointing control and as auxiliary science instruments (the Contribution to ARIEL Spectroscopy of Exoplanets [CASE] mission concept). The FGSs would add spectrophotometric wavelength coverage from 0.55 to 1.90 microns, which would provide information on the albedos of planets and the presence of aerosols. U.S. scientists would benefit from the CASE mission by participating in the planning, execution, and exploitation of the ARIEL survey.

___________________

⁷ For more information on ARIEL, see European Space Agency, Science and Technology, “Ariel,” http://sci.esa.int/ariel/, and the Ariel Space Mission website at https://ariel-spacemission.eu/.

Finding: By conducting a statistical survey of exoplanet atmospheres, the European ARIEL mission will provide broader context for more focused JWST observations. The U.S. exoplanet community would benefit from participation in ARIEL.

Giant Segmented Mirror Telescopes

Transit spectroscopy observations are challenging from the ground because of the inherent variability of Earth’s atmosphere, its opaqueness at many infrared wavelengths, the thermal emission at mid-infrared wavelengths, and the lack of continuous observability of the night sky at most locations on Earth imposed by its daily rotation. However, the photon-collecting power of the upcoming GSMTs will be a factor of 15-36 larger than that of JWST. Ground-based telescopes offer a more stable instrument environment at lower cost than space telescopes, more flexible instrument setups, and the potential to deploy new instruments on a short time scale. Therefore, ground-based observatories have certain advantages over space-based platforms, despite the challenge of observing on Earth and, in particular, through Earth’s atmosphere.

Accurate calibration of transmission or secondary eclipse spectra can be obtained by simultaneously observing a nearby reference star (Bean et al., 2010a), but this is unlikely to rival future space-based observations, except for very faint targets. A particularly powerful technique involves high-dispersion spectroscopy using spectrographs with resolving powers of R > 25,000 (Figure 4.9; Snellen et al., 2010). While this technique does not preserve spectroscopic broadband features, it allows accurate instantaneous calibration of Earth’s atmosphere and measures the contrasts of atomic and molecular lines in the planet spectrum. The signal of many lines can be combined to boost the signal of a particular molecule, which is identified through cross correlation of the observed spectrum against a model planet spectrum. The technique can be used for transmission spectroscopy, but can also target the emission and reflection spectra from nontransiting planets. The change in the radial component of the orbital velocity of the planet can be used to disentangle the stellar and telluric components of the spectrum from that of the planet. This technique can further be used to constrain the orbital inclination of nontransiting planets, thereby

allowing a measurement of the true mass (rather than the minimum mass) of a planet detected only by the radial velocity method.

The GSMTs will provide a major step forward for this method. The atmospheres of rocky planets in the habitable zones of nearby M dwarfs are within reach, including the detection of carbon dioxide, methane, water, and molecular oxygen (Maiolino et al., 2013; Rodler and Lopez-Morales, 2014). More generally, the method is complementary to space-based observations in the following senses:

• The reliance on molecular transitions results in unambiguous molecular identifications.
• The sensitivity to Doppler effects means it can measure the atmospheric dynamics and spin of the planet (e.g., Snellen et al., 2010, 2014; Rauscher and Kempton, 2014).
• Lines of different oscillator strengths probe different altitudes in a planet atmosphere, providing a unique measure of the atmospheric temperature structure. For dayside emission this can be used to map the variation of the atmospheric temperature with longitude.
• The method probes atmospheres at lower pressures than low-dispersion spectroscopy, making it less affected by clouds and hazes (e.g., Kempton et al., 2014).
• The method could allow for the detection of different molecular isotopologue ratios, such as HDO/H₂O, providing unique insight in the formation history of exoplanets.

The following relevant instruments are currently envisaged for the GSMTs: For the GMT, the Giant Magellan Telescope-Consortium Large Earth Finder will be a high-resolution, highly stable, fiber-fed visible light échelle spectrograph operating from 350 to 950 nm with spectral resolutions up to 120,000, and the Giant Magellan Telescope Near-Infrared Spectrograph will cover 1-5 microns with a resolution up to 100,000. TMT could select a second-generation instrument, High-Resolution Optical Spectrometer or Near-Infrared Echelle Spectrometer, with high-dispersion capabilities in the optical or near infrared, respectively. On the European ELT, the Mid-Infrared E-ELT Imager and Spectrograph will contain an R = 100,000 Integral Field Unit for the L and M bands, and the High Resolution Imaging Spectrometer, an optical-near-IR high-dispersion spectrograph, is currently under study as a second-generation instrument. Novel instrument designs (e.g., Ben-Ami et al., 2018) focused on this method promise significant increases by restricting the wavelength range and boosting throughput.

Finding: GMT and TMT, equipped with high-resolution optical and infrared spectrographs, will be powerful tools for studying the atmospheres of transiting and nontransiting close-in planets, and have the potential to detect molecular oxygen in temperate terrestrial planets transiting the closest and smallest stars.

EXOPLANET MASSES

As discussed in Chapter 3, knowledge of the masses of exoplanets is essential both as a diagnostic of planetary nature and as a tracer of planetary origins. Mass is the most fundamental property because of its role in planetary structure and evolution; it governs a planet’s ability to form and retain an atmosphere, the strength of its dynamical interactions with other bodies, and its initial gravitational energy.

Combined with the transit technique, precise masses constrain the bulk compositions and interior structures of transiting exoplanets by yielding their densities. Kepler/K2, TESS, PLATO, and the Characterising Exoplanets Satellite (CHEOPS) missions will deliver thousands of transiting planets over the coming decade, giving the opportunity to expand studies of the planet bulk composition in terms of host star mass, age, composition, and planetary orbital separation. These missions will also identify potentially habitable planets, which are compelling targets for mass measurements in order to reveal which ones are rocky.

Masses are also needed to constrain surface gravities and atmospheric scale heights, which are key boundary conditions for interpreting spectra obtained through either the transit or direct imaging techniques. Thus, exoplanet mass measurements are important for the success of JWST, ARIEL, the GSMT exoplanet studies, and a direct imaging mission. The scientific return of direct imaging will benefit tremendously if the target masses

can be determined, because direct imaging generally does not yield masses or radii, leaving such information to be extracted indirectly from spectra and with a heavy dependence on atmospheric and/or evolutionary models.

Direct imaging missions would benefit from precursor observations that not only measure planetary masses but also identify targets with planets first and map their orbits. The reason for this is that a blind direct imaging survey for planets, and Exo-Earths especially, generally spends a lot of time missing planets because they are inside the inner working angle of the instrument or are in crescent phase and are too hard to detect (Stark et al., 2014). Radial velocity measurements could predetermine the most promising candidates and, if performed contemporaneously (Savransky et al., 2009), would save observing time by targeting epochs when companions are known to be maximally separated from their stars (Crepp et al., 2016). The desire to measure the masses of giant planets beyond the snow line, to measure the masses for directly imaged giant planets, and to illuminate the formation histories of planetary systems that include directly imaged terrestrials motivates radial velocity monitoring over decades (Montet et al., 2014; Howard and Fulton, 2016).

The radial velocity technique remains the principle method for measuring planet masses because of its wide applicability. The transit timing variation (TTV) technique is useful for the subset of compact, multiplanet systems that are in or near resonance. Astrometry has not been widely used to date, but results from Gaia are eagerly anticipated.

Radial Velocities

Since the last decadal survey researchers have entered the era of extremely precise radial velocities (EPRVs), with detections of planetary signals of 1 m/s. For Sun-like stars, this corresponds to planets with masses as small as several times that of Earth, but only for orbital periods of several days. Although impressive (approximately 0.001 pixel Doppler shift recovery), this level of precision is insufficient to detect the gravitational perturbation induced by an Earth-Sun analogue.

For rocky planets in the habitable zones of mid to late M dwarfs, the RV signal is several meters per second, and a number of such measurements have been made (e.g., Anglada-Escude et al., 2016), including one transiting example (LHS1140b; Dittmann et al., 2017). A number of specialized instruments focused on discovery and mass determination for M-dwarf planets are recently commissioned or coming online in the next year (see Fischer et al., 2016; Wright and Robertson, 2017). These facilities will play an essential role in determining the density and surface gravities of the potentially habitable planets transiting the closest small stars, which will be scrutinized by JWST and the GSMTs.

Despite the efforts of many teams, the state of the art in Doppler precision has experienced only marginal improvements in the last 5-10 years, reaching an apparent plateau in performance just beneath 1 m/s for bright stars (Fischer et al., 2016). Earth-mass planets in the habitable zones of Sun-like stars require an improvement in radial velocity precision to the cm/s level (Figure 4.10). This is beyond the reach of the expected performance of instruments being built today, including the high-precision Doppler spectrograph under development by the NASA-NSF Exoplanet Observational Research partnership (Schwab et al., 2016). Furthermore, even if spectrographs reach this internal precision, mass measurements will likely be limited by two noninstrumental effects: stellar variability and contamination from telluric lines.

Overcoming Stellar Variability

The surfaces of Sun-like stars are mottled by numerous effects, including convective granulation and magnetic phenomena such as spots, plage, and faculae. As these effects evolve in time, and as features rotate into and out of view on the differentially rotating surface, these inhomogeneities create spurious Doppler shifts that severely complicate the interpretation of RV time-series measurements (e.g., Dumusque et al., 2014). A distinguishing factor between stellar variability and the purely translational shift caused by an orbiting planet is that surface features create changes in the detailed shapes of absorption lines, in ways that often vary with wavelength (e.g., Queloz et al., 2001). Future EPRV instruments need to therefore have sufficient spectral resolution (R > 100,000 at a minimum, and R > 200,000 ideally), spectral coverage, SNR, and observing cadence to measure, and subsequently correct for, stellar variability (Fischer et al., 2016; Davis et al., 2017).

Small telescopes (D < 4 m) will not be able to satisfy all of the above requirements since they likely cannot achieve the requisite SNR at the required dispersion in the required integration time. Moreover, measurements carried out with sufficient cadence to explicitly resolve the various time scales of stellar variability will enable separating these signals from planetary signals. For example, pressure-mode (p-mode) oscillations have a characteristic time scale of 5 minutes for the Sun (Koen et al., 2003). Current observing strategies purposefully alias the signal by targeting the nodes of pulsation time series, or they simply average over them by exposing for significantly longer than the characteristic time scale of these oscillations. Resolving this signal would enable removing it more effectively, but requires short integration times. This, coupled with the need for high SNR at high spectral resolution, motivates the use of larger aperture telescopes. Dedicated facilities should be considered, as acquiring sufficient observing time on large-aperture telescopes typically proves difficult (Plavchan et al., 2014).

The standard approach of tracking stellar activity using familiar activity indicators (e.g., changes in the equivalent widths of the Ca H and K lines) has generally proven to have limited value. New methods should be developed that attempt to empirically model and correct for stellar variability in time. Theorists studying magnetohydrodynamics, stellar activity, stellar astrophysics, and heliophysics should work closely with EPRV survey teams to model absorption line profiles (Wright and Sigurdsson, 2018). Observing the Sun simultaneously with spatially resolved spectroscopy and in hemisphere-integrated light (e.g., Haywood et al., 2016; Dumusque et al., 2015) offers the opportunity to uniquely identify the surface features that result in apparent RV variations, and thus a path for comparing theoretical calculations to disk-integrated measurements. The goal of such work would be to identify metrics that could be used to correct the RVs for other stars. A more fundamental understanding of how stellar variability behaves in time and as a function of wavelength and stellar properties should be sought. Further, novel statistical methods employing principal component analysis and machine learning should be developed and

refined to disentangle the signal of the Doppler shifts of absorption lines due to orbiting planets from the variability of absorption lines due to stars (Davis et al., 2017) or the atmosphere (see the next section). Such work may find that contemporaneous photometry is an essential input into the analysis (e.g., Haywood et al., 2014) to push down to a precision of several centimeters per second.

Telluric Contamination

Spectral contamination from telluric lines is a problem for EPRV measurements because their variation in strength and position over time scales ranging from minutes to seasons can lead to spurious radial velocity shifts much larger than the sought-after planetary signals. There is a growing sense in the EPRV community that small telluric lines, so-called micro-tellurics, are a limiting factor for ground-based observation. Macroscopic telluric lines are easily identified and masked in the radial velocity measurement process in the optical, although they remain a challenge for near-infrared radial velocities (Bean et al., 2010b). By contrast, micro-tellurics require higher spectral resolution and SNRs to identify, and ultimately may be so prevalent that they cannot all be masked even if they could all be identified. Micro-tellurics affect between 4 and 10 percent of pixels, depending on the (variable) water content of the atmosphere at depths that matter for the requisite SNR of up to 300. New data-driven techniques are emerging that may circumvent this problem by modeling the lines directly. The higher spectral resolution instruments on larger telescopes motivated by the stellar variability problem will also help the telluric contamination issue. Measurements from space might be a final option if the telluric contamination problem cannot be solved. However, as described above, large apertures (D > 4 m) and significant time investment are required to solve the pernicious challenge of stellar variability, even for space-based measurements.

Instrument Stability

Sub-milli-Kelvin thermal stability and vacuum levels below P = 10^–6 Torr are needed to generate sub-meter-per-second single measurement precision (Mahadevan et al., 2014). The opto-mechanical footprint of seeing-limited spectrographs that generate R > 100,000 for 8-10 m class and larger telescopes are a challenge to stabilize at this level, as the beam diameter grows with the size of the telescope. Novel technological methods to calibrate instrument drift, including combinations of a Fabry-Pérot etalon, laser-frequency combs, and emission lamps, have continued to advance. Using adaptive optics to inject starlight into small fibers would help keep EPRV instruments to manageable sizes. Diffraction-limited systems that use single-mode fibers show significant promise (Crass et al., 2018, white paper). The same technology could be used for high-dispersion coronagraphy of directly imaged planets (Jovanovic et al., 2018, white paper).

Rigorous Error Budgets

EPRV instruments present a complicated systems engineering challenge. Interpreting comprehensive error budgets to make informed design decisions is hindered by the fact that many sources of uncertainty have comparable values (Halverson et al., 2016). While provisional error budgets have been put together for recent instruments designed to generate sub-meter-per-second precision, these frameworks have largely not yet been validated empirically nor do they possess demonstrated predictive power. Reliable and quantitatively rigorous error budgets will be needed to move beyond the current generation of instruments that are coming online.

Collaboration

The EPRV community has historically been divided into numerous small teams that compete against one another for resources and generally do not share data, and yet it is widely recognized that properly addressing stellar variability requires more coordination and more resources than any one group of investigators can marshal. This approach is nonoptimal scientifically, tends to result in duplication of work in a field that has limited resources, and should be enhanced with a more collaborative and sustainable framework that encourages an open network

of interinstrument sharing of spectra, data analysis methods, design concepts, and hardware development efforts, while at the same time protecting proprietary periods for data acquired by individual investigators.

Summary of RV Needs and the Motivation for an EPRV Initiative

The EPRV field needs an initiative that fosters community engagement and activities that promote collaboration and combines efforts and expertise for Doppler instrumentation, follow-up measurements, survey planning and execution, and data analysis techniques. Such an initiative should also strategically encourage the free exchange of ideas between the above-mentioned subdisciplines in stellar astrophysics, heliophysics, instrumentation, and statistics for overcoming the effects of stellar variability. An EPRV initiative could undertake the coordinated, sustained effort to tackle the myriad of error terms that currently limit RV precision.

Efforts to understand and address the subtle physical effects that manifest at the 1 cm/s level (10-sigma detection of Earth analogue) have thus far been attempted piecemeal by individual investigators. In addition to standard principal investigator (PI)-led programs, a comprehensive effort at the national level with serious capital investment is required to address the multifaceted problem inherent to measuring ultraprecise stellar RVs. A dedicated initiative is needed that organizes and coordinates the efforts of exoplanet researchers from diverse scientific backgrounds in hardware, software, operations, and related topics.

Rather than a brick-and-mortar institute located at a single physical address, the initiative should be a strategic virtual program that benefits from the contributions of researchers originating from different geographic regions and areas of expertise. The initiative should rely on NASA’s experience in developing rigorous error budgets, performing trade studies, systematically working to solve challenging technical problems, and overseeing and managing numerous research teams under a common programmatic goal.

Although EPRVs were highlighted previously in the 2010 Decadal Survey, the resources allocated for ground-based support of Kepler, TESS, and forthcoming space missions have been insufficient (Plavchan et al., 2015). The scientific return from space transit missions will be increased dramatically if more resources are committed to these efforts before, during, and after launch. Ground-based measurements should be carried out leading up to the launch of a direct imaging space mission, irrespective of whether the coronagraph technology used involves an internal or an external occulter. Importantly, NASA has focused its precious resources on measuring the masses of discovered planets (which the committee emphasizes is an important goal), but not on the challenge of improving the current state of the art of EPRV itself. Here, the committee advocates that together NASA and NSF take on the grand challenge of achieving the precision required to measure the masses of terrestrial planets orbiting Sun-like stars. The goal of the field should be to reach σ = 10 cm/s single measurement precision from the ground in the near term; ultimately, control of systematics at the level of roughly 1 cm/s is needed to study Earth/Sun analogue systems. While such measurements will likely be done from the ground, they are inextricably linked to the scientific success of the numerous current and proposed NASA missions, including Kepler/K2, TESS, and a future large imaging mission. NASA has tackled formidable technology challenges in the past in pursuit of its scientific goals; here, the committee advocates that this same coordinated effort be aimed at RVs.

A next-generation EPRV initiative building on the success of the NASA and NSF collaboration that yielded the NEID project is urgently needed. Research teams should work together on comprehensive databases that self-consistently combine the longitudinal coverage and high-cadence measurements from multiple instruments and telescopes forming a global network (Brown et al., 2013). One of the top priorities for such an initiative would be the development of a comprehensive and rigorous EPRV error budget that is empirically validated in the laboratory and at the telescope, and has predictive power for arbitrary spectrograph designs. The error budget should consider changes in RV stability over long periods of time and wavelength calibration standards. Mitigation schemes for stellar variability and differential telluric absorption are an immediate need and should be explored by theorists, statisticians, observers, heliophysicists, and instrument builders. Comprehensive studies across both the visible and near-infrared portions of the spectrum should be performed to compare wavelength calibration, stellar variability correction, injection of starlight into tiny fibers (including single mode), and compensation for differential telluric absorption. The EPRV initiative should culminate in the development and use of next-generation facilities that deliver cm/s radial velocity measurements. Ultimately, pushing RV precision and accuracy to several cm/s

is a grand challenge in this field that requires a centrally coordinated approach to systematically address each source of noise.

Finding: The radial velocity method will continue to provide essential mass, orbit, and census information to support both transiting and directly imaged exoplanet science for the foreseeable future.

Finding: Radial velocity measurements are currently limited by variations in the stellar photosphere, instrumental stability and calibration, and spectral contamination from telluric lines. Progress will require new instruments installed on large telescopes, substantial allocations of observing time, advanced statistical methods for data analysis informed by theoretical modeling, and collaboration between observers, instrument builders, stellar astrophysicists, heliophysicists, and statisticians.

Recommendation: NASA and NSF should establish a strategic initiative in extremely precise radial velocities (EPRVs) to develop methods and facilities for measuring the masses of temperate terrestrial planets orbiting Sun-like stars.

