No Clear, Direct Evidence for Multiple Protoplanets Orbiting LkCa 15: LkCa 15 bcd are Likely Inner Disk Signals
Thayne Currie, Christian Marois, Lucas Cieza, Gijs Mulders, Kellen Lawson, Claudio Caceres, Dary Rodriguez-Ruiz, John Wisniewski, Olivier Guyon, Timothy Brandt, N. Jeremy Kasdin, Tyler Groff, Julien Lozi, Jeffrey Chilcote, Klaus Hodapp, Nemanja Jovanovic, Frantz Martinache, Nour Skaf, Wladimir Lyra, Motohide Tamura, Ruben Asensio-Torres, Ruobing Dong, Carol Grady, Misato Fukagawa, Derek Hand, Masahiko Hayashi, Thomas Henning, Tomoyuki Kudo, Masayuki Kuzuhara, Jungmi Kwon, Michael McElwain, Taichi Uyama
DDraft version May 14, 2019
Typeset using L A TEX twocolumn style in AASTeX61
NO CLEAR, DIRECT EVIDENCE FOR MULTIPLE PROTOPLANETS ORBITING LKCA 15:LKCA 15 bcd ARE LIKELY INNER DISK SIGNALS
Thayne Currie,
1, 2, 3
Christian Marois,
4, 5
Lucas Cieza, Gijs D. Mulders, Kellen Lawson, Claudio Caceres,
9, 10
Dary Rodriguez-Ruiz, John Wisniewski, Olivier Guyon,
2, 12, 13, 14
Timothy D. Brandt, N. Jeremy Kasdin, Tyler D. Groff, Julien Lozi, Jeffrey Chilcote, Klaus Hodapp, Nemanja Jovanovic, Frantz Martinache, Nour Skaf,
2, 22
Wladimir Lyra,
23, 24
Motohide Tamura,
14, 25, 26
Ruben Asensio-Torres, Ruobing Dong, Carol Grady,
3, 17
Misato Fukagawa, Derek Hand, Masahiko Hayashi, Thomas Henning, Tomoyuki Kudo, Masayuki Kuzuhara, Jungmi Kwon, Michael W. McElwain, and Taichi Uyama NASA-Ames Research Center, Moffett Blvd., Moffett Field, CA, USA Subaru Telescope, National Astronomical Observatory of Japan, 650 North A‘oh ¯ o k ¯ u Place, Hilo, HI 96720, USA Eureka Scientific, 2452 Delmer Street Suite 100, Oakland, CA, USA National Research Council of Canada Herzberg, 5071 West Saanich Rd, Victoria, BC, V9E 2E7, Canada University of Victoria, 3800 Finnerty Rd, Victoria, BC, V8P 5C2, Canada N´ucleo de Astronom´ıa, Facultad de Ingenier´ıa y Ciencias, Universidad Diego Portales, Av Ej´ercito 441, Santiago, Chile ”Department of the Geophysical Sciences, The University of Chicago, Chicago, IL 60637, USA Homer L. Dodge Department of Physics, University of Oklahoma, Norman, OK 73071, USA Departamento de Ciencias Fisicas, Facultad de Ciencias Exactas, Universidad Andres Bello. Av. Fernandez Concha 700, Las Condes,Santiago, Chile N´ucleo Milenio Formaci´on Planetaria - NPF, Universidad de Valpara´ıso, Av. Gran Breta˜na 1111, Valpara´ıso, Chile Department of Physics and Astronomy, Rochester Institute of Technology, Rochester, NY, USA Steward Observatory, University of Arizona, Tucson, AZ 85721, USA College of Optical Sciences, University of Arizona, Tucson, AZ 85721, USA Astrobiology Center of NINS, 2-21-1, Osawa, Mitaka, Tokyo, 181-8588, Japan Department of Physics, University of California, Santa Barbara, Santa Barbara, California, USA Department of Mechanical Engineering, Princeton University, Princeton, NJ, USA NASA-Goddard Space Flight Center, Greenbelt, MD, USA Department of Physics, University of Notre Dame, South Bend, IN, USA Institute for Astronomy, University of Hawaii, 640 North A‘oh ¯ o k ¯ u Place, Hilo, HI 96720, USA Department of Astronomy, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA 91125 Universit´e Cˆote d’Azur, Observatoire de la Cˆote d’Azur, CNRS, Laboratoire Lagrange, France Imperial College London, Kensington, London SW7 2AZ, UK Department of Physics and Astronomy, California State University Northridge, 18111 Nordhoff Street, Northridge CA 91130, USA Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA Department of Astronomy, Graduate School of Science, The University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan National Astronomical Observatory of Japan, 2-21-2, Osawa, Mitaka, Tokyo 181-8588, Japan Department of Astronomy, Stockholm University, AlbaNova University Center, SE-106 91 Stockholm, Sweden Max Planck Institut fur Astronomie, Konigstuhl 17, 69117 Heidelberg, Germany ISAS/JAXA, 3-1-1 Yoshinodai, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan
ABSTRACTTwo studies utilizing sparse aperture-masking (SAM) interferometry and H α differential imaging have reportedmultiple jovian companions around the young solar-mass star, LkCa 15 (LkCa 15 bcd): the first claimed directdetection of infant, newly formed planets (“protoplanets”). We present new near-infrared direct imaging/spectroscopyfrom the Subaru Coronagraphic Extreme Adaptive Optics (SCExAO) system coupled with the Coronagraphic High Corresponding author: Thayne [email protected],[email protected] a r X i v : . [ a s t r o - ph . E P ] M a y Angular Resolution Imaging Spectrograph (CHARIS) integral field spectrograph and multi-epoch thermal infraredimaging from Keck/NIRC2 of LkCa 15 at high Strehl ratios. These data provide the first direct imaging look at thesame wavelengths and in the same locations where previous studies identified the LkCa 15 protoplanets, and thusoffer the first decisive test of their existence.The data do not reveal these planets. Instead, we resolve extended emission tracing a dust disk with a brightness andlocation comparable to that claimed for LkCa 15 bcd. Forward-models attributing this signal to orbiting planets areinconsistent with the combined SCExAO/CHARIS and Keck/NIRC2 data. An inner disk provides a more compellingexplanation for the SAM detections and perhaps also the claimed H α detection of LkCa 15 b.We conclude that there is currently no clear, direct evidence for multiple protoplanets orbiting LkCa 15, al-though the system likely contains at least one unseen jovian companion. To identify jovian companions aroundLkCa 15 from future observations, the inner disk should be detected and its effect modeled, removed, and shown tobe distinguishable from planets. Protoplanet candidates identified from similar systems should likewise be clearlydistinguished from disk emission through modeling. Keywords: planetary systems, stars: T Tauri, stars: individual: LkCa 15 INTRODUCTIONYoung, 1–10
