aa r X i v : . [ a s t r o - ph . E P ] J a n Astronomy & Astrophysicsmanuscript no. aanda © ESO 2021January 22, 2021
A search for a 5th planet around HR 8799 using the star-hoppingRDI technique at VLT/SPHERE
Z. Wahhaj , , J. Milli , , C. Romero , , L. Cieza , , A. Zurlo , , , A. Vigan , E. Peña , G. Valdes , F. Cantalloube , J.Girard , and B. Pantoja European Southern Observatory, Alonso de Córdova 3107, Vitacura, Casilla 19001, Santiago, Chilee-mail: [email protected] Université Grenoble Alpes, CNRS, IPAG, F-38000 Grenoble, France Núcleo de Astronomía, Facultad de Ingeniería y Ciencias, Universidad Diego Portales, Av. Ejercito 441, Santiago, Chile Escuela de Ingeniería Industrial, Facultad de Ingeniería y Ciencias, Universidad Diego Portales, Av. Ejercito 441, Santiago, Chile Aix Marseille Univ, CNRS, CNES, LAM, Marseille, France Max Planck Institute for Astronomy, Königstuhl 17, 69117 Heidelberg, Germany Space Telescope Science Institute, 3700 San Martin Dr., Baltimore, MD 21218, USA Department of Physics and Astronomy, Bucknell University, Lewisburg, PA 17837Received June 29, 2020; accepted December 20, 2020
ABSTRACT
Context.
The direct imaging of extrasolar giant planets demands the highest possible contrasts ( ∆ H &
10 magnitudes) at the smallestangular separations ( ∼ . ′′ ) from the star. We present an adaptive optics observing method, called star-hopping , recently o ff eredas standard queue observing (service mode) for the SPHERE instrument at the VLT. The method uses reference di ff erence imaging(RDI) but unlike earlier works, obtains images of a reference star for PSF subtraction, within minutes of observing the target star. Aims.
We aim to significantly gain in contrast over the conventional angular di ff erencing imaging (ADI) method, to search for a fifthplanet at separations less than 10 au, interior to the four giant planets of the HR 8799 system. The most likely semi-major axes allowedfor this hypothetical planet, estimated by dynamical simulations in earlier work, were 7.5 and 9.7 au within a mass range of 1–8 M Jup . Methods.
We obtained 4.5 hours of simultaneous low resolution integral field spectroscopy (R ∼
30, Y–H band with IFS) and dual-bandimaging (K1 and K2-band with IRDIS) of the HR 8799 system, interspersed with observations of a reference star. The reference starwas observed for about one-third of the total time, and generally needs to be of similar brightness ( ∆ R . . o . The hops between stars were made every 6–10 minutes, with only 1 minute gaps in on-sky integration per hop. Results.
We did not detect the hypothetical fifth planet at the most plausible separations, 7.5 and 9.7 au, down to mass limits of 3.6and 2.8 M Jup respectively, but attained an unprecedented contrast limit of 11.2 magnitudes at 0.1 ′′ . We detected all four planets withhigh signal-to-noise ratios. The YJH spectra for planets c , d were detected with redder H-band spectral slopes than found in earlierstudies. As noted in previous works, the planet spectra are matched very closely by some red field dwarfs. Finally, comparing thecurrent locations of the planets to orbital solutions, we found that planets e and c are most consistent with coplanar and resonantorbits. We also demonstrated that with star-hopping RDI, the contrast improvement at 0.1 ′′ separation can be up to 2 magnitudes. Conclusions.
Since ADI, meridian transit and the concomitant sky rotation are not needed, the time of observation can be chosenfrom within a 2—3 times larger window. In general, star-hopping can be used for stars fainter than R = . Key words. exoplanets – adaptive optics
1. Introduction
Radial velocity (RV) surveys have revealed to us the exo-planet population orbiting within ∼ . // exoplanet.eu / ). Direct imaging and inter-ferometry are the only methods that allow us to obtain spectra ofexoplanets separated by more than a few au from their host stars(Bonnefoy et al. 2014, 2016). Direct imaging is also the only https: // github.com / zwahhaj / starhopping. technique that captures protoplanetary disks in the act of form-ing planets (Keppler et al. 2018; Müller et al. 2018; Ha ff ert et al.2019). Moreover, it has shown us fully formed planetary systemswith their left-over dusty planetesimal disks (Lagrange et al.2012), and captured these dust-producing rocky disks at vari-ous stages over their lifetime (e.g., Boccaletti et al. 2018, 2020;Wahhaj et al. 2016).Studies of systems like HR 8799 with its four planets can of-fer us a glimpse at possible early (age <
30 Myrs) architectures(Marois et al. 2010), perhaps at a stage prior to major planet-migration and scattering (Chatterjee et al. 2008; Crida 2009;Raymond et al. 2010). However, the extrasolar Jupiter and Sat-urn analogs are mostly still hidden from us, orbiting in the glareof their parent stars between 5 and 10 au (Fernandes et al. 2019).Fortunately, a giant planet at an age of 30 Myr can be a hundred
Article number, page 1 of 11 & Aproofs: manuscript no. aanda times brighter than at 300 Myr (e.g., Allard et al. 2012a). Withdirect imaging, we are trying to detect the younger component ofthis hidden population, bridging the unexplored gap to connectto the RV and transit exoplanet populations closer in. In fact,some of the state-of-the-art direct imaging surveys have nearlycompleted and yielded a few more giant planets, fainter and or-biting closer to their stars than in earlier surveys, but mostlythey report that the regions beyond 10 au rarely have planetsmore massive than 3–5 M Jup . (Nielsen et al. 2019; Chauvin et al.2017; Macintosh et al. 2015).Especially for gound-based instruments, the success of thedirect imaging technique, imaging dozens of exoplanets and pro-toplanetary disks has been mainly due to angular and spectraldi ff erence imaging (ADI, SDI and ASDI; Liu 2004; Marois et al.2006; Sparks & Ford 2002; Wahhaj et al. 2011). Without pointspread function (PSF) di ff erencing, within minutes we hit a wallin terms of sensitivity because of quasi static speckles in adaptiveoptics images. Speckles essentially mimic astronomical pointsources, integrating more like signal than noise. The ADI, SDIand other related techniques decouple the speckles from the realsignal, allowing them to be isolated and subtracted. However,these techniques are hampered by the self-subtraction problem(Marois et al. 2006). Since the decoupling of speckles and astro-nomical signal is never complete, there is inevitably some self-subtraction of signal. This can be manageable for planets mod-erately separated from the star, where we just lose sensitivitydepending on the subtraction algorithm used (e.g., Wahhaj et al.2013, 2015). However, for planet-star separations of 1–2 reso-lution elements and extended structures like circumstellar disks,the signal can be completely subtracted or the morphology sig-nificantly altered or completely masked (Milli et al. 2012).Reference di ff erence imaging (RDI), a possible solution,has been routinely used in space telescope observations (e˙g˙Weinberger et al. 1999; Choquet et al. 2016), as the PSF is quitestable over successive orbits of the telescope. However, RDI isnot often used in ground-based observing where PSFs changesignificantly over hours. This is because, prior to extreme AO,the PSF of other stars could not closely match the target PSFsin speckle similarity, especially if the reference star images werenot obtained the same night as the science target. Nevertheless,impressive ground-based results on quite a few targets have beenachieved (Lagrange et al. 2009; Xuan et al. 2018; Ruane et al.2019; Bohn et al. 2020). In the more recent e ff orts, referencePSFs were obtained 30 mins to hours apart and the telescope op-erator would have to manually setup the guiding for each targetchange, costing significant human e ff ort and photon dead-time.Starting recently at VLT / SPHERE, we now o ff er fast automatedRDI available in queue mode for the first time, requiring only aone minute gap for each target change, a technique monikeredstar-hopping RDI. To demonstrate the power of this new observ-ing mode, and to look for new planets closer to the star, we tar-geted HR 8799, the home of the four giants.HR 8799 is a young main-sequence star (age 20–160 Myrs; Cowley et al. 1969; Moór et al. 2006; Marois et al.2008; Hinz et al. 2010; Zuckerman et al. 2011; Baines et al.2012) at a distance of 41.29 ± ∼