Transit Timing Variations

Multiplanet systems are known to be common (Lissauer et al., 2012; Batalha et al., 2013); they also tend to form compact orbital configurations, making TTVs a powerful method for studying gravitational interactions and estimating exoplanet masses (Holman and Murray, 2005; Agol et al., 2005). In the TTV technique, the masses of planets can be deduced by observing the signature of planet-planet interactions leading to perturbed (i.e., non-Keplerian) orbits. The orbital perturbation is seen through changes in the time of transit of planets. In a number of cases, such as the TRAPPIST-1 system, the parent stars of terrestrial planetary systems are too faint or spin too rapidly to be studied with RVs (Gillon et al., 2017). TTV studies will continue to be relevant for TESS objects of interest, but the limited baseline (compared to Kepler) will result in degeneracies in planet mass and eccentricity (e.g., Lithwick et al., 2012). These can be ameliorated to some extent with observations of subsequent transits by other facilities, including CHEOPS, and in some cases RV can disambiguate and refine the results derived from TTV measurements, and vice versa.

Astrometry

Astrometry is less sensitive to stellar variability than the Doppler method (Shao et al., 2018, white paper). Given the reasonably large field of view required to optimize the number of reference images used for calibration and pixel sampling, astrometry also offers a significant target multiplexing advantage compared to single-object techniques. Astrometry removes the orbital inclination ambiguity of radial velocity measurements and is also sensitive to planets with larger orbital separations (given a sufficiently long survey baseline), complementing the parameter space explored by Kepler and radial velocities.

The most precise astrometric observations conducted to date are of order 10 µas for bright (G < 15) stars from the Gaia mission (Lindegren, 2018). This level of precision is sufficient to detect gas giant planets amenable to direct imaging follow-up with GSMTs, including young planets. The precision of astrometric measurements conducted from the ground are an order of magnitude larger (Ghez et al., 2008; Sahlmann et al., 2014). Reaching the sensitivities needed to detect Earth-mass planets requires observations with precisions at roughly the σ = 0.1 µas level, which are generally believed to only be obtainable from space.

Astrometry has a spotty history involving exoplanet false positive detections and the cancellation of the Space Interferometry Mission. Only two “high-confidence” exoplanets have been discovered with this technique (Muterspaugh et al., 2010), neither of which has been independently confirmed. Despite its intrinsic merit, astrometry has not been viable as a search technique, given that a space mission is needed for large samples and low-mass planets. It has also been somewhat supplanted due to the anticipated detection of terrestrial planets using the EPRV method. Although not an immediate priority for the exoplanet community, astrometry is being considered for the

proposed LUVOIR mission. TRLs and hardware maturation plans should ideally be studied in parallel with the above-mentioned ground-based EPRV efforts.

Finding: High-precision, narrow-angle astrometry could play a role in the identification and mass measurement of Earth-like planets around Sun-like stars, particularly if the radial velocity technique is ultimately limited by stellar variability.

THE NEED FOR DETAILED STELLAR CHARACTERIZATION

An understanding of exoplanets is inextricably linked to an understanding of the stars they orbit. Moreover, the ability to detect planets and the precision with which researchers can determine their properties is often limited by knowledge of the star.

Connecting the Exoplanet and Stellar Astrophysics Communities

As described in the section “Exoplanet Masses,” earlier in this chapter, variations in the stellar photosphere are currently the dominant source of uncertainty for RV determinations of the planet mass, and as shown below, the same is often true for transit studies of the exoplanet atmospheres. As described earlier, in order to achieve the RV precision to detect an Earth-mass planet orbiting a Sun-like star, increased observational cadence, spectral resolution, SNRs, and stellar activity monitoring are required to disentangle stellar from planetary signals. This approach should include close collaboration with the solar and stellar astrophysics communities, including theorists, modelers, and observers (Wright and Sigurdsson, 2018, white paper).

Fortunately, it is not a one-way street: the high quality of the exoplanet-grade data has proven to be of great value to the stellar astrophysics community (e.g., Brogaard et al., 2018; Handberg et al., 2017; Campante et al., 2015). However, the use of such data by the stellar astrophysics community is impeded in part by the fact that RV data are not always public, or, even when public, can require highly specialized knowledge that is often unavailable to non-team members, yet critical for the proper analysis of the data. The RV community is also plagued with important and increasingly recognized data analysis challenges (e.g., Dumusque, 2016; Dumusque et al., 2017) that can influence the measurement of the stellar signal. Moreover, RV observations are usually taken at precisions that are of great interest to stellar astrophysicists, yet at cadences and over time spans that are more optimized for exoplanet detection. As discussed in the section “Exoplanet Masses,” leveraging the stellar expertise that is already available, including at the instrument design level, increases the likelihood that the stellar variability barriers to the RV detection of Earth-mass planets are ameliorated. Examples of specific areas of possible collaboration between the RV and stellar astrophysics communities include (1) using three-dimensional (3D) solar photosphere models in order to assess which spectral lines are most diagnostic of solar-like variability; (2) observing the Sun as a star in RVs (e.g., Dumusque et al., 2015; Haywood et al., 2016); (3) combining 3D hydrodynamic simulations of stellar photospheres with high-spectral-resolution observations of specific absorption lines of stars with spectral types other than solar, in order to assess the variations to which those lines are most sensitive (e.g., Dravins et al., 2018); and (4) guidance from stellar modelers to optimize the spectral coverage at the instrument design stage in order to best monitor instrinsic stellar variations.

Stellar Variability and Surface Heterogeneity Across the Electromagnetic Spectrum

Heterogeneities on stellar photospheres are ubiquitous. At optical and infrared wavelengths, such regions of various temperatures, and thus differing local emission spectra, will corrupt the wavelength-dependent transit measurements of any planet (Figure 4.11).

For example, dark spots on the stellar surface that intersect the transit chord result in a shallower transit, while spots that do not intersect the chord result in a deeper transit (e.g., Pont et al., 2013). The opposite is true for bright regions, such as faculae on the photosphere and plages on the chromosphere. As researchers push transmission spectroscopy to study the atmospheres of ever smaller planets, understanding and correcting for these effects

become ever more important. At UV and X-ray wavelengths, the active regions have greater temperature contrasts to their surroundings and are more extended than active regions in the optical and IR; even for giant planets these inhomogeneities can measurably impact the transit signal (Llama and Shkolnik, 2015).

There are several areas of concern. First, researchers need to pay attention to the variability of the stellar active regions over the duration of the observations used to construct the transmission spectrum of the planet. This can be due to short-term (minutes to hours) changes in the photosphere caused by flares and long-term changes (days to months) due to stellar rotation. Very long term (over years) changes, such as stellar magnetic cycles (Duncan et al., 1991; Baliunas et al., 1995; Jeffers et al., 2018), may also add uncertainty when compiling transit (and precise RV data) acquired over months to years. This can be somewhat mitigated by collecting many transits in the same bandpass, in order to average over the effects of such stellar inhomogeneities.

Second, the stellar spectrum of the transit chord is not identical to the stellar spectrum of the integrated disk, which will add error to the measured size of the planet at a particular wavelength (Figure 4.11; Zellem et al., 2017; Rackham et al., 2018). In some applications, these effects will be negligible, but in others, such as planets transiting active stars and M dwarfs, they will preclude an accurate measurement of the atmosphere. Both variability and surface inhomogeneity become increasingly problematic at short wavelengths, such as in the far-UV (FUV), where chromospheric emission lines such as Lyman-alpha are being used to study the extended atmospheres of transiting exoplanets (e.g., Vidal-Madjar et al., 2003; Lecavelier Des Etangs et al., 2010; Kulow et al., 2014; Ehrenreich et al., 2015). These concerns also directly impact plans to search at infrared wavelengths for atmospheric biosignature gases in the atmospheres of HZ planets. Although in the infrared these effects are lessened, they are not entirely eliminated (e.g., Rackham et al., 2018).

Impact of High-Energy Stellar Radiation on Exoplanet Atmospheres

A better understanding of the high-energy emission from host stars will allow for better interpretation of atmospheric features in exoplanetary spectra, and will inform the target strategy for an imaging mission. The near-UV (NUV; 200-300 nm) and FUV (100-200 nm) flux modifies the photochemistry of the atmosphere, potentially photodissociating important diagnostic molecules such as water, methane, and carbon dioxide (Figure 4.12). The extreme UV (EUV; 10-90 nm) also photoionizes, heats, and inflates the planet’s upper atmosphere. The combination of this, with pick-up by the high-energy protons from the stellar wind, controls the mass loss rate of the atmosphere (Koskinen et al., 2010; Tilley et al., 2017).

Lammer et al. (2007) considered the impact of EUV emission on terrestrial planetary atmospheres and concluded that the atmosphere of an unmagnetized planet can be completely eroded in its first billion years. Magnetic fields of exoplanets may play a role in the protection of planetary atmospheres from erosion, but meaningful measurement limits are restricted to hot Jupiters (e.g., Zarka et al., 1997; Grießmeier et al., 2007; Shkolnik et al., 2008; Zarka et al., 2014; van Haarlem et al., 2013; and see reviews by Lazio et al., 2016; Shkolnik and Llama, 2017) and free-floating planetary mass objects (Kao et al., 2016).

Together, the FUV and EUV can produce hazes in reducing atmospheres (Zerkle et al., 2012) and ozone (O₃) in oxidizing atmospheres (Segura et al., 2003, 2005), both of which may strongly affect the observed planet spectrum (e.g., Bean et al., 2011; Kreidberg et al., 2014; Linsky, 2014; Shkolnik and Barman, 2014). Photochemical models of exoplanetary atmospheres require realistic inputs of high-energy stellar fluxes, slopes, variability, and evolution.

For habitable-zone terrestrial planets orbiting M dwarfs, the high-energy photon and proton flux is even more critical to understand, as the fluxes are at least five times stronger than the fluxes received at 1 AU from a solar-type star (France et al., 2016). The UV flux emitted during the super-luminous pre-main-sequence phase of M stars drives water loss and photochemical O₂ buildup for terrestrial planets within the HZ (Luger and Barnes, 2015). This phase can persist for up to a billion years for the lowest mass M stars (Figure 4.13; e.g., Stelzer et al., 2013; Shkolnik and Barman, 2014; Schneider and Shkolnik, 2018). The slope of the high-energy spectrum of M dwarfs is also very different than for the Sun, with FUV to NUV flux ratios being greater than 1000 times that of the Sun (Figure 4.14; France et al., 2012; Miles and Shkolnik, 2017). This is problematic, as it enables several photochemical pathways to creating false positive biosignatures in the form of abiotic production of O₂ and O₃. In some scenarios, false negatives of biosignatures are also a possibility through the photodissociation of the biosignature gases (Meadows, 2017).

Finding: Stellar UV emission impacts planetary habitability as well as the interpretation of putative atmospheric biosignature gases.

Measuring and Predicting Ultraviolet Emission from Stars

Photoevaporation models of protoplanetary disks (e.g., Clarke et al., 2001; Owen et al., 2010) and photochemical planet atmosphere models for all types of exoplanets, from Earths to Jupiters, need input stellar UV fluxes (Segura et al., 2010; Line et al., 2010; Kaltenegger et al., 2011; Hu et al., 2012; Kopparapu et al., 2012; Moses et al., 2013) across planet formation and evolution time scales. Photometric data from the Galaxy Evolution Explorer demonstrated that median UV flux appears to stay at the “saturation” level, an empirical maximum emission also observed in X-rays, for a few hundred million years for early M stars, and out to a Gyr for late M stars (Figure 4.14). Furthermore, UV emission spans one to two orders of magnitude at every age, likely due to flaring events and intrinsic differences in activity from star to star. Existing HST UV spectra of old, relatively inactive, M-dwarf planet hosts reveal a variety of flaring activity within the 0.5-2 hour exposures (France et al., 2013; Loyd and France, 2014), with some emission lines flaring by as much as a factor of 10. Using round-the-clock visible-light monitoring, Davenport et al. (2014) analyzed 11 months of Kepler data of the active M4 star GJ 1243 and found over 6000 individual flaring events averaging 19 flares per day. Even for the slow-rotator and HZ planet host, Proxima Centauri, Davenport et al. (2016) report a strong optical light flare rate of about 2 per day, with weaker flares predicted to occur 63 times per day and super-flares occurring at the rate of approximately 5 per year (Howard et al., 2018; MacGregor et al., 2018). Note that the associated particle flux during these flares is completely unconstrained at this time, but will be important in understanding the combined effects of photons plus particles on planetary atmospheres (e.g., Tilley et al., 2017).

Current access to stellar FUV and NUV spectra is solely available with HST, whose sensitivity to these wavelengths is degrading over time. The Russian-led World Space Observatory, a 1.7-meter UV telescope, is aiming for a 2023 launch, but there is no such planned NASA mission, ultimately leaving the U.S. astrophysical community without access to the UV for many years. Small satellites can help fill this UV gap. NASA recently funded the development of two new small UV telescopes housed in CubeSats: the Colorado Ultraviolet Transit Experiment, an NUV low-resolution spectrometer (Fleming et al., 2018), and the Star-Planet Activity Research CubeSat, an FUV and NUV photometer (Shkolnik et al., 2018). Both experiments are dedicated to the temporal characterization of high-energy emission from exoplanet host stars.

Finding: Once HST ceases operation, researchers will essentially lose the ability to gather UV spectra of exoplanet host stars, which will limit the ability to interpret spectra of the planetary atmospheres and to understand their habitability.

EUV fluxes are not accessible due to the attenuation of the interstellar medium. For these, reliance on scaling laws from the FUV and X-ray is insufficient given the wide range of quiescent and flare emission levels. These facts necessitate a detailed grid of upper-atmosphere models, across stellar mass, age, and flare state, in order to predict the EUV flux and variability. Atmosphere models (e.g., Phoenix; Hauschildt et al., 1997; Allard et al., 2001) substantially underpredict the UV emission from low-mass stars and do not model the upper atmosphere, which is brightest at high-energy wavelengths. Since access to the NUV and FUV is limited, and the EUV is inaccessible, there is a need for upper-atmosphere models that can accurately predict these fluxes. Existing models could be modified to include the nonlocal thermodynamic equilibrium radiative transfer prescriptions needed for predicting the upper atmosphere emission; efforts to build such empirically guided models are just beginning (Peacock et al., 2015; Fontenla et al., 2016).

System Ages for Planetary Formation and Evolution

Planets evolve throughout their lifetimes, and hence knowledge of their ages is an important input to interpreting their properties. The ages of stars are estimated from several techniques, each of which has observational and intrinsic astrophysical regimes in which it works best. Stellar age determination methods include asteroseismology,

cluster or young moving group membership, lithium absorption, surface gravity, isochrone fitting, and rotation-activity relations (Soderblom, 2010).

For stars younger than 300 Myr, ages can be measured with relatively high precision. This is valuable in particular to the discovery and characterization of self-luminous giant exoplanets with direct imaging, for which the age of the system determines the inferred mass of the companion through planetary model atmospheres (e.g., Bowler, 2016). Finding planets around young stars using the RV and transit methods is intrinsically difficult due to their high levels of magnetic activity and rapid rotation, although transit surveys have recently succeeded in some young open clusters (David et al., 2016, 2018; Livingston et al., 2018; Mann et al., 2016, 2017; Pepper et al., 2017). However, the vast majority of planets orbit inactive, slowly rotating, and therefore older (>1 Gyr) main sequence stars, and this regime is where many empirical age-dating methods break down (Sodorblom, 2010). Asteroseismology can provide precise age measurements for older main sequence stars (e.g., Brown et al., 1994; Chaplin et al., 2014; Campante et al., 2015), but only if they are sufficiently bright, and it requires long baselines, rapid (less than approximately 1 minute) cadence, and high-precision photometry. These considerations, and the drop in amplitude with decreasing stellar mass, make asteroseismology generally unusable for main sequence stars much less massive than the Sun given currently achievable photometric precision. Gyrochronology, the evolution of stellar rotation with age, is another opportunity to measure ages for solar-mass stars for ages less than about 5 Gyr (e.g., Barnes et al., 2016), but has its own empirical inconsistencies, as well (e.g., Angus et al., 2015).

Finding: Understanding of exoplanets is limited by measurements of the properties of the parent stars, including stellar mass, radius, distance, binarity, rotation period, age, composition, emergent spectrum, and variability.

PLANET FORMATION

Planet formation studies require multiple techniques across multiple wavelengths to trace both the gas and dust over the scales relevant to planetary architectures. Because planets exist at semimajor axes from several hundredths of an AU to several hundreds of AU, and because planets may form from material that migrates over the full sizes of disks, the relevant disk temperatures span 1800 K (the first temperature at which solids condense) to 20 K (where ices in outer disks play critical roles in the formation chemistry). The disparate temperatures immediately imply the need for observational wavelengths spanning from the optical to the radio. From the Solar System, one spatial scale emerges as particularly important—namely, the division at 4 AU between primarily rocky material and the regime where ice and gas dominate. This physical scale needs to be observed around young stars. The nearest large groups of stars younger than 10 Myr sit at 140 pc, so the spatial resolution needs to be at least 10 mas.

A central goal of planet formation studies is to be able to predict the properties of the exoplanet population, including mass, density, multiplicity, and semimajor axes, from the range of initial conditions of the disks. Because solids migrate through disks, the history of planet formation impacts the final population synthesis. This section provides an overview of the important paths toward understanding planet formation.

Finding: An understanding of planet formation requires a census of protoplanetary disks, young planets, and mature planetary systems across a wide range of planet-star separations.

Key inputs are the initial disk mass, disk size, and disk lifetime. Only the last of these is well constrained currently, based on studies of disk accretion and disk continuum emission from young stars in clusters and associations of varying age.

The bulk of the disk mass is in molecular hydrogen, which is difficult to measure because it lacks strong rovibrational transitions as a consequence of having no dipole moment. Tracer gases such as CO show strong depletions from the gas phase (e.g., Kama et al., 2016; McClure et al., 2016; Schwarz et al., 2018), either due to photodissociation or condensation and incorporation into icy material. Dust disk masses are uncertain due to uncertainty in dust opacities, varying opacities from different-size grains, and high optical depths. Furthermore, because small solids drift with respect to the gas in disks, solid-to-gas ratios likely vary with distance from the

star. This process may significantly deplete solids in the outer regions of mature disks, where dust is most likely to be optically thin and dust disk masses are most easily measured. Direct measurements of the bulk gas over time are the only way to measure the properties that set the time scale for planet formation and dictate the formation and migration physics of large bodies. The European Space Agency’s Herschel Space Telescope measured masses for only three disks (Bergin et al., 2013; McClure et al., 2016). A future cold (4 K) telescope that is five times larger than Herschel could detect many hundreds of disks at least as massive as the minimum mass solar nebula (Pontoppidan et al., 2018, white paper).

Finding: Understanding the time scale and mechanism, including turbulence generation, for the dispersal of disk gas is key to understanding the final chemistry of planets and architectures of systems. Measurement of HD is the only direct method for directly measuring disk masses. The fundamental rotational transition of HD is at 112 microns, and therefore its detection requires a large, cold-space telescope.