M yr -old jovian protoplanets embeddedin disks around newly born stars provide a crucial linkbetween the first stages of planet formation and theproperties of directly imaged, fully formed planets or-biting 10–100
M yr old stars (e.g. Marois et al. 2008b,2010a). LkCa 15, a solar-mass T Tauri star and mem-ber of the 1–3
M yr old Taurus–Auriga star-forming re-gion (Kenyon et al. 2008), is a superb laboratory forstudying planet formation and searching for protoplan-ets. The star is surrounded by an accreting, gas-richprotoplanetary disk with multiple dust components: hot( T eff = 1400 K ), sub-au scale dust producing broadbandnear-infrared (NIR) excess and cooler massive outerdust, which are separated by a solar system-scale cav-ity plausibly created by jovian protoplanets (Espaillatet al. 2007; Thalmann et al. 2010; Andrews et al. 2011;Dodson-Robinson and Salyk 2011; Dong and Fung 2017;Alencar et al. 2018).Using sparse aperture masking interferometry (SAM;Tuthill et al. 2006) of LkCa 15, Kraus & Ireland (2012)reported the detection of one protoplanet located withinan ostensibly cleared gap in dust emission. Also usingSAM, Sallum et al. (2015b) then identified three pro-toplanets within ρ ∼ . (cid:48)(cid:48)
15 ( ≈
25 au) (LkCa 15 bcd),one of which was recovered in H α (LkCa 15 b). Thus,LkCa 15 appeared to show evidence for multiple jovianprotoplanets: the first such system ever reported.However, the closure phase signals of disks in SAMdata can mimic those of protoplanets (Cieza et al. 2013;Kraus et al. 2013). LkCa 15’s circumstellar environmentas seen in scattered light is complex, including a brightouter dust wall (Thalmann et al. 2010, 2014). Addition-ally, inner dust disk material is now resolved at opticalwavelengths and NIR polarimetry out to LkCa 15 bcd-like separations (Oh et al. 2016a; Thalmann et al. 2016).Depending on this dust disk’s brightness and spatial ex-tent in (a) total intensity at (b) the longer wavelengthswhere LkCa 15 bcd were identified (2.2–3.8 µm ), it couldinstead be the signal masquerading as these protoplan-ets. However, previous 2.2–3.8 µm total intensity datalack the image quality/sensitivity to probe these regions(Thalmann et al. 2014).In this Letter, we use multi-epoch direct imaging ob-servations of LkCa 15 obtained from the Subaru Coron-agraphic Extreme Adaptive Optics (SCExAO) projectcoupled with the Coronagraphic High Angular Reso-lution Imaging Spectrograph (CHARIS) in the near-infrared ( JHK /1.1–2.4 µm ; Groff et al. 2015; Groff etal. 2017; Jovanovic et al. 2015a) and Keck/NIRC2 inthe thermal infrared ( L p /3.78 µm ). These data providethe first direct imaging look at the same wavelengths and in the same locations where previous studies iden-tified the LkCa 15 protoplanets ( K , L p ) and thus offerthe first decisive test of their existence. OBSERVATIONS AND DATA REDUCTION2.1.
SCExAO/CHARIS JHK DirectImaging/Spectroscopy
We observed LkCa 15 on UT 2017 September 07and UT 2018 January 8 using SCExAO coupled withCHARIS operating in low-resolution ( R ∼ JHK filters simultaneously( t int = 31 and 19 minutes). All data were acquired inangular differential imaging mode (ADI; Marois et al.2006). For the September data, given our modest par-allactic angle rotation (∆PA = 60 o ), we also observeda nearby, near-color matched star (V819 Tau) as a con-temporaneous point-spread function (PSF) reference .Conditions were excellent (0 . (cid:48)(cid:48) . (cid:48)(cid:48) V -band seeing). De-spite LkCa 15’s and V819 Tau’s optical faintness ( R ∼ ≈
70% Strehl in H band. For the January data, condi-tions were poorer and we did not observe a PSF referencestar, but the parallactic angle motion was larger (120 o ).Spectral extraction utilized the cube rectificationpipeline from Brandt et al. (2017) and basic image pro-cessing was performed as in Currie et al. (2018a,b). Amodel spectral energy distribution (SED) reproducingLkCa 15’s broadband photometry provided spectropho-tometric calibration . No coronagraphs or satellite spotswere used; all stellar PSFs were unsaturated.2.2. Keck/NIRC2 L p Direct Imaging
First, we reduced multiple LkCa 15 high-contrastimaging data sets from the Keck Observatory Archivewith more than 2 λ / D parallactic angle rotation at LkCa15 bcd’s reported angular separation, selecting 2009 V819 Tau has a marginal unresolved infrared (IR) excess long-wards of 10–15 µm (Furlan et al. 2009). However, we findno hint of a disk in SCExAO/CHARIS data nor in a separateKeck/NIRC2 L p data set. Subaru/HiCIAO H -band polarime-try data show that V819 Tau is a non-detection for any disk (J.Hashimoto, pvt. comm). For our purposes, V819 Tau is effectivelya bare stellar photosphere. LkCa 15 exhibits small-amplitude variability at optical andmid-IR (MIR) wavelengths (Espaillat et al. 2011; Rodriguez etal. 2017), with a peak-to-peak value of ∼ JHK bands, let alone at a level that could affect ourconclusions.