30 Myrs suggest that the planetmasses are 5–7 M Jup (Marois et al. 2010; Currie et al. 2011; Sudol & Haghighipour 2012). Interior and exterior to the plan-ets, warm dust at 6–10 au and an exo-Kuiper Belt beyond 100 auhave been detected (Sadakane & Nishida 1986; Su et al. 2009;Hughes et al. 2011; Matthews et al. 2014; Booth et al. 2016).Thus, it is likely that the planets formed in a circumstellar disk,instead of directly from a protostellar cloud as in binary or mul-tiple star formation. However, it is currently a theoretical chal-lenge to form so many massive planets in a single system.The total system architecture and stability, consideringthe age, mass and debris disk formation history have beenstudied in some detail (see Go´zdziewski & Migaszewski2009, 2014, 2018; Reidemeister et al. 2009; Su et al.2009; Fabrycky & Murray-Clay 2010; Moro-Martín et al.2010; Galicher et al. 2011; Marleau & Cumming 2014;Matthews et al. 2013; Booth et al. 2016; Konopacky et al. 2016;Wilner et al. 2018; Geiler et al. 2019). HR 8799 is a star of the λ Bootis type (indicating an iron poor atmosphere), and alsoa γ Dor variable, indicating small surface-pulsations perhapsalso due to some accretion-associated chemical peculiarity(Saio et al. 2018; Saio 2019; Takata et al. 2020). Spectra of theplanets has been obtained in the NIR bands with Keck / OSIRIS(Barman et al. 2011, 2015; Konopacky et al. 2013), Project1640 at Palomar (Oppenheimer et al. 2013), VLT / NACO(Janson et al. 2010), GPI (Ingraham et al. 2014) and SPHERE(Zurlo et al. 2016; Bonnefoy et al. 2016). The comparison ofthe spectra to brown dwarfs, cool field objects and currentatmospheric models suggest patchy thin and thick cloudsof uncertain height, non-equilibrium chemistry, and a dustylow-gravity atmosphere (Marois et al. 2008; Currie et al. 2011;Madhusudhan et al. 2011; Skemer et al. 2012; Marley et al.2012; Morley et al. 2012; Apai et al. 2013; Buenzli et al. 2015).Given the theoretical challenge in explaining such a massivemulti-planet and debris disk system with detailed and specificinformation, and the prospect of finding additional planets(Go´zdziewski & Migaszewski 2014, 2018) the system deservesa deeper look. We describe our SPHERE study of HR 8799in the following sections. The reduction software used in thispaper can be found online .
2. Observations
The goal of star-hopping on VLT / SPHERE is to switch fromrecording adaptive optics corrected images of the science starto the reference star with only a ∼ ∼
10 minutes without much loss in photoncollecting e ffi ciency, and ensuring minimal change in the PSFshape in the elapsed time. We do not provide an exact calcula-tion for the optimum hopping frequency as it depends stronglyon how the seeing and coherence time vary over the observation.However, we found in our observations that PSF similarity drops ∼
2% every 10 minutes (see Section 3.3). This is significant as thesensitivity reached depends non-linearly on the PSF subtractionquality. Thus, we recommend observing the science target for 10minutes, then hopping to the reference star and observing it for5 mins, repeating the cycle as needed.To preserve PSF similarity and for time-e ffi ciency, the AOloops would not be re-optimized when changing stars, and thus https: // github.com / zwahhaj / starhoppingArticle number, page 2 of 11. Wahhaj et al.: Searching for HR 8799 f with Star-hopping RDI on SPHERE the reference star would need to have an R-magnitude (mag)within 1 mag of the science star, to ensure similar AO perfor-mance. While, we do not have strong constraints on the color ofthe reference star, again similar brightness (within 1 mag) in theobserving wavelength is important. This is because the adaptiveoptics performance need to be similar and the signal-to-noiseof the reference images need to be comparable or better. Also,the reference star would need to be within 1 to 2 degrees of thescience star, so that the main mirror’s shape-changes at the newpointing would not result in large changes in PSF properties. For-tunately, for the vast majority of stars fainter than R ∼ R ∼ starhop and hopback which are only responsible for mov-ing the telescope between the two stars and store relevant setupinformation so that subsequent hops can be made automatically.Thus a typical observing sequence would be: 1) Normal acqui-sition of science star with desired instrumental mode and setup,2) An observing template lasting a few minutes, 3) Acquisitionof reference star a few degrees away, with the starhop template,4) Another observing template, 5) Quick return to the sciencestar using the hopback template lasting ∼ hopback template again, 8) As many iterations of steps 4 to 7 asdesired.All three types of acquisitions constitute a full preset of thetelescope, i.e., the primary mirror’s shape and the secondary’spointing are set by a look-up table, then a guide star is selected(automatic for hopback ) for accurate pointing corrections, con-tinuous active optics corrections for the main mirror shape areactivated using the guide star. However, human operators onlyassist with the first (normal) acquisition and the starhop acqui-sition, especially in the selection of the guide star and relatedsetup. The starhop template stores all parameters required forthese setups for the first star, moves (presets) to the second star,lets the operator assist in the second acquisition, and then storesall the parameters for the second acquisition. Small telescopeo ff sets for fine-centering made by the operator when position-ing the star on the instrument detector, are also recorded. Thus,the hopback template already has the relevant parameters savedand can automatically hop back and forth between the two stars,taking only ∼ We observed HR 8799 as part of a director’s discretionary time(DDT) proposal, to test the performance limits of star-hoppingwith RDI on SPHERE. The SPHERE instrument (Beuzit et al.2019), installed at the Nasmyth Focus of unit telescope 3 (UT3)at the VLT, is a state-of-the-art high-contrast imager, polarime-ter and spectrograph, designed to find and characterize exo-planets. It employs an extreme adaptive optics system, SAXO(Fusco et al. 2005, 2006; Petit et al. 2012; Sauvage et al. 2016),with 41 ×
41 actuators (1377 active in the pupil) for wavefrontcontrol, a low read noise EMCCD running at 1380 Hz, a fast(800 Hz bandwidth) tip-tilt mirror (ITTM) for pupil stabiliza-tion, extremely smooth toric mirrors (Hugot et al. 2012), anda di ff erential tip-tilt loop for accurate centering in the NIR.This system can deliver H-band strehl ratios for bright stars(R <
9) of up to 90% and continue to provide AO correction for stars as faint as R =
14 mags. SPHERE also provides coro-nagraphs for di ff raction suppression, including apodized Lyotcoronagraphs (Soummer 2005) and achromatic four-quadrantsphase masks (Boccaletti et al. 2008). It is comprised of threesubsystems: the infrared dual-band imager and spectrograph(IRDIS; Dohlen et al. 2008), an integral field spectrograph (IFS;Claudi et al. 2008) and the Zimpol imaging polarimeter (ZIM-POL; Schmid et al. 2018).We observed HR 8799 in the IRDIFS extended mode(Zurlo et al. 2014), where IRDIS K1 and K2-band imagesand IFS Y–H spectra are obtained simultaneously (Vigan et al.2010). The IRDIFS data was obtained in three 1.5 hour observ-ing blocks (OBs), one block on the night of October 31, 2019 andtwo contiguous blocks on the night of November 1, 2019. Weused the N_ALC_YJ_S coronagraph with a central obscurationof radius 73 mas, which is not ideal for the maximum contrast inK-band but ensures that any object at 100 mas separation wouldnot be partially obscured. With IRDIS we used 8s exposures,while with IFS we used 32s. We also obtained short-exposure un-saturated non-coronagraphic observations of the primary star forflux calibration, which we will call FLUX observations hence-forth. The datasets can be found in the ESO archive under pro-gram ID 2103.C-5076(A) and container IDs: 2622640, 2623891and 2623923. Each container represents a separation epoch, con-sisting of several OBs alternating between HR 8799 and the ref-erence star. The reference star, HD 218381 (spectral type K0vs F0V for HR 8799), is separated 0.55 o from HR 8799 and is0.52 mag fainter than it in R-band but 0.75 mag brighter in H-band. In total, we had 1440 IRDIS exposures for HR 8799 and830 for the reference star. With IFS, we had 190 exposures forHR 8799 and 114 for the reference star. The observing condi-tions were average, with a coherence time of 4.7 ± ± ′′ , and a windspeed of 2.1–7.7 m / s without thelow-wind e ff ect (Milli et al. 2018). The total sky-rotations were23.8 o on the first night and 53.4 o on the second night.
3. Data Reduction
Since our main motivation is to achieve sensitivities to fainterplanets than earlier observations, we begin by estimating the de-tection limits of our data set and post-processing method. Thedetection limits are estimated by comparison to simulated plan-ets which undergo the same reduction processes as the real plan-ets. The measurement and analysis of the real planets in the sys-tem are presented afterwards. For the basic reduction calibra-tions, we used SPHERE pipeline version 0.36.0 and scripts by(Vigan et al. 2015, http: // astro.vigan.fr / tools.html) The IFS datasets from all 3 epochs were combined to form a cube of 7254images, 186 images in each of the 39 wavelength channels. Ineach image, 16 simulated companions were inserted with o ff setswrt. to the star, given by separations: 0.1” to 1.6” with steps of0.1” and position angles increments of 90 o with each step. Thesimulated companions were made from the FLUX exposures ofthe primary appropriately scaled in intensity. Since these sourceswere given constant chromatic contrast, i.e. the same spectra asthe host star, we did not apply any spectral di ff erencing in thereduction described below. The contrasts of these sources werechosen to be roughly 2 mags brighter than a preliminary contrastlimit estimate for the data set. The reference PSF data set con-sisted of 4446 images. All science and reference images wereunsharp-masked, i.e., each image was convolved with a Gaus-sian of FWHM 0.1” (roughly twice the image resolution) and Article number, page 3 of 11 & Aproofs: manuscript no. aanda subtracted from the original to produce an image where mostlarge scale spatial features like di ff use stellar light has been re-moved (e˙g˙ Racine et al. 1999; Wahhaj et al. 2013).A diagonally oriented stripe pattern was found in all the IFSimages, which we were unable to remove in the basic calibratedimages. A zero-valued image passed through the basic calibra-tion also yielded this pattern, found to be independent of thechannel wavelength. Thus the pattern is likely an artefact of thepipeline. The output pattern image was bad-pixel cleaned andunsharp-masked to prepare it to be subtracted from the scienceimages. Two annular regions were defined to optimize PSF sub-traction, i.e., minimize the residual RMS in each region. Thesetwo annuli had inner and outer radii of 0.075” and 0.67”, and0.67” and 1.33” respectively. The science images were median-combined without de-rotation to reveal the background stripepattern more clearly. Then we obtained the best intensity-scaledpattern images for the inner and outer annuli, which we in turnsubtracted from each science image, to perform a preliminaryremoval of the pattern. Next, for each science image, we com-puted the best linear combination of reference images that re-duced the RMS in the two annular regions separately, similar tothe LOCI algorithm (Lafrenière et al. 2007), but a much simplerversion since optimization is done only over the two large annuli.We then took the di ff erence of the science image and this com-posite reference image, and further applied an azimuthal profileremoval filter as described in Wahhaj et al. (2013). All the dif-ference images were median-combined again to check for anyresidual striped pattern, and remove it again by the same proce-dure as before.Generally, we see a consistent but modest improvement incontrast ( ∼ ff er in PSF morphology, we also recommendstudying reductions without applying such filters, even when try-ing to detect faint point sources. −0.5 0.0 0.5−0.50.00.5 a rcs e c ond s IFS Y−H 0.3" −0.5 0.0 0.5 a rcs e c ond s IRDIS K1+K2 −100 −67 −33 0 33 67 100 −100 −67 −33 0 33 67 100
Fig. 1.
Left: An IFS Y–H band reduced image showing simulated plan-ets which are recovered with high SNR. The source recovered closestto the star indicates a contrast limit of 11.2 mags at 0.1 ′′ projected sep-aration. Right: An IRDIS K1 + K2 band reduced image also showingsimulated planets at the same separations, all recovered with high SNR.The same contrast at 0.1 ′′ was reached with IRDIS also. The planetswere inserted into the basic calibrated data (flat-fielded, dark-subtractedand bad pixel corrected) All real planets have been masked out. Thecolor scale is linear with intensity. Next, the images were derotated to align the sky with Northup and East left orientation and median-combined. A signal-to-noise map is made for the final reduced image (Figure 1), wherethe pixels in annular rings of width 4 pixels are divided by therobust standard deviation in that region. The robust value is taken σ c on t r a s t f o r f l a t s pe c t r u m s ou r c e s ( m agn i t ude s ) IRDIS K1+K2−bandIFS YJH−band
Fig. 2.
Contrast limits achieved in the IFS and IRDIS data sets,estimated by flux comparison to simulated planets recovered post-reduction. to mitigate the e ff ect of the simulated planets on the RMS. Thesignal-noise-ratio of each recovered simulated planet was thencompared to its input contrast to calculate the 5 σ -contrast limitachieved at the separation, like so Contrast = InputContrast × S NR /
5. The 5 σ -contrasts achieved in this RDI-only reduction at0.1”, 0.2”, 0.4” and 0.8” separations were 11.2, 13.5, 14.4 and 15mags, corresponding to mass limits of 6.5, 3.1, 2.3 and 1.8 M Jup respectively, as estimated from BT-Settl models assuming an ageof 30 Myrs (Allard et al. 2012b). The contrast curve is shown inFigure 2. The reduction showing only the real planets (withoutsimulated planet insertions) is shown in Figure 3. No new planetsare detected.
The IRDIS reductions with simulated planets were done in asimilar way to the IFS reductions. Since there were less imagesto process, we opted to use a more sophisticated but also morecomputation intensive reduction method. The simulated planetswere inserted in the basic calibrated data at the same o ff sets withrespect to the star as before. The planets inserted were ∼ σ detection limit. For this exercise, we did notcorrect the relative rotational o ff set between IFS and IRDIS, sothe PAs of the real HR 8799 planets do not agree between thetwo reduced images in Figure 1. There were 1443 good scienceimages in the three datasets combined and 828 reference images.The images were first unsharp-masked. Next, we calculatedthe residual rms between all pairs of science and reference im-ages, after intensity scaling to minimize the rms between 70 masand 270 mas. For each science image, the best 16 reference im-ages (more would worsen signal loss) were linearly combinedby LOCI for subtraction to minimize the residual rms separatelyin annular rings covering the whole image. Each target annulus,where the subtraction was actually done, had width 200mas. Butthe reference annuli, where LOCI tried to minimize the residual Article number, page 4 of 11. Wahhaj et al.: Searching for HR 8799 f with Star-hopping RDI on SPHERE −1.0 −0.5 0.0 0.5 1.0−1.0−0.50.00.51.0 a r cseconds IFS Y−H 0.4" −1.0 −0.5 0.0 0.5 1.0 a r cseconds IRDIS K1+K2 −10 −7 −3 0 3 7 10 −10 −7 −3 0 3 7 10
Fig. 3.
IFS and IRDIS images from star-hopping RDI reductions shown with same scale and orientation (North is up, East is left). Left: SNRmap of the IFS Y–H band reduced image, showing only the real planets. The azimuthal filtering creates the dark negative arcs around the planets.They are more pronounced in the IFS reduction as more images were combined here than for the IRDIS reduction. Right: SNR map of the IRDISK1 + K2 band reduced image, showing only the real planets. The star, at the center of the black circle, is masked by the coronagraph. No newplanets are detected in the newly probed region around 0.1 ′′ separation above the contrast limit of 11.2 mags. rms, started 25mas outside the target annuli and extended out-wards to the cover the rest of the image. This was done to miti-gate over-subtraction and signal loss. We chose these parametersmostly by trial and error. The azimuthal filtering, de-rotation andcombination of all the di ff erence images, and the contrast limitestimates were done in the same way as in the IFS reduction.The final reduced images (with and without simulated compan-ions) and the contrast performance are shown in Figures 1, 3and 2, respectively. The IRDIS contrast limit is 11.2 mags at0.1 ′′ which is equal to the IFS limits, but IFS fares ∼ For a comparison of typical ADI and RDI IRDIS observationswe use only the first of the three data sets, totalling 1.5 hoursof execution time, since this is slightly longer that the typicalobservation length (1 hour) at the VLT. The data set constitutes481 science images and for RDI, 276 reference images. The to-tal sky rotation in the sciences images was 24 o . We performed3 di ff erent ADI-based reductions which we call ADI-LOCI-F1 , ASDI-LOCI-F10 and
ASDI-PCA-F10 . The
ADI-LOCI-F1 is thesame as the RDI reduction in terms of reference image selec-tion and reference sector size and the use of LOCI, except thatthe references were restricted to those with more relative ro-tation than one-half FWHM (found by trial). The
ASDI-LOCI-F10 reduction (ASDI is Angular and Spectral Di ff erence Imag-ing) was performed on a data set with simulated companionswhich were made 10 times fainter (thus labeled F10 ) in the K2channel than in K1 channel, allowing aggressive spectral di ff er-encing and a potential contrast gain over ADI. Since referenceimages could have companions both spectrally and rotationallydisplaced, only the combined displacement need to be more than one-half FWHM. The ASDI-PCA-F10 reduction was performedon the same data set as that of
ASDI-LOCI-F10 . The reductionparameters were again optimized by trial and error. We usedprincipal component analysis (PCA) to construct the subtractionPSFs with 5 components (See Soummer et al. 2012). However,for each science image, and for each annular sub-component ofthe image (same as the reductions above) only selected subsetsof the science images were chosen as input for the PCA – resid-ual rms were calculated after subtracting all science image pairs,the best 30 matches (with least rms) that had more relative ro-tation than one-half FWHM were chosen, if less than 30 appro-priate matches were found then the relative rotation criteria wasrelaxed to down to one-fourth FWHM, but no further, to allowinput images for the PSF construction. This more selective ap-proach to PCA helps to reduce the signal self-subtraction ex-pected in ADI, and our tests supported this assumption, yieldingsignificantly better results than PCA alone.The RDI reduction (see top of section 3.2) was repeated forthe same 1.5 hour data set used in the ADI reduction. In Fig-ure 4 we compare the RDI and the
ADI-LOCI-F1 reduction. Thesimulated planets inside 0.3 ′′ separation are much better recov-ered in the star-hopping RDI reduction. In the ADI reduction,the innermost planet at 0.1 ′′ is not recovered at all, while theone at 0.2 ′′ is barely recovered. Contrast curves were calculatedfrom the signal to noise ratio of the recovered simulated plan-ets as before. The contrast improvement of RDI over the threeADI reductions, more than 2 mags at 0.1 ′′ separation, is shownin Figure 5 as a di ff erence between the two contrast curves. Theimprovement will of course vary with the total amount of skyrotation in the science images.Figure 7 illustrates why star-hopping RDI performs so muchbetter than ADI. It shows the residual fractional rms (RFR)for each science image as a function of relative rotation, i.e., Article number, page 5 of 11 & Aproofs: manuscript no. aanda −0.4 −0.2 0.0 0.2 0.4−0.4−0.20.00.20.4 a rcs e c ond s Star−hoppingRDIK1+K2 0.3" −0.4 −0.2 0.0 0.2 0.4 a rcs e c ond s ADI −100 −67 −33 0 33 67 100 −100 −67 −33 0 33 67 100
Fig. 4.
Comparison of star-hopping RDI versus ADI reductions ofIRDIS K1 + K2 band data injected with flat spectrum simulated plan-ets. The inner two simulated planets are not successfully recovered inthe ADI reduction, while they are clearly detected in the RDI reduc-tions. The third simulated planet is recovered significantly better in thestar-hopping RDI reduction. All real planets have been masked out. Thecolor scale is linear with intensity. RD I c on t r a s t i m p r o v e m en t, K − band ( M ag ) RDI − ADI, K1/K2 flux = 1, LOCIRDI − ASDI, K1/K2 flux = 10, LOCIRDI /LOCI − ASDI /PCA, K1/K2 flux = 10
Fig. 5.
RDI contrast improvement over ADI or ASDI, estimated fromthe SNR of recovered simulated companions from an IRDIS data set.The star-hopping RDI technique yields detections limits more than 2mags fainter than ADI at 0 ′′ .1 separation from the target star. The greenline shows the case for a K1 / K2 companion flux ratio of 10, and verysimilar algorithms for RDI and ASDI, except that the ASDI reduction isfine-tuned to minimize self-subtraction. The blue line similarly showsRDI − ADI di ff erence for equal K1, K2 flux. The red line shows the RDIimprovment against the best PCA-based ADI reduction for a K1 / K2 fluxratio of 10. The LOCI and PCA reductions are described in section 3.3. the remaining rms between 0.1 ′′ –0.3 ′′ separations after sub-traction of another science or reference star image, divided bythe original rms in each science image. Specifically, RFR i = RMS ( s i − o j ) / RMS ( s i ), where s i is a science image, o j is anotherscience or reference star image and RMS is computed between0.1 ′′ –0.3 ′′ separations. The RFRs post-RDI subtraction had a 2 σ range of 0.32–0.78. We see that although the science images pro-vide better-matched PSFs in general, the images that can be usedwith minimal self-subtraction are much fewer and much poorermatches than the RDI reference set. Thus, the reference star im-ages constitute a superior set for constructing subtraction PSFs.In Figure 6 we show that artificially increasing the field ro-tation for the RDI reduction (1.5 hour data set) before coaddingthe images does not improve the contrast significantly. Thus thespeckle residuals are comparable to white noise as more rotationdoes not seem to result in additional smoothing. We estimate no − s i g m a C on t r a s t ( M ag ) Fig. 6.
Artificially increasing the field rotation for an RDI reductionbefore coadding the images does not improve the contrast significantly(see section 3.3). The legend gives the total rotation of the reductionfor each contrast curve. At small separations (0.1–0.2 ′′ ) we see no im-provement, as contrasts are not correlated with rotation angles. At largerseparations, we see a maximum of 0.5 mag improvement between theminimum and maximum rotations, 2 o and 143 o , but only 0.3 mag im-provement between 18 o and 143 o . improvement at 0.1 ′′ , ∼ o to 20 o , and 0.5 mag improvementat 1 ′′ , when comparing rotations of 140 o to 2 o . The reductionswere done by mutliplying the actual position angles of the im-ages by specific factors that would achieve total field rotaionsof 2 o to ∼ o (distributed logarithmically), before coadding theimages.During star-hopping tests on the night of August 8, 2019, weobtained 8 images for each of a pair of stars, HD 196963 andHD 196081, which are separated by ∼ o . Since this pair hasa much larger angular separation, we can use the RFR from thisdata set to gauge whether there is significant degradation in PSFsimilarity. Fortunately, the 2 σ range of the RFR was 0.33–0.53,indicating that star-hopping is still very e ff ective for such largeseparations. It should be noted that the coherence time was only1.9–2.1 msec for these observations, compared to 2.5–7.2 msecfor the HR 8799 observations. Although we have low statisticsfor such a performance, these results shows that even in poor toaverage conditions, star-hopping RDI can be e ff ective for a pairof stars separated by almost 2 o . JH -bandspectra fromIFS The spectra of planets c , d and e were extracted with an aperturesize of 3 pixels for all IFS channels. The spectra for planets d and e were corrected for flux loss by comparing them to three flatcontrast sources (uniform contrast across wavelength) per planetinserted at the planets’ separations, but at di ff erent PAs (o ff setfrom the planets by 30 o to 270 o ). These simulated planets are justthe IFS FLUX exposures scaled appropriately in intensity. Theywere inserted at 10 mags of contrast, which is somewhat brighterthan the real planets. Since planet c was detected at the edge ofthe IFS detector where simulated planets could not be inserted,we used the same comparison sources for planets c and d . Thesimulated planets undergo the same reduction process as the realplanets, and their fluxes are extracted using the same aperturesizes, and thus their systematic fractional flux error are the same. Article number, page 6 of 11. Wahhaj et al.: Searching for HR 8799 f with Star-hopping RDI on SPHERE F r a c t i ona l R M S i n d i ff e r en c e i m age SCI−REF (usable)SCI−SCI (not usable)SCI−SCI (usable, ∆ PA> 20 deg)
Fig. 7.
The comparison of PSF similarity between reference-scienceand science-science pairs. The residual fractional rms of di ff erence im-ages are plotted as a function of relative position angle / rotational o ff -set. The black dots represent science-science subtractions, the blue dotsrepresent science-reference subtractions, the red dots represent science-science di ff erences with acceptable self-subtraction. For the science-reference points, the relevant quantity is the time di ff erence, which inour case has an almost linear relationship to the PA di ff erence. We verified this by checking that the spectrum recovered fromthe simulated companions did indeed have a uniform contrast.Thus the planet spectra is calculated as B PR ( λ ) = F PR ( λ ) F PS ( λ ) × − B S ( λ ) (1)where F PR and F PS are the real and simulated planet aperturefluxes respectively, and B S is the stellar spectra. Here, the frac-tional flux losses for the real planet are fully accounted for in theratio, F PR ( λ ) / F PS ( λ ).The flux corrected spectra for planets d and e are shownin Figure 8 along with that of the particularly red L6 object2MASS J2148 + H -band, in com-parison to typical late L-types. Although not as red, the dustydwarves of the field population also have redder than averagespectra(see Zurlo et al. 2016; Stephens et al. 2009; Gagné et al.2014). It should be noted that the spectra do di ff er somewhatin shape from earlier publications, (e.g., Zurlo et al. 2016). Thiscould be because the spectra we present here are the first notto be e ff ected by signal self-subtraction due to ADI or SDI pro-cessing. The most notable di ff erences from earlier spectra (seeFigure 9) are less defined peaks at 1.1 µ m, and for planet d in2019, a gentler slope towards 1.6 µ m. The absence of the peak at1.1 µ m is quite common among observed late L-type (see Figure3 of Bonnefoy et al. 2016, for example), and also seen in thespectra of 2MASS J2148 + µ m are very similar to planet e in 2019. Although, the higher fluxes at 1.6 µ m are rarer amongsuch L-types, it would explain the earlier discrepancy betweenIRDIS and IFS fluxes near the H -band (Zurlo et al. 2016). Wecould not estimate an accurate flux normalization for the spectraof planet c as it was detected near the edge of the detector, sowe show its spectra normalized to 1 at 1.25 µ m in Figure 10. Wedo not pursue this further, as accurate JH -band photometry hasalready been provided in past publications. However, the shapeof the planet’s spectra is reliably detected and show’s an even −6 −6 −6 −6 −5 F l u x / ( S t e ll a r f l u x a t . m i c r on s ) −6 −6 −6 −6 −5 F l u x / ( S t e ll a r f l u x a t . m i c r on s ) planet eplanet d2MASS J2148+4003, L6 Fig. 8.
The spectra for planets d and e compared with that of the L6 ob-ject, 2MASS J2148 + µ m to show the contrast atthat wavelength. The L6 object spectra was scaled to match planet e at1.25 µ m. The shaded regions indicate the 1 σ error ranges of the spec-tra. The wavelength range 1.37–1.45 µ m which is dominated by telluriclines is not shown. −6 −6 −6 −6 −5 F l u x / ( S t e ll a r f l u x a t . m i c r on s ) −6 −6 −6 −6 −5 F l u x / ( S t e ll a r f l u x a t . m i c r on s ) planet e, RDI, 2019planet e, ASDI, 2016planet d, RDI, 2019planet d, ASDI, 2016 Fig. 9.
The RDI-extracted spectra for planets d and e in 2019 comparedwith their ADI-extracted spectra from 2016 as reported in Zurlo et al.(2016). The 2016 planet spectra to the 2019 have been matched at1.25 µ m for easier comparison for their respective shapes. The shadedregions indicate the 1 σ error ranges of the spectra. The wavelengthrange 1.37–1.45 µ m which is dominated by telluric lines is not shown. redder J − H color than planets d and e . Although such red spec-tra are not common, a very similar slope (flux doubling between1.25 and 1.6 µ m) was seen in the L7 object, VHS J125601257 b(Gauza et al. 2015). This L7 object, a planetary candidate com-panion to a brown dwarf, is also thought to have a dusty atmo-sphere with thick clouds (see Bonnefoy et al. 2016, for a dis-cussion). Booth et al. (2016), using the ALMA millimeter array, detecteda broad debris ring, extending from ∼
145 au to ∼
430 au with aninclination of 40 ± o and a position angle of 51 ± o . Prior to this,Su et al. (2009) inferred from the spectral energy distribution of Article number, page 7 of 11 & Aproofs: manuscript no. aanda F l u x no r m a li z ed a t . m i c r on s F l u x no r m a li z ed a t . m i c r on s planet c2MASS J2148+4003, L6VHS J125601257 b, L7 Fig. 10.
The RDI-extracted spectra for planets c compared with that ofthe L6 object, 2MASS J2148 + µ m which is dominated by telluric lines is not shown. the system that a planetesimal belt extending from 100 and 300au separation was the source of blow-out grains extending out to ∼ ′′ and the outer radius could be as far as 11 ′′ from the star.It is expected that RDI reductions would be a major im-provement over ADI for detections of disks with large angularextents, as self-subtraction in these cases is a severe problemfor ADI. To detect the disk, we repeated the IRDIS RDI reduc-tion without simulated companions or any prior image filtering(used to enhance speckle subtraction), as these remove all ex-tended emission. We only used the K1-band images as the K2-band have much higher background. Detecting disks which areclose to azimuthally symmetric in the plane of the sky, and ex-tended over several arcseconds is a challenge very di ff erent fromplanet recovery, as the expected signal area is most of the im-age and the background area is perhaps non-existent. The imagesectors used for PSF subtraction cannot be small, as this wouldremove extended signal. So, we used one large annulus extend-ing from 0.4 ′′ to 2 ′′ separations to cover most of the PSF halo.The final reduction is shown in Figure 11, but no disk emissionwas detected down to a 5 σ contrast of 14.1 magnitudes beyond2.5 ′′ separations. The non-detection is not surprising given themarginal detection of the much brighter 49 Cet debris disk withSPHERE (Choquet et al. 2017). The fractional disk luminosityof HR 8799 is 8 × − (Su et al. 2009) versus 9 × − for 49 Cet(Moór et al. 2015). The inner radius of the disks start at roughly2 ′′ separation for both (Choquet et al. 2017; Booth et al. 2016),with expected physical separations of 100–150 au. The two starshave similar spectral types (F0–A1) with very similar H -bandmagnitudes (5.3–5.5 mag). The photometry of the four planets were extracted by compari-son with simulated planets in a similar way to the IFS spectra.For each of the four planets, three simulated planets were in-serted into the dataset with a contrast of 10 mags, at the sameseparation as the real planets, but with large PA o ff sets (30 to270 o ). The relative aperture photometry was done similar to IFS,but with aperture radius 4 pixels, because of the larger FWHMin the K-band. The recovered photometry are all brighter than −6 −4 −2 0 2 4 6−6−4−20246 a r cseconds IRDIS K1 5"
Fig. 11.
A IRDIS reduction without any prior image filtering to searchfor an extended circumstellar disk beyond angular separations of 2.5 ′′ (to > ′′ ) from the star. ALMA observations by Booth et al. (2016) in-dicate that the disk should have a position angle of 51 ± o and an incli-nation of 40 ± o . We do not detect any disk down to a contrast limit of14.1 magnitudes. Some faint thermal emission from the detector back-ground is seen in the lower right, but not in the expected orientation ofthe known disk. North is up and East is to the left. the Zurlo et al. (2016) measurements by about 0.1 mag (see Ta-ble 1). The standard deviation in the contrasts estimates for thethree reference simulated planets is less than 0.03 mags. Thedominant contrast uncertainty comes from the measurement ofthe AO-corrected stellar PSF core flux, which is measured onlyonce every 1.5 hours. The IRDIS data set for science images were separately re-duced by the
SPHERE data center (Delorme et al. 2017) whichtreated it as an ordinary pupil-tracking sequence. The data cen-ter applied the optimal distortion correction methods consistentwith Maire et al. (2016), to produce a basic-calibrated data setwith high astrometric fidelity (3–4 mas). These images werethen reduced using the high-contrast imaging algorithm, AN-DROMEDA (Cantalloube et al. 2015), to produce astrometricmeasurements (see Table 1) for the four known HR 8799 planets.We also compared the recovered coordinates for the real planetsbetween the RDI and ADI reductions, and found that the planetlocations agreed to within 2.7 mas, smaller than the errors esti-mated in Table 1.An exhaustive orbital fitting e ff ort is being currently under-taken by Zurlo et al. (in preparation) including all extant astrom-etry. Moreover, extensive work has been done to find orbital so-lutions to the prior astrometry for this system, so we just com-pare our latest measurements to the viable orbits computed byWang et al. (2018). From millions of orbits generated by a montecarlo method, they generated 3 sets of solutions: 1) the orbits areforced to be coplanar and have 1:2:4:8 orbital commensurabili- Article number, page 8 of 11. Wahhaj et al.: Searching for HR 8799 f with Star-hopping RDI on SPHERE ties, 2) no coplanarity but with low eccentricity and period com-mensurabilities as before, 3) with no additional constraints. InFigure 12, we overlay our astrometry on orbital solution sets 1and 3. Although, the latest points are consistent with both sets ofsolutions, planets e and c fall close to the expected position in thedynamically stable set, but a bit far from the mean expected loca-tion in the unconstrained set of orbits. Thus, the coplanar orbitswith period commensurabilities are favored in our comparisons.Survival of the four planets and even a hypothetical fifthplanet is possible for the lifetime of the system ( >
30 Myrs), butrequires the period commensurabilities mentioned above. In fact,this was needed even when only planets b , c and d were known(Go´zdziewski & Migaszewski 2009; Reidemeister et al. 2009;Fabrycky & Murray-Clay 2010; Marshall et al. 2010). Such dy-namical models envision that the four planets were formed atlarger separations and migrated inwards. This would allow thevery similar chemical compositions indicated by their spectra,as opposed to more variation expected if they had formed insitu(Marois et al. 2010).The most likely semi-major axes allowed for the hypothet-ical inner planet f , estimated by Go´zdziewski & Migaszewski(2014, 2018) were 7.5 au and 9.7 au, with dynamical constraintson the masses of 2–8 M Jup and 1.5–5 M Jup respectively. TheIFS contrasts we achieved at these separation were 13.05 and13.86 mags, corresponding to estimated masses of 3.6 M Jup and2.8 M Jup respectively (assuming an age of 30 Myr), from theBT-Settl models (Allard et al. 2012b). Thus, the planet may stillexist with a mass of 2–3.6 M Jup at 7.5 au or 1.5–2.8 M Jup at10 au.
4. Conclusions
In this paper, we successfully used the new star-hopping RDItechnique to detect all four known planets of the HR 8799 sys-tem, and significantly improved on the contrast limits attainedpreviously with ADI, at separations less than 0.4 ′′ . This tech-nique of moving quickly to a reference star to capture a similarAO PSF for di ff erencing, with only a 1 minute gap in photon col-lection, can now be used in service mode at the VLT with all theobserving modes available on the SPHERE instrument. Usingstar-hopping RDI, we demonstrated the contrast improvement at0.1 ′′ separation can be up to 2 mags, while at larger separationsthe improvement can be ≈ = o apart. The technique provides significantcontrast improvement mainly because of two reasons: the usablePSF, those without significant self-subtraction or flux loss fromPSF subtraction 1) occur closer in time and thus are more similarto the target image than in ADI and 2) are more numerous thanin ADI as they are spread uniformly over the whole sequence,rather than only available after significant sky rotation. The ben-efit for extended object like disks will be the most impactful,as in ADI the self-subtraction artefacts can result in significantchange in their apparent morphology.In our SPHERE observations of HR 8799, we did not detectplanet f at the most plausible locations, 7.5 and 9.7 au, downto mass limits of 3.6 and 2.8 M Jup , respectively. Also, we did −2 −1 0 1 2arcseconds−2−1012 a r cs e c ond s −2 −1 0 1 2 −2−1012 Fig. 12.
Top: The November 1, 2019 epoch astrometry overlaid asgray diamonds on the most dynamically stable orbital solutions fromWang et al. (2018) (see their Figure 4), where coplanarity and 1:2:4:8period commensurabilities were imposed. The black dots represent ear-lier measured astrometry for the four planets. Bottom: Same pointsoverlaid on the orbital solutions without the additional constraints. The2019 locations for planets e and c are more consistent with the dynami-cal stable family of orbits. not detect any new candidate companions, even at the small-est observable separation, 0.1 ′′ or ≈ M Jup in K1 + K2-band (6.5 M Jup in JHK-band using BT-Settl models from Allard et al. 2012a).However, we detected all 4 planets in K1 + K2-band with SNR of41, 83, 96 and 47 for planets e , d , c and b , respectively. The YJHspectra for planets c , d , e were detected with very red colors.Our spectra of planet c has higher SNR than earlier observations(P1640, Oppenheimer et al. 2013; Pueyo et al. 2015). Planets c Article number, page 9 of 11 & Aproofs: manuscript no. aanda
Table 1.
Astrometry and photometry of the four HR 8799 planets. planet ρ (mas) σ ρ (mas) PA σ PA SNR ∆ K1 (mag) ∆ K2 (mag) Mass ( M J )e 406 4 302.72 o o
41 10.8 ± ± + − d 686 4 231.38 o o
83 10.7 ± ± + − c 958 3 335.86 o o
96 10.8 ± ± + − b 1721 4 69.05 o o
47 11.89 ± ± + − The mass estimates are from the PHOENIX BT-Settl atmospheric models (Bara ff e et al. 2015), assuming an age of 30 + − Myrs. However, themost dynamically stable orbital solutions from Wang et al. (2018) set much tighter limits: a mass of 5 . ± . M J for planet b , and 7 . ± . M J for the other planets. and d spectra have some di ff erences with respect to earlier obser-vations. Particularly, the spectral slope is redder in the H-band,which is significant as that part of the spectra has the highestSNR. This could be due to real evolution of the atmosphere ofthe planets over the past few years. Previous work has alreadyshown that the spectra are di ffi cult to find close matches with cur-rent compositional models due to inadequate understanding ofcloud properties and non-equilibrium chemistry (Bonnefoy et al.2016). However, the spectra are matched very closely by somered field dwarfs and a planetary mass companion to a browndwarf (VHS J125601257 b; Gauza et al. 2015). We disk notdetect the debris disk seen by ALMA (Booth et al. 2016), butthis is not surprising given that the much brighter debris diskof a comparable system, 49 Cet, was only marginally detectedby SPHERE (Choquet et al. 2017). Finally, comparing the cur-rent locations of the planets to orbital solutions from Wang et al.(2018), we found that planets e and c are more consistent withcoplanar and resonant orbits than without such restrictions.In summary, the star-hopping RDI technique significantlyboosts SPHERE’s detection capabilities both for planets and cir-cumstellar disks, and should contribute to high-impact exoplanetscience, as the technique is brought to other telescope facilities. Acknowledgements.
This work has made use of the the SPHERE Data Centre,jointly operated by OSUG / IPAG (Grenoble), PYTHEAS / LAM / CESAM (Mar-seille), OCA / Lagrange (Nice), Observatoire de Paris / LESIA (Paris), and Ob-servatoire de Lyon. We would especially like to thank Nadege Meunier at theSPHERE data center in Grenoble for the distortion-corrected reductions usedfor the astrometric measurements, Bartosz Gauza for providing the spectra ofVHS J125601.92-125723.9 b, Jason Wang for allowing us to use the dynam-ical modeling figures from his publication, and Matias Jones, Florian Rodler,Benjamin Courtney-Barrer, Francisco Caceres and Alain Smette at the VLT fortechnical help during the various phases of the development of the star-hoppingtechnique.
References
Allard, F., Homeier, D., & Freytag, B. 2012a, Philosophical Transactions of theRoyal Society of London Series A, 370, 2765Allard, F., Homeier, D., Freytag, B., & Sharp, C. M. 2012b, in EAS PublicationsSeries, Vol. 57, EAS Publications Series, ed. C. Reylé, C. Charbonnel, &M. Schultheis, 3–43Apai, D., Radigan, J., Buenzli, E., et al. 2013, ApJ, 768, 121Baines, E. K., White, R. J., Huber, D., et al. 2012, ApJ, 761, 57Bara ff e, I., Homeier, D., Allard, F., & Chabrier, G. 2015, A&A, 577, A42Barman, T. S., Konopacky, Q. M., Macintosh, B., & Marois, C. 2015, ApJ, 804,61Barman, T. S., Macintosh, B., Konopacky, Q. M., & Marois, C. 2011, ApJ, 733,65Bell, C. P. M., Mamajek, E. E., & Naylor, T. 2015, Proceedings of the Interna-tional Astronomical Union, 10, 41–48Beuzit, J. L., Vigan, A., Mouillet, D., et al. 2019, A&A, 631, A155Boccaletti, A., Chauvin, G., Baudoz, P., & Beuzit, J. L. 2008, A&A, 482, 939Boccaletti, A., Di Folco, E., Pantin, E., et al. 2020, A&A, 637, L5Boccaletti, A., Sezestre, E., Lagrange, A. M., et al. 2018, A&A, 614, A52Bohn, A. J., Kenworthy, M. A., Ginski, C., et al. 2020, MNRAS, 492, 431Bonnefoy, M., Marleau, G. D., Galicher, R., et al. 2014, A&A, 567, L9 Bonnefoy, M., Zurlo, A., Baudino, J. L., et al. 2016, A&A, 587, A58Booth, M., Jordán, A., Casassus, S., et al. 2016, MNRAS, 460, L10Buenzli, E., Saumon, D., Marley, M. S., et al. 2015, ApJ, 798, 127Cantalloube, F., Mouillet, D., Mugnier, L. M., et al. 2015, A&A, 582, A89Chatterjee, S., Ford, E. B., Matsumura, S., & Rasio, F. A. 2008, ApJ, 686, 580Chauvin, G., Desidera, S., Lagrange, A. M., et al. 2017, A&A, 605, L9Choquet, É., Milli, J., Wahhaj, Z., et al. 2017, ApJ, 834, L12Choquet, É., Perrin, M. D., Chen, C. H., et al. 2016, ApJ, 817, L2Claudi, R. U., Turatto, M., Gratton, R. G., et al. 2008, in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series, Vol. 7014, Proc. SPIE,70143ECowley, A., Cowley, C., Jaschek, M., & Jaschek, C. 1969, AJ, 74, 375Crida, A. 2009, in SF2A-2009: Proceedings of the Annual meeting of the FrenchSociety of Astronomy and Astrophysics, ed. M. Heydari-Malayeri, C. Reyl’E,& R. Samadi, 313Currie, T., Burrows, A., Itoh, Y., et al. 2011, ApJ, 729, 128Delorme, P., Meunier, N., Albert, D., et al. 2017, in SF2A-2017: Proceedings ofthe Annual meeting of the French Society of Astronomy and Astrophysics,ed. C. Reylé, P. Di Matteo, F. Herpin, E. Lagadec, A. Lançon, Z. Meliani, &F. Royer, DiDohlen, K., Langlois, M., Saisse, M., et al. 2008, in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series, Vol. 7014, Proc. SPIE,70143LDong, S. & Zhu, Z. 2013, ApJ, 778, 53Fabrycky, D. C. & Murray-Clay, R. A. 2010, ApJ, 710, 1408Fernandes, R. B., Mulders, G. D., Pascucci, I., Mordasini, C., & Emsenhuber, A.2019, ApJ, 874, 81Fusco, T., Petit, C., Rousset, G., Conan, J. M., & Beuzit, J. L. 2005, OpticsLetters, 30, 1255Fusco, T., Rousset, G., Sauvage, J. F., et al. 2006, Optics Express, 14, 7515Gagné, J., Lafrenière, D., Doyon, R., Malo, L., & Artigau, É. 2014, ApJ, 783,121Gaia Collaboration. 2018, VizieR Online Data Catalog, I / ff ert, S. Y., Bohn, A. J., de Boer, J., et al. 2019, Nature Astronomy, 3, 749Hinz, P. M., Rodigas, T. J., Kenworthy, M. A., et al. 2010, ApJ, 716, 417Howard, A. W., Marcy, G. W., Johnson, J. A., et al. 2010, Science, 330, 653Hughes, A. M., Wilner, D. J., Andrews, S. M., et al. 2011, ApJ, 740, 38Hugot, E., Ferrari, M., El Hadi, K., et al. 2012, A&A, 538, A139Ingraham, P., Marley, M. S., Saumon, D., et al. 2014, ApJ, 794, L15Janson, M., Bergfors, C., Goto, M., Brandner, W., & Lafrenière, D. 2010, ApJ,710, L35Keppler, M., Benisty, M., Müller, A., et al. 2018, A&A, 617, A44Konopacky, Q. M., Barman, T. S., Macintosh, B. A., & Marois, C. 2013, Science,339, 1398Konopacky, Q. M., Marois, C., Macintosh, B. A., et al. 2016, AJ, 152, 28Lafrenière, D., Marois, C., Doyon, R., Nadeau, D., & Artigau, É. 2007, ApJ, 660,770Lagrange, A. M., Boccaletti, A., Milli, J., et al. 2012, A&A, 542, A40Lagrange, A. M., Gratadour, D., Chauvin, G., et al. 2009, A&A, 493, L21Liu, M. C. 2004, Science, 305, 1442Looper, D. L., Kirkpatrick, J. D., Cutri, R. M., et al. 2008, ApJ, 686, 528Macintosh, B., Graham, J. R., Barman, T., et al. 2015, Science, 350, 64Madhusudhan, N. 2019, ARA&A, 57, 617Madhusudhan, N., Burrows, A., & Currie, T. 2011, ApJ, 737, 34Maire, A. L., Bonnefoy, M., Ginski, C., et al. 2016, A&A, 587, A56Marleau, G. D. & Cumming, A. 2014, MNRAS, 437, 1378Marley, M. S., Saumon, D., Cushing, M., et al. 2012, ApJ, 754, 135 Article number, page 10 of 11. Wahhaj et al.: Searching for HR 8799 f with Star-hopping RDI on SPHERE
Marois, C., Lafrenière, D., Doyon, R., Macintosh, B., & Nadeau, D. 2006, ApJ,641, 556Marois, C., Macintosh, B., Barman, T., et al. 2008, Science, 322, 1348Marois, C., Zuckerman, B., Konopacky, Q. M., Macintosh, B., & Barman, T.2010, Nature, 468, 1080Marshall, J., Horner, J., & Carter, A. 2010, International Journal of Astrobiology,9, 259Matthews, B., Kennedy, G., Sibthorpe, B., et al. 2013, The Astrophysical Journal,780, 97Matthews, B., Kennedy, G., Sibthorpe, B., et al. 2014, ApJ, 780, 97Mayor, M., Marmier, M., Lovis, C., et al. 2011, arXiv e-prints, arXiv:1109.2497Milli, J., Kasper, M., Bourget, P., et al. 2018, in Society of Photo-Optical In-strumentation Engineers (SPIE) Conference Series, Vol. 10703, Proc. SPIE,107032AMilli, J., Mouillet, D., Lagrange, A. M., et al. 2012, A&A, 545, A111Moór, A., Ábrahám, P., Derekas, A., et al. 2006, ApJ, 644, 525Moór, A., Kóspál, Á., Ábrahám, P., et al. 2015, MNRAS, 447, 577Morley, C. V., Fortney, J. J., Marley, M. S., et al. 2012, ApJ, 756, 172Moro-Martín, A., Malhotra, R., Bryden, G., et al. 2010, ApJ, 717, 1123Müller, A., Keppler, M., Henning, T., et al. 2018, A&A, 617, L2Nielsen, E. L., De Rosa, R. J., Macintosh, B., et al. 2019, AJ, 158, 13Oppenheimer, B. R., Baranec, C., Beichman, C., et al. 2013, ApJ, 768, 24Petit, C., Sauvage, J. F., Sevin, A., et al. 2012, in Society of Photo-Optical In-strumentation Engineers (SPIE) Conference Series, Vol. 8447, Proc. SPIE,84471ZPueyo, L., Soummer, R., Ho ff mann, J., et al. 2015, ApJ, 803, 31Racine, R., Walker, G. A. H., Nadeau, D., Doyon, R., & Marois, C. 1999, PASP,111, 587Raymond, S. N., Armitage, P. J., & Gorelick, N. 2010, ApJ, 711, 772Reidemeister, M., Krivov, A. V., Schmidt, T. O. B., et al. 2009, A&A, 503, 247Ruane, G., Ngo, H., Mawet, D., et al. 2019, AJ, 157, 118Sadakane, K. & Nishida, M. 1986, Publications of the Astronomical Society ofthe Pacific, 98, 685Saio, H. 2019, MNRAS, 487, 2177Saio, H., Bedding, T. R., Kurtz, D. W., et al. 2018, MNRAS, 477, 2183Sauvage, J.-F., Fusco, T., Petit, C., et al. 2016, Journal of Astronomical Tele-scopes, Instruments, and Systems, 2, 025003Schmid, H. M., Bazzon, A., Roelfsema, R., et al. 2018, A&A, 619, A9Skemer, A. J., Hinz, P. M., Esposito, S., et al. 2012, ApJ, 753, 14Soummer, R. 2005, ApJ, 618, L161Soummer, R., Pueyo, L., & Larkin, J. 2012, ApJ, 755, L28Sparks, W. B. & Ford, H. C. 2002, ApJ, 578, 543Stephens, D. C., Leggett, S. K., Cushing, M. C., et al. 2009, ApJ, 702, 154Su, K. Y. L., Rieke, G. H., Stapelfeldt, K. R., et al. 2009, The AstrophysicalJournal, 705, 314Su, K. Y. L., Rieke, G. H., Stapelfeldt, K. R., et al. 2009, ApJ, 705, 314Sudol, J. J. & Haghighipour, N. 2012, ApJ, 755, 38Takata, M., Ouazzani, R. M., Saio, H., et al. 2020, A&A, 635, A106Torres, C. A. O., Quast, G. R., Melo, C. H. F., & Sterzik, M. F. 2008, YoungNearby Loose Associations, ed. Reipurth, B., 757Vigan, A., Gry, C., Salter, G., et al. 2015, MNRAS, 454, 129Vigan, A., Moutou, C., Langlois, M., et al. 2010, MNRAS, 407, 71Wahhaj, Z., Cieza, L. A., Mawet, D., et al. 2015, A&A, 581, A24Wahhaj, Z., Liu, M. C., Biller, B. A., et al. 2011, ApJ, 729, 139Wahhaj, Z., Liu, M. C., Biller, B. A., et al. 2013, ApJ, 779, 80Wahhaj, Z., Milli, J., Kennedy, G., et al. 2016, A&A, 596, L4Wang, J. J., Graham, J. R., Dawson, R., et al. 2018, AJ, 156, 192Weinberger, A. J., Becklin, E. E., Schneider, G., et al. 1999, ApJ, 525, L53Wilner, D. J., MacGregor, M. A., Andrews, S. M., et al. 2018, ApJ, 855, 56Xuan, W. J., Mawet, D., Ngo, H., et al. 2018, AJ, 156, 156Zuckerman, B., Rhee, J. H., Song, I., & Bessell, M. S. 2011, ApJ, 732, 61Zurlo, A., Vigan, A., Galicher, R., et al. 2016, A&A, 587, A57Zurlo, A., Vigan, A., Mesa, D., et al. 2014, A&A, 572, A85mann, J., et al. 2015, ApJ, 803, 31Racine, R., Walker, G. A. H., Nadeau, D., Doyon, R., & Marois, C. 1999, PASP,111, 587Raymond, S. N., Armitage, P. J., & Gorelick, N. 2010, ApJ, 711, 772Reidemeister, M., Krivov, A. V., Schmidt, T. O. B., et al. 2009, A&A, 503, 247Ruane, G., Ngo, H., Mawet, D., et al. 2019, AJ, 157, 118Sadakane, K. & Nishida, M. 1986, Publications of the Astronomical Society ofthe Pacific, 98, 685Saio, H. 2019, MNRAS, 487, 2177Saio, H., Bedding, T. R., Kurtz, D. W., et al. 2018, MNRAS, 477, 2183Sauvage, J.-F., Fusco, T., Petit, C., et al. 2016, Journal of Astronomical Tele-scopes, Instruments, and Systems, 2, 025003Schmid, H. M., Bazzon, A., Roelfsema, R., et al. 2018, A&A, 619, A9Skemer, A. J., Hinz, P. M., Esposito, S., et al. 2012, ApJ, 753, 14Soummer, R. 2005, ApJ, 618, L161Soummer, R., Pueyo, L., & Larkin, J. 2012, ApJ, 755, L28Sparks, W. B. & Ford, H. C. 2002, ApJ, 578, 543Stephens, D. C., Leggett, S. K., Cushing, M. C., et al. 2009, ApJ, 702, 154Su, K. Y. L., Rieke, G. H., Stapelfeldt, K. R., et al. 2009, The AstrophysicalJournal, 705, 314Su, K. Y. L., Rieke, G. H., Stapelfeldt, K. R., et al. 2009, ApJ, 705, 314Sudol, J. J. & Haghighipour, N. 2012, ApJ, 755, 38Takata, M., Ouazzani, R. M., Saio, H., et al. 2020, A&A, 635, A106Torres, C. A. O., Quast, G. R., Melo, C. H. F., & Sterzik, M. F. 2008, YoungNearby Loose Associations, ed. Reipurth, B., 757Vigan, A., Gry, C., Salter, G., et al. 2015, MNRAS, 454, 129Vigan, A., Moutou, C., Langlois, M., et al. 2010, MNRAS, 407, 71Wahhaj, Z., Cieza, L. A., Mawet, D., et al. 2015, A&A, 581, A24Wahhaj, Z., Liu, M. C., Biller, B. A., et al. 2011, ApJ, 729, 139Wahhaj, Z., Liu, M. C., Biller, B. A., et al. 2013, ApJ, 779, 80Wahhaj, Z., Milli, J., Kennedy, G., et al. 2016, A&A, 596, L4Wang, J. J., Graham, J. R., Dawson, R., et al. 2018, AJ, 156, 192Weinberger, A. J., Becklin, E. E., Schneider, G., et al. 1999, ApJ, 525, L53Wilner, D. J., MacGregor, M. A., Andrews, S. M., et al. 2018, ApJ, 855, 56Xuan, W. J., Mawet, D., Ngo, H., et al. 2018, AJ, 156, 156Zuckerman, B., Rhee, J. H., Song, I., & Bessell, M. S. 2011, ApJ, 732, 61Zurlo, A., Vigan, A., Galicher, R., et al. 2016, A&A, 587, A57Zurlo, A., Vigan, A., Mesa, D., et al. 2014, A&A, 572, A85