Disk sizes can be observed in dust, gas tracers, and hydrogen deuteride (HD), and both by spatially resolving the disk and by looking at velocity-resolved spectral lines. Atacama Large Millimeter/Submillimeter Array (ALMA) can pursue submillimeter emissions to trace large grains and kinematically resolved gas. HST, JWST, and large ground-based telescopes (e.g., SPHERE on the VLT, GPI on Gemini, and instruments to come on the GSMTs) have made and will make images in scattered light that show the extent of small dust grains, which are coupled strongly to gas and which trace the outermost reaches of disks.

Disk lifetimes can be refined by studying the gas-to-dust ratio as a function of radial location in the disk. This will require a combination of techniques that can trace both gas kinematics and location, as well as dust size and location. Large ground-based optical and infrared telescopes and GSMT can play a leading role in this with spatially resolved, high-resolution imaging and spectroscopy that can trace both gas kinematics with spectral resolution and gas location with spatial resolution. ALMA and even higher spatial resolution interferometers can also play a big role in both gas kinematics and dust continuum levels in disk midplanes.

Measurements of disk chemistry, particularly in the midplane of disks, are required to understand the diversity of compositions that exoplanets may have (Figure 4.14). JWST will make great strides on the warm molecular layer and gas composition in optically thin inner holes, particularly for water and organic molecules, following on successes from Spitzer (e.g., Carr and Najita, 2008) and large ground-based telescopes (e.g., Salyk et al., 2008; Mandell et al., 2012). ALMA is making outstanding progress on snowline locations and chemistry more generally in the cool, outer disk (e.g., Qi et al., 2013; Walsh et al., 2016). Midplane chemistry in the 1-5 AU region is challenging because of very high optical depths. However, it is there that planetesimals will form, so the midplane (and the circulation between the midplane and upper layers of the disk that brings in new chemical constituents) will dictate planetary compositions. Ultimately, it would be fruitful to connect studies of Solar System comets, asteroids, and meteorites including timing of differentiation, isotopic trends, and relationships between high- and low-temperature condensates with conditions measured in protoplanetary disks (Figure 4.15).

In the future, an enhanced interferometer could be essential for this problem, as mm to cm wavelengths would provide for new molecular tracers and higher spatial resolution; a cold space telescope would permit us to trace warm water vapor and its incorporation into solids. Astrochemical theory and laboratory measurements will be essential for correctly interpreting observations including improved understanding of chemical networks at various temperature and pressure regimes, grain opacities, porosity, size distributions, and the processes of growth and sticking.

Finding: Disk chemistry in the midplane affects the compositions of the resulting planets. A cold space telescope, high-resolution spatially resolved infrared spectroscopy, and ground-based mm interferometry would enable significant advances and permit meaningful comparisons to studies of comets, asteroids, and meteorites from the Solar System.

Observations of young planets in disks will provide the ground truth for the otherwise indirectly inferred time scales of planet formation and permit studies of the dynamical interactions between disks and planets. From RV

and direct imaging studies, there are indications of massive planets around very young stars (less than 5 Myr old; Kraus and Ireland, 2012; Sallum et al., 2015; Donati et al., 2016; Johns-Krull et al., 2016; Keppler et al., 2018). These systems will be ideal places to test the physics that ties planets to their disks, but more need to be found. Direct imaging at key wavelengths is essential and will be greatly enabled by the GSMTs. In particular, H-alpha emission would verify the presence of a massive, compact body (e.g., Wagner et al., 2018), and infrared emission would determine its luminosity and, by inference, its mass (e.g., Morzinski et al., 2015).

ALMA has resolved ring-like structures in a number of protoplanetary disks (Figure 4.16). Further progress on the imaging of disk structure at a variety of wavelengths will be essential to figure out how these structures formed. JWST, ALMA, and GSMTs are all complementary because they have different sensitivities to grain size. The size distribution and composition of grains inside and outside of the gaps will show if pressure bumps or condensation fronts are the more likely causes of gaps, or whether planetary-size objects form them. Observations at higher spatial and spectral resolution of gas, combined with theory, may uncover if rings and gaps are caused by planets and, if so, what planetary properties can be inferred. Conversely, if the physics of planet-disk interaction is well understood, the size of gaps can be used to infer the presence of planets not massive or bright enough to be directly detected. An exciting forefront is looking at planet-gas interactions in disks. ALMA observations have uncovered the first of these (Pinte et al., 2018; Teague et al., 2018). High-resolution spectroscopy (R > 60,000) in the infrared is necessary to pursue the same type of interactions in warmer gas and with specific molecules.

For slightly older systems, researchers should better characterize exozodiacal and cold debris disks. Detailed images of belt locations (only now becoming available with ALMA) will allow modeling of the system architectures. Measurement of dust compositions will allow constraints on the compositions of planets. JWST will make progress on this, but a large cold space telescope would allow the detection of Kuiper-belt levels of dust around nearby stars without confusion from background sources.

The ability to detect asteroidal-like belts may be particularly interesting for assessing the existence and habitability of any rocky planets. Ground-based interferometers (Very Large Telescope Interferometer [VLTI] and LBTI have detected hot inner dust but have not had the spatial resolution or sensitivity to look for clumps in the distribution that would be expected from planetesimals in resonance with planets. Detection of dust clumps that orbit stars would uncover planets otherwise hard to detect.

Finding: The detection of young planets in disks will provide the ground truth for the time scale of planet formation and permit studies of the dynamical interaction between disks and planets. With the high spatial resolution of the GMT and TMT, researchers will be able to search the inner parts of planet-forming systems.

At the end of its prime mission in 2018, LBTI will provide limits that will just reach a level that helps inform the size of a direct imaging mission (see the section “The Case for Imaging,” earlier in this chapter). Furthermore, LBTI measures the thermal emission of the dust, whereas it is the amount of light that dust scatters that determines the background for direct imaging measurements. The conversion from thermal emission to scattered light surface brightness depends on the composition and size of the dust grains, which have yet to be measured. The WFIRST coronagraph could make progress on measuring scattered light exozodiacal levels and constrain dust clumpiness. LBTI could be extended to push down to Solar System levels of dust and explore the relationship between cold and warm dust. More theoretical and observational work on systems with detected dust could enable a better understanding of how and where remnant planetesimals produce dust.

The process of disk dispersal sets the time limit for planet formation and changes the composition of the disk including the gas-to-dust ratio that affects migration and planetary composition. Therefore, understanding the time scale and mechanism of disk dispersal is key to understanding the final chemistry of planets and architectures of

systems. Four types of observations constrain the angular momentum loss of disks: (1) signatures (such as H-alpha emission) of accretion onto the central star; (2) wind tracers such as Doppler-shifted and broadened forbidden lines of atomic species that show outflowing gas; (3) broadened molecular lines arising in disks where the broadening is from turbulent motion of the gas; and (4) X-ray/UV measurements of young stars over time. The first of these is a well-developed technique, although more synoptic observations are needed to elucidate how much mass is accreted in a steady state versus outbursts, and how those accretion bursts affect the other properties of the disk. Direct measurements of disk photoevaporation or winds would benefit from spatial resolutions afforded by ground-based telescopes with adaptive optics than can work at red visible wavelengths. More measurements of turbulence are needed to understand the turbulence distribution, in disks at different evolutionary stages, at closer distances to the stars, and at different scale heights. Only then can researchers test whether magnetorotational instability models can drive accretion and whether turbulent concentration of dust works to form planetesimals rapidly. More capable spatially resolved, high-resolution spectroscopy from the infrared to millimeter wavelengths will be essential.

The X-ray luminosities of young stars over time may control their disk dissipation rates and help set the disk chemistry through their penetrating ionization of disk gas. New facilities with higher sensitivity would be able to measure X-ray luminosities and spectra for very low mass stars (below 0.5 M_sun) that were too faint for Chandra (e.g., Prisinzano et al., 2008; Kastner et al., 2016).

Finding: An understanding of planet formation would permit the inference of some planetary properties that cannot be directly probed. A better understanding of the link between stellar composition, disk composition, planetary composition, and atmospheric composition is necessary to translate observations of planetary atmospheres into statements about bulk planetary compositions.

THEORY OF EXOPLANET EVOLUTION, INTERIORS, SURFACES, AND ATMOSPHERES

Theory and observation in the exoplanet field are inexorably coupled. While for some areas of exoplanets theoretical modeling has moved well ahead of the current data, there are a number of areas in which the available data, although limited, are already charging ahead of theoretical investigations. Some outstanding challenges in the theory and modeling of exoplanets that need to be addressed in order to achieve the two goals described in Chapter 3 are discussed below. As outlined here, achieving these goals requires collaboration between research groups with different and overlapping expertise.

Specific Scientific Challenges in the Theory of Exoplanets

Planetary Evolution

The long-standing question of whether giant planets can form through disk instability as well as core accretion has given way to a more nuanced question of the accretion histories of these objects. Planets that form through large-scale quasi-adiabatic processes will form with high entropy and therefore high initial luminosity; those that accrete most of their mass through a compact shock, as scientists think Jupiter perhaps did, will have a low initial luminosity. This record of a planet’s entropy of formation and subsequent cooling history is thus fossilized in its present day luminosity for planets less than about 100 million years of age. Extracting meaningful constraints relies on comparing the observed mass and emitted flux to models of the planet formation process. A second question relates to the inflated radii of many hot Jupiters. The range of theories broadly bifurcates into those that provide an additional energy source over the lifetime of the planet, and those that trap in heat from formation and slow the planetary cooling process. Recent advances include applying statistical techniques to the full population of known hot Jupiters (Thorngren and Fortney, 2018), which conclude that Ohmic heating (Batygin and Stevenson, 2010) is the likely mechanism for transferring heat from the atmosphere into the interior. Modeling the magnetic environment of a close-in exoplanet is challenging and poorly constrained by observations. Magnetohydrodynamic

models detailing the interactions between magnetic fields and the planet’s atmosphere and interior are currently underdeveloped.

For small rocky planets, the big questions that need to be addressed by theoretical modeling include the histories of volatile delivery, and atmospheric outgassing and escape, which are mitigated by both the host star irradiation and the dynamical history, including collisions. As described previously in the section “Stellar Composition, Planet Formation, and the Delivery of Volatiles” in Chapter 3, evidence suggests that some volatile-rich Earth-size planets may have formed in outer orbits and then migrated inward. Future key work will focus on models for terrestrial planet formation and migration around different types of stars, which can illuminate the initial composition of the terrestrial planet, and interactions between other components of the system that can affect volatile delivery.

XUV and particle fluxes from the host star can strip a terrestrial planetary atmosphere completely (Garcia-Sage et al., 2017; Dong et al., 2018) or modify the composition of the atmosphere via escape (Schaefer et al., 2016) and photochemistry (Segura et al., 2005; Rugheimer et al., 2015). Ocean loss can also be driven by the star, which will also modify the atmosphere (Luger and Barnes, 2015). These compositional modifications will, in turn, affect the planetary climate and stellar UV flux incident on the planetary surface (Meadows et al., 2018). Tidal deformation produced by gravitational interaction between a star and planet on an elliptical orbit could also possibly heat the planetary interior, drive off a planetary ocean, enhance tectonic activity, and shut down the magnetic dynamo (Driscoll and Barnes, 2015; Meadows and Barnes, 2018). It is likewise essential to couple such studies to the processes of atmospheric outgassing. It will be important to use atmospheric, interior, and orbital evolution models, as well as climate, photochemistry, radiative transfer, and interior and tidal heating models to understand the history of the production of secondary atmospheres and the sculpting of these atmospheres by environmental factors.

The anomalously low occurrence rates for highly irradiated planets around 1.5 R_E from Kepler data appear to indicate that small gas-rich planets fail to survive on close-in orbits (Fulton et al., 2017; Fulton and Petigura, 2018), and provides urgent motivation for renewed theoretical work on this problem.

Planetary Interiors

Models of interior structure are the primary mechanism through which researchers convert mass and radius data to an understanding of the bulk composition of the planets. Degeneracies are a well-known problem in interior structure modeling. As a result key properties such as the core mass fraction or the total thickness of a hydrogen-rich gas layer cannot be uniquely constrained. This is a fundamental issue that cannot be overcome with improved interior structure models, but clever and innovative workarounds may exist. For example, detailed measurements of the composition of an exoplanet atmosphere may indicate the makeup of deeper regions of the planet’s interior, or the presence of a magnetic field may be tied to the structure of the planet’s core. For these reasons innovation in interior structure modeling and its relation to other exoplanet observables is of significant value to make headway.

The equations of state (EOS) of high-pressure planet-forming materials (ices, rock, and even H/He gas) are not always well known over the relevant range of parameter space. Remedying this shortcoming requires additional laboratory experiments and ab initio calculations (see below). Furthermore, mixing of materials between layers is typically treated simplistically in exoplanet models (full mixing of H₂O and H/He layers, e.g., Nettelmann et al., 2010, and Valencia et al., 2013; or no mixing at all, e.g., Rogers and Seager, 2010). Such differences in treatment can result in significant differences in inferred planetary composition. Improved self-consistent treatments of mixing could remove a key source of uncertainty in interpreting exoplanetary mass-radius measurements.

Planetary Surfaces

In the quest for discovering habitable exoplanets, remote sensing of planetary surfaces is highly desirable, to establish both the presence of liquid water and the local conditions under which a potential biosphere is operating. However, aerosols or thick atmospheres will challenge researchers’ ability to detect planetary surfaces. Furthermore, degeneracies in exoplanet spectra may make it impossible to credibly confirm the presence of a planetary surface. Related outstanding theoretical work includes the following:

• Which planets should have a well-defined surface?
• How do researchers model surface fluxes of constituent gases, both sources and sinks? These can include volcanism, biology, and the carbonate-silicate cycle.
• How do researchers establish the relative covering fractions of oceans and continents, and their distributions over the planetary surface using mapping techniques, inversions, or retrievals? In the case of continents, what are their surface materials and elevation profiles? How do these, in turn, affect atmospheric circulation?
• How prevalent are planets with exotic surfaces, such as lava worlds or desert worlds?

Atmosphere Modeling

Significant effort has been put into the modeling of exoplanet atmospheres, but substantial gaps and multiple disparate computational approaches remain. To plan and interpret the data that researchers envision obtaining (see the sections “The Case for Imaging” and “Opportunities to Characterize Planets Through Transits,” earlier in this chapter), the shortcomings of current modeling efforts need to be addressed.

Aerosols

Aerosols are a commonplace in transmission spectra (Sing et al., 2016; Crossfield and Kreidberg, 2017), yet accurate modeling of aerosols is a challenge, even for solar system planets. The difficulties lie in the broad range of chemical and physical processes that need to be well understood for a successful aerosol model. These include the aerosol formation process (the direct condensation of gases, or photochemically induced formation of large molecules), the coagulation and particle nucleation processes, and the atmospheric transport of aerosol particles. Furthermore, the composition of the aerosols will be difficult to ascertain. These challenges are compounded by the fact that most known exoplanets are much hotter than solar system objects and are therefore expected to have exotic aerosols. Understanding aerosols has been a primary bottleneck in interpreting exoplanet spectra. Concerted modeling efforts are needed that attack the aerosol problem from many different angles, with a focus on the overlap and feedback between the different approaches.

Gas-Phase Chemistry

There are various treatments of exoplanet atmospheric chemistry. Thermochemical equilibrium approaches are computationally cheap and may be appropriate for hot atmospheres, in which reaction rates are fast. Chemical kinetics codes, which perform time-dependent calculations of reaction rates, photolysis reactions, and vertical mixing and evolve the atmosphere to a steady state, are computationally intensive, and are typically only applied in one-dimensional (1D) models. Chemical relaxation schemes that attempt to account for some aspects of the chemical kinetics approach have been applied sparingly in 3D general circulation models (GCMs; e.g., Cooper and Showman, 2006; Tsai et al., 2017). None of these computational approaches are able to include all possible chemical species or reactions, so in all cases choices are made for which atmospheric constituents require direct tracking. In some situations, important physics and chemistry can be left out. For example, it was recently realized that thermal dissociation of water and ionization processes lead to substantial H– abundances in the atmospheres of ultrahot Jupiters. The previously unmodeled effect of H^– opacity on the radiative transport and resulting thermal emission spectra for these planets is substantial (Arcangeli et al., 2018; Kreidberg et al., 2018; Parmentier et al., 2018; Mansfield et al., 2018). The validity and relative merits of differing approaches need to be explored across the full parameter space.

Modeling Approaches

Exoplanet modeling efforts bifurcate into 1D versus 3D models, and forward versus retrieval models. Each approach has its own merits and shortcomings. Briefly, 1D models can typically treat radiative and chemical processes in more detail but cannot treat any spatial inhomogeneities that exist in the atmosphere. Forward models include detailed physics and chemistry at differing levels of self-consistency, whereas retrieval codes recover the

properties that are present in an atmosphere without necessarily linking those to a physical explanation. It will be beneficial to perform comparisons between different classes of models to determine the situations in which the various simplifications taken by each approach remain appropriate and where the agreement between disparate models breaks down.

Spectral Retrievals

While spectral retrievals are one of the specific types of modeling approaches, they deserve their own discussion because of their recent rise in popularity within the exoplanet community. Statistical techniques are applied to determine best-fit parameters and confidence intervals based on the level of agreement between observed spectra and large suites of forward models run in a Monte Carlo framework. A variety of approaches have been applied, including those that assume thermochemical equilibrium and those that assume well-mixed atmospheres but allow the abundances of individual molecules to vary independently. A range of prescriptions for the thermal structure have also been adopted. In general, it has not been possible for exoplanet retrievals to be fully coupled between the atmospheric chemistry, thermal structures, and cloud properties. This can result in physically implausible best-fit scenarios. Therefore, careful attention needs to be paid toward the range of allowed solutions and whether they are physically motivated. Furthermore, exoplanet retrievals are typically 1D and do not account for 3D atmospheric dynamics or variability. Retrieval codes rarely use line-by-line calculations because these are computationally intensive when running many models. In the era of JWST and beyond, as the quality of exoplanet spectroscopy improves, retrieval techniques will be called upon to include additional layers of complexity and self-consistency, motivating substantial commitment to further development in this area.

Overarching Challenges

The Need for Coupled Models

As described above, there are a great number of modeling approaches being applied to the characterization of specific aspects of exoplanets, including their physical structure, and the chemistry and physical conditions of their atmospheres. In principle, all of the aforementioned modeling approaches from the section “Specific Scientific Challenges in the Theory of Exoplanets” should be coupled, since there are important feedbacks between them. For example, the atmosphere forms the upper boundary of the planet’s interior, and the interior forms the lower boundary for the atmosphere. It is important to track sources and sinks at the boundaries (both energetic and chemical ones), but this is rarely done self-consistently. In the atmosphere, dynamics can alter thermal structures and carry a system away from chemical equilibrium. Similarly, the 1D chemical kinetics codes that track photolysis reactions and vertical mixing can produce abundance patterns that differ substantially from chemical equilibrium, and those changes in composition have feedback on the radiative processes in an atmosphere. The formation of aerosols (clouds and haze) are also dependent on the local chemistry and stellar irradiation, and the formation of clouds in turn alters the energy transport.

While an ambitious end goal may be to produce an exoplanet model that fully couples together all of the chemical, mechanical, and energetic processes that shape a planet and its atmosphere, this is not technically feasible in the foreseeable future. Indeed, this is one of the reasons for the popularity of spectral retrieval codes in exoplanet characterization. Retrieval approaches acknowledge the unknown and uncertain aspects of forward modeling and provide another avenue for determining atmospheric and system properties. Instead, an approach of successively coupled models is needed, with a focus on determining which regimes of coupling are the most pressing for interpretation of observational data.

Model Validation

Exoplanet models are often applied to portions of parameter space where precious little data are available and that extend far from the well-studied regime of Solar System planets. For this reason, model validation is neces-

sary. In some cases, this might mean verification of model outputs, as compared directly against experimental or observational data (e.g., validations with observations of Solar System planets). Other situations in which a well-calibrated benchmark data set is not available may necessitate individual research groups performing cross-comparisons of their models directly with one another.

Laboratory Measurements and Ab Initio Calculations

Theoretical calculations are only as good as the physics and chemistry provided as inputs. A current barrier in modeling the interiors and atmospheres of exoplanets is the lack of availability of laboratory and ab initio data. In many cases, the available data are extrapolated to situations far from their originally intended use. A white paper detailing the specific laboratory data needed by the exoplanet atmosphere modeling community (Fortney et al., 2016) described the following needs:

1. Molecular opacity line lists with parameters for a diversity of broadening gases, which are needed for radiative transfer calculations of atmospheres with diverse compositions;
2. Extended databases for collision-induced absorption and dimer opacities, which are also needed to model atmospheres of diverse composition;
3. High-spectral-resolution opacity data for relevant molecular species, which are needed as cross-correlation templates for interpreting ground-based high-dispersion spectra;
4. Laboratory studies of haze and condensate formation and optical properties, which are needed to predict and interpret the properties of aerosols in exoplanet atmospheres;
5. Significantly expanded databases of chemical reaction rates, which are needed to predict the compositions of exoplanet atmospheres that stray from chemical equilibrium; and
6. Measurements of gas photo-absorption cross sections at high temperatures, which are also needed to predict the nonequilibrium composition of irradiated atmospheres.

Additional laboratory and ab initio calculations are also required by researchers modeling exoplanet interiors, including improved EOS data. The set of expertise needed to respond to all of these needs is diverse and spans multiple research communities.

Finding: The limited laboratory and ab initio data covering the parameter space relevant to exoplanets is a barrier to accurate models of exoplanet atmospheres and interiors. Mechanisms to increase collaboration between exoplanet astronomers and experimental physicists and chemists would help overcome this barrier.

More Flexible and Modern Codes

Many exoplanet modeling tools have heritage from codes developed decades ago in now-outdated coding languages, with hard-wired parameters, and under significant computer memory limitations. Because of this, many of these codes are not easily adapted to new planetary parameters, are proprietary to specific research groups, are hard to read, and often do not take advantage of modern coding techniques. Yet these same codes are often well cited and produce consistent research results. There is a need to modernize old codes and to produce new, fast, flexible, well-documented, and publicly available exoplanet modeling tools. Such software packages lower the threshold of entry for new members of the field and observers with less modeling background, and these tools can be adapted readily as new research results become available.

Improved Computing Resources

The following areas would benefit from the further development of national-scale high-performance computing resources:

• Dynamical simulations of planet formation and evolution with many bodies;
• Hydrodynamic simulations of protoplanetary disk formation and evolution, including simulations tied directly to the star formation process and nonideal magnetohydrodynamic simulations required to study disk accretion;
• Hydrodynamic simulations of planet formation, requiring resolution of a large range of physical scales;
• Hydrodynamic simulations exploring planet disk interactions;
• Hybrid simulations of planet formation capable of simultaneously exploring more physical processes;
• Multidimensional spectral retrievals;
• Coupled models of atmospheric escape and photoionization; and
• Large-scale grids of GCMs.

Finding: Theoretical models are essential to plan and interpret observations of exoplanets, and are enabled by robust support via individual investigator grants.

Throughout Chapter 4, the committee has reached a number of findings that demonstrate that NASA has not fully realized the scientific yield of its exoplanet missions because the required theoretical, laboratory, and ground-based work to interpret exoplanet data has not been undertaken.

Recommendation: NASA should support a robust individual investigator program that includes grants for theoretical, laboratory, and ground-based telescopic investigations; otherwise, the full scientific yield of exoplanet missions will not be realized.

ASTROBIOLOGY

Observations and Studies to Support the Search for Habitable Environments and Life

The search for life begins with the search for habitable environments. Required observations include not just those that determine if a given planet is in fact habitable, but observations of nonhabitable planets that can serve as a baseline for comparison, as well as suites of observations of HZ planets that can be used to study the diversity of outcomes of terrestrial planet evolution, including the habitable-zone concept itself, as a function of planetary parameters and host star type.

Observations for Habitability Assessment and Biosignature Searches

The habitability assessment of potentially habitable planets will include a sequence of observations to determine the following:

1. The planetary mass, which in most cases will be done with radial velocities (see the section “Exoplanet Masses,” earlier in this chapter). If the planet is transiting as well, photometric and RV observations can be combined to determine the size and bulk density of the planet. Combined with a knowledge of the stellar elemental abundances, this will constrain the interior composition.
2. Whether or not the planet has an atmosphere. For example, in the case of a planet transiting a nearby M dwarf, this could be accomplished with JWST transit observations (e.g., Morley et al., 2018; Lustig-Yaeger et al., 2017).
3. The nature of that atmosphere (bulk composition and greenhouse gas census), and surface (for direct imaging), including possible surface pressure and temperature. These will be challenging to obtain, but could be pursued with transmission spectroscopy, thermal phase curves, secondary eclipses, infrared photometry, or direct imaging spectroscopy (Meadows et al., 2018). These measurements could be compromised by the presence of cloud cover.

4. Orbit and rotational period. For directly imaged planets that may not have had their mass and orbit determined via RV or transit in step 1, then the orbit can be obtained from multiple direct images of the planet, or from RV or astrometric measurements of the host star. The rotational period, which can indicate if the planet is synchronously rotating and governs climate regime and the likelihood of atmospheric collapse (Turbet et al., 2016, 2018), may be derivable if multiwavelength, time-resolved direct imaging observations can be used to map an inhomogeneous planetary surface (Pallé et al., 2008; Cowan et al., 2009; Fujii et al., 2017).
5. Whether glint or polarization signals from an ocean are observed. These signals should be sought at multiple points on the orbit to capture the phase dependence of the glint signal (Robinson et al., 2010, 2014), and may be easier to detect if the planet is simultaneously mapped to identify the spatial extent of the ocean (Tovar et al., 2017).
6. Once planetary mass, size, rotation rate, the presence of an atmosphere, and a census of greenhouse gases are known, whether coupled climate-photochemical models can be used to constrain surface condition solutions.
7. Whether biosignature gases, surface signatures, or seasonal variability of atmospheric gases are present. Atmospheric gases will likely be the most readily detectable biosignatures, but for the very best targets, additional observations to search for surface reflectivity signatures, or to search for seasonal variations in gases (Olson et al., 2018), may be attempted to assess any biosignatures detected (or not detected) in the context of observational discriminants for potential false positives (false negatives) that are present in the planetary environment.

An alternative approach to the assessment of an individual planet for habitability and life would be a statistical approach that looks at particular properties that are more readily observable, such as the presence of CO₂ in the planetary atmosphere and attempts to make that observation for a number of different planets as a function of other parameters such as orbital distance and stellar type. This technique could be used to make an initial observational test for the location of the habitable zone by determining when and where exoplanet terrestrial atmospheres become dominated by CO₂ (e.g., Bean et al., 2017). Similarly, thermal phase curves could potentially identify those planets that lack an atmosphere, providing insight into the persistence of atmospheres on planets orbiting M dwarfs.

Ultimately, the assessment of whether or not a planet is habitable will need to be embedded in the context of the outcomes of terrestrial exoplanet evolution, and so it is critically important to also gather information on planets that are not likely to be habitable—for example, exo-Venuses—to better understand the range of atmospheric compositions expected. Of particular interest are the observationally accessible close-in M-dwarf planets (e.g., GJ1132b; Berta-Thompson et al., 2015) that may have experienced ocean loss and could exhibit an O₂-dominated atmosphere (Luger and Barnes, 2015; Schaefer et al., 2016). Systems containing multiple terrestrial planets will provide an excellent opportunity to compare the evolutionary outcomes at a range of irradiations.

Exoplanet habitability studies will benefit from knowledge of processes and characteristics observed in the Solar System’s terrestrial planets, such as atmospheric loss processes on Mars, and ocean loss and catalytic photochemistry on Venus. Understanding terrestrial planet evolution from the Solar System perspective will likely play a crucial role to identify processes that affect habitability.

Developing Complementary Facilities to Characterize Potentially Habitable Planets

Many of the above measurements are extremely challenging, and the number of different types of stellar and planetary characteristics that impact habitability and biosignature detection speaks to developing a complementary approach using multiple techniques and instruments.

For the M-dwarf opportunity, target selection for potentially habitable planets will include several targets already known, including TRAPPIST-1 (Gillon et al., 2017), LHS1140 (Dittmann et al., 2017), and Proxima Centauri b (Anglada-Escudé et al., 2016), as well as a number of planets expected from TESS (Sullivan et al., 2015), and ground-based transit and RV searches. The planets orbiting Sun-like stars that will be studied by direct imaging missions could be discovered by the EPRV or astrometric methods, or discovered by the missions themselves; either way, advanced EPRV or astrometry will be essential to measure the masses of the directly imaged planets.

JWST will likely obtain transmission spectroscopy, secondary eclipse, or phase curves on promising M-dwarf targets to first establish the presence of an atmosphere, and then proceed to rudimentary atmospheric composition, planetary temperature, and possibly day-night temperature difference (Morley et al., 2017; Lustig-Yaeger et al., 2017; Kreidberg and Loeb, 2016; Meadows et al., 2018). Transmission spectroscopy does not probe to the planetary surface. Hence, it will not be as sensitive to water vapor as direct imaging observations, but it is more sensitive to trace gases in the stratosphere due to the longer slant path. Although these observations will be challenging, they will provide one of the first chances to observe habitable-zone planets at infrared wavelengths.

High-resolution spectroscopy, coupled with transit or extreme adaptive optics and coronagraphy, will use transmission spectroscopy (Rodler and Lopez-Morales, 2014), or the radial velocity of reflected light from the exoplanet, to shift exoplanet absorption away from absorption by similar molecules in Earth’s atmosphere (Snellen et al., 2015). The search for O₂ and CH₄ in the atmosphere of Proxima Centauri b may be possible in the next 5 years with the VLT using these techniques (Lovis et al., 2017), and the GSMTs may be able to perform similar observations for several other nearby M-dwarf planets. Mid-infrared direct imaging with GSMTs may detect warm exoplanets orbiting very nearby Sun-like stars, providing information on planetary temperature and complementing direct imaging observations with space-based telescopes.

A direct imaging mission promises the most capability for habitability and biosignature searches of habitable-zone terrestrial planets orbiting FGKM stars from ultraviolet to near-infrared wavelengths. Direct imaging can probe to the planetary surface and so is more definitive for habitability detection than transmission, which cannot probe the near-surface atmospheres. Consequently, direct imaging is more sensitive to lower atmosphere water vapor and surface-generated biosignatures. Direct imaging will potentially enable atmosphere detection, atmospheric composition determination (including the deep atmosphere), surface mapping for rotation and obliquity, detection of surface inhomogeneity (or not), detection of ocean glint, and a broad survey of greenhouse gases for FGK stars.

The significant interdisciplinary work described in Chapter 3 will be needed to underpin target selection, measurement requirements, data analysis, spectral retrieval, and habitability, and biosignature assessment will be optimally implemented via broad interdisciplinary collaboration, which can be enabled via the mechanisms described in the following section.

MECHANISMS TO ACHIEVE INTERDISCIPLINARITY

As described in Chapter 3 and this chapter, the study of exoplanets requires a strongly interdisciplinary approach. The detection of exoplanets has been undertaken within the field of observational astronomy, but researchers are now moving into a new era where the characterization and interpretation of data on exoplanet environments will require an interdisciplinary synthesis of observations, computer modeling, data from the Sun and solar system, and laboratory measurements. This is especially true of studies of terrestrial exoplanets that aim to identify signs of habitability and life, where observations and interpretations of these secondary atmospheres will require expertise from both observational astronomy and many of the Solar System planetary sciences, including terrestrial geophysics, geochemistry, photochemistry and atmospheric science, heliophysics, and Earth science and biology.

Consequently, mechanisms to encourage research endeavors that span and interconnect these disciplines should be strongly encouraged. At NASA, research within the Science Mission Directorate is stove-piped into four divisions: Astrophysics, Planetary Science, Heliophysics, and Earth Science. The Exoplanets Research Program (XRP) is jointly funded by the Astrophysics and Planetary Science Divisions, allowing scientific cross-collaboration between these two communities. The NASA Astrobiology Program, run out of the Planetary Science Division, includes but is not limited to the NASA Astrobiology Institute (NAI), and this program has fostered interdisciplinary collaboration and cooperation, including some work on exoplanet habitability and biosignatures. The NAI helped to build the astrobiology community by funding large, interdisciplinary research efforts and encouraging teams to collaborate with one another.

NASA recently implemented a new interdisciplinary and cross-divisional research coordination network, the Nexus for Exoplanet Systems Science (NExSS). NExSS was conceived as a large-scale experiment in managing

and catalyzing exoplanetary science that integrates the astronomical, terrestrial, planetary, and heliophysical sciences. NExSS is intended to inform and enhance science from upcoming NASA missions such as TESS, JWST, and the potential future LUVOIR, HabEx, and OST telescope concepts that could be used to search for signs of habitability and life on exoplanets. The NExSS research coordination network connects and leverages research from several research and analysis competitions (with awards of different sizes from NAI-scale large collaborative groups to individual PI research in the XRP) across cooperating NASA divisions, breaking down interdisciplinary and interdivisional barriers to engage in systems science study of exoplanets. One of its key measures of success will be how well it integrates the larger scientific community into its activities. Unfortunately, there is a perception in the community of exclusivity because it is unclear how teams have been chosen to participate, and there has never been a call for proposals to participate explicitly in NExSS.

NExSS has facilitated new, community-inclusive activities and products that transcend the output of the constituent research teams. These NExSS-led activities include the Upstairs-Downstairs Workshop (2016) on the impact of terrestrial planet interiors on planetary atmospheric and surface conditions, essentially working to understand the formation and nature of secondary outgassed atmospheres, which was jointly supported by the NAI and NSF; the Exoplanet Biosignatures Workshop (2016), which produced six community scientific publications that greatly advanced understanding of the significance of false positives and agnostic biosignatures, and fostered the development of the comprehensive framework for biosignature assessment (Kiang et al., 2018; Schwieterman et al., 2018; Meadows et al., 2018; Catling et al., 2018; Walker et al., 2018; Fujii et al., 2018); the Habitable Worlds 2017: A System Science Workshop, which had over 150 attendees as well as webcasting, and strong participation from Earth scientists, planetary scientists and heliophysicists, in addition to astronomers and exoplanet scientists. A NExSS-led community group developed a Laboratory Astrophysics Gap List of needed laboratory studies to be able to interpret exoplanet spectra (Fortney et al., 2016), and mobilized and coordinated the community to contribute numerous white papers to both the National Academy of Sciences (NAS) Astrobiology and NAS Exoplanet studies. NExSS PIs and their collaborators also contributed to plans for utilization of current space telescopes by bringing together the U.S. and international research communities (in almost equal numbers) to win 23 percent of JWST Early Release Science. These proposals will provide the initial characterization of JWST performance for studies of exoplanets, and will provide the first steps toward habitable-zone planet characterization and biosignature searches.

Finding: The search for life outside the Solar System is a fundamentally interdisciplinary endeavor. The Nexus for Exoplanet Systems Science (NExSS) research coordination network encourages the cross-disciplinary and cross-divisional collaborations needed to support NASA exoplanet research and missions.

However, NExSS is currently comprised of research efforts funded via other programs at NASA and has very modest funding of its own, sufficient to cover only community activities like workshops and conferences. This model is not optimal for selecting proposals that are closely aligned with NExSS goals, as the teams selected need, first and foremost, to be responsive to the particular call under which they were funded, and responsiveness to NExSS goals is not a criterion for selection in the parent funding program. Additionally, without a NExSS-specific call for participation, teams have been selected from different relevant programs in a process that is not as transparent as an open call, although PIs now have the option to self-select membership into NExSS for the NASA Exobiology and Habitable Worlds programs. Consequently, the NExSS research coordination networks would be strengthened by the provision of program status, with sufficient funding to enable open competitive selection of teams funded directly by the NExSS program.

Recommendation: Building on the NExSS model, NASA should support a cross-divisional exoplanet research coordination network that includes additional membership opportunities via dedicated proposal calls for interdisciplinary research.

REDUCING BARRIERS TO SCIENTIFIC EXCELLENCE

The committee was charged with surveying the status of the field of exoplanet science (see charge in Appendix A), and no such survey would be complete without consideration of the state of the field’s workforce. The search for life on other worlds is both a profound and a profoundly difficult endeavor. Maximizing excellence and ensuring success of exoplanet science depends on marshaling, developing, and supporting all available talent. As a growing field, exoplanetary astronomy is particularly dependent on the effective development and retention of junior scientists because it is now putting into place the cohort that will provide senior leadership for many decades.

While studies have not been published specifically for the field of exoplanets, the committee has examined recent reports that provide evidence of systemic barriers in in science, technology, engineering, and mathematics fields including astronomy. The Exoplanet Science Strategy, therefore, includes a strategy for developing and maintaining human capital, including addressing demographics and standards of professional conduct, and identifies areas requiring further research.

Equity and Inclusion

Several recent reports provide evidence that women and people of color are underrepresented in the professional astronomy workforce. Only 26 percent of all members are female and just 5.3 percent are members of underrepresented minority groups (URMs⁸; Pold et al., 2016). Over the last 10 years, women earned approximately 32 percent of the Ph.D.s in astronomy, and this number has been essentially flat over that period.⁹ According to a recent study, perceived discrimination in a field significantly deters women from choosing that course of study in university (Ganley et al., 2017). In 2013, women in astronomy comprised 26 percent of assistant professors, 19 percent of associate professors, and 14 percent of full professors (Hughes, 2014).

Concrete recommendations on how to address this underrepresentation can be found in three studies highlighted here. The report of the 2015 Inclusive Astronomy (Nashville) Conference (hereafter referred to as IA2015), endorsed by the U.S. American Astrological Society (AAS), analyzes current barriers to access for URMs in astronomy and provides a wide-ranging set of recommendations for removing them.¹⁰ The National Academies recent report titled, Sexual Harassment of Women: Climate, Culture, and Consequences in Academic Sciences, Engineering, and Medicine (NASEM, 2018; hereafter referred to as SHW2018) lists evidence-based practices to achieve gender parity, most of which come from the substantial research NSF has funded through its ADVANCE program.¹¹ The Best Practices Guide of the AAS Committee for Sexual-Orientation and Gender Minorities in Astronomy provides ways to improve the climate for LGBT+ scientists, who are also marginalized, that apply to improving equity and inclusion for all.¹²

Adopting the recommendations presented by these three reports are important first steps toward addressing the barriers raised by homogeneity, bias, harassment, and discrimination in the astrophysics and planetary science workforce.¹³ The committee therefore fully endorses those recommendations.

___________________

⁸ URMs are defined in this report as Hispanic or Latino, Black or African American, American Indian or Alaska Native, or Native Hawaiian or other Pacific Islander.

⁹ See American Institute of Physics, “Percent of Bachelor’s Degrees and Doctorates in Astronomy Earned by Women, Classes 1986 through 2016,” 2016, https://www.aip.org/statistics/data-graphics/percent-bachelors-degrees-and-doctorates-astronomy-earned-women-classes.

¹⁰ See AAS Group Wiki, “Inclusive Astronomy: The Nashville Recommendations,” American Astronomical Society, https://tiki.aas.org/tiki-index.php?page=Inclusive_Astronomy_The_Nashville_Recommendations.

¹¹ See National Science Foundation, “ADVANCE: Organizational Change for Equity in STEM Academic Professions (ADVANCE),” https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=5383.

¹² See American Astronomical Society, LGBT+ Inclusivity in Physics and Astronomy: A Best Practices Guide, 2nd Ed., 2018, https://sgma.aas.org/sites/sgma.aas.org/files/LGBTInclusivityPhysicsAstronomy-BestPracticesGuide2ndEdn_small.pdf.

¹³ See Chapter 6 in SHW2018 (NASEM, 2018).

Harassment

The AAS Code of Ethics defines harassment as “behaviors that, if engaged in because of race, religion, color, gender, age, national origin, disability, marital status, sexual orientation, gender identity expression, or any other protected class, may give rise to a hostile work environment,” and the committee adopts this definition here.¹⁴

SHW2018 reports that, “In the best meta-analysis to date on sexual harassment prevalence, Ilies and colleagues (2003) reveal that 58 percent of female academic faculty and staff experienced sexual harassment.” Although SHW2018 does not deal with other forms of harassment, it recognizes that “women who have multiple marginalities . . . experience certain kinds of harassment at greater rates than other women.” The Clancy et al. (2017) survey of 474 astronomers and planetary scientists found that “39 percent of respondents report experiencing verbal harassment at their current position, and 9 percent report experiencing physical harassment,” while also reporting that “women of color experienced the most hostile environment.”

SHW2018 notes that male-dominated organizations are more likely to have sexual harassment within them so that “Two important steps in correcting this problem are achieving critical masses of women at every level and changing policies and practices that are impeding the ability for women to enter and advance in academia.” It goes on to say, “Gender parity, specifically among faculty, is especially important, given that faculty lead and set the tone in labs, medical teams, classrooms, departments, and schools.”

New codes of ethics of professional societies, including the AAS, reflect a broad realization that unethical environments prevent scientific excellence. The American Geophysical Union’s 2017 Scientific Integrity and Professional Ethics Policy recognizes harassment as a form of scientific misconduct, as does the SHW2018 report, which then goes on to recommend that institutions “Move beyond legal compliance to address culture and climate.”

Finding: To maximize scientific potential and opportunities for excellence, institutions and organizations can enable full participation by a diverse workforce by taking concrete steps to eliminate discrimination and harassment and to proactively recruit and retain scientists from underrepresented groups.

In society at large, retaliation against complainants is common.¹⁵ SHW2018 notes that fear of retaliation and negative career outcomes prevent women from reporting harassment and recommends that funders, such as NASA and NSF, support research into mechanisms for protecting targets from retaliation. This committee endorses the SHW2018 finding that “systems and policies that support targets of sexual harassment and provide options for informal and formal reporting can reduce the reluctance to report harassment as well as reduce the harm sexual harassment can cause the target.”

As SHW2018 finds, when reporting systems are not comprehensive, well advertised, and used to solve problems, informal “whisper networks” “are used to warn women away from particular programs, labs, or advisors,” which “automatically [reduces women’s] options and chances for career success.”¹⁶ Such networks also exist among other underrepresented groups. Furthermore, “Confidentiality agreements in settlements [can] shield harassment cases from view and make it possible for perpetrators to seek new jobs and keep problems secret” (Cantalupo and Kidder, 2017).

Discrimination and pervasive harassment within the greater scientific workforce likely affect the exoplanet community and serve as barriers to the participation of people from certain demographic groups. SHW2018 cites many studies to show that harassment causes scientists to be less productive and leave their jobs. This committee wants junior scientists to find the exoplanet field hospitable and to continue to contribute their talents to it.

___________________

¹⁴ See American Astronomical Society, “AAS Code of Ethics,” last updated October 11, 2017, https://aas.org/ethics#bullying.

¹⁵ See Cortina and Magley (2003) and U.S. Equal Opportunity Employment Commission, “Enforcement and Litigation Statistics,” https://www.eeoc.gov/eeoc/statistics/enforcement/.

¹⁶ SHW2018 (NASEM, 2018) cites S. Meza, “What Is a Whisper Network? How Women Are Taking Down Bad Men in the #MeToo Age,” Newsweek, November 22, 2017, http://www.newsweek.com/what-whisper-network-sexual-misconduct-allegations-719009.

Areas Needing Further Research

A variety of approaches to education and remediation may reduce harassment, discrimination, and other forms of abuse, not limited to illegal behavior. The committee suggests that the following areas merit further study:

1. Research into the most effective professional sanctions for different forms of abuse. SHW2018 recommends that academic institutions could consider a range of disciplinary actions.
2. Mechanisms to destigmatize reporting. IA2015 recommends that the astronomical community “create and highly publicize a robust reporting procedure to address all relevant dimensions of identity and social experience.” Such publicity might encourage reporting of incidents for which education may be a more appropriate response than punitive sanctions. Lowering the perceived bar for reporting problems may generate additional constructive opportunities for change, while increasing the likelihood that patterns of problematic behavior will be reported.
3. Training of numerous people within academia to provide support and resources to victims of harassment and discrimination, including advice both about how to stop or report the abuse and about the career implications of reporting. The number of trained people required to meet this need without posing an undue burden on each individual merits evaluation.
4. Demographic and climate surveys that reveal problems that may limit an institution’s success at recruiting and supporting a diverse workforce and track the efficacy of newly implemented procedures. Surveys should be evaluated by people who appreciate barriers to survey completion, particularly for members of groups whose identities can readily be discerned.
5. Examination of telescope and agency funding allocation processes for bias. For example, the Space Telescope Science Institute analyzed Hubble Space Telescope proposals and found that male PIs had a higher success rate (Reid, 2014).
6. Development of leadership training programs. SHW2018 particularly emphasized the importance of strong leadership for setting cultural standards. Leaders have the power and responsibility to build a culture of inclusivity within their group, department, or organization.

This list is not intended to be comprehensive, and the committee recognizes that a panel dedicated to studying workforce development may identify additional important items.

Decadal Survey Endorsement

IA2015 recommended that all stakeholders, including policy makers and funding agencies, “act proactively as well as reactively,” not “wait for a problem to arise to attempt to fix it.” It lists actions that can be taken by funding agencies to remove barriers to access and create inclusive environments with effective leaders.¹⁷

Finding: Development and dissemination of concrete recommendations to improve equity and inclusion and combat discrimination and harassment would be valuable for building the creative, interdisciplinary teams needed to maximize progress in exoplanet science over the coming decades.

The committee endorses the IA2015 recommendation that, “The decadal survey should address issues of policy making and leadership diversity imbalances as recommendations that can be acted upon by policy makers.” To achieve this goal, the Astronomy and Astrophysics and the Planetary Science Decadal Surveys will need to consult with experts beyond the astrophysics and planetary science communities and with members of underrepresented and marginalized groups.

As the exoplanet community continues to research and implement the recommendations cited above, these

___________________

¹⁷ See AAS Group Wiki, “Inclusive Astronomy: The Nashville Recommendations,” American Astronomical Society, https://tiki.aas.org/tiki-index.php?page=Inclusive_Astronomy_The_Nashville_Recommendations.

efforts will enable a diverse and productive cohort of scientists to accomplish the exciting, interdisciplinary, and profound goals of this report.

REFERENCES

Agol, E., J. Steffen, S. Re’em, and W. Clarkson. 2005. On detecting terrestrial planets with timing of giant planet transits. Monthly Notices of the Royal Astronomical Society 359(2):567.

Allard, F., P.H. Hauschildt, D.R. Alexander, A. Tamanai, and A. Schweitzer. 2001. The limiting effects of dust in brown dwarf model atmospheres. Astrophysical Journal 556(1):357.

Andrews, S.M., D.J. Wilner, Z. Zhu, T. Birnstiel, J.M. Carpenter, L.M. Pérez, X.-N. Bai, K.I. Öberg, A.M. Hughes, A. Isella, and L. Ricci. 2016. Ringed substructure and a gap at 1 AU in the nearest protoplanetary disk. Astrophysical Journal 820(2):40.

Angel, R., and A. Burrows. 1995. Seeking planets around nearby stars. Nature 374(6524):678.

Anglada-Escude, G., P.J. Amado, J. Barnes, Z.M. Berdiñas, R.P. Bulter, G.A.L. Coleman, I. de la Cueva, et al. 2016. A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536:437.

Angus, R., S. Aigrain, D. Foreman-Mackey, and A. McQuillan. 2015. Calibrating gyrochronology using Kepler asteroseismic targets. Monthly Notices of the Royal Astronomical Society 450(2):1787.

Arcangeli, J., J.-M. Désert, M.R. Line, J.L. Bean, V. Parmentier, K.B. Stevenson, L. Kreidberg, J.J. Fortney, M. Mansfield, and A.P. Showman. 2018. H opacity and water dissociation in the dayside atmosphere of the very hot gas giant WASP-18b. Astrophysical Journal Letters 855(2):L30.

Atkinson, D., D. Hall, C. Baranec, I. Baker, S. Jacobson, and R. Riddle. 2014. Observatory deployment and characterization of SAPHIRA HgCdTe APD arrays. Proceedings of SPIE 9154:915419.

Baliunas, S.L., R.A. Donahue, W.H. Soon, J.H. Horne, J. Frazer, L. Woodard-Eklund, M. Bradford, et al. 1995. Chromospheric variations in main-sequence stars. Astrophysical Journal, Part 1 438(1):269.

Barclay, T., J. Pepper, and E.V. Quintana. 2018. A revised exoplanet yield from the Transiting Exoplanet Survey Satellite (TESS). Submitted to the Astrophysical Journal; abstract at 2018arXiv180405050B.

Barnes, S.A., J. Weingrill, D. Fritzewski, K.G. Strassmeier, and I. Platais. 2016. Rotation periods for cool stars in the 4 Gyr old open cluster M67, the solar-stellar connection, and the applicability of gyrochronology to at least solar age. Astrophysical Journal 823(1):16.

Batalha, N.E., N.K. Lewis, M.R. Line, J. Valenti, and K. Stevenson. 2018. Strategies for constraining the atmospheres of temperate terrestrial planets with JWST. Astrophysical Journal Letters 856(2):L34.

Batalha, N.M., J.F. Rowe, S.T. Bryson, T. Barclay, C.J. Burke, D.A. Cadwell, J.L. Christiansen, et al. 2013. Planetary candidates observed by Kepler. III. Analysis of the first 16 months of data. Astrophysical Journal Supplement Series 204(2):24.

Batygin, K., and D.J. Stevenson. 2010. Inflating hot Jupiters with Ohmic dissipation. Astrophysical Journal Letters 714(2):L238.

Bean, J.L., D.S. Abbot, and E.M.-R. Kempton. 2017. A statistical comparative planetology approach to the hunt for habitable exoplanets and life beyond the Solar System. Astrophysical Journal Letters 841(2):L24.

Bean, J.L., J.-M. Désert, P. Kabath, B. Stalder, S. Seager, E.M.-R. Kempton, Z.K. Berta, D. Homeier, S. Walsh, and A. Seifahrt. 2011. The optical and near-infrared transmission spectrum of the super-Earth GJ 1214b: Further evidence for a metal-rich atmosphere. Astrophysical Journal 743(1):92.

Bean, J.L., E.M.-R. Kempton, and D. Homeier. 2010a. A ground-based transmission spectrum of the super-Earth exoplanet GJ 1214b. Nature 468:669.

Bean, J.L., A. Seifahrt, H. Hartman, H. Nilsson, G. Wiedemann, A. Reiners, S. Dreizler, and T.J. Henry. 2010b. The CRIRES search for planets around the lowest-mass stars. I. High-precision near-infrared radial velocities with an ammonia gas cell. Astrophysical Journal 713(1):410.

Bean, J.L., K.B. Stevenson, N.M. Batalha, Z. Berta-Thompson, L. Kreidberg, N. Crouzet, B. Benneke, et al. 2018. The transiting exoplanet community early release science program for JWST. Submitted to the Publications of the Astronomical Society of the Pacific; https://arxiv.org/pdf/1803.04985.pdf.

Ben-Ami, S., M. Lopez-Morales, and A. Szentgyorgyi. 2018. A Fabry Perot based instrument for biomarkers detection. Proceedings of SPIE 10702, Ground-based and Airborne Instrumentation for Astronomy VII, 107026N (8 July 2018); doi:10.1117/12.2313445.

Bennett, D.P., and S.H. Rhie. 2002. Simulation of a space-based microlensing survey for terrestrial extrasolar planets. Astrophysical Journal 574(2):985.

Bennett, D.P., J. Anderson, and B.S. Gaudi. 2006. Characterization of gravitational microlensisng planetary host stars. Astrophysical Journal 660(1):781.

Bennett, D.P., J. Anderson, J.-P. Beaulieu, I. Bond, E. Cheng, K. Cook, S. Friedman, et al. 2009. A census of exoplanets in orbits beyond 0.5 AU via space-based microlensing. White paper submitted to the Science Frontier Panel, National Research Council, Washington, D.C.

Bennett, D.P., J. Anderson, J.-P. Beaulieu, I. Bond, E. Cheng, K. Cook, S. Friedman, et al. 2010. “Completing the Census of Exoplanets with the Microlensing Planet Finder (MPF).” RFI Response for the Astr2010 Program Prioritization Panel: The Basis for the Exoplanet Program of the WFIRST Mission. https://arxiv.org/ftp/arxiv/papers/1012/1012.4486.pdf.

Bergin, E.A., L.I. Cleeves, U. Gorti, K. Zhang, G.A. Blake, J.D. Green, S.M. Andrews, N.J. Evans II, et al. 2013. An old disk still capable of forming a planetary system. Nature 493:644.

Berta-Thompson, Z.K., J. Irwin, D. Charbonneau, E.R. Newton, J.A. Dittmann, N. Astudillo-Defru, X. Bonfils, et al. 2015. A rocky planet transiting a nearby low-mass star. Nature 527:204-207.

Bowler, B.P. 2016. Imaging extrasolar giant planets. Publications of the Astronomical Society of the Pacific 128(968):102001.

Brogaard, K., C.J. Hansen, A. Miglio, D. Slumstrup, S. Frandsen, J. Jessen-Hansen, M.N. Lund, D. Bossini, A. Thygesen, G.R. Davies, W.J. Chaplin, T. Arentoft, H. Bruntt, F. Grundahl, and R. Handberg. 2018. Establishing the accuracy of asteroseismic mass and radius estimates of giant stars. I. Three eclipsing systems at [Fe/H]^- -0.3 and the need for a large high-precision sample. Monthly Notices of the Royal Astronomical Society 476(3):21.

Brown, R.A. 2004. Obscurational completeness. Astrophysical Journal 607(2):1003-1013.

Brown, T.M., J. Christensen-Dalsgaard, B. Weibel-Mihalas, and R.L. Gilliland. 1994. The effectiveness of oscillation frequencies in constraining stellar model parameters. Astrophysical Journal, Part 1 427(2):1013.

Brown, T.M., N. Baliber, F.B. Bianco, M. Bowman, B. Burleson, P. Conway, M. Crellin, et al. 2013. Las Cumbres Observatory Global Telescope Network. Publications of the Astronomical Society of the Pacific 125(931):1031.

Bryan, M.L., B. Benneke, H.A. Knutson, K. Batygin, and B.P. Bowler. 2018. Constraints on the spin evolution of young planetary-mass companions. Nature Astronomy 2:138.

Burke, C.J., J.L. Christiansen, F. Mullaly, S. Seader, D. Huber, J.F. Rose, J.L. Coughlin, et al. 2015. Terrestrial planet occurrence rates for the Kepler GK Dwarf sample. Astrophysical Journal 809(1):8.

Calchi Novati, S., A. Gould, A. Udalski, J.W. Menzies, I.A. Bond, Y. Shvartzvald, R.A. Street, et al. 2015. Pathway to the galactic distribution of planets: Combined Spitzer and ground-based microlens parallax measurements of 21 single-lens events. Astrophysical Journal 804(1):20.

Campante, T.L., T. Barclay, J.J. Swift, D. Huber, V.Zh. Adibekyan, W. Cochran, C.J. Burke, et al. 2015. An ancient extrasolar system with five sub-Earth-size planets. Astrophysical Journal 799(2):17.

Cantalupo, N.C., and W. Kidder. 2017. A systematic look at a serial problem: Sexual harassment of students by university faculty. Utah Law Review 2018:671.

Carr, J.S., and J.R. Najita. 2008. Organic molecules and water in the planet formation region of young circumstellar disks. Science 319(5869):1504.

Cash, W. 2006. Detection of Earth-like planets around nearby stars using a petal-shaped occulter. Nature 442(7098):51.

Cassan, A., D. Kubas, J.-P. Bealieu, M. Dominik, K. Horne, J. Greenhill, J. Wambsganss, et al. 2012. One or more bound planets per Milky Way star from microlensing observations. Nature 481:167.

Catling, D.C., J. Krissansen-Totton, N.Y. Kiang, D. Crisp, T.D. Robinson, S. DasSarma, A. Rushby, et al. 2018. Exoplanet biosignatures: A framework for their assessment. Astrobiology doi:10.1089/ast.2017.1737.

Chaplin, W.J., S. Basu, D. Huber, A. Serenelli, L Casagrande, V. Silva Aguirre, W.H. Ball, et al. 2014. Asteroseismic fundamental properties of solar-type stars observed by the NASA Kepler mission. Astrophysical Journal Supplemental Series 210(1):1.

Clancy, K.B.H., K.M.N. Lee, E.M. Rodgers, and C. Richey. 2017. Double jeopardy in astronomy and planetary science: Women of color face greater risks of gendered and racial harassment. Journal of Geophysical Research Planets 122(7):1610.

Clarke, C.J., A. Gendrin, and M. Sotomayor. 2001. The dispersal of circumstellar discs: The role of the ultraviolet switch. Monthly Notices of the Royal Astronomical Society 328(2):485.

Cleeves, L.I., E.A. Bergin, C.M. O’D Alexander, F. Du, D. Graninger, K. Öberg, and T.J. Harries. 2014. The ancient heritage of water ice in the Solar System. Science 345(6204):1590.

Cleeves, L.I., E.A. Bergin, C.M. O’D Alexander, F. Du, D. Graninger, K. Öberg, and T.J. Harries. 2016. Exploring the origins of deuterium enrichments in solar nebular organics. Astrophysical Journal 819(1):13.

Coffeen, D.L. 1979. Polarization and scattering characteristics in the atmospheres of Earth, Venus, and Jupiter. Journal of the Optical Society of America 69:1051.

Collins, K.A., K.I. Collins, J. Pepper, J. Labadie-Bartz, K. Stassun, S.B. Gaudi, D. Bayliss, et al. 2018. The KELT follow-up network and transit false positive catalog: Pre-vetted false positives for TESS. Submitted to Astronomy and Astrophysics; arXiv:1803.01869.

Cooper, C.S., and A.P Showman. 2006. Dynamics and disequilibrium carbon chemistry in hot Jupiter atmospheres, with application to HD 209458b. Astrophysical Journal 649(2):1048-1063.

Cortina, L.M., and V.J. Magley. 2003. Raising voice, risking retaliation: Events following interpersonal mistreatment in the workplace. Journal of Occupational Health Psychology 8(4):247-265.

Cowan, N.B., E. Agol, V.S. Meadows, T. Robinson, T.A. Livengood, D. Deming, C.M. Lisse, M.F. A’Hearn, D.D. Wellnitz, S. Seager, and D. Charbonneau. 2009. Alien maps of an ocean-bearing world. Astrophysical Journal 700(2):915.

Cowan, N.B., T. Greene, D. Angerhausen, N.E. Batalha, M. Clampin, K. Colón, I.J.M. Crossfield, et al. 2015. Characterizing transiting planet atmospheres through 2025. Publications of the Astronomical Society of the Pacific 127(949):311.

Crepp, J.R., E.J. Gonzales, E.B. Bechter, B.T. Montet, J.A. Johnson, D. Piskorz, A.W. Howard, and H. Isaacson. 2016. The trends high-contrast imaging survey. VI. Discovery of a mass, age, and metallicity benchmark brown dwarf. Astrophysical Journal 831(2):136.

Crill, B.P., and N. Siegler. 2017. Space technology for directly imaging and characterizing exo-Earths. Proceedings of SPIE 10398:103980H.

Crossfield, I.J.M. 2013. On high-contrast characterization of nearby, short-period exoplanets with giant segmented-mirror telescopes. Astronomy and Astrophysics 551:A99.

Crossfield, I.J.M., and L. Kreidberg. 2017. Trends in atmospheric properties of Neptune-size exoplanets. Astronomical Journal 14(6):261.

Crossfield, I.J.M., B. Biller, J.E. Schlieder, N.R. Deacon, M.Bonnefoy, D. Homeier, F. Allard, et al. 2014. A global cloud map of the nearest known brown dwarf. Nature 505:654.

Davenport, J.R.A., D.M. Kipping, D. Sasselov, J.M. Matthews, and C. Cameron. 2016. MOST observations of our nearest neighbor: Flares on Proxima Centauri. Astrophysical Journal Letters 829(2):L31.

Davenport, J.R.A., S.L. Hawley, L. Hebb, J.P. Wisniewski, A.F. Kowalski, E.C. Johnson, M. Malatesta, et al. 2014. Kepler flares. II. The temporal morphology of white-light flares on GJ 1243. Astrophysical Journal 797(2):122.

David, T.J., I.J. Crossfield, B. Benneke, E.A. Petigura, E.J. Gonzales, J.E. Schlieder, L. Yu, et al. 2018. Three small planets transiting the bright young field star K2-233. Astronomical Journal 155(5):222.

David, T.J., L.A. Hillenbrand, E.A. Petigura, J.M. Carpenter, I.J.M. Crossfield, S. Hinkley, D.R. Ciardi, et al. 2016. A Neptune-size transiting planet closely orbiting a 5-10-million-year-old star. Nature 534(7609):658.

Davis, A.B., J. Cisewski, X. Dumusque, D.A. Fischer, and E.B. Ford. 2017. Insights on the spectral signatures of stellar activity and planets from PCA. Astrophysical Journal 846(1).

Des Marais, D.J., M.O. Harwit, K.W. Jucks, J.F. Kasting, D.N.C. Lin, J.I. Lunine, J. Schneider, S. Seager, W.A. Traub, and N.J. Woolf. 2002. Remote sensing of planetary properties and biosignatures on extrasolar terrestrial planets. Astrobiology 2(2):153.

Dittmann, J.A., J.M. Irwin, D. Charbonneau, X. Bonfils, N. Astudillo-Defru, R.D. Haywood, Z.K. Berta-Thompson, et al. 2017. A temperate rocky super-Earth transiting a nearby cool star. Nature 544(7650):333.

Donati, J.F., C. Moutou, L. Malo, C. Baruteau, L. Yu, E. Hébrard, G. Hussain, et al. 2016. A hot Jupiter orbiting a 2-million-year-old solar-mass T Tauri star. Nature 534(7609):662.

Dong, C., M. Jin, M. Lingam, V.S. Airapetian, Y. Ma, and B. van der Holst. 2018. Atmospheric escape from the TRAPPIST-1 planets and implications for habitability. Proceedings of the National Academy of Sciences of the United States of America 115(2):260.

Dravins, D., M. Gustavsson, and H.-G. Ludwig. 2018. Spatially resolved spectroscopy across stellar surfaces. III. Photospheric Fe I lines across HD 189733A (K1 V). Submitted to Astronomy and Astrophysics; https://arxiv.org/pdf/1806.00012.pdf.

Dressing, C.D., and D. Charbonneau. 2015. The occurrence of potentially habitable planets orbiting M dwarfs estimated from the full Kepler dataset and an empirical measurement of the detection sensitivity. Astrophysical Journal 807(1):45.

Driscoll, P.E., and R. Barnes. 2015. Tidal heating of Earth-like exoplanets around M stars: Thermal, magnetic, and orbital evolutions. Astrobiology 15(9):739.

Dumusque, X. 2016. Radial velocity fitting challenge. I. Simulating the data set including realistic stellar radial-velocity signals. Astronomy and Astrophysics 593:A5.

Dumusque, X., A. Glenday, D.F. Phillips, N. Buchschacher, A. Collier Cameron, M. Cecconi, D. Charbonneau, et al. 2015. HARPS-N observes the Sun as a star. Astrophysical Journal Letters 814(2):L21.

Dumusque, X., F. Borsa, M. Damasso, R.F. Díaz, P.C. Gregory, N.C. Hara, A. Hatzes, et al. 2017. Radial-velocity fitting challenge. II. First results of the analysis of the data sets. Astronomy and Astrophysics 598:A133.

Dumusque, X., I. Boisse, and N.C. Santos. 2014. SOAP 2.0: A tool to estimate the photometric and radial velocity variations induced by stellar spots and plages. Astrophysical Journal 796(2):132.

Duncan, D.K., A.H. Vaughan, O.C. Wilson, G.W. Preston, J. Frazer, H. Lanning, A. Misch, et al. 1991. CA II H and K measurements made at Mount Wilson Observatory, 1966-1983. Astrophysical Journal Supplement Series 76:383.

Ehrenreich, D., V. Bourrier, P.J. Wheatley, A. Lecavelier des Etangs, G. Hébrard, S. Udry, X. Bonfils, X. Delfosse, J.-M. Désert, D.K. Sing, and A. Vidal-Madjar. 2015. A giant comet-like cloud of hydrogen escaping the warm Neptune-mass exoplanet GJ 436b. Nature 522(7557):459.

Ertel, S., D. Defrere, P. Hinz, B. Mennesson, G.M. Dennedy, W.C. Danchi, C. Gelino, et al. 2018. The HOSTS survey-exozodiacal dust measurements for 30 stars. Astronomical Journal 155(5):194.

ESA (European Space Agency). 2017. Plato Definition Study Report. ESA-SCI(2017)1. April 4. http://sci.esa.int/plato/59252plato-definition-study-report-red-book/.

Fabrycky, D.C., and R.A. Murray-Clay. 2010. Stability of the directly imaged multiplanet system HR 8799: Resonance and masses. Astrophysical Journal 710:1408.

Feng, Y., J.J. Fortney, and M.R. Line. 2016. The impact of non-uniform thermal structure on the interpretation of exoplanet emission spectra. American Astronomical Society, DPS meeting #48, id.212.03.

Fischer, D.A., G. Anglada-Escude, P. Arriagada, R.V. Baluev, J.L. Bean, F. Bouchy, L.A. Buchhave, et al. 2016. State of the field: Extreme precision radial velocities. Publications of the Astronomical Society of the Pacific 128(964):066001.

Fleming, B.T., K. France, N. Nell, R. Kohnert, K. Pool, A. Egan, L. Fossati, et al. 2018. Colorado Ultraviolet transit experiment: A dedicated CubeSat mission to study exoplanetary mass loss and magnetic fields. Journal of Astronomical Telescopes, Instruments, and Systems, volume 4 id#014004.

Fontenla, J.M., J.L. Linsky, J. Witbrod, K. France, A. Buccino, P. Mauas, M. Vieytes, and L.M. Walkowicz. 2016. Semi-empirical modeling of the photosphere, chromosphere, transition region, and corona of the M-dwarf host star GJ 832. Astrophysical Journal 830(2):154.

Fortney, J.J., M.S. Marley, D. Saumon, and K. Lodders. 2008. Synthetic spectra and colors of young giant planet atmospheres: Effects of initial conditions and atmospheric metallicity. Astrophysical Journal 683(2):1104.

Fortney, J.J., T.D. Robinson, S. Domagal-Goldman, D.S. Amundsen, M. Brogi, M. Claire, D. Crisp, et al. 2016. The need for laboratory work to aid in the understanding of exoplanetary atmospheres. White paper initiated within the NASA Nexus for Exoplanetary System Science (NExSS).

France, K., C.S. Froning, J.L. Linksy, A. Roberge, J.T. Stocke, F. Tian, R. Bushinsky, J.-M. Désert, P. Mauas, M. Vieytes, and L.M. Walkowicz. 2013. The ultraviolet radiation environment around M dwarf exoplanet host stars. Astrophysical Journal 763(2):149.

France, K., J.L. Linsky, F. Tian, C.S. Froning, and A. Roberge. 2012. Time-resolved ultraviolet spectroscopy of the M-dwarf GJ 876 exoplanetary system. Astrophysical Journal Letters 750(2):L32.

France, K., R.O.P. Loyd, A. Youngblood, A. Brown, C.P. Schneider, S.L. Hawley, C.S. Froning, et al. 2016. The MUSCLES treasury survey. I. Motivation and overview. Astrophysical Journal 820(2):89.

Fujii, Y., D. Angerhausen, R. Deitrick, S. Domagal-Goldman, J.L. Grenfell, Y. Hori, S.R. Kane, et al. 2018. Exoplanet biosignatures: Observational prospects. Astrobiology doi:10.1089/ast.2017.1733.

Fujii, Y., A.D. Del Genio, and D.S. Amundsen. 2017. NIR-driven moist upper atmospheres of synchronously rotating temperature terrestrial exoplanets. Astrophysical Journal 848(2):100.

Fulton, B.J., and E.A. Petigura. 2018. The California Kepler survey VII. Precise planet radii leveraging Gaia DR2 reveal the stellar mass dependence of the planet radius gap. Astrophysical Journal, in press.

Fulton, B.J., E.A. Petigura, A.W. Howard, H. Isaacson, G.W. Marcy, P.A. Cargile, L. Hebb, et al. 2017. The California-Kepler survey. III. A gap in the radius distribution of small planets. Astronomical Journal 154(3):109.

Ganley, C.M., C.E. George, J.R. Cimpian, and M.B. Makowski. 2017. Gender equity in college majors: Looking beyond the STEM/non-STEM dichotomy for answers regarding female participation. American Educational Research Journal 55(3):453.

Garcia-Sage, K., A. Glocer, J.J. Drake, G. Gronoff, and O. Cohen. 2017. On the magnetic protection of the atmosphere of Proxima Centauri b. Astrophysical Journal Letters 844:L13.

Garrett, D., D. Savransky, and B. Macintosh. 2017. A simple depth-of-search metric for exoplanet imaging surveys. Astronomical Journal 154(2):47.

Ghez, A.M., S. Salim, N.N. Weinberg, J.R. Lu, T. Do, J.K. Dunn, K. Matthews, et al. 2008. Measuring distance and properties of the Milky Way’s central supermassive black hole with stellar orbits. Astrophysical Journal 689(2):1044.

Gillon, M., A.H.M.J. Triaud, B.O. Demory, E. Jehin, E. Agol, K.M. Deck, S.M. Lederer, et al. 2017. Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542:456.

Gilmozzi, R., and J. Spyromilio. 2007. The European extremely large telescope (E-ELT). Messenger 127:11.

Gordon, I.E., L.S. Rothman, C. Hill, R.V. Kochanov, Y. Tan, P.F. Bernath, M. Birk, et al. 2017. The HITRAN2016 molecular spectroscopic database. Journal of Quantitative Spectroscopy and Radiative Transfer 203:3.

Green, J., P. Schechter, C. Baltay, R. Bean, D. Bennett, R. Brown, C. Conselice, et al. 2012. Wide-Field InfraRed Survey Telescope (WFIRST) Final Report. arXiv:1208.4012.

Grießmeier, J.-M., P. Zarka, and H. Spreeuw. 2007. Predicting low-frequency radio fluxes of known extrasolar planets. Astronomy and Astrophysics 475:359.

Guyon, O. 2005. Limits of adaptive optics for high-contrast imaging. Astrophysical Journal 629(1):592.

Guyon, O., and J. Males. 2017. Adaptive optics predictive control with empirical orthogonal functions (EOFs). Accepted by Astronomical Journal; arXiv:1707.00570.

Guyon, O., E.A. Pluzhnik, M.J. Kuchner, B. Collins, and S.T. Ridgeway. 2006. Theoretical limits on extrasolar terrestrial planet detection with coronagraphs. Astrophysical Journal Supplemental Series 167(1):81.

Halverson, S., R. Terrien, S. Mahadevan, A. Roy, C. Bender, G.K. Stefánsson, A. Monson, et al. 2016. A comprehensive radial velocity error budget for next generation Doppler spectrometers. Proceedings of SPIE 9908:99086P.

Handberg, R., K. Brogaard, A. Miglio, D. Bossini, Y. Elsworth, D. Slumstrup, G.R. Davies, and W.J. Chaplin. 2017. NGC 6819: Testing the asteroseismic mass scale, mass loss and evidence for products of non-standard evolution. Monthly Notices of the Royal Astronomical Society 472(1):979.

Hauschildt, P.H., E. Baron, and F. Allard. 1997. Parallel implementation of the PHOENIX generalized stellar atmosphere program. Astronomical Journal 483(1):390.

Haywood, R.D., A. Collier Cameron, Y.C. Unruh, C. Lovis, A.F. Lanza, J. Llama, M. Deleuil, et al. 2016. The Sun as a planet-host star: Proxies from SDO images for HARPS radial-velocity variations. Monthly Notices of the Royal Astronomical Society 457(4):3637.

Hoeijmakers, H.J., H. Schwarz, I.A.G Snellen, R.J. de Kok, M. Bonnefoy, G. Chauvin, A.M. Lagrange, and J.H. Girard. 2018. Medium-resolution integral-field spectroscopy for high-contrast exoplanet imaging: Molecule maps of the beta Pictoris system with SINFONI. Astronomy and Astrophysics arXiv:1802.09721.

Holman, M.J., and N.W. Murray. 2005. The use of transit timing to detect terrestrial-mass extrasolar planets. Science 307(5713):1288.

Howard, A.W., and B.J. Fulton. 2016. Limits on planetary companions from Doppler surveys of nearby stars. Publications of the Astronomical Society of the Pacific 128(969):114401.

Howard, W.S., M.A. Tilley, H. Corbett, A. Youngblood, R.O. Parke Loyd, J.K. Ratzloff, N.M. Law, et al. 2018. The first naked-eye superflare detected from Proxima Centauri. Astrophysical Journal Letters 860(2):6.

Hu, R., S. Seager, and W. Bains. 2012. Photochemistry in terrestrial exoplanet atmospheres. I. Photochemistry model and benchmark cases. Astrobiology Journal 761(2):166.

Huang, C.X., A. Shporer, D. Dragomir, M. Fausnaugh, A.M. Levine, E.H. Morgan, T. Nguyen, G.R. Ricker, M. Wall, D.F. Woods, and R.K. Vanderspek. 2018. Expected yields of planet discoveries from the TESS primary and extended missions. Earth and Planetary Astrophysics 1807:11129.

Hughes, A.M. 2014. “The 2013 CSWA Demographic Survey: Portrait of a Generation of Women in Astronomy.” From January 2014 Status: A Report on Women in Astronomy; http://womeninastronomy.blogspot.com/2014/03/the-2013-cswa-demographics-survey.html.

Inclusive Astronomy. 2015. “Inclusive Astronomy 2015 Recommendations (or the “Nashville Recommendations”).” Council of the American Astronomical Society, Nashville, Tenn., June 17-19. https://docs.google.com/document/d/1JipEb7xz7kAh8SH4wsG59CHEaAJSJTAWRfVA1MfYGM8/edit?pref=2&pli=1#.

Jeffers, S.V., M. Mengel, C. Moutou, S.C. Marsden, J.R. Barnes, M.M. Jardine, P. Petit, J.H.M.M Schmitt, V. See, and A.A. Vidotto. 2018. The relation between stellar magnetic field geometry and chromospheric activity cycles II: The rapid 120 day magnetic cycle of Tau Bootis. Accepted by the Monthly Notices of the Royal Astronomical Society.

Johns, M., P. McCarthy, K. Raybould, A. Bouchez, A. Farahani, J. Filgueira, G. Jacoby, S. Shectman, and M. Sheehan. 2012. Giant Magellan Telescope: Overview. Proceedings of the SPIE 8444:84441H.

Johns-Krull, C.M., J.N. McLane, L. Prato, C.J. Crockett, D.T. Jaffe, P.M. Hartigan, C.A. Beichman, et al. 2016. A candidate young massive planet in orbit around the classical T Tauri star CI Tau. Astrophysical Journal 826(2):206.

Jovanovic, N., C. Schwab, O. Guyon, J. Lozi, N. Cvetojevic, F. Martinache, S. Leon-Saval, et al. 2017. Efficient injection from large telescopes into single-mode fibres: Enabling the era of ultra-precision astronomy. Astronomy and Astrophysics 604:A122.

Jovanovic, N., R. Martinache, O. Guyon, C. Clergeon, G. Singh, T. Kudo, V. Garrel, K. Newman, D. Doughty, and J. Lozie. 2015. The Suburu Coronagraphic Extreme Adaptive Optics System: Enabling high-contrast imaging on solar-system scales. Publications of the Astronomical Society of the Pacific 127(955):890.

Kaltenegger, L., A. Segure, and S. Mohanty. 2011. Model spectra of the first potentially habitable super-earth—G1581d. Astrophysical Journal 733(1):35.

Kama, M., S. Bruderer, M. Carney, M. Hogerheijde, E.F. can Dishoeck, D. Fedele, A. Baryshev, et al. 2016. Observations and modelling of CO and [C I] in protoplanetary disks: First detections of [C I] and constraints on the carbon abundance. Astronomy and Astrophysics 588:A108.

Kao, M.M., G. Hallinana, J.S. Pineda, I. Escala, A. Burgasser, S. Bourke, and D. Stevenson. 2016. Auroral radio emission from late L and T dwarfs: A new constraint on dynamo theory in the substellar regime. Astrophysical Journal 818(1):24.

Kastner, J.H., D.A. Principe, K. Punzi, B. Stelzer, U. Gorti, I. Pascucci, and C. Argiroffi. 2016. M stars in the TW Hya Association: Stellar x-rays and disk dissipation. Astronomical Journal 152(1):3.

Kempton, E.M.-R., J.L. Bean, D.R. Louie, D. Deming, D.D.B. Koll, M. Mansfield, M. Lopez-Morales, et al. 2018. A framework for prioritizing the TESS planetary candidates most amendable to atmospheric characterization. Submitted to Publications of the Astronomical Society of the Pacific; https://arXiv.org/pdf/1805.03671.pdf.

Kempton, E.M.-R., R. Perna, and K. Heng. 2014. High resolution transmission spectroscopy as a diagnostic for Jovian exoplanet atmospheres: Constraints from theoretical models. Astrophysical Journal 795(1):24.

Keppler, M., M. Benisty, A. Müller, Th. Henning, R. van Boekel, F. Cantalloube, C. Ginski, et al. 2018. Discovery of a planetary-mass companion within the gap of the transition disk around PDS 70. Earth and Planetary Astrophysics 1806:11568.

Kiang, N.Y., S. Domagal-Goldman, M.N. Parenteau, D.C. Catling, Y. Fujii, V.S. Meadows, E.W. Schwieterman, and S.I. Walker. 2018. Exoplanet biosignatures: At the dawn of a new era of planetary observations. Astrobiology doi:10.1089/ast.2018.1862.

Kim, S.-L., C.-U. Lee, B.-G. Park, D.-J. Kim, S.-M. Cha, Y. Lee, C. Han, M.-Y. Chun, and I. Yuk. 2016. KMTNET: A network of 1.6 m wide-field optical telescopes installed at three southern observatories. Journal of Korean Astronomical Society 49(1):37.

Koen, C., L.A. Balona, K. Khadaroo, I. Laine, A. Prinsloo, B. Smith, and C.D. Laney. 2003. Pulsations in β Pictoris. Monthly Notices of the Royal Astronomical Society 344(4):1250.

Konopacky, Q.M., T.S. Barman, B.A. Macintosh, and C. Marois. 2013. Detection of carbon monoxide and water absorption lines in an exoplanet atmosphere. Science 339(6126):1398.

Kopparapu, R.K., J.F. Kasting, and K.J. Zahnle. 2012. A photochemical model for the carbon-rich planet WASP-12b. Astrophysical Journal 745(1):77.

Kopparapu, R.K., E. Hébrard, R. Belikov, N.M. Batalha, G.D. Mulders, C. Stark, D. Teal, S. Domagal-Goldman, and A. Mandell. 2018. Exoplanet classification and yield estimates for direct imaging missions. Astrophysical Journal 856(2):122.

Koskinen, T.T., R.V. Yelle, P. Lavvas, and N.K. Lewis. 2010. Characterizing the thermosphere of HD209458b with UV transit observations. Astrophysical Journal 723(1):116.

Kraus, A.L., and M.J. Ireland. 2011. LkCa 15: A young exoplanet caught in formation? Astrophysical Journal 745(1):5.

Kreidberg, L., and A. Loeb. 2016. Prospects for characterizing the atmosphere of Proxima Centauri b. Astrophysical Journal Letters 832(1):L12.

Kreidberg, L., J.L. Bean, J.-M. Désert, B. Benneke, D. Deming, K.B. Stevenson, S. Seager, Z. Berta-Thompson, A. Seifahrt, and D. Homeier. 2014. Clouds in the atmosphere of the super-Earth exoplanet GJ 1214b. Nature 505:69.

Kreidberg, L., M.R. Line, V. Parmentier, K.B. Stevenson, T. Louden, M. Bonnefoy, J.K. Faherty, et al. 2018. Global climate and atmospheric composition of the ultra-hot Jupiter WASP-103b from HST and Spitzer phase curve observations. Available at https://arxiv.org/pdf/1805.00029.pdf.

Kulow, J.R., K. France, J. Linsky, and R.O.P. Loyd. 2014. Lyalpha transit spectroscopy and the neutral hydrogen tail of the hot Neptune GJ 436b. Astrophysical Journal 786(2):132.

Lammer, H., H.I.M. Lichtenegger, Y.N. Kulikov, J.-M. Grießmeier, N. Terada, N.V. Erkaev, H.K. Biernat, M.L. Khodachenko, I. Ribas, T. Penz, and F. Selsis. 2007. Coronal mass ejection (CME) activity of low mass M stars as an important factor of the habitability of terrestrial exoplanets. II. CME-induced ion pick up of Earth-like exoplanets in close-in habitable zones. Astrobiology 7(1):185.

Lazio, T.J.W., E. Shkolnik, G. Hallinan, and the Planetary Habitability Study Team. 2016. Planetary Magnetic Fields: Planetary Interiors and Habitability. Report to the W.M. Keck Institute for Space Studies. http://kiss.caltech.edu/papers/magnetic/papers/Magnetosphere.pdf.

Lecavelier Des Etangs, A., D. Ehrenreich, A. Vidal-Madjar, G.E. Ballester, J.-M. Désert, R. Ferlet, G. Hébrard, D.K. Sing, K.-O. Tchakoumegni, and S. Udry. 2010. Evaporation of the planet HD 189733b observed in H I Lyman-alpha. Astronomy and Astrophysics 514:A72.

Lindegren, L. 2018. The Tycho-Gaia astrometric solution. Proceedings of the International Astronomical Union 12(Symposium s330):41.

Line, M.R., M.C. Liang, and Y.L. Yung. 2010. High-temperature photochemistry in the atmosphere of HD 189733b. Astrophysical Journal 717(1):496.

Linksy, J. 2014. The radiation environment of exoplanet atmospheres. Challenges 5:351-373.

Lissauer, J.J., G.W. Marcy, J.F. Rowe, S.T. Bryson, E. Adams, L.A. Buchhave, D.R. Ciardi, et al. 2012. Almost all of Kepler’s multi-planet candidates are planets. Astrophysical Journal 750(2):112.

Lithwick, Y., J. Xie, and Y. Wu. 2012. Extracting planet mass and eccentricity from TTV data. Astrophysical Journal 761(2):122.

Livingston, J.H., F. Dai, T. Hirano, D. Gandolfi, G. Nowak, M. Endl, S. Velasco, et al. 2018. Three small planets transiting a Hyades star. The Astronomical Journal 155(3):115.

Llama, J., and E.L. Shkolnik. 2015. Transiting the Sun: The impact of stellar activity on x-ray and ultraviolet transits. Astrophysical Journal 802(1):41.

Louie, D.R., D. Deming, L. Albert, L.G. Bouma, J. Bean, and M. Lopez-Morales. 2018. Simulated JWST/NIRISS transit spectroscopy of anticipated tess planets compared to select discoveries from space-based and ground-based surveys. Publications of the Astronomical Society of the Pacific 130(986):044401.

Lovis, C., I. Snellen, D. Mouillet, F. Pepe, F. Wildi, N. Astudillo-Defru, J.-L. Beuzit, et al. 2017. Atmospheric characterization of Proxima b in coupling the SPHERE high-contrast imager to the ESPRESSO spectrograph. Astronomy and Astrophysics 599:A16.

Loyd, R.O.P., and K. France. 2014. Fluctuations and flares in the ultraviolet line emission of cool stars: Implications for exoplanet transit observations. Astrophysical Journal Supplement 211(1):9.

Luger, R., and R. Barnes. 2015. Extreme water loss and abiotic O-2 buildup on planets throughout the habitable zones of M dwarfs. Astrobiology 15:119.

Lupu, R.E., M.S. Marley, N. Lewis, M. Line, W.A. Traub, and K. Zahnle. 2016. Developing atmospheric retrieval methods for direct imaging spectroscopy of gas giants in reflected light. I. Methane abundances and basic cloud properties. Astronomical Journal 152(6):217.

Lustig-Yaeger, L., G. Tovar, Y. Fujii, E.W. Schweiterman, and V.S. Meadows. 2017. “Mapping Surfaces and Clouds on Terrestrial Exoplanets Observed with Next-Generation Coronagraph-Equipped Telescopes.” Presentation to the Astrobiology Science Conference 2017. Paper 3558. LPI Contrib. No. 1965. https://www.hou.usra.edu/meetings/abscicon2017/pdf/3558.pdf.

MacGregor, M.A., A.J. Weinberger, D.J. Wilner, A.F. Kowalski, and S.R. Cranmer. 2018. Detection of a millimeter flare from Proxima Centauri. Astrophysical Journal Letters 855(1):6.

Mahadevan, S., L.W. Ramsey, R. Terrien, S. Halverson, A. Roy, F. Hearty, E. Levi, et al. 2014. The habitable-zone planet finder: A status update on the development of a stabilized fiber-fed near-infrared spectrograph for the Hobby-Eberly telescope. Proceedings of SPIE 9147:91471G.

Maiolino, R., M. Haehnelt, M.T. Murphy, D. Queloz, L. Origlia, J. Alcala, Y. Alibert, et al. 2013. “A Community Science Case for E-ELT HIRES.” White paper submitted for E-ELT HIRES. https://arxiv.org/pdf/1310.3163.pdf.

Malbet, F., J.W. Yu, and M. Shao. 1995. High-dynamic-range imaging using a deformable mirror for space coronography. Publications of the Astronomical Society of the Pacific 107(710):386.

Males, J.R., and O. Guyon. 2018. Ground-based adaptive optics coronagraphic performance under closed-loop predictive control. Journal of Astronomical Telescopes, Instruments, and Systems 4(1):019001.

Mandell, A.M., J. Bast, E.F. van Dishoeck, G.A. Blake, C. Salyk, M.J. Mumma, and G. Villanueva. 2012. First detection of near-infrared line emission from organics in young circumstellar disks. Astrophysical Journal 747(2):92.

Mann, A.W., E. Gaidos, A. Vanderburg, A.C. Rizzuto, M. Ansdell, J.V. Medina, G.N. Mace, A.L. Kraus, and K.R. Sokal. 2017. Zodiacal Exoplanets in Time (ZEIT). IV. Seven transiting planets in the Praesepe cluster. Astronomical Journal 153(2):64.

Mann, A.W., E.R. Newton, A.C. Rizzuto, J. Irwin, G.A. Feiden, E. Gregory, E. Gaidos, et al. 2016. Zodiacal exoplanets in Time (ZIET). III. A short-period planet orbiting a pre-main-sequence star in the Upper Scorpius OB Association. The Astronomical Journal 152(3):61.

Mansfield, M., J.L. Bean, M.R. Line, V. Parmentier, L. Kreidberg, J.-M. Desert, J.J. Fortney, K.B. Stevenson, J. Arcangeli, and D. Dragomir. 2018. A HST/WFC3 thermal emission spectrum of the hot Jupiter HAT-P-7b. https://arxiv.org/pdf/1805.00424.pdf.

Mawet, D., J.R. Delorme, N. Jovanovic, J.K. Wallace, R.D. Bartos, P.L. Wizinowich, M. Fitzgerald, et al. 2017. A fiber injection unit for the Keck planet imager and characterizer. Proceedings of the SPIE 10400:1040029.

McClure, M.K., E.A. Bergin, L.I. Cleeves, E.F. van Dishoeck, G.A. Blake, N.J. Evans II, J.D. Green, Th. Henning, K.I. Öberg, K.M. Pontoppidan, and C. Salyk. 2016. Mass measurements in protoplanetary disks from hydrogen deuteride. Astrophysical Journal 831(2):167.

Meadows, V.S. 2017. Reflections on O₂ as a biosignature in exoplanetary atmospheres. Astrobiology 17(10):1022.

Meadows, V.S., and R.K. Barnes. 2018. Factors affecting exoplanet habitability. Pp. 1-24 in Handbook of Exoplanets (H. Deeg and J. Belmonte, eds.). Springer, Cham, Switzerland.

Meadows, V.S., G.N. Arney, E.W. Schweiterman, J. Lustig-Yaeger, A.P. Lincowski, T. Robinson, S.D. Domagal-Goldman, et al. 2018. The habitability of Proxima Centauri b: Environmental states and observational discriminants. Astrobiology 18(2):133.

Meeker, S.E., B.A. Mazin, A.B. Walter, P. Strader, N. Fruitwala, C. Bockstiegel, P. Szypryt, et al. 2018. DARKNESS: A microwave kinetic inductance detector integral field spectrograph for high-contrast astronomy. Publications of the Astronomical Society of the Pacific 130(988):065001.

Miles, B.E., and E.L. Shkolnik. 2017. HAZMAT. II. Ultraviolet variability low-mass stars in the GALEX archive. Astronomical Journal 154(2):67.

Millar-Blanchaer, M.A., M.D. Perrin, L.-W. Hung, M.P. Fitzgerald, J.J. Wang, J. Chilcote, S. Bruzzone, and P.G. Kalas. 2016. GPI observational calibrations XIV: Polarimetric contrasts and new data reduction techniques. Proceedings of SPIE 9908:990836.

Montet, B.T., J.R. Crepp, J.A. Johnson, A.W. Howard, and G.W. Marcy. 2014. The trends high-contrast imaging survey. IV. The occurrence rate of giant planets around M dwarfs. Astrophysical Journal 781(1):28.

Morley, C.V., A.J. Skemer, K.N. Allers, M.S. Marley, J.K. Faherty, C. Visscher, S.A. Beiler, et al. 2018. An L band spectrum of the coldest brown dwarf. Astrophysical Journal 858(2):97.

Morley, C.V., L. Kreidberg, Z. Rustamkulov, T. Robinson, and J.J. Fortney. 2017. Observing the atmospheres of known temperate Earth-size planets with JWST. Astronomical Journal 850(2):121.

Morzinski, K.M., L.M. Close, K.M. Morzinski, Z. Wahhaj, M.C. Liu, A.J. Skemer, D. Kopon, et al. 2015. Magellan adaptive optics first-light observations of the exoplanet β Pic b. II. 3-5 μm direct imaging with MagAO+Clio, and the empirical bolometric luminosity of a self-luminous giant planet. Astrophysical Journal 815(2):24.

Moses, J.I., N. Madhusudhan, C. Visscher, and R.S. Freedman. 2013. Chemical consequences of the C/O ratio on hot Jupiters: Examples from WASP-12b, CoRoT-2b, XO-1b, and HD 189733b. Astrophysical Journal 763(1):25.

Muterspaugh, M.W., B.F. Lane, S.R. Kulkarni, M. Konacki, B.F. Burke, M.M. Colavita, M. Shao, W.I. Hartkopf, A.P. Boss, and M. Williamson. 2010. The phases differential astrometry data archive. V. Candidate substellar companions to binary systems. Astronomical Journal 140(6):1657.

NASA. 2015. Exo-S: Starshade Probe-Class Exoplanet Direct Imaging Mission Concept: Final Report. March. https://exoplanets.nasa.gov/internal_resources/788/.

NASA. 2017. WFIRST Independent External Technical/Management/Cost Review (WIETR). October 19. https://www.nasa.gov/sites/default/files/atoms/files/wietr_final_report_101917.pdf.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2018. Sexual Harassment of Women: Climate, Culture, and Consequences in Academic Sciences, Engineering, and Medicine. The National Academies Press, Washington, D.C.

Nemati, B., J.E. Krist, and B. Mennesson. 2017. Sensitibity of the WFIRST coronagraph performance to key instrument parameters. Proceedings of the SPIE 10400:1040007.

Nettelmann, N., U. Kramm, R. Redmer, and R. Neuhauser. 2010. Interior structure models of GJ 436b. Astronomy and Astrophysics 523:A26.

NRC (National Research Council). 2010. New Worlds, New Horizons in Astronomy and Astrophysics. The National Academies Press, Washington, D.C.

NRC. 2016. New Worlds, New Horizons: A Midterm Assessment. The National Academies Press, Washington, D.C.

NSF (National Science Foundation). 2018. Important Notice No. 144: Harassment. https://www.nsf.gov/pubs/issuances/in144.jsp.

Öberg, K.I., and E.A. Bergin. 2016. Excess C/O and C/H in outer protoplanetary disk gas. Astrophysical Journal Letters 831(2):L19.

Olson, S.L., E.W. Schwieterman, C.T. Reinhard, A. Ridgwell, S.R. Kane, V.S. Meadows, and T.W. Lyons. 2018. Atmospheric seasonality as an exoplanet biosignature. Astrophysical Journal Letters 858(2):L14.

Owen, J.E., B. Ercolano, C.J. Clarke, and R.D. Alexander. 2010. Radiation-hydrodynamic models of X-ray and EUV photo-evaporating protoplanetary discs. Monthly Notices of the Royal Astronomical Society 401(3):1415.

Pallé, E., E.B. Ford, S. Seager, P. Montanés-Rodríguez, and M. Vazquez. 2008. Identifying the rotation rate and the presence of dynamic weather on extrasolar Earth-like planets from photometric observations. Astrophysical Journal 676:1319.

Parmentier, V., M.R. Line, J.L. Bean, M. Mansfeild, L. Kreidberg, R. Lupu, C. Visscher, et al. 2018. From thermal dissociation to condensation in the atmospheres of ultra hot Jupiters: WASP-121b in context. https://arXiv.org/pdf/1805.00096.pdf.

Peacock, S., T.S. Barman, and E. Shkolnik. 2015. “HAZMAT II: Modeling the Evolution of Extreme-UV Radiation from M Stars.” Presentation to the American Astronomical Society, Meeting #225, id#138.26. American Astronomical Society, Washington, D.C.

Penny, M., J.E. Rodriguez, T. Beatty, and G. Zhou. 2018. “Alpha Elements’ Effects on Planet Formation and the Hunt for Extragalactic Planets.” Presentation to the American Astronomical Society, Meeting #231, id#427.01. American Astronomical Society, Washington, D.C.

Pepper, J., E. Gillen, H. Parviainen, L.A. Hillenbrand, A.M. Cody, S. Aigrain, J. Stauffer, et al. 2017. A low-mass exoplanet candidate detected by K2 transiting the Praesepe M dwarf JS 183. Astronomical Journal 153(4):177.

Pinte, C., D.J. Price, F. Menard, G. Duchene, W.R.F. Dent, T. Hill, I. de Gregorio-Monsalvo, A. Hales, and D. Mentiplay. 2018. Kinematic evidence for an embedded protoplanet in a circumstellar disc. Accepted for publication in the Journal of Astrophysics Letters.

Plavchan, P., C. Bilinski, and T. Currie. 2014. Investigation of Kepler objects of interest stellar parameters from observed transit durations. Astronomical Society of the Pacific 126(935):34-47.

Plavchan, P., D. Latham, S. Gaudi, J. Crepp, X. Dumusque, G. Furesz, A. Vanderburg, et al. 2015. Radial velocity prospects current and future: A white paper report prepared by the Study Analysis Group 8 for the Exoplanet Program Analysis Group (ExoPAG). Instrumentation and Methods for Astrophysics, Earth and Planetary Astrophysics 1503:01770.

Pold, J., R. Ivie, I. Momcheva, and the AAS Demographic Committee. 2016. Workforce Survey of 2016 US AAS Members Summary Results. American Astronomical Society. https://aas.org/files/aas_members_workforce_survey.pdf.

Poleski, R., and J. Yee. 2018. Microlensing model fitting with Mulens Model. Submitted to Astronomy and Computing. https://arxiv.org/pdf/1803.01003.pdf.

Pont, F., D.K. Sing, N.P. Gibson, S. Aigrain, G. Henry, and N. Husnoo. 2013. The prevalence of dust on the exoplanet HD 189733b from Hubble and Spitzer observations. Monthly Notices of the Royal Astronomical Society 432(4):2917.

Poyneer, L.A., B.A. Macintosh, and J.-P Véran. 2007. Fourier transform wavefront control with adaptive prediction of the atmosphere. Journal of the Optical Society of America A 24(9):2645.

Prisinzano, L., G. Micela, E. Flaccomio, J.R. Stauffer, T. Megeath, L. Rebull, M. Robberto, K. Smith, E.D. Feigelson, N. Grosso, and S. Wolk. 2008. X-ray properties of protostars in the Orion Nebula. Astrophysical Journal 677(1):401.

Qi, C., K.I. Oberg, D.J. Wilner, P. D’Alessio, E. Bergin, S.M. Andrews, G.A. Blake, M.R. Hogerheijde, and E.F. van Dishoeck. 2013. Imaging of the CO snow line in a solar nebula analog. Science 341(6146):630.

Quanz, S.P., I. Crossfield, M.R. Meyer, A. Schmalzl, and J. Held. 2015. Direct detection of exoplanets in the 3-10 μm rand with E-ELT/METIS. International Journal of Astrobiology 14(2):279.

Queloz, D., G.W. Henry, J.P. Sivan, S.L. Baliunas, J.L. Beuzit, R.A. Donahue, M. Mayor, D. Naef, C. Perrier, and S. Udry. 2001. No planet for HD 166435. Astronomy and Astrophysics 379:279.

Rackham, B.V., D. Apai, and M.S. Giampapa. 2018. The transit light source effect: False spectral features and incorrect densities for M-dwarf transiting planets. Astrophysical Journal 853(2):122.

Rauscher, B.J., E.R. Canavan, S.H. Moseley, J.E. Sadleir, and T. Stevenson. 2016. Detectors and cooling technology for direct spectroscopic biosignature characterization. Journal of Astronomical Telescopes, Instruments, and Systems 2:041212.

Rauscher, E., and E.M.R. Kempton. 2014. The atmospheric circulation and observable properties of nonsynchronously rotating hot Jupiters. Astrophysical Journal 790(1):79.

Reid, N.I. 2014. Gender-correlated systematics in HST proposal selection. Publications of the Astronomical Society of the Pacific 126(944):923.

Ricker, G.R., J.N. Winn, R. Vanderspek, D.W. Latham, G.A. Bakos, J.L. Bean, Z.K. Berta-Thompson, et al. 2015. Transiting Exoplanet Survey Satellite (TESS). Journal of Astronomical Telescopes, Instruments, and Systems 1:014003.

Robinson, T.D., K.R. Stapelfeldt, and M.S. Marley. 2016. Characterizing rocky and gaseous exoplanets with 2 m class space-based coronagraphs. Publications of the Astronomical Society of the Pacific 128(960):025003.

Robinson, T.D., L. Maltagliati, M.S. Marley, and J.J. Fortney. 2014. Titan solar occultation observations reveal transit spectra of a hazy world. Proceedings of the National Academy of Sciences of the United States of America 111(25):9042.

Robinson, T.D., V.S. Meadows, and D. Crisp. 2010. Detecting oceans on extrasolar planets using the glint effect. Astrophysical Journal Letters 721(1):L67.

Rodler, F., and M. López-Morales. 2014. Feasibility studies for the detection of O₂ in an Earth-like exoplanet. Astrophysical Journal 781(1):54.

Rogers, L.A., and S. Seager. 2010. A framework for quantifying the degeneracies of exoplanet interior composition. Astrophysical Journal 712(2):974.

Rugheimer, S., L. Keltenegger, A. Segura, J. Linsky, and S. Mohanty. 2015. Effect of UV radiation on the spectral fingerprints of Earth-like planets orbiting M stars. Astrophysical Journal 809(1):57.

Sahlmann, J., P.F. Lazorenko, D. Ségransan, E.L. Martín, M. Mayor, D. Queloz, and S. Udry. 2014. Astrometric planet search around southern ultracool dwarfs. I. First results, including parallaxes of 20 M8-L2 dwarfs. Astronomy and Astrophysics 565(A20):19.

Sallum, S., K.B. Follette, J.A. Eisner, L.M. Close, P. Hinz, K. Kratter, J. Males, et al. 2015. Accreting protoplanets in the LkCa 15 transition disk. Nature 527:342.

Salyk, C., K.M. Pontoppidan, G.A. Blake, F. Lahuis, E.F. van Dishoeck, and N.J. Evans II. 2008. H₂O and OH gas in the terrestrial planet-forming zones of protoplanetary disks. Astrophysical Journal Letters 676(1):L49-L52.

Sanders, G.H. 2013. The Thirty Meter Telescope (TMT): An international observatory. Journal of Astrophysics and Astronomy 34(2):81.

Savransky, D., J.D.N. Kasdin, and B.A. Belson. 2009. The utility of astrometry as a precursor to direct detection. Proceedings of SPIE 7440:74400B.

Schaefer, L., R.D. Wordsworth, Z. Berta-Thompson, and D. Sasselov. 2016. Predictions of the atmospheric composition of GJ 1132b. Astrophysical Journal 829(2):63.

Schneider, A.C., and E.L. Shkolnik. 2018. HAZMAT. III. The UV evolution of mid- to late-M stars with GALEX. Astronomical Journal 155(3):122.

Schwab, C., A. Rakich, Q. Gong, S. Mahadevan, S.P. Halverson, A. Roy, R.C. Terrien, et al. 2016. Design of NEID, an extreme precision Doppler spectrograph for WIYN. Proceedings of the SPIE 9908:99087H.

Schwarz, K.R., E.A. Bergin, L.I. Cleeves, K. Zhang, K.I. Oberg, G.A. Blake, and D. Anderson. 2018. Unlocking CO depletion in protoplanetary disks. I. The warm molecular layer. Astrophysical Journal 856(1):85.

Schwieterman, E.W., N.Y. Kiang, M.N. Parenteau, C.E. Harman, S. DasSarma, T.M. Fisher, G.N. Arney, et al. 2018. Exoplanet biosignatures: A review of remotely detectable signs of life. Astrobiology doi:10.1089/ast.2017.1729.

Seager, S., M. Turnbull, W. Sparks, M. Thomson, S.B. Shaklan, A. Roberge, M. Kuchner, et al. 2015. The Exo-S probe class starshade mission. Proceedings of the SPIE 9605:96050W.

Segura, A., J.F. Kasting, V.S. Meadows, M. Cohen, J. Scalo, D. Crisp, R.A. Butler, and G. Tinetti. 2005. Biosignatures from Earth-like planets around M dwarfs. Astrobiology 5(6):706.

Segura, A., K. Krelove, J.F. Kasting, D. Sommerlatt, V.S. Meadows, D. Crisp, M. Cohen, and E. Mlawer. 2003. Ozone concentrations and ultraviolet fluxes on Earth-like planets around other stars. Astrobiology 3(4):689.

Segura, A., L.M. Walkowicz, V.S. Meadows, J. Kasting, and S. Hawley. 2010. The effect of a strong stellar flare on the atmospheric chemistry of an Earth-like planet orbiting an M dwarf. Astrobiology 10(7):751.

Selsis, F., J.F. Kasting, B. Levrard, J. Paillet, I. Ribas, and X. Delfosse. 2007. Habitable planets around the star Gliese 581? Astronomy and Astrophysics 476(3):1373.

Shkolnik, E., D.A. Bohlender, G.A.H. Walker, and A.C. Cameron. 2008. The on/off nature of star-planet interactions. Astrophysical Journal 676(1):628.

Shkolnik, E.L., and J. Llama. 2017. Signatures of star-planet interactions. Pp. 1-17 in Handbook of Exoplanets (H. Deeg and J. Belmonte, eds.). Springer, Cham, Switzerland.

Shkolnik, E.L., and T.S. Barman. 2014. HAZMAT. I. The evolution of far-UV and near-UV emission from early M stars. Astronomical Journal 148(4):64.

Shkolnik, E.L., D. Ardila, T. Barman, M. Beasley, J.D. Bowman, V. Gorjian, D. Jacobs, et al. 2018. Monitoring the high-energy radiation environment of exoplanets around low-mass stars with SPARCS (Star-Planet Activity Research CubeSat). Presentation to the American Astronomical Society, Meeting #231, id#228.04.

Shvartzvald, Y., G. Bryden, A. Gould, C.B. Henderson, S.B. Howell, and C. Beichman. 2017. UKIRT microlensing surveys as a pathfinder for WFIRST: The detection of five highly extinguished low -| b| events. Astronomical Journal 153(2):61.

Sing, D.K., J.J. Fortney, N. Nikolov, H.R. Wakeford, T. Kataria, T.M. Evans, S. Aigrain, et al. 2016. A continuum from clear to cloudy hot-Jupiter exoplanets without primordial water depletion. Nature 529(7584):59.

Snellen, I.A.G., B.R. Brandl, R.J. de Kok, M. Brogi, J. Birkby, and H. Schwarz. 2014. Fast spin of the young extrasolar planet β Pictoris b. Nature 509:63.

Snellen, I.A.G., R. de Kok, J.L. Birky, B. Brandl, M. Brogi, C. Keller, M. Kenworthy, H. Schwarz, and R. Stuik. 2015. Combining high-dispersion spectroscopy with high contrast imaging: Probing rocky planets around our nearest neighbors. Astronomy and Astrophysics 576:A59.

Snellen, I.A.G., R.J. de Kok, E.J.W. de Mooji, and S. Albrecht. 2010. The orbital motion, absolute mass and high-altitude winds of exoplanet HD 209458b. Nature 465:1049.

Snellen, I.A.G., J.M. Désert, L.B.F.M. Waters, T. Robinson, V. Meadows, E.F. van Dishoeck, B. Brandl, et al. 2017. Detecting Proxima b’s atmosphere with JWST targeting CO₂ at 15 μm using a high-pass spectral filtering technique. Astronomical Journal 154(77).

Soderblom, D.R. 2010. The ages of stars. Annual Review of Astronomy and Astrophysics 48:581.

Spergel, D., N. Gehrels, C. Baltay, D. Bennett, J. Breckinridge, M. Donahue, A. Dressler, et al. 2015. Science Definition Team and WFIRST Study Office Report: Wide-Field InfrarRed Survey Telespoce—Astrophysics Focused Telescope Assets WFIRST-AFTA. https://arXiv.org/ftp/arxiv/papers/1503/1503.03757.pdf.

Spitzer, L. 1963. Star formation. Pp. 39-53 in Origin of the Solar System (R. Jastrow and A.G.W. Cameron eds.). Academic Press, New York.

Stark, C.C., A. Roberge, A. Mandell, M. Clampin, S.D. Domagal-Goldman, M.W. McElwain, and K.R. Stapelfeldt. 2015. Lower limits on aperture size for an ExoEarth detecting coronagraphic mission. Astrophysical Journal 802(2):149.

Stark, C.C., A. Roberge, A. Mandell, and T.D. Robinson. 2014. Maximizing the ExoEarth candidate yield from a future direct imaging mission. Astrophysical Journal 795(2):122.

Stelzer, B., A. Marino, G. Micela, J. López-Santiago, and C. Liefke. 2013. The UZ and X-ray activity of the M dwarfs within 10 pc of the Sun. Monthly Notices of the Royal Astronomical Society 431(3):2063.

Stevenson, K.B., N.K. Lewis, J.L. Bean, C. Beichman, J. Fraine, B.M. Kilpatrick, J.E. Krick, et al. 2016. Transiting exoplanet studies and community targets for JWST’s early release science program. Publications of the Astronomical Society of the Pacific 128(976):094401.

Sullivan, P.W., J.N. Winn, Z.K. Berta-Thompson, D. Charbonneau, D. Deming, C.D. Dressing, D.W. Latham, et al. 2015. The transiting exoplanet survey satellite: Simulations of planet detections and astrophysical false positives. Astrophysical Journal 809(1):99.

Teague, R., J. Bae, E.A. Bergin, T. Birnstiel, and D. Foreman-Mackey. 2018. A kinematical detection of two embedded Jupiter mass planets in HD 163296. Accepted for publication in the Journal of Astrophysics Letters.

Thorngren, D.P., and J.J. Fortney. 2018. Bayesian analysis of hot-Jupiter radius anomalies: Evidence for Ohmic dissipation? Astronomical Journal 155(5):214.

Tilley, M.A., A. Segura, V.S. Meadows, S. Hawley, and J. Davenport. 2017. Modeling repeated M-dwarf flaring at an Earthlike planet in the habitable zone: I. Atmospheric effects for an unmagnetized planet. Submitted to Astrobiology. https://arXiv.org/pdf/1711.08484.pdf.

Tinetti, G., P. Drossart, P. Eccleston, P. Hartogh, J. Leconte, G. Micela, M. Ollivier, et al. 2017. “The Science of ARIEL.” European Planetary Science Congress 2017. EPSC2017-713. https://www.epsc2017.eu/.

Tovar, G., B. Montet, and J.A. Johnson. 2017. Understanding Activity Cycles of Solar Type Stars with Kepler. American Astronomical Society Meeting #229, id.154.14. American Astronomical Society, Washington, D.C.

Trauger, J.T., and W.A. Traub. 2007. A laboratory demonstration of the capability to image an Earth-like extrasolar planet. Nature 446(7137):771.

Tsai, S.-M., D. Kitzmann, J.R. Lyons, J. Mendonca, S.L. Grimm, and K. Heng. 2017. Towards consistent modeling of atmospheric chemistry and dynamics in exoplanets: Validation and generalization of chemical relaxation method. Astrophysical Journal 862(1):31

Turbet, M., E. Bolmont, J. Leconte, F. Selsis, F. Forget, F. Selsis, G. Tobie, A. Caldas, J. Naar, and M. Gillon. 2018. Modeling climate diversity, tidal dynamics and the fate of volatiles on TRAPPIST-1 planets. Astronomy and Astrophysics 612:A86.

Turbet, M., J. Leconte, F. Selsis, E. Bolmont, F. Forget, I. Ribas, S.N. Raymond, and G. Anglada-Escudé. 2016. The habitability of Proxima Centauri b II. Possible climates and observability. Astronomy and Astrophysics 596:A112.

Valencia, D., T. Guillot, V. Parmentier, and R.S. Freedman. 2013. Bulk composition of GJ 1214b and other sub-Neptune exoplanets. Astrophysical Journal 775(1):10.

van Haarlem, M.P., M.W. Wise, A.W. Gunst, G. Heald, J.P. McKean, J.W.T. Hessels, A.G. de Bruyn, et al. 2013. LOFAR: The Low-Frequency Array. Astronomy and Astrophysics 556:A2.

van Holstein, R.G., F. Smik, J.H. Girard, J. de Boer, C. Ginski, C.U. Keller, D.M. Stam, et al. 2017. Combing angular differential imaging and accurate polarimetry with SPHERE/IRDIS to characterize young giant exoplanets. Proceedings of the SPIE 10400:1040015.

Vidal-Madjar, A., A. Lecavelier des Etangs, J.-M. Désert, G.E. Ballester, R. Ferlet, G. Hébrard, and M. Mayor. 2003. An extended upper atmosphere around the extrasolar planet HD 209458b. Nature 422(6928):143.

Wagner, K., K.B. Follette, L.M. Close, D. Apai, A. Gibbs, M. Keppler, A. Müller, et al. 2018. Magellan adaptive optics imaging of PDS 70: Measuring the mass accretion rate of a young giant planet within a gapped disk. Earth and Planetary Astrophysics, Solar and Stellar Astrophysics 1807:10766.

Walker, S.I., W. Bains, L. Cronin, S. DasSarma, S. Denielache, S. Domagal-Goldman, B. Kacar, et al. 2018. Exoplanet biosignatures: Future directions. Astrobiology doi:10.1089/ast.2017.1738.

Walsh, C., R.A. Loomis, K.I. Oberg, M. Kama, M.L.R. van‘t Hoff, T.J. Millar, Y. Aikawa, E. Herbst, S.L.W. Weaver, and H. Nomura. 2016. First detection of gas-phase methanol in a protoplanetary disk. Astrophysical Journal Letters 823(1):L10.

Wang, J., D. Mawet, G. Ruane, R. Hu, and B. Benneke. 2017. Observing exoplanets with high dispersion coronagraphy. I. The scientific potential of current and next-generation large ground and space telescopes. Astronomical Journal 153(4):183.

Williams, D.M., and E. Gaidos. 2008. Detecting the glint of starlight on the oceans of distant planets. Icarus 195(2):927.

Wright, J.T., and P. Robertson. 2017. The third workshop on extremely precise radial velocities: The new instruments. Research Notes of the American Astronomical Society 1(1):51.

Wright, J.T., and S. Sigurdsson. 2018. Advances in precise radial velocimetry from cross-disciplinary work in heliophysics, stellar astronomy, and instrumentation. White paper submitted to the Committee on Exoplanet Science Strategy, National Academies of Sciences, Engineering, and Medicine, Washington, D.C.

Yee, J.C., M. Albrow, R.K. Barry, D. Bennett, G. Bryden, S.-J. Chung, B.S. Gaudi, et al. 2014. “Preparing for the WFIRST Microlensing Survey.” Presentation by the NASA Exoplanet Exploration Study Analysis Group 11. https://arXiv.org/pdf/1409.2759.pdf.

Zarka, P., J. Lazio, and G. Hallinan. 2014. “Magnetospheric Radio Emissions from Exoplanets with the SKA.” Presentation to Advancing Astrophysics with the Square Kilometer Array, June 9-13, Giardini Naxos, Italy. Proceedings of Science Online. http://pos.sissa.it/cgi-bin/reader/conf.cgi?confid=215, id.174.

Zarka, P., J. Queinnec, B.P. Ryabov, V.B. Ryabov, V.A. Shevchenka, A.V. Arkhipov, H.O. Rucker, L. Denis, A. Gerbault, P. Dierich, and C. Rosolen. 1997. Ground-based high sensitivity radio astronomy at decameter wavelengths. Pp. 101 in Planetary Radio Emission IV (H.O. Rucker, S.J. Bauer, and A. Lecacheux, eds.). Austrian Academy of Sciences Press, Vienna, Austria.

Zellem, R.T., M.R. Swain, G. Roudier, E.L. Shkolnik, M.J. Creech-Eakman, D.R. Ciardi, M.R. Line, A.R. Iyer, G. Bryden, J. Llama, and K.A. Fahy. 2017. Forecasting the impact of stellar activity on transiting exoplanet spectra. Astrophysical Journal 844(1):27.

Zerkle, A.L., M.W. Calire, S.D. Domagal-Goldman, J. Farquhar, and S.W. Poulton. 2012. A bistable organic-rich atmosphere on the Neoarchaean Earth. Nature Geoscience 5:359.

Zhu, W., and A. Gould. 2016. Augmenting WFIRST microlensing with a ground-based telescope network. Journal of the Korean Astronomical Society 49(3):93.