November 21 L p data (PI: L. Hillenbrand; ∆PA =132 o .5, t int = 5.4 minutes). These data have the high-est quality of those taken without a coronagraph thatmay partially occult LkCa 15 bcd and are contempora-neous with the first aperture-masking detection reportedin Kraus & Ireland (2012). Second, we obtained NIRC2data on 2017 December 9 and 10 for 17.6 and 13.8 min-utes with 150 o and 160 o parallactic motion. LkCa 15was observed continuously through transit on the firstnight; on the second night, we alternated between itand a diskless PSF reference star (V1075 Tau). All datawere acquired in ADI mode using the narrow camerawith various dither patterns.Keck/NIRC2’s adaptive optics (AO) system deliveredmedian Strehl ratios of 0.79 and 0.77–0.79 in L p for the2009 November and two 2017 December data sets, asmeasured by a modified (for the appropriate pixel scale)observatory-supplied routine nirc2strehl.pro. Stars wereunsaturated in all images. Basic processing followed pre-vious steps used for thermal-IR data with our well-testedbroadband imaging pipeline (Currie et al. 2011, 2014b),including a linearity correction, sky subtraction, distor-tion correction and bad pixel interpolation, image regis-tration, and flux normalization.2.3. PSF Subtraction
Because of the complex astrophysical scene within ρ ∼ . (cid:48)(cid:48) Lo-cally Optimized Combination of Images (LOCI) and
Karhunen-Lo´eve Image Projection (KLIP) algorithmsplus successors (Currie et al. 2012; Lafreni`ere et al. 2007;Marois et al. 2010b, 2014; Soummer et al. 2012). Addi-tionally, at very small angles, morphological biasing ofan astrophysical source in ADI due to self-subtractioncan be severe.Therefore, we adopted the following approach. First,for data sets obtained with a suitable PSF referencestar, we performed reference star differential imaging(RDI) using KLIP and the Adaptive Locally OptimizedCombination of Images algorithm (A-LOCI; Currie et al.2012), where we equate the region used to construct aweighted reference PSF (the optimization zone) and theregion over which this PSF is subtracted (the subtrac-tion zone) with the outer radius set to the visible PSFhalo, beyond the angles covered by LkCa 15’s disk struc- tures ( ρ ≈ . (cid:48)(cid:48) . (cid:48)(cid:48) δ ≈ λ / D , while truncating the covariance matrix’s diagonalterms with singular value decomposition. To better sup-press residual speckles with the 2018 January CHARISdata, we performed a classical SDI reduction (median-combination of channels rescaled by wavelength) on theADI/A-LOCI residuals . DETECTION OF THE LKCA 15 INNER DUSTDISK AND NON-DETECTION OF LKCA 15 bcdFigure 1 shows the SCExAO/CHARIS near-IR imagesin broadband (a median-combination of all channels)and in K band (top panels) and Keck/NIRC2 L p images(bottom panels). All data easily resolve the forward-scattering side of the crescent-shaped outer dust diskwall (e.g. Thalmann et al. 2010, 2014). However, no dataset reveals direct evidence for LkCa 15 bcd. Instead, alldata resolve another crescent-shaped extended structureinterior to the outer disk wall, consistent with the wallof an inner dust disk previously only seen in polarizedlight (Thalmann et al. 2015; Oh et al. 2016a).Inspection of individual CHARIS data cubes andNIRC2 images shows that this extended inner disk emis-sion cannot be explained by residual speckle noise thatis preserved when images are derotated and combined(for CHARIS and NIRC2) or wavelength-collapsed (forCHARIS). RDI-reduced images obtained using a rangeof principal components (for KLIP) or a range of SVDcutoffs (for A-LOCI) all recover the same structure.For CHARIS, the inner disk is visible in most indi-vidual channels, especially those covering the H and K passbands. Furthermore, ADI and ASDI-reducedimages (January 2018 CHARIS data and two of thethree Keck/NIRC2 data sets) also show negative self-subtraction footprints of this inner disk .We further confirmed that we could have detectedLkCa 15 bcd-like planets in absence of disk emission. To CHARIS’s large bandpass enables SDI while only partiallyannealing point sources at LkCa 15 bcd-like separations. A separate ASDI reduction of the 2017 September CHARISdata and reduction of other data sets not considered here – anADI reduction of archival 2016 October K s SCExAO/HiCIAOdata, and ADI reductions of additional archival Keck/NIRC2 M p and L p data from 2012 and 2015 – likewise show a detection ofthe inner disk, not planets, albeit with more residual speckle con-tamination and/or poorer sensitivity. Figure 1.
LkCa 15 images from SCExAO/CHARIS (top panels) and Keck/NIRC2 (bottom panels). The data are processedusing different combinations of ADI, SDI, and RDI using the A-LOCI and KLIP algorithms. All images reveal spatially-extendedemission consistent with disk emission, not planets. A white dashed circle shows a radius of 0 . (cid:48)(cid:48)
06. LkCa 15 bcd’s positions(circled) from Sallum et al. (2015b) trace the edge of the inner disk. The vertical bars show the intensity scale in units of mJynormalized to one FWHM. empirically assess our sensitivity to point sources, we in-jected and attempted to recover model planets with anearly L dwarf-like spectrum into our raw LkCa 15 datareduced with RDI (September 2017 CHARIS data andDecember 2017 NIRC2 data). We considered the half-field of view opposite the peak brightness of the innerdisk and at a range of angular separations . We variedthe brightnesses of these planets with respect to the starto be equal to or fainter than that for LkCa 15 bc at K and L p as reported by Sallum et al. (2015b): ∆ K ∼ L p ∼ Typically, contrast curves are derived numerically based onthe radial noise profile (e.g. Marois et al. 2008a; Currie et al. 2011).However, at small angles relevant for this study, corrections to thenominal 3–5 σ limits due to finite sample sizes (Mawet et al.2014) are significant. In particular, the contrast penalty to achievea Gaussian noise-equivalent 5 σ limit at 1–2 λ /D with a False Pos-itive Fraction (FPF) of ∼ × − is prohibitively large for ahalf-field of view (see Figure 6 in Mawet et al. 2014). Settingthe FPF to 1.35 × − as recommended by Mawet et al. (2014)for the smallest angles, equivalent to the FPF for a 3 σ detectionin Gaussian statistics, shows that planets with brightnesses com-parable LkCa 15 bcd would in fact be recovered at the > σ leveldespite residual disk emission. While residual disk emission atsmall angles causes the true noise to be overestimated, a substan-tial positive skew in the noise profile (which itself is uncertain dueto finite sample sizes) can cause the FPF to be underestimated(Marois et al. 2008a; Currie et al. 2014a). For all these compli-cations, we opt for a more direct, empirical approach of injectingand recovering planets with known contrasts. inner disk emission, planets with separations compa-rable to LkCa 15 bcd ( ρ ∼ . (cid:48)(cid:48) . (cid:48)(cid:48)
1) are detected atLkCa 15 bcd-like contrasts (∆ K ∼ L p ∼
5, 5.9) and visible as point sources. The contrasts ofthese recovered planets are similar to limits achievedfor diskless stars with SAM in Kraus et al. (2011) andLacour et al. (2011). Planets even fainter than proposedfor LkCa 15 bcd – ∆ K ∼ L p ∼ ρ ∼ . (cid:48)(cid:48) . (cid:48)(cid:48) ρ (cid:38) . (cid:48)(cid:48) K , ∆ L p ∼ .Comparisons between our images and SAM resultsstrongly suggest that this inner disk emission is the sameastrophysical source previously interpreted as the LkCa15 bcd protoplanets. For both CHARIS and NIRC2data, the inner disk emission extends from ρ ∼ . (cid:48)(cid:48)
07 to ρ ∼ . (cid:48)(cid:48)
25 ( r proj ≈ ρ ∼ . (cid:48)(cid:48) . (cid:48)(cid:48)
1, respectively ( r proj ≈ ∼ o , which is roughly the same position angle range forLkCa 15 bcd reported in Sallum et al. (2015b). Planet For the 2009 NIRC2 L p data reduced with ADI/A-LOCI, theforward-scattering peak of the inner disk severely self-subtractspoint sources injected into the data at ρ ∼ . (cid:48)(cid:48)
1: thus, injectingplanets into these data as performed for our RDI-reduced datasets substantially underestimates our true sensitivity in absenceof a disk. Nevertheless, planets with LkCa 15 bc-like contrastsare still detectable at LkCa 15 bcd-like separations as well.
Figure 2.
September 2017 SCExAO/CHARIS (left) and December 2017 Keck/NIRC2 (right) images with planets (circled)injected into the raw data prior to PSF subtraction with RDI/KLIP and RDI/A-LOCI. In regions lacking bright extendedemission, planets with positions and brightnesses comparable to that reported for LkCa 15 bcd are easily recovered. Note alsothat the injected planets are comparable in brightness to the extended emission consistent with an inner disk. At slightly widerseparations ( ρ ∼ . (cid:48)(cid:48) positions reported in Sallum et al. (2015b) (circles in the2018 January CHARIS data) trace this emission. Theaggregate flux density for LkCa 15 bcd from Sallum etal. (2015b) is ≈ ± ± K and L p , respectively. Over the same range of position an-gles/separations reported for LkCa 15 bcd, the summedinner disk flux densities in the CHARIS K band andNIRC2 L p data reduced using RDI are the same, withinuncertainties : ≈ ≈ JHK , λ o = 1.63 µm ), the peak brightness of the inner component isabout 30% higher than the peak of the outer compo-nent. At K -band ( λ o = 2.18 µm ), the peak brightnessof the inner disk is about 1.75 times than the outer disk,while at L p the inner disk is more than twice as bright asthe outer disk. The physical origin of these differenceswill be addressed in § FORWARD-MODELING OF LKCA 15 IMAGES:A FORWARD-SCATTERING INNER DUSTDISK, NOT MULTIPLE ORBITING PLANETSWe now compare the LkCa 15 images to forward-models (Marois et al. 2010b) for LkCa 15 bcd and aninner disk. Our analysis adopts the approaches in Pueyo(2016) and Currie et al. (2018b) for KLIP and A-LOCI,using the eigenvalues/eigenvectors in KLIP or coeffi- As we found in the immediate preceding analysis, RDI pro-cessing induces only modest signal loss for point sources and disksat LkCa 15 bcd-like separations: the throughput-corrected fluxdensity for the inner disk still matches that reported for LkCa 15bcd combined together. cients in A-LOCI drawn from the real data and apply-ing them to synthetic planet/disk signals injected intoempty data cubes/images. Our goal is to (1) confirmthat the emission we interpret as an inner disk cannot bereproduced by properties previously attributed to LkCa15 bcd and (2) then explore the general properties ofthis inner disk.We focused on the highest-quality data easily amenableto forward-modeling at wavelengths where LkCa 15 bcdwere identified ( K , L p ). Thus, we considered the K -band portion of the 2017 September SCExAO/CHARISdata processed with RDI/KLIP, the 2009 NovemberNIRC2 L p data processed with ADI/A-LOCI, and the2017 December NIRC2 L p data processed with RDI/A-LOCI. 4.1. Planet Forward-Modeling
We produced forward-models of (a) all three planets(LkCa 15 bcd) and (b) just the two identified in Sal-lum et al. 2015b from multiple epochs (LkCa 15 bc),(1) at the planets’ last reported positions in Sallum etal. (2015b) in November 2014-February 2015, and (2)at the planets’ estimated positions in 2009 November,2017 September, and 2017 December. To predict theplanets’ positions in multiple epochs, we adopted theSallum et al. astrometry and the
Gaia second data re-lease (DR2) distance to LkCa 15 (158.9 pc ), assumingthat the planets are on circular orbits in the same planeas the outer disk ( i ∼ o , P A minor ∼ o ; Thalmannet al. 2014, 2015; Oh et al. 2016a). Their deprojectedorbital separations in 2014 November-2015 February are ∼ ≈ o yr − in the orbital plane.We adopted the Sallum et al. (2015b) L p photometryfor LkCa 15 bcd. In K , we also adopted their LkCa 15 Figure 3.
Comparisons between our observed data (left panels) and forward-models of LkCa 15bcd (middle/right panels) forthe September 2017 SCExAO/CHARIS data (top) and November 2009 and December 2017 Keck/NIRC2 data (bottom). Thepredicted positions for LkCa 15 bcd in November 2009 and September/December 2017 are ρ ∼ . (cid:48)(cid:48) o and ρ ∼ . (cid:48)(cid:48) o for LkCa 15 b; ρ ∼ . (cid:48)(cid:48) o and ρ ∼ . (cid:48)(cid:48) o for LkCa 15 c; and ρ ∼ . (cid:48)(cid:48)
1, PA=39 o and ρ ∼ . (cid:48)(cid:48) o for LkCa 15 d. bc photometry. LkCa 15 d has no claimed detection in K from Sallum et al. (2015b). We assumed that LkCa15 d’s K - L p colors are similar to LkCa 15 bc’s and thusadopted ∆ K = 7.Figure 3 shows forward-models of the LkCa 15 planetsfor CHARIS K -band (top panels) and NIRC2 L p (bot-tom panels). The emission’s apparent brightness in theCHARIS data is comparable to the combined brightnessproposed for LkCa 15 bcd. However, LkCa 15 bcd wouldbe clearly distinguishable as separate point sources in K , whereas the CHARIS data instead show a contin-uous structure. Thus, the SCExAO/CHARIS data areinconsistent with planets being responsible for this emis-sion.At L p , LkCa 15 bcd’s PSFs are partially blended However, due to orbital motion, the aggregate emissionfrom LkCa 15 bc(d) should rotate clockwise by ∼ o between 2009 and 2017: the emission centroid, mea-sured in the forward-modeled planet images from regionswithin 50% of the peak intensity, changes by ≈ λ / D .In contrast, the measured center of mass for this emis-sion in the real λ / D , implying a static morphology over 8 yr. On the other hand, a forward-model including only LkCa15 bc, resembling the reconstructed images from 2009 NovemberSAM data (Kraus & Ireland 2012; Sallum et al. 2016), is mor-phologically inconsistent with our real 2009 November data, asit would reveal the planets as separate point sources. The SAMimage reconstructions in some cases are therefore not faithfullyreproducing the spatial distribution of astrophysical signals nearLkCa 15.
Thus, the Keck/NIRC2 data are inconsistent with plan-etary orbital motion.4.2.
Disk Forward-Modeling
To explore the general properties of inner disk emis-sion previously attributed to LkCa 15 bcd, we pro-duced and then forward-modeled synthetic scattered-light disk images with SCExAO/CHARIS using the MC-Max3D radiative transfer code (Min et al. 2009), adopt-ing the formalism from Mulders et al. (2010, 2013). Ourapproach considered three spatially extended compo-nents: (1) an optically thick (sub-)au scale hot com-ponent responsible for the NIR broadband excess and10 µm silicate feature, (2) a warm component responsi-ble for the inner disk resolved with SCExAO/CHARISand Keck/NIRC2, and (3) the optically-thick outerdisk, which has been resolved in optical/NIR scatteredlight (e.g. Thalmann et al. 2014, 2016) and with (sub-)millimeter data (Andrews et al. 2011; Isella et al. 2014).Following Thalmann et al. (2014, 2016), we envisionedthat components 1 and 2 shadow and may be slightlymisaligned with the outer disk (component 3). We ex-plored a small range of component parameters, settlingon a fiducial model with properties listed in Table 1.Except for a few Spitzer /IRS channels probing the un-resolved sub-au component, the model fits LkCa 15’sentire SED from the optical to millimeter to within ∼ L p image with our fidu- Table 1.
Disk Model ParametersParameter ValueGlobal ParametersDistance 158.9 pcT eff L (cid:63) L (cid:12) R (cid:63) R (cid:12) M (cid:63) M (cid:12) A V θ ) 60 o Dust Size Power Law, p a i ) 50 o o o Inner radius, R in (au) 0.12 20 55Outer radius, R out (au) 3 40 160Disk wall radius, R w (au) 0.12 25 82.5Wall shape ( w ) flat/vertical rounded/0.3 rounded/0.25 M dust ( M (cid:12) ) 5 × − × − × − Radial surface density power law ( (cid:15) ) 1 0.5 1Minimum dust size ( a min , µm ) 0.1 0.6 0.1Maximum dust size ( a max , µm ) 0.25 1000 1000Scale height at inner radius, H o , in p gas Note —The disk component surface density follows Σ (
R < R w ) ∝ R − (cid:15) × exp(-( − R/R exp w ) ) and Σ ( R ≥ R w ) ∝ R − (cid:15) . The wall shape parameter defines thespatial scale over which the disk surface density increases from R in to R w . See Mulderset al. (2010, 2013) and Thalmann et al. (2014) for detailed explanations of MCMax3Dterminology. cial model. The PSF-subtracted model reproduces thebrightness and morphology of the inner/outer disk com-ponents: the subtraction residuals do not reveal anyemission consistent with LkCa 15 bcd. The peak pixelintensity at positions covering LkCa 15 bcd (circled)is always less than 1/3 (1/4) that predicted for LkCa15 b(c). Residuals at ρ (cid:46) . (cid:48)(cid:48) . For example, weak residuals for NIRC2 just exterior to LkCa15 c’s predicted position correspond to the forward-scattering peakof the outer disk, not the inner disk. Modified models may bettermatch the combined LkCa 15 data: e.g. faint negative (positive)
While a wide range of models match either theSCExAO/CHARIS or Keck/NIRC2 data, the combineddata point toward different grain properties for the threedisk components. A larger minimum dust grain size forthe resolved inner disk vs. resolved outer disk ( ∼ µm vs. 0.1 µm ) better reproduces the inner disk’s red-der color and more pronounced forward-scattering peak.While unresolved, the sub-au disk component requiressubmicron-sized grains to reproduce the 10 µm silicatefeature (see also Espaillat et al. 2007). A future paper residuals on the east (west) for the inner/outer disk may be elim-inated by introducing pericenter offsets (Thalmann et al. 2016). Figure 4.
Comparing the LkCa 15 SCExAO/CHARIS broadband image (top) and Keck/NIRC2 L p image (bottom) to syntheticdisk models. The left panels show the real data. The middle panels show the forward-modeled image; the right panel showsthe residual image (real data minus model). The residual image reveals no evidence for embedded planets at LkCa 15 bcd’slocations (circles). The model is produced as-is, not re-scaled in flux to minimize residuals in any dataset. will thoroughly analyze LkCa 15’s disk structures andderive best-fit parameters.Our modeling also (a) implies that LkCa 15’s diskstructures should be detectable in optical total inten-sity imaging and (b) is consistent with the millimeterdetection of the outer disk and non-detection of the in-ner disk. At 0.65 µm , the inner disk’s continuum signalcompared to the star (convolved with a gaussian and in-tegrated within 1.5–2 FWHM) near the reported LkCa15 b position in H α is just slightly lower than LkCa 15b’s reported H α contrast (∆F ∼ × − ), as is theforward-scattering peak of the outer disk ( ρ ∼ . (cid:48)(cid:48) ρ ∼ . (cid:48)(cid:48)
08) is comparable in contrast to LkCa15 b ((5–8.5) × − ). At 7 mm, the model reproducesthe outer disk edge’s typical intensity, with a character-istic brightness of ∼ µ Jy beam − for a beam size of0 . (cid:48)(cid:48)
15; for a 0 . (cid:48)(cid:48)
07 beam, it accurately predicts that theinner disk (0.5 µ Jy beam − ) would be undetected givena 1 σ noise floor of 3.6 µ Jy beam − . DISCUSSIONInstead of protoplanets, our direct images of LkCa 15obtained with SCExAO/CHARIS show extended, un-resolved inner disk emission. Forward-modeling showsthat the SCExAO data were capable of distinguishingbetween disk emission and point sources with K bandphotometry and astrometry reported for LkCa 15’s plan-ets by Sallum et al. (2015b). While Kraus & Ireland(2012) also identify concentrated emission sources inSAM data, they use a binary (LkCa 15 A+ compan-ions) light distribution model for image reconstruction, which is valid only if the brightness distribution resem-bles point sources. Our data show that it does not.On the other hand, the inner disk signal is comparableto the total flux density reported for LkCa 15 bcd fromSallum et al. (2015b) at K and L p . Thus, we emphasizethat the Sallum et al. (2015b) SAM data likely detectedthe inner disk at multiple wavelengths. Furthermore,the gaps and misalignments between LkCa 15’s resolveddisk structures, as well as a warp inferred from the sub-au component (Alencar et al. 2018), may be evidencefor unseen jovian planets (Dong and Fung 2017), whichcould be detected with future facilities (e.g. the ThirtyMeter Telescope ; Skidmore et al. 2015).Our Keck/NIRC2 L p data obtained between 2009 and2017 reveal this emission to be static. Based on SAMdata taken over a shorter timescale, Sallum et al. (2015b,2016) argued that LkCa 15 bcd astrometry reveals ev-idence for orbital motion, although different compo-nents are detected in different epochs and the combinedastrometry appears consistent with stationary sourcesgiven large error bars. While the evaluation of our datais straightforward, several factors may complicate thisaspect of SAM data interpretation for LkCa 15. Forexample, variable u − v coverage between epochs can in-duce apparent astrometric offsets when a binary modelis assumed in the image reconstruction process (C. Cac-eres 2019, in preparation). Instead of bare stellar pho-tospheres, the calibrators used for LkCa 15 in Sallumet al. (2015b) and especially Kraus & Ireland (2012) in-clude multiple stars with bright resolved disk emissionon the same spatial scale as LkCa 15’s disk: some are0also highly variable (e.g. GM Aur, UX Tau; Tanii et al.2012; Oh et al. 2016b).Another common argument is that LkCa 15 bcd aretoo red to be consistent with scattered-light disk emis-sion (Kraus & Ireland 2012; Ireland and Kraus 2014).However, for a system with a pre-transitional disk struc-ture like LkCa 15, (a) scattering can be extremely redbecause (b) the sub-au dust component contributes sig-nificantly to the NIR broadband flux and intercepts (andthen re-emits) a significant fraction of the starlight (Mul-ders et al. 2013; Currie et al. 2017b). The light thatLkCa 15’s 20 au scale disk “sees” is then far redder thanthe star. Indeed, our fiducial disk model successfully re-produces the brightness of the inner dust disk at K and L p . While Sallum et al. (2015b) argued that a disk can-not explain LkCa 15 bc(d) in current SAM data, theyuse a very simple inclined disk model, not a radiativetransfer model. Additionally, from inspection of theirFigure 8, the inner component of this model appears tohave semimajor and semiminor axes of ∼ . (cid:48)(cid:48)
08 and ∼ . (cid:48)(cid:48)
05, which are inconsistent with the larger, spatiallyresolved and extended disk as resolved at K and L p inthis study (0 . (cid:48)(cid:48) . (cid:48)(cid:48) H α detection for LkCa 15 b, which tech-nically remains a candidate companion. However, theyhelp strengthen arguments voicing strong skepticism.As LkCa 15 A itself is bright in H α due to accretionits disk structures should have an elevated H α luminos-ity. Mendigutia et al. (2018) recently found that LkCa15’s spectroastrometric signature at H α is inconsistentwith that of a planet but consistent with a disk. Theyrule out H α emission from a LkCa 15 b unless the can-didate has an H α contrast fainter than 5.5 mags or acontinuum contrast brighter than 6 mags: the H α pho-tometry and continuum upper limits from Sallum et al.(2015b) are just barely consistent with these spectroas-trometric limits. Their predicted emitting region for H α is ρ ∼ . (cid:48)(cid:48) . (cid:48)(cid:48)
16, consistent with our resolved images ofLkCa 15’s inner disk.Furthermore, SPHERE/ZIMPOL data (Thalmann etal. 2015) and our modeling show that both the inner diskand outer disk are bright, modest-contrast structuresand should be detectable at optical wavelengths coveringthe MagAO H α observations. Yet Sallum et al. (2015b)did not report a disk detection with MagAO, implyingthat their H α planet detection may instead be spuriousor a misidentified, partially subtracted piece of the H α -bright disk. Their quoted position for LkCa 15 b in H α isconspicuously close to the inner disk’s major axis. Giventhe MagAO observations’ poor field rotation (1.5 λ / D at0 . (cid:48)(cid:48)
1) and negligibly small rotation gap (5 o or ∼ λ /D at 0 . (cid:48)(cid:48) α data are proprietary, not public, preventing any in-dependent verification that the planet hypothesis is pre-ferred. The public availability of archival Keck/NIRC2data presented here was crucial in assessing evidence forplanets orbiting LkCa 15 from aperture masking.In summary, we rule out the proposed LkCa 15 bcdprotoplanets as being primarily responsible for emissionseen at small angles in SAM data because the emission(a) would be resolved as separate point sources in theSCExAO data (when it is not) and (b) would rotate be-tween 2009 and 2017 Keck/NIRC2 data due to the plan-ets’ orbital motion (which it does not). Our results alsostrengthen the argument from Mendigutia et al. (2018)that H α data also likely identifies a disk, not LkCa 15b.Thus, there is currently no clear, direct evidence formultiple protoplanets orbiting LkCa 15. While the sys-tem shows indirect evidence for at least one unseen jo-vian planet, the bright inner dust disk impedes the de-tection of this companion(s). To confirm jovian com-panions around LkCa 15 from future observations, theinner disk should be resolved and its effect modeled,removed, and shown to be distinguishable from plan-ets. Protoplanet candidates identified from similar sys-tems should likewise be clearly distinguished from diskemission through multi-wavelength and/or multi-epochmodeling (e.g. Keppler et al. 2018).Distinguishing between disk emission and bona fideprotoplanets will continue to be a key challenge for thefield of direct imaging (e.g. Cieza et al. 2013; Kraus etal. 2013; Sallum et al. 2015a; Ligi et al. 2018; Rich et al.2019; Christiaens et al. 2019, this work). Acknowledgements – We thank Michiel Min forgraciously sharing the MCMax3D code and the Sub-aru and NASA/Keck Time Allocation Committees fortheir generous support. We thank the anonymous ref-eree for a careful, thoughtful review. Laurent Pueyo,Jun Hashimoto, Christian Thalmann, Catherine Espail-lat, Nienke van der Marel, Hannah Jang-Condell, Ge-off Bower, and Scott Kenyon provided helpful com-ments and/or additional, independent assessments ofthis manuscript. We emphasize the pivotal culturalrole and reverence that the summit of Maunakea has1always had within the Hawaiian community. We aremost fortunate to conduct scientific observations fromthis mountain. T.C. was supported by a NASA Se-nior Postdoctoral Fellowship and NASA/Keck grantLK-2663-948181; L.C. was supported by CONICYT-FONDECYT grant No. 1171246. C.C. acknowledgessupport from project CONICYT PAI/Concurso Na- cional Insercion en la Academia, convocatoria 2015, folio79150049. MT is supported by JSPS KAKENHI grantNos. 18H05442 and 15H02063. This work utilized theKeck Observatory Archive (KOA), which is operated bythe W. M. Keck Observatory and the NASA ExoplanetScience Institute (NExScI), under contract with the Na-tional Aeronautics and Space Administration.REFERENCES
Alencar, S. H. P., Bouvier, J., Donati, J.-F., et al., 2018,A&A, 620, 195Andrews, S., Rosenfeld, K., Wilner, D. J., Bremer, M.,2011, ApJ, 742, L5Brandt, T. D., Rizzo, M., Groff, T., et al., 2017, JATIS, 3,8002Christiaens, V., Casassus, S., Absil, O., et al., 2019,MNRAS in press, arxiv:1905.01860Cieza, L., Lacour, S., Schreiber, M., et al., 2013, ApJ, 762,L12Currie, T., Burrows, A., Itoh, Y., et al., 2011, ApJ, 729, 128Currie, T., Debes, J., Rodigas, T., et al., 2012, ApJ, 760,L32Currie, T., Burrows, A., Girard, J., et al., 2014a, ApJ, 795,133Currie, T., Daemgen, Debes, J., et al., 2014b, ApJ, 780, L30Currie, T., Cloutier, R., Brittain, S., et al., 2015, ApJ, 814,L27Currie, T., Brittain, S., Grady, C., et al., 2017a, RNAAS, 1,40Currie, T., Guyon, O., Tamura, M., et al., 2017b, ApJ, 836,L15Currie, T., Brandt, T. D., Uyama, T., et al., 2018b, AJ,156, 291Currie, T., Kasdin, N. J., Groff, T., et al., 2018a, PASP,130, 044505Dodson-Robinson, S., Salyk, C., 2011, ApJ, 738, 131Dong, R., Fung, J., 2017, ApJ, 835, 146Espaillat, C., Calvet, N., D’Alessio, P., et al., 2007, ApJ,670, L135Espaillat, C., Furlan, E., D’Alessio, P., et al., 2011, ApJ,728, 49Furlan, E., Forrest, W. J., Sargent, B. A., et al., 2009, ApJ,706, 1194Groff, T. D., Kasdin, N. J., Limbach, M., et al., 2015,Proc. SPIE, 9605, 96051Groff, T. D., Chilcote, J., Brandt, T. D., et al., 2017,Proc. SPIE, 10400, 1040016 Ireland, M., Kraus, A., 2014,
Exploring the Formation andEvolution of Planetary Systems , IAU Symposium, ed.Booth, M. and Matthews, B. C. and Graham, J. R., 299,199Jovanovic, N., Martinache, F., Guyon, O., et al., 2015,PASP, 127, 890Kenyon, S. J., Gomez, M., Whitney, B., 2008, Handbook ofStar-Forming RegionsKeppler, M., Benisty, M., Muller, A., et al., 2018, A&A,617, 44Kraus, A., Ireland, M., Martinache, F., Hillenbrandt, L.,2011, ApJ, 731, 8Kraus, A., Ireland, M., 2012, ApJ, 745, 5Kraus, S., Ireland, M., Sitko, M., et al., 2013, ApJ, 768, 80Isella, A., Chandler, C. J., Carpenter, J. M., Perez, L. M.,Ricci, L., 2014, ApJ, 788, 129Lacour, S., Tuthill, P., Amico, P., et al., 2011, A&A, 532,L72Lafreni´ere, D., Marois, C., Duyon, R., et al., 2007, ApJ,660, 770Ligi, R., Vigan, A., Gratton, R., et al., 2018, MNRAS, 437,1773Marois, C., Lafreni´ere, D., Duyon, R., al., 2006, ApJ, 641,556Marois, C., Lafreni´ere, D., Macintosh, B., Doyon, R.,2008a, ApJ, 673, 647Marois, C., Macintosh, B., Barman, T., et al., 2008b,Science, 322, 1348Marois, C., Zuckerman, B., Konopacky, Q., et al., 2010a,Nature, 468, 1080Marois, C., Macintosh, B., Veran, J.-P., 2010b, Proc. SPIE,7736, 52Marois, C., Correia, C., Galicher, R., et al., 2014,Proc. SPIE, 9148, 91480Mendigutia, I., Oudmaijer, R. D.; Schneider, P. C., et al.2018, A&A, 618, L9Mawet, D., Milli, J., Wahhaj, Z., et al., 2014, ApJ, 792, 97Min, M., Dullemond, C. P., Dominik, C., de Koter, A., andHovenier, J. W., 2009, A&A, 497, 155Mulders, G., Dominik, C., Min, M., 2010, A&A, 512, 112