An ALMA survey of λ Orionis disks: from supernovae to planet formation
Megan Ansdell, Thomas J. Haworth, Jonathan P. Williams, Stefano Facchini, Andrew Winter, Carlo F. Manara, Alvaro Hacar, Eugene Chiang, Sierk van Terwisga, Nienke van der Marel, Ewine F. van Dishoeck
DDraft version October 2, 2020
Typeset using L A TEX twocolumn style in AASTeX63
An ALMA survey of λ Orionis disks: from supernovae to planet formation
Megan Ansdell,
1, 2
Thomas J. Haworth, Jonathan P. Williams, Stefano Facchini, Andrew Winter, Carlo F. Manara, Alvaro Hacar, Eugene Chiang, Sierk van Terwisga, Nienke van der Marel, andEwine F. van Dishoeck Flatiron Institute, Simons Foundation, 162 Fifth Ave, New York, NY 10010, USA NASA Headquarters, 300 E Street SW, Washington, DC 20546, USA Astronomy Unit, School of Physics and Astronomy, Queen Mary University of London, London E1 4NS, United Kingdom Institute for Astronomy, University of Hawai‘i at M¯anoa, Honolulu, HI, USA European Southern Observatory, Karl-Schwarzschild-Str. 2, D-85748 Garching bei M¨unchen, Germany Astronomisches Rechen-Institut, Zentrum f¨ur Astronomie der Universit¨at Heidelberg, M¨onchhofstraße 12-14, 69120 Heidelberg, Germany Leiden Observatory, Leiden University, P.O. Box 9513, 2300-RA Leiden, The Netherlands Department of Astronomy, University of California at Berkeley, Berkeley, CA 94720, USA Max-Planck-Institute for Astronomy, K¨onigstuhl 17, 69117 Heidelberg, Germany Department of Physics & Astronomy, University of Victoria, Victoria, BC, V8P 1A1, Canada (Accepted 18 Sept. 2020)
Submitted to AAS JournalsABSTRACTProtoplanetary disk surveys by the Atacama Large Millimeter/sub-millimeter Array (ALMA) arenow probing a range of environmental conditions, from low-mass star-forming regions like Lupus tomassive OB clusters like σ Orionis. Here we conduct an ALMA survey of protoplanetary disks in λ Orionis, a ∼ ∼ M ⊕ ). We assess how massive OB stars impact planet formation, in particularfrom the supernova that may have occurred ∼ λ Orionis; studying theseeffects is important as most planetary systems, including our Solar System, are likely born in clusterenvironments. We find that the effects of massive stars, in the form of pre-supernova feedback and/or asupernova itself, do not appear to significantly reduce the available planet-forming material otherwiseexpected at the evolved age of λ Orionis. We also compare a lingering massive “outlier” disk in λ Orionisto similar systems in other evolved regions, hypothesizing that these outliers host companions in theirinner disks that suppress disk dispersal to extend the lifetimes of their outer primordial disks. Weconclude with numerous avenues for future work, highlighting how λ Orionis still has much to teachus about perhaps one of the most common types of planet-forming environments in the Galaxy.
Keywords: protoplanetary disks; planet formation; supernovae; OB stars; millimeter astronomy INTRODUCTIONThousands of diverse exoplanetary systems have nowbeen discovered (e.g., see review in Winn & Fabrycky2015), yet how they all formed remains unclear due toour still-incomplete understanding of the evolution ofthe progenitor protoplanetary disks. These disks are
Corresponding author: Megan Ansdell; @megaparsec808 ; (cid:135) @mansdell [email protected] traditionally thought to evolve through viscous accretion(e.g., Lynden-Bell & Pringle 1974), where turbulence re-distributes angular momentum and drives material ontothe central star. However other processes, both internaland external to the disks, can also significantly influencetheir evolution. In particular, the external influences ofmassive OB stars on the disks of surrounding lower-massstars is important to study as many planetary systems,including our Solar System, are likely born in clusterenvironments (e.g., Adams 2010; Winter et al. 2020). a r X i v : . [ a s t r o - ph . S R ] S e p Ansdell et al.
In stellar clusters, the ultraviolet (UV) emission fromthe OB stars induces thermal winds from nearby disks,which effectively remove planet-forming material fromtheir outer regions in a process called “external photo-evaporation” (e.g., Hollenbach et al. 1994). Theoreti-cal work suggests that external photoevaporation canseverely shorten disk lifetimes and truncate outer diskradii in cluster environments, whereas stellar encountersplay a relatively insignificant role in sculpting disk pop-ulations (e.g., Scally & Clarke 2001; Concha-Ram´ırezet al. 2019), even in moderate UV environments (e.g.,Facchini et al. 2016; Haworth et al. 2018; Winter et al.2018). Moreover, external photoevaporation of the outerdisk is theorized to dominate over viscous spreading un-der realistic cluster conditions and dust grain growthprescriptions (e.g., Clarke 2007; Facchini et al. 2016;Winter et al. 2018). While direct detection of photo-evaporative disk winds is observationally challenging, ahandful of cases exist (e.g., Henney & O’Dell 1999; Rigli-aco et al. 2009). Meanwhile, indirect evidence of exter-nal photoevaporation from observations of its expectedimpact on other disk properties is growing (e.g., Mannet al. 2014; Kim et al. 2016; Guarcello et al. 2016; Ans-dell et al. 2017; van Terwisga et al. 2019b).Still, our understanding of how disk evolution is al-tered when one of these rapidly evolving OB stars in-evitably goes supernova remains limited, due in partto the lack of recent supernova events in nearby star-forming regions (SFRs) available for study. This makesthe λ Orionis cluster a key target, as a supernova isthought to have occurred in the core of the region ∼ ∼ λ Orionis may therefore provide a rare snapshotof an evolved SFR post-supernova, with the remainingOB stars, lower-mass stellar population, and remnantmolecular cloud all still present. Supernovae are the-orized to strip significant amounts of mass from disksaround nearby stars via ram pressure (e.g., Close & Pit-tard 2017) as well as expose them to enhanced cosmicray ionization rates that accelerate carbon processingof CO into other molecules (e.g., Eistrup et al. 2016;Schwarz et al. 2018a; Bosman et al. 2018), thereby pro-viding predictions to test against observations.A partial survey of the disk population in λ Orio-nis was conducted at millimeter wavelengths by Ans-dell et al. (2015) with JCMT/SCUBA-2. Observationsof disks at these longer wavelengths are particularlyuseful because any optically thin continuum emissioncan be related to the available planet-forming solids inthe disk (e.g., Hildebrand 1983; Andrews 2015), whilevarious molecular lines can probe gas content and/or chemistry (e.g., Miotello et al. 2017; van Terwisga et al.2019b; Miotello et al. 2019; Booth & Ilee 2020). How-ever, due to the evolved age ( ∼ ∼
400 pc) of λ Orionis, combined with the limited sen-sitivity of JCMT/SCUBA-2, Ansdell et al. (2015) de-tected only one disk in their survey. Fortunately, the At-acama Large Millimeter/sub-millimeter Array (ALMA)now provides the high sensitivity required to efficientlysurvey the disk population in λ Orionis to dust masssensitivities that are commonly achieved for the nearby( ∼
150 pc) SFRs like Lupus (Ansdell et al. 2016, 2018),Chamaeleon I (Pascucci et al. 2016), ρ Ophiuchus (Ciezaet al. 2019), and Upper Sco (Barenfeld et al. 2016). Thispaper presents the results of such an ALMA survey.We begin in Section 2 by discussing the formation andevolution of λ Orionis, as the cluster’s history is centralto our analysis. In Section 3, we describe our sample ofprotoplanetary disks and their host star properties. OurALMA survey of these disks and the key observationalresults are presented in Section 4, while the basic diskproperties are derived in Section 5. We then examine theimplications for our understanding of how OB stars af-fect disk evolution in Section 6, which also discusses howmassive “outlier” disks in evolved regions like λ Orionismay improve our knowledge on the pathways of planetformation and disk dispersal. We summarize our find-ings and provide avenues for future work in Section 7. THE λ ORIONIS CLUSTERThe formation, evolution, and current state of the λ Orionis cluster has been studied and debated in manyworks (e.g., Maddalena & Morris 1987; Cunha & Smith1996; Dolan & Mathieu 2001; Hern´andez et al. 2009;Bayo et al. 2011; Mathieu 2015; Kounkel et al. 2018).In this section, we briefly summarize the existing obser-vations of the cluster, then assume a commonly adoptedbut still-debated interpretation, which has implicationsfor our later analysis.As illustrated in Figure 1, λ Orionis currently consistsof several hundred low-mass members centered on a coreof OB stars. The most massive is the O8III star, λ Ori,the brightest in the head of the Orion constellation. Al-though λ Ori is the only O-type star in the cluster, it hasa B0V companion at 1900 au projected separation, andthis binary system is accompanied by nine other B-typestars (some close binaries themselves) in a dense ∼ ∼
30 pc radius ring (orpossibly shell; Lee et al. 2015), which is centered on theclump of OB stars, encompasses the lower-mass popu-lation, and is rapidly expanding at ∼
14 km s − (e.g., LMA Survey of λ Orionis Disks h m m m m m Right Ascension D e c li n a t i o n D i s k D u s t M a ss ( M ) Figure 1.
An IRAS 100- µ m image of the λ Orionis cluster. The white dashed box outlines the region surveyed for disks with
Spitzer in Hern´andez et al. (2009, 2010), which we use as the basis of our ALMA survey sample selection (Section 3.1). Theinset zooms in on this region: the large circles are our ALMA detections color-coded by dust mass (Section 5.1), while the smallgray circles are the non-detections (Section 4.2); the dotted lines represent 1 pc and 3 pc radial distances from the λ Ori systemat the center of the cluster. Gray arrows show the
Gaia
DR2 proper motions of the cluster members identified by Kounkel et al.(2018) in the LSR frame with the median cluster value subtracted; a reference vector of magnitude 1 mas yr − and a scale barof 5 pc are shown in the lower right corner. The small white stars are the locations of the B-type stars in the region, and thelarger white star is the location of the λ Ori system, which contains the only O-type star in the region (Section 2).
Ansdell et al.
Maddalena & Morris 1987; Lang et al. 2000). As shownin Figure 1, many of the λ Orionis members also haveproper motions directed radially away from the clustercenter, with the more distant stars moving away thefastest (Kounkel et al. 2018).The age of the lower-mass stellar population in λ Ori-onis is ∼ ∼ ∼ not consider the youngerdark Barnard 30 and Barnard 35 clouds, located alongthe edge of the ring, as part of the λ Orionis cluster.We adopt the common interpretation in the literatureof these observations, which is that an O-type star ex-ploded as a supernova ∼ λ Orionis mem-bers have been attributed to a “single-trigger” expan-sion caused by the supernova (Kounkel et al. 2018),such kinematics can result from any mechanism thatremoves the interstellar gas and its gravitational poten-tial (Winter et al. 2019), and outward acceleration mayeven be aided by gravitational feedback of the dispersedgas (Zamora-Avil´es et al. 2019). While these and othercaveats are discussed further in Section 6.1, the massivestars in λ Orionis, whether through pre-supernova feed-back and/or a supernova event itself, have played a keyrole in shaping the region. SAMPLE3.1.
Disk Sample Selection
Like most SFRs, the disk census in λ Orionis is basedon targeted
Spitzer observations, which can identifystars exhibiting excess emission above the stellar photo-sphere at near-infrared (IRAC 3.6, 4.5, 5.8, and 8.0 µ m;Fazio et al. 2004) and/or mid-infrared (MIPS 24 µ m;Rieke et al. 2004) wavelengths where dust emits effi-ciently. Hern´andez et al. (2009) used Spitzer data tostudy the intermediate-mass population in λ Orionis,finding 29 members earlier than F5 but only 10 bear-ing disks; they classified the nine sources with moder-ate infrared excess as debris disks and the one source with large infrared excess as an optically thick disk.Hern´andez et al. (2010) then studied the lower-masspopulation in λ Orionis, finding 436 members down tothe substellar limit but only 49 with disks; they groupedthe disks according to their spectral energy distributions(SEDs)—optically thick disks had the largest excesses,evolved disks had smaller excesses, and (pre-)transitiondisks exhibited signs of inner disk clearings. We ex-clude from our sample the nine debris disks, as theseare likely second-generation dust disks (Wyatt 2008),resulting in an initial sample of 50 primordial (or proto-planetary) disks. We use the same naming conventionsas Hern´andez et al. (2009, 2010) in this work.As shown in Figure 1, the region surveyed by
Spitzer (white box) only covers ∼ Gaia (Gaia Collaborationet al. 2016) Data Release 2 (DR2; Gaia Collaborationet al. 2018) combined with spectroscopic data from theApache Point Observatory Galactic Evolution Experi-ment (APOGEE; Majewski et al. 2017) revealed a pop-ulation of radially expanding λ Orionis members thatextend well beyond this area (Kounkel et al. 2018). Sur-veys for disks in these outer regions have not yet beenconducted, making it possible that our ALMA samplebased solely on
Spitzer data is incomplete. Neverthe-less, the low protoplanetary disk fraction in λ Orionisinferred from the
Spitzer data is consistent with the ∼ Spitzer observations for the sample selection isa general (albeit moderate) problem facing the ALMAdisk demographic literature, as
Gaia continues to revealmissed or interloping stellar populations in young SFRs(e.g., Manara et al. 2018b; Galli et al. 2020; Luhman &Esplin 2020). We also note that the radially expandingpopulation missed by
Spitzer is unlikely to represent ayounger population formed as a consequence of the su-pernova and/or feedback from the massive OB stars, asDolan & Mathieu (1999, 2001) found no evidence fortriggered or sequential star formation in the region.We identify interlopers in the
Spitzer sample by us-ing distances from
Gaia
DR2 to find contaminant back-ground sources—a method that has been previously ap-plied to refine membership in the Lupus clouds (Man-ara et al. 2018b). While the individual
Gaia
DR2 par-allaxes of λ Orionis members remain imprecise due tothe cluster’s distance, they are sufficient for identify-ing clear interlopers. We therefore remove five sources(LO 1310, 2357, 2404, 5042, and 7517) from our sample,
LMA Survey of λ Orionis Disks Table 1.
Source Properties
Source 2MASS ID R.A. Decl. SpT M (cid:63) v LSR
Ref. † µ α µ δ ( M (cid:12) ) (km s − ) (mas yr − ) (mas yr − )LO 65 J05331515+0950301 05:33:15.15 +09:50:30.05 M4 0.19 11.5 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± † References for stellar spectral types (SpT) and radial velocities ( v LSR ): (1) Bayo et al. (2011), (2) Hern´andez et al. (2010), (3)Hern´andez et al. (2009), (4) Kounkel et al. (2018), (5) Maxted et al. (2008), (6) Sacco et al. (2008).
Ansdell et al. as the lower bounds of their estimated
Gaia
DR2 dis-tances from Bailer-Jones et al. (2018) are >
480 pc, mak-ing them likely background sources. Indeed, LO 1310and LO 2357 have radial velocities in the local stan-dard of rest (LSR) reference frame of v LSR = 1 . −
35 km s − , respectively, which differ significantly fromthe average cluster value of 12 km s − (Kounkel et al.2018); LO 2404, 5042, and 7517 do not have known ra-dial velocities. We also remove LO 3710, found to be anon-member by Bayo et al. (2011) due to a discrepantsurface gravity. Of these interlopers, only LO 7517is detected by our ALMA survey at marginal (3.4 σ )significance; given its estimated Gaia
DR2 distance of3744 +405 − pc, it is likely a background galaxy.Table 1 gives our final sample of the 44 protoplanetarydisks analyzed in the remainder of this work.3.2. Host Star Properties
The available host star properties for our disk sampleare provided in Table 1. Stellar spectral types (SpT)were mostly determined by Bayo et al. (2011) frommoderate-resolution optical and near-infrared spectra.However, 14 stars have only photometric spectral typesfrom Hern´andez et al. (2010), who interpolated R − J col-ors onto the spectral type sequence using the standard R − J colors from Kenyon & Hartmann (1995). Althoughthe R − J colors were not corrected for reddening, thereddening toward λ Orionis is low at E ( B − V ) ≈ . ± µ RA )and declination ( µ Dec ) are from
Gaia
DR2. Radial ve-locities are also provided and taken from various liter-ature sources, translated into the LSR reference frame( v LSR ) using the source coordinates. The source coordi-nates in right ascension (R.A.) and declination (Decl.)are the fitted positions for our ALMA detections (Sec-tion 4.2) or the Two Micron All-Sky Survey (2MASS;Skrutskie et al. 2006) positions for the non-detections.As previously mentioned, the λ Orionis cluster istoo distant for reliable individual
Gaia
DR2 parallaxes.This is further complicated for young, disk-hosting starswhose variability from disk scattered light and/or sur-rounding nebulosity can result in poor fits to the
Gaia
DR2 single-star astrometric model, leading to higher as-trometric noise. Thus we do not provide distance esti-mates for each source in our sample, but rather rely onthe average distance of the non-disk-bearing membersin λ Orionis from
Gaia
DR2, which is well-constrainedto 404 ± ±
50 pc based on main-sequencefitting to the massive OB stars (Dolan & Mathieu 2001;Mathieu 2015). We therefore adopt a distance of 400 pcfor λ Orionis in the remainder of this work.We estimate stellar masses ( M (cid:63) ) from these SpT val-ues and 2MASS J -band magnitudes, following the meth-ods of Ansdell et al. (2017). Each target is placed onthe HertzsprungRussel (HR) diagram by converting SpTto stellar effective temperature and J -band magnitudeto stellar luminosity using the relations from Herczeg& Hillenbrand (2015) and a distance of 400 pc. Es-timates of M (cid:63) are then found by comparing the posi-tions on the HR diagram to the evolutionary modelsof Baraffe et al. (2015). For the one intermediate-massstar, HD 245185, we instead use the evolutionary modelsof Siess et al. (2000). We do not provide M (cid:63) estimatesfor the two sources without 2MASS data (LO 7957 andLO 7951). The typical M (cid:63) uncertainties, propagatedfrom the uncertainties on SpT and J -band magnitude,are 0.1–0.2 M (cid:12) . ALMA OBSERVATIONS & RESULTS4.1.
ALMA Observations
The ALMA observations used in this work were takenin Cycle 5 under program 2017.1.00466.S (PI: Ansdell).The program allowed for a range of array configura-tions to maximize the probability of survey completion,and the observations were conducted over 13 executionblocks (see Appendix A for information on the arrayconfiguration and weather conditions for each executionblock). All sources in our sample were observed duringeach execution block, thus the sensitivities and synthe-sized beams are uniform across the sample.The spectral setup was identical for each executionblock: four spectral windows were centered on 247.96,245.46, 232.97, and 230.51 GHz, each with usable band-widths of 1.88 GHz, for a bandwidth-weighted meancontinuum frequency of 239.36 GHz (1.25 mm). Thelast spectral window covered the CO J = 2 − ∼ − ) spectral resolution topreserve the maximum possible continuum bandwidthwhile allowing for the possibility of detecting moleculargas emission.On-source integration times were ∼ σ dust mass constraints would be compa-rable to the limits reached in the ALMA surveys of morenearby SFRs (e.g., Barenfeld et al. 2016; Ansdell et al.2016; Pascucci et al. 2016), assuming a linear relationbetween millimeter flux and dust mass (e.g., Hildebrand1983). Data were pipeline calibrated by NRAO staffusing the Common Astronomy Software Applications( CASA ) package (McMullin et al. 2007) version 5.4.0.
LMA Survey of λ Orionis Disks HD 245185 LO 4187 LO 1624 LO 4407 LO 4155 LO 6866 LO 4163LO 4531 LO 4520 LO 5267 LO 3360 LO 1152 LO 65 LO 4255
Figure 2.
ALMA 1.25 mm continuum images of the 14 detected disks in our λ Orionis sample, ordered by decreasing fluxdensity (as reported in Table 2). The 2 (cid:48)(cid:48) × (cid:48)(cid:48) images are centered on the source, scaled to their maximum value, and clippedbelow 1.5 σ for clarity. The typical beam size of ∼ (cid:48)(cid:48) is shown in the first panel by the white ellipse. The pipeline included flux, bandpass, and gain calibra-tions (see Appendix A for a list of the calibrators). Weassume an absolute flux calibration uncertainty of 10%,similar to other ALMA disk surveys (e.g., Barenfeldet al. 2016; Ansdell et al. 2016; Cazzoletti et al. 2019).4.2.
ALMA Continuum Results
We create continuum images from the calibrated visi-bilities by averaging over the continuum channels usingthe split task in
CASA , then cleaning with a Briggsrobust weighting parameter of +0 . tclean task. This results in a median continuum rms of 34 µ Jyand beam size of 0 . (cid:48)(cid:48) × (cid:48)(cid:48) . We do not perform self-calibration as the detected sources are faint, with a me-dian signal-to-noise ratio of 15 (see Table 2).In most cases, we measure continuum flux densities byfitting point source models to the visibility data with the uvmodelfit task in CASA . The point source model hasthree free parameters: integrated flux density ( F . ),right ascension offset from the phase center (∆ α ), anddeclination offset from the phase center (∆ δ ). For thenon-detections, we fix ∆ α and ∆ δ to zero when running uvmodelfit to avoid spurious detections (the phase off-sets are typically only 0.05 (cid:48)(cid:48) for the detections, muchsmaller than the beam size). In three cases, the sourcesare resolved, therefore we use an elliptical Gaussianmodel instead, which has three additional parameters:full-width-half-max along the major axis ( a ), aspect ra-tio of the axes ( r ), and position angle (P . A . ). With theunderlying assumption that these models describe thedata appropriately, we multiply the uncertainties on allthe fitted parameters by the factor needed to produce areduced χ of 1 (typically a factor of two).Table 2 reports the F . values for all sources,along with their statistical uncertainties (i.e., not includ- ing the 10% flux calibration error), while Figure 2 showsthe continuum images for the ≥ σ detections. Only 14of the 44 sources are detected at ≥ σ significance andonly 3 of the detections are marginally resolved. Weconservatively identify resolved sources as those wherethe ratio of a to its uncertainty is greater than five. Theresolved sources are HD 245185, LO 4187, and LO 4407with the following fitted elliptical Gaussian parame-ters, respectively: a = 0 . (cid:48)(cid:48) ± (cid:48)(cid:48) , 0.113 (cid:48)(cid:48) ± (cid:48)(cid:48) ,and 0.123 (cid:48)(cid:48) ± (cid:48)(cid:48) ; r = 0 . ± ± ± . A . = 70 . ◦ ± ◦ , − . ◦ ± ◦ , and85.5 ◦ ± ◦ .We stack the images of the 30 non-detections toconstrain the average continuum flux for the individ-ually undetected sources, finding a tentative detectionof 0 . ± .
006 mJy (3.2 σ ) in the stacked image. Weverify this result by calculating the mean continuumflux density and standard error on the mean for the 30non-detections using the values in Table 2, which gives0 . ± .
005 mJy (4.0 σ ).4.3. ALMA CO Results
We extract CO channel maps from the calibratedvisibilities by first subtracting the continuum using the uvcontsub task in
CASA . To search for CO emission,we follow the general procedure of Ansdell et al. (2017).In short, we extract an initial spectrum for each sourceto identify candidate detections with emission exceed-ing 3 × the channel rms near the expected v LSR of thecluster (12 km s − ; Kounkel et al. 2018). These candi-dates are visually inspected and any emission is cleanedwith a Briggs robust weighting parameter of +0 . tclean . Zero-moment maps are then created by sum-ming the channels ± − from the systemic veloc-ity unless clear emission was seen beyond these limits. Ansdell et al.
Table 2.
ALMA Disk Fluxes
Source F . F (mJy) (mJy km s − )LO 65 0.117 ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± < ± ± ± < ± < ± < ± < ± < ± ± ± < ± ± ± < ± < ± ± ± < ± < ± < ± < ± < ± < ± < ± < ± < ± ± ± < ± < ± < ± < The integrated CO line fluxes ( F ) are measuredusing a curve-of-growth aperture photometry method,with errors ( E ) estimated by taking the standarddeviation of fluxes measured within the same-sized aper-ture placed randomly within the field of view but awayfrom the source.Only five sources (HD 245185, LO 4155, LO 4187,LO 4407, LO 6866) are detected with F ≥ × E , though two of these (LO 4155, LO 4407) aremarginal (3–4 σ ) detections. One other source, LO 1624,shows emission at ∼ × the rms in the single channel atits known v LSR , but is undetected in its zero-momentmap as summing over several channels dilutes the emis-sion. For the non-detections, we construct zero-momentmaps by integrating ± − from their known v LSR when available (Table 1), else from the average v LSR ofthe cluster (12 km s − ; Kounkel et al. 2018). Table 2gives the F values for the detections and upper lim-its of 3 × the rms in the zero-moment maps for the non-detections. Figure 3 shows the zero-moment and first-moment maps for the five detections, while Appendix Bpresents the CO spectra for all sources in our sample.Figure 4 then shows ALMA Band 6 CO emission asa function of the Band 6 continuum for the continuum-detected disks in the evolved λ Orionis cluster (thiswork), the young Lupus clouds (Ansdell et al. 2018),and the middle-aged σ Orionis cluster (Ansdell et al.2017; Ansdell et al., in prep), all scaled to 150 pc. Wedo not show other notable ALMA disk surveys, as theywere conducted in Band 7 (e.g., Barenfeld et al. 2016;Pascucci et al. 2016) or do not have published Band 6 CO fluxes (e.g., Cieza et al. 2019). Interestingly, theroughly linear correlation between continuum and COflux holds over the first ∼ λ Orionissurvey can be explained by the low continuum emission.This roughly linear correlation between the (some-what) optically thin millimeter continuum flux and theoptically thick CO flux is potentially due to more mas-sive dust disks having more extended gas disks (e.g.,Barenfeld et al. 2016). Moreover, since there is an ob-served linear relationship between the millimeter contin-uum luminosity ( L mm ) and emitting surface area ( R )of protoplanetary disks (Tripathi et al. 2017; Andrewset al. 2018a), and Figure 4 implies that the CO lu-minosity ( L CO ) is proportional to L mm , while we alsoexpect L CO to be proportional to the CO emittingsurface area ( R ) if it is optically thick, we can predictthat R CO ∝ R mm , which is indeed seen in observations(Ansdell et al. 2018; Trapman et al. 2020). LMA Survey of λ Orionis Disks H D CO (0th) CO (1st) k m / s L O k m / s L O k m / s L O k m / s L O k m / s Figure 3.
Our ALMA Band 6 observations of the five COdetections in our sample (Section 4.3). The first columnshows the 1.25 mm continuum maps with 4 σ , 20 σ , and 150 σ contours. The middle column shows the CO zero-momentmaps, scaled to their maximum value and clipped below 2 σ for clarity. The final column shows the CO first-momentmaps within the 3 σ contours of the zero-moment maps. Im-ages are 2 (cid:48)(cid:48) × (cid:48)(cid:48) and the typical beam size is given in the firstpanel by the dashed ellipse.5. DISK PROPERTIES5.1.
Dust Masses
Under the simplified assumption that dust emissionfrom a protoplanetary disk at millimeter wavelengths isoptically thin and isothermal, the observed continuumflux density at a given frequency ( F ν ) can be directlyrelated to the mass of the emitting dust ( M dust ), as es-tablished in Hildebrand (1983): M dust = F ν d κ ν B ν ( T dust ) ≈ . × F . , (1)where M dust is in Earth masses and F . is in mJy.Here we use B ν ( T dust ) as the Planck function for a char-acteristic dust temperature of T dust = 20 K, the median ALMA Band 6 Continuum Flux (mJy) C O F l u x ( J y k m / s ) Lupus (2 Myr) Orionis (3 Myr) Orionis (5 Myr)
Figure 4. CO flux as a function of continuum flux inALMA Band 6 for the continuum-detected disks in Lupus(Ansdell et al. 2018), σ Orionis (Ansdell et al. 2017, Ansdellet al., in prep), and λ Orionis (this work). Fluxes are normal-ized to 150 pc and downward-facing triangles are 3 σ upperlimits on CO flux. Approximate ages of each region areprovided for reference. The four λ Orionis CO detectionsfollow the roughly linear correlation seen in younger regions,and the lingering bright disk in λ Orionis is HD 245185. for Taurus disks (Andrews & Williams 2005). We takethe dust grain opacity, κ ν , as 2.3 cm g − at 230 GHzand use an opacity power-law index of β d = 1 . β = 0 .
4. For the distance, d , we use the cluster’s av-erage Gaia
DR2 value of 400 pc (Section 3.2), and takethe F . measurements from Table 2.With this approach, the median M dust of the contin-uum detections in our λ Orionis survey is only ∼ M ⊕ .These dust masses may be underestimated, however,if the observed millimeter continuum emission is (par-tially) optically thick (e.g., Andrews & Williams 2005;Zhu et al. 2019) and/or if the temperature of the dust inthe outer disk (where we assume most of the disk massis located) is lower than 20 K. Nevertheless, employingthis simplified relation with these caveats in mind pro-vides the most practical approach given the faint andunresolved emission from the disks in our observations.Figure 5 shows the M dust estimates for the continuumdetections in our survey in increasing order. It also in-cludes the median 3 σ upper limit of ∼ M ⊕ for theindividual non-detections in our survey, as well as themean detection of ∼ M ⊕ found when stacking thesenon-detections (Section 4.2).0 Ansdell et al. S t a c k e d N D × R M S L O L O L O L O L O L O L O L O L O L O L O L O L O H D D u s t M a ss ( M ) OMC-2 (1 Myr)Lupus (2 Myr) Orionis (3 Myr)Upper Sco (5 Myr) Orionis (5 Myr)
Figure 5.
Dust masses of the 14 continuum-detected disks in our λ Orionis survey (white circles) in increasing order (sourcenames are given on the x axis). The white downward triangle is the median 3 σ upper limit for individual non-detections (ND),while the white star shows their detected average dust mass from our stacking analysis (Section 5.1). Comparable disk dustmass populations (see Section 5.2) from OMC-2 (purple crosses), Lupus (blue diamonds), σ Orionis (orange x’s), and UpperSco (green squares) are also shown, illustrating the smooth distribution in the younger regions in contrast with the steep risetoward the high-mass outlier disks in the older populations (see Sections 5.2 and 6.3). The colored solid lines are polynomialfits to guide the eye for each region; approximate ages of each region are also provided for reference.
Comparisons to Other Regions At ∼ λ Orionis provides an important pointof comparison to the several younger disk populations,as well as the similarly aged Upper Sco disk population,that have been previously surveyed by ALMA. Figure 6compares the M dust cumulative distribution for λ Orio-nis (this work) to that of OMC-2 (van Terwisga et al.2019a), Lupus (Ansdell et al. 2018), σ Orionis (Ans-dell et al. 2017), and Upper Sco (Barenfeld et al. 2016).The M dust values are uniformly calculated using Equa-tion 1 with the reported wavelengths of the surveys andtypical Gaia
DR2 distances of ∼
150 pc for Lupus andUpper Sco and ∼
400 pc for σ Orionis and OMC-2. Forthe approximate ages of the regions, we adopt the valuesused in the ALMA surveys or reported in more recentanalyses of the protoplanetary disk populations (e.g.,Andrews 2020). The cumulative distributions are con-structed using the Kaplan-Meier Estimator (with thePython lifelines package; Davidson-Pilon et al. 2020)to account for upper limits, as in previous works (e.g.,Barenfeld et al. 2016; Ansdell et al. 2017; Cieza et al.2019; van Terwisga et al. 2019b; Cazzoletti et al. 2019). Figure 6 illustrates that λ Orionis follows the generaldecay in the overall disk dust mass population with agethat has been previously reported (e.g., Ansdell et al.2016; Barenfeld et al. 2016; Pascucci et al. 2016; Ciezaet al. 2019; van Terwisga et al. 2019b). Moreover, the M dust distribution in λ Orionis is statistically indistin-guishable from that of the similarly aged Upper Sco as-sociation. Both of these findings suggest that, if a super-nova did occur relatively recently in λ Orionis, it did nothave a significant impact on disk dust mass evolution inthe region. We discuss this further in Section 6.1.A reliable comparison of the M dust distributions inