Potential Exoplanet Direct-Imaging Science with the
Potential Exoplanet Direct-Imaging Science with the WFIRST Coronagraph Instrument (CGI) N. J. Kasdin (CGI Adjutant Scientist) Eugene Higgins Professor Mechanical and Aerospace Engineering Princeton University [email protected], 609-258-5673 V. Bailey*, B. Menneson*, J. Trauger, S. Hildebrandt*, M. Frerking, J. Rhodes, E. Cady, B. Kern, L. Moustakas, A J E. Riggs JPL/Caltech M. Marley* NASA Ames B. Nemati* University of Alabama, Huntsville J. Debes*, N. Lewis, A. Bellini, J. Girard, J. Kakirai, M. Perrin, L. Pueyo, R. Soummer, C. Stark, R. van der Marel STScI
A. Roberge*, D. Bennett, A. Bhattacharya, K. Carpenter, W. Danchi, T. Groff, J. Kruk, S. Malhotra, A. Mandell, M. McElwain, E. Quintana, C. Ranc, M. Rizzo, J. Schlieder, N. Zimmerman, NASA GSFC R. Akeson, S. Ramirez, T. Meshkat, G. Helou, R. Paladini, C.Gelino, M. Yagouf, S. Laine, D. Gelino, S. Calchi Novati, D. Ciardi, P. Lowrance Caltech/IPAC M. Penny Ohio State University
H. Jang-Condell* University of Wyoming
M. Meyer University of Michigan
D. Savransky* Cornell University
A. Boccaletti, P. Baudoz, T. Bhowmik, D. Rouan, T. Schmidt Paris Observatory
M. Turnbull SETI Institute S. Kane UC Riverside *Principle contributors.
T. Henning, A. Maire, H. Avenhaus MPIA
M. Tamura University of Tokyo/ABC/NAOJ T. Sumi Osaka University N. Murakami University of Hokaido J. Nishikawa NAOJ J. Kwon JAXA/ISAS M. Langlois, M. N’Diaye, A. Vigan, K. Dohlen CNRS G. Chauvin FCLA/IPAG S. Messina INAF-Catania J. Baudino Oxford/AOPP M. Janson Stockholm University
WFIRST CGI Direct Imaging Science
Introduction One of the two instruments on NASA’s Wide-Field Infrared Survey Telescope (WFIRST) is a coronagraph technology demonstrator. The Coronagraph Instrument (CGI) will demonstrate in space, for the first time, key technologies necessary for future exo-Earth imaging missions, including: high actuator count deformable mirrors; low-noise, single photon-counting detectors in the visible; new coronagraph masks and architectures; low-resolution integral field spectroscopy; advanced algorithms for wavefront sensing and control; high-fidelity integrated spacecraft and coronagraph modeling; and post-processing approaches to extract images and spectra. The importance of this technology demonstration to lowering the design and implementation risk of the next generation of large telescopes cannot be overstated. More details on these technologies and their specific value to future missions can be found in the white paper by Bailey et al. While CGI is nominally manifested as a technology demonstrator, NASA is committed to also making CGI available to the general astronomy community as a science instrument assuming a successful completion of its technology demonstration in the first 1.5 years of mission operations. In this white paper, we describe the potential high-contrast science that CGI may achieve and how this science will advance community goals in exoplanet astronomy and inform future observations. The scientific yield described here uses an integration time estimator that models the detector performance and incorporates planet photon noise, zodi and exo-zodi levels, and our current predictions of coronagraph performance, as described in the Bailey et al. whitepaper. With the project scheduled to enter Phase B in April 2018, further refinements to the integrated model are expected along with accompanying improvements in our best estimates of performance. These current predictions show that CGI will provide over two orders of magnitude improvement in detection and spectroscopic capability from what is currently possible on the ground (Figure 1). And while the sample of planets reachable by CGI is small, the potential science and the opportunity to refine our extraction approaches is manifest. In the remainder of this white paper we describe the following science potentially achievable by CGI: visible light spectroscopy and photometry of known young, selfluminous planets and known radial velocity planets; a blind search for undiscovered Jupiters and miniNeptunes; and visible light imaging of protoplanetary and debris disks. The observations described in this paper assume a set of four CGI filters: two 10% bandwidth filters centered at 575nm and 825nm, and Figure 1: Predicted CGI performance on a V=5 star, in two 18% bandwidth filters, centered the context of known planetary systems and current at 660nm and 760nm. The filter set instrumentation. See Bailey et al. whitepaper Appendix A for is sensitive to molecular bands of a further details on assumptions and data sources. variety of strengths as well as the Page 1
WFIRST CGI Direct Imaging Science continuum flux level, providing a range of information to place constraints on atmospheric structure and composition. Potential observing modes are described in the Bailey et al white paper.
Visible Light Spectra of Known Self-Luminous Giant Planets About two dozen young, low-mass companions to main sequence stars have been discovered by coronagraphic imaging methods (see review by Bowler & Nielsen 2018). These observations shed light on the formation of planets at orbital radii >10 au. To date, near- and mid-infrared spectroscopy and photometry have provided some constraints on the atmospheric abundance of C and O-bearing species and clouds on these planets, as well as on planet mass and temperature (e.g., Konopacky et al. (2013); Rajan et al. (2017)). Optical spectroscopy and photometry would help place constraints on the composition and atmospheric structure of these planets. The optical spectra are sculpted by strong, broad absorption features of gaseous Na and K (Figure 2). The abundance of these species provides additional constraints on metallicity, temperature, and gravity (and thus planet mass). The wavelength dependence of cloud and haze opacity is strongly sensitive to particle size (e.g., Zahnle et al. 2016). High altitude photochemical haze particles are typically small and thus invisible in the near-infrared, but they Figure 2: Planet to star contrast as a function of wavelength for can have an outsized impact at optical young, self-luminous giant planets with and without clouds. The wavelengths (e.g., Gao et al. 2017). effective temperatures of the star and planet are 6000K and 1100K. Estimated flux ratios for several young planets are shown in Figure 1. CGI should be able to obtain photometry of several of these objects with favorable flux ratios and spectroscopy of a few.
Characterization of Known RV Gas Giants in Reflected Visible Light Much of exoplanet science has focused on studying the atmospheres of transiting planets, particularly the hot Jupiters. While very interesting in their own right, these highly irradiated planets are not representative of the mature extrasolar giant planets found at larger orbital distances. Hundreds of such gas giant planets, more akin to our own Jupiter, have been discovered by radial velocity surveys, yet we know very little about their masses or atmospheric composition. CGI will, for the first time, allow further study and characterization of these planets. Several images spaced out over a planetary orbit will resolve the (sin i) degeneracy inherent in RV detections and reveal the planets’ true masses. CGI photometry in multiple passbands along with spectroscopy will characterize their atmospheres for the first time. Clouds, hazes, gravity, and atmospheric metallicity all play important roles in controlling giant planet spectra in reflected light. These effects are well understood from studies of solar system giants and two decades of theoretical modeling (e.g., Marley et al. 1999; Burrows et al. 2004; Cahoy et al. 2010) and thus we are well equipped to interpret CGI data on these planets. Retrieval studies on simulated WFIRST datasets (Lupu et al. 2016; Nayak et al. 2017; Lacy et al. 2018) Page 2
WFIRST CGI Direct Imaging Science demonstrate that even with broad band photometry, WFIRST users will be able to distinguish the presence or absence of cloud layers and atmospheric methane (Figure 3), providing clues into atmospheric temperature and composition. With R~50 spectra and photometry, limits can be placed on atmospheric abundance of key absorbers such as CH4 and H2O. While the WFIRST CGI sample will be small, it is crucial to test the end-to-end modeling and interpretation of planets detected in reflected light before much larger scale missions set out to find and characterize potentially habitable planets. Understanding the extent to which the contributions of clouds, hazes, and other effects can be discerned from reflected light data will give confidence that future missions studying more challenging targets will be properly prepared to succeed. A white paper by Marley & Lewis discusses this point in further detail.
Figure 3: The effects of various atmospheric parameters on the geometric albedo spectra of a gas giant planet. Clockwise from top left, the impact of varying atmospheric methane mixing ratio, surface gravity, cloud top pressure, and cloud single scattering albedo. Simultaneous detection of multiple methane features as well as the continuum flux level permits many of the apparent degeneracies to be resolved with sufficient quality data.
Blind Search Discovery of Gas Giants and Mini-Neptunes The WFIRST CGI will also be capable of searching for new planets as well as characterizing previously discovered ones. The performance modeling described above can also be used to study the overall sensitivity of CGI to various planet classes. CGI will be most sensitive to giant planets greater than 5 Earth radii between 1 and 10 au, detecting these at a rate of greater than 50% in systems where they are present, but will also have some sensitivity down to 2 Earth radii, particularly between 0.5 and 2 au. If a population of high albedo (p > 0.7) 1.5-2 Earth radii planets exists Page 3
WFIRST CGI Direct Imaging Science around and within 1 au, as described by Kaltenegger and Sasselov (2011), then CGI could potentially detect one or more habitable zone mini-Neptunes/Super-Earths. Otherwise, it is likely that all new CGI discoveries will be gas and ice giants within 10 au of their host star. While the total number of new planets detected by CGI would likely be small (on the order of 10, depending on assumptions made – see Figure 4 and Savransky et al., 2016), carrying out an exoplanet imaging survey from space would be hugely valuable to the next generation of exoplanets missions. A blind search for new planets would include prioritizing targets under the various constraints of the real mission and scheduling revisits to confirm planet candidates. Lessons learned from the CGI survey would be directly applicable to all future exoplanets imagers and may even influence their designs and mission formulations. Figure 4: Depth of search (summed completeness assuming one planet per star at each location in the semi-major axis-Radius phase space) for the WFIRST CGI assuming optimal utilization of 3 months of integration time (96 total targets). Giant planets (>5 Earth radii) are found most frequently (>50%) while the CGI is barely sensitive to 2 Earth-radii planets (0.5%). Albedos were assumed to be uniformly 0.2 for planets below 1.4 Earth radii and 0.5 for larger planets.
Visible Light Imaging of Protoplanetary and Debris Disks and Protoplanets In addition to the potential exoplanet science, CGI will be capable of imaging protoplanetary and debris disks. Studying these disks is key to advancing our understanding of the formation and evolution of planetary systems. A combination of photometry, polarization properties, and scattered light brightness as a function of scattering angle can provide constraints on dust grain shape, composition, and grain size distribution (Perrin et al., 2015 and Rodiagas et al., 2015). Imaging of known circumstellar disks represents a guaranteed science return from WFIRST. Figure 5 shows the predicted performance of the CGI imaging mode (for a V=5 star), compared to a selection of protoplanetary and debris disks previously observed in visible scattered light. CGI has the potential to characterize known disks in at least two photometric bands, in both total intensity and polarized intensity with unprecedented fidelity, and it has the ability to detect disks that are orders of magnitude fainter than previously observed. Imaging of protoplanetary disks and protoplanets yields important constraints on when and where planets form. Imaging will constrain the properties of protoplanetary disk material and provide insights into how protoplanetary disks form, evolve, and eventually dissipate. Many protoplanetary disks feature large scale structure: spiral arms, annular gaps, inner holes, and brightness asymmetries. Some of these features can be explained by the formation and growth of planets. The proposed H-alpha filter may have particular value, since H-alpha is a tracer of gas accretion. Detection of localized H-alpha emission could indicate the presence of an accreting Page 4
WFIRST CGI Direct Imaging Science protoplanet, directly probing the planet formation process and distinguishing protoplanets from other clumpy disk features (Sallum et al. 2015). Debris disks – the product of collisions between planetesimals – represent a later stage of planet formation. Many debris disks appear to be shaped into narrow rings, indicating belts of planetesimals shepherded by planets that have already formed. A growing number of observations for nearby debris disks show variability in scattered light. The nearby disk around the M0V star AU Mic shows features that move radially on yearly timescales (Boccaletti et al., 2015), and a CGI img pred. number of protoplanetary disks also show variable shadow features (Debes et al., 2017; Pinilla et al., 2015; Wolff et al., 2016). These variable disk features in particular provide Figure 5: Predicted CGI disk imaging performance in the context dynamical timescales for possible of HST detection limits and surface brightness profiles of previously-resolved disks and exozodi predictions. Surface hidden planets and open a new and brightness units are flux ratio per resolution element. exciting window into dust disk processes and planetary formation.
Community Data Challenges The goals of the Exoplanet Data Challenge (EDC) are to provide quantitative feedback to the specifications and requirements of the CGI and to leverage and develop expertise in the broader community. For example, previous cycles informed the choice of IFS resolution and bandpass. The EDC team produces and publicly releases simulated data; it then evaluates the efficacy of pipeline and instrument design through blind spectral retrieval studies. Community-driven, independent analysis is welcome in the second stage. Future cycles will incorporate the latest instrument parameters and will add dust emission (exo-zodi) and background objects in the images. The EDC builds valuable experience, ensuring that the community will be ready to interpret CGI data as soon as observations begin.
References Boccaletti, Aet al. (2015) Nature, 526, 230 Bowler, B. & Nielsen, E. (2018) in press. Burrows, A. et al. (2004) ApJ, 609, 407 Cahoy, K. et al. (2010) ApJ, 724, 189 Debes, J. H., et al. (2017) ApJ, 835, 205 Gao, P., et al. (2017) ApJ, 152, 217 Kaltenegger, L and Sasselov, D. (2011) ApJ 736, 2 Konopacky, Q. et al., (2013) Science 339, 1398 Lacy, B. et al. (2018) arXiv:1801.08964 Lupu, R. et al. (2016) ApJ, 152, 217
Marley, M. et al. (1999) ApJ, Nayak, M. et al. (2017) PASP 129, 034401 Perrin et al., (2015) ApJ, 799, 182 Pinilla, Pet al., (2015), A&A, 584, L4 Rajan, A. et al. (2017) Astron. J. 154, 10 Rodiagas et al. (2015) ApJ 798 96 Sallum et al. (2015) Nature 527, 342 Savransky, D., et al. (2016) Proc. SPIE 9904 Wolff, S. G., et al., (2016), ApJL, 818, L15 Zahnle, K., et al. (2016) ApJ, 824, 137
Page 5
A. Roberge*, D. Bennett, A. Bhattacharya, K. Carpenter, W. Danchi, T. Groff, J. Kruk, S. Malhotra, A. Mandell, M. McElwain, E. Quintana, C. Ranc, M. Rizzo, J. Schlieder, N. Zimmerman, NASA GSFC R. Akeson, S. Ramirez, T. Meshkat, G. Helou, R. Paladini, C.Gelino, M. Yagouf, S. Laine, D. Gelino, S. Calchi Novati, D. Ciardi, P. Lowrance Caltech/IPAC M. Penny Ohio State University
H. Jang-Condell* University of Wyoming
M. Meyer University of Michigan
D. Savransky* Cornell University
A. Boccaletti, P. Baudoz, T. Bhowmik, D. Rouan, T. Schmidt Paris Observatory
M. Turnbull SETI Institute S. Kane UC Riverside *Principle contributors.
T. Henning, A. Maire, H. Avenhaus MPIA
M. Tamura University of Tokyo/ABC/NAOJ T. Sumi Osaka University N. Murakami University of Hokaido J. Nishikawa NAOJ J. Kwon JAXA/ISAS M. Langlois, M. N’Diaye, A. Vigan, K. Dohlen CNRS G. Chauvin FCLA/IPAG S. Messina INAF-Catania J. Baudino Oxford/AOPP M. Janson Stockholm University
WFIRST CGI Direct Imaging Science
Introduction One of the two instruments on NASA’s Wide-Field Infrared Survey Telescope (WFIRST) is a coronagraph technology demonstrator. The Coronagraph Instrument (CGI) will demonstrate in space, for the first time, key technologies necessary for future exo-Earth imaging missions, including: high actuator count deformable mirrors; low-noise, single photon-counting detectors in the visible; new coronagraph masks and architectures; low-resolution integral field spectroscopy; advanced algorithms for wavefront sensing and control; high-fidelity integrated spacecraft and coronagraph modeling; and post-processing approaches to extract images and spectra. The importance of this technology demonstration to lowering the design and implementation risk of the next generation of large telescopes cannot be overstated. More details on these technologies and their specific value to future missions can be found in the white paper by Bailey et al. While CGI is nominally manifested as a technology demonstrator, NASA is committed to also making CGI available to the general astronomy community as a science instrument assuming a successful completion of its technology demonstration in the first 1.5 years of mission operations. In this white paper, we describe the potential high-contrast science that CGI may achieve and how this science will advance community goals in exoplanet astronomy and inform future observations. The scientific yield described here uses an integration time estimator that models the detector performance and incorporates planet photon noise, zodi and exo-zodi levels, and our current predictions of coronagraph performance, as described in the Bailey et al. whitepaper. With the project scheduled to enter Phase B in April 2018, further refinements to the integrated model are expected along with accompanying improvements in our best estimates of performance. These current predictions show that CGI will provide over two orders of magnitude improvement in detection and spectroscopic capability from what is currently possible on the ground (Figure 1). And while the sample of planets reachable by CGI is small, the potential science and the opportunity to refine our extraction approaches is manifest. In the remainder of this white paper we describe the following science potentially achievable by CGI: visible light spectroscopy and photometry of known young, selfluminous planets and known radial velocity planets; a blind search for undiscovered Jupiters and miniNeptunes; and visible light imaging of protoplanetary and debris disks. The observations described in this paper assume a set of four CGI filters: two 10% bandwidth filters centered at 575nm and 825nm, and Figure 1: Predicted CGI performance on a V=5 star, in two 18% bandwidth filters, centered the context of known planetary systems and current at 660nm and 760nm. The filter set instrumentation. See Bailey et al. whitepaper Appendix A for is sensitive to molecular bands of a further details on assumptions and data sources. variety of strengths as well as the Page 1
WFIRST CGI Direct Imaging Science continuum flux level, providing a range of information to place constraints on atmospheric structure and composition. Potential observing modes are described in the Bailey et al white paper.
Visible Light Spectra of Known Self-Luminous Giant Planets About two dozen young, low-mass companions to main sequence stars have been discovered by coronagraphic imaging methods (see review by Bowler & Nielsen 2018). These observations shed light on the formation of planets at orbital radii >10 au. To date, near- and mid-infrared spectroscopy and photometry have provided some constraints on the atmospheric abundance of C and O-bearing species and clouds on these planets, as well as on planet mass and temperature (e.g., Konopacky et al. (2013); Rajan et al. (2017)). Optical spectroscopy and photometry would help place constraints on the composition and atmospheric structure of these planets. The optical spectra are sculpted by strong, broad absorption features of gaseous Na and K (Figure 2). The abundance of these species provides additional constraints on metallicity, temperature, and gravity (and thus planet mass). The wavelength dependence of cloud and haze opacity is strongly sensitive to particle size (e.g., Zahnle et al. 2016). High altitude photochemical haze particles are typically small and thus invisible in the near-infrared, but they Figure 2: Planet to star contrast as a function of wavelength for can have an outsized impact at optical young, self-luminous giant planets with and without clouds. The wavelengths (e.g., Gao et al. 2017). effective temperatures of the star and planet are 6000K and 1100K. Estimated flux ratios for several young planets are shown in Figure 1. CGI should be able to obtain photometry of several of these objects with favorable flux ratios and spectroscopy of a few.
Characterization of Known RV Gas Giants in Reflected Visible Light Much of exoplanet science has focused on studying the atmospheres of transiting planets, particularly the hot Jupiters. While very interesting in their own right, these highly irradiated planets are not representative of the mature extrasolar giant planets found at larger orbital distances. Hundreds of such gas giant planets, more akin to our own Jupiter, have been discovered by radial velocity surveys, yet we know very little about their masses or atmospheric composition. CGI will, for the first time, allow further study and characterization of these planets. Several images spaced out over a planetary orbit will resolve the (sin i) degeneracy inherent in RV detections and reveal the planets’ true masses. CGI photometry in multiple passbands along with spectroscopy will characterize their atmospheres for the first time. Clouds, hazes, gravity, and atmospheric metallicity all play important roles in controlling giant planet spectra in reflected light. These effects are well understood from studies of solar system giants and two decades of theoretical modeling (e.g., Marley et al. 1999; Burrows et al. 2004; Cahoy et al. 2010) and thus we are well equipped to interpret CGI data on these planets. Retrieval studies on simulated WFIRST datasets (Lupu et al. 2016; Nayak et al. 2017; Lacy et al. 2018) Page 2
WFIRST CGI Direct Imaging Science demonstrate that even with broad band photometry, WFIRST users will be able to distinguish the presence or absence of cloud layers and atmospheric methane (Figure 3), providing clues into atmospheric temperature and composition. With R~50 spectra and photometry, limits can be placed on atmospheric abundance of key absorbers such as CH4 and H2O. While the WFIRST CGI sample will be small, it is crucial to test the end-to-end modeling and interpretation of planets detected in reflected light before much larger scale missions set out to find and characterize potentially habitable planets. Understanding the extent to which the contributions of clouds, hazes, and other effects can be discerned from reflected light data will give confidence that future missions studying more challenging targets will be properly prepared to succeed. A white paper by Marley & Lewis discusses this point in further detail.
Figure 3: The effects of various atmospheric parameters on the geometric albedo spectra of a gas giant planet. Clockwise from top left, the impact of varying atmospheric methane mixing ratio, surface gravity, cloud top pressure, and cloud single scattering albedo. Simultaneous detection of multiple methane features as well as the continuum flux level permits many of the apparent degeneracies to be resolved with sufficient quality data.
Blind Search Discovery of Gas Giants and Mini-Neptunes The WFIRST CGI will also be capable of searching for new planets as well as characterizing previously discovered ones. The performance modeling described above can also be used to study the overall sensitivity of CGI to various planet classes. CGI will be most sensitive to giant planets greater than 5 Earth radii between 1 and 10 au, detecting these at a rate of greater than 50% in systems where they are present, but will also have some sensitivity down to 2 Earth radii, particularly between 0.5 and 2 au. If a population of high albedo (p > 0.7) 1.5-2 Earth radii planets exists Page 3
WFIRST CGI Direct Imaging Science around and within 1 au, as described by Kaltenegger and Sasselov (2011), then CGI could potentially detect one or more habitable zone mini-Neptunes/Super-Earths. Otherwise, it is likely that all new CGI discoveries will be gas and ice giants within 10 au of their host star. While the total number of new planets detected by CGI would likely be small (on the order of 10, depending on assumptions made – see Figure 4 and Savransky et al., 2016), carrying out an exoplanet imaging survey from space would be hugely valuable to the next generation of exoplanets missions. A blind search for new planets would include prioritizing targets under the various constraints of the real mission and scheduling revisits to confirm planet candidates. Lessons learned from the CGI survey would be directly applicable to all future exoplanets imagers and may even influence their designs and mission formulations. Figure 4: Depth of search (summed completeness assuming one planet per star at each location in the semi-major axis-Radius phase space) for the WFIRST CGI assuming optimal utilization of 3 months of integration time (96 total targets). Giant planets (>5 Earth radii) are found most frequently (>50%) while the CGI is barely sensitive to 2 Earth-radii planets (0.5%). Albedos were assumed to be uniformly 0.2 for planets below 1.4 Earth radii and 0.5 for larger planets.
Visible Light Imaging of Protoplanetary and Debris Disks and Protoplanets In addition to the potential exoplanet science, CGI will be capable of imaging protoplanetary and debris disks. Studying these disks is key to advancing our understanding of the formation and evolution of planetary systems. A combination of photometry, polarization properties, and scattered light brightness as a function of scattering angle can provide constraints on dust grain shape, composition, and grain size distribution (Perrin et al., 2015 and Rodiagas et al., 2015). Imaging of known circumstellar disks represents a guaranteed science return from WFIRST. Figure 5 shows the predicted performance of the CGI imaging mode (for a V=5 star), compared to a selection of protoplanetary and debris disks previously observed in visible scattered light. CGI has the potential to characterize known disks in at least two photometric bands, in both total intensity and polarized intensity with unprecedented fidelity, and it has the ability to detect disks that are orders of magnitude fainter than previously observed. Imaging of protoplanetary disks and protoplanets yields important constraints on when and where planets form. Imaging will constrain the properties of protoplanetary disk material and provide insights into how protoplanetary disks form, evolve, and eventually dissipate. Many protoplanetary disks feature large scale structure: spiral arms, annular gaps, inner holes, and brightness asymmetries. Some of these features can be explained by the formation and growth of planets. The proposed H-alpha filter may have particular value, since H-alpha is a tracer of gas accretion. Detection of localized H-alpha emission could indicate the presence of an accreting Page 4
WFIRST CGI Direct Imaging Science protoplanet, directly probing the planet formation process and distinguishing protoplanets from other clumpy disk features (Sallum et al. 2015). Debris disks – the product of collisions between planetesimals – represent a later stage of planet formation. Many debris disks appear to be shaped into narrow rings, indicating belts of planetesimals shepherded by planets that have already formed. A growing number of observations for nearby debris disks show variability in scattered light. The nearby disk around the M0V star AU Mic shows features that move radially on yearly timescales (Boccaletti et al., 2015), and a CGI img pred. number of protoplanetary disks also show variable shadow features (Debes et al., 2017; Pinilla et al., 2015; Wolff et al., 2016). These variable disk features in particular provide Figure 5: Predicted CGI disk imaging performance in the context dynamical timescales for possible of HST detection limits and surface brightness profiles of previously-resolved disks and exozodi predictions. Surface hidden planets and open a new and brightness units are flux ratio per resolution element. exciting window into dust disk processes and planetary formation.
Community Data Challenges The goals of the Exoplanet Data Challenge (EDC) are to provide quantitative feedback to the specifications and requirements of the CGI and to leverage and develop expertise in the broader community. For example, previous cycles informed the choice of IFS resolution and bandpass. The EDC team produces and publicly releases simulated data; it then evaluates the efficacy of pipeline and instrument design through blind spectral retrieval studies. Community-driven, independent analysis is welcome in the second stage. Future cycles will incorporate the latest instrument parameters and will add dust emission (exo-zodi) and background objects in the images. The EDC builds valuable experience, ensuring that the community will be ready to interpret CGI data as soon as observations begin.
References Boccaletti, Aet al. (2015) Nature, 526, 230 Bowler, B. & Nielsen, E. (2018) in press. Burrows, A. et al. (2004) ApJ, 609, 407 Cahoy, K. et al. (2010) ApJ, 724, 189 Debes, J. H., et al. (2017) ApJ, 835, 205 Gao, P., et al. (2017) ApJ, 152, 217 Kaltenegger, L and Sasselov, D. (2011) ApJ 736, 2 Konopacky, Q. et al., (2013) Science 339, 1398 Lacy, B. et al. (2018) arXiv:1801.08964 Lupu, R. et al. (2016) ApJ, 152, 217
Marley, M. et al. (1999) ApJ, Nayak, M. et al. (2017) PASP 129, 034401 Perrin et al., (2015) ApJ, 799, 182 Pinilla, Pet al., (2015), A&A, 584, L4 Rajan, A. et al. (2017) Astron. J. 154, 10 Rodiagas et al. (2015) ApJ 798 96 Sallum et al. (2015) Nature 527, 342 Savransky, D., et al. (2016) Proc. SPIE 9904 Wolff, S. G., et al., (2016), ApJL, 818, L15 Zahnle, K., et al. (2016) ApJ, 824, 137
Page 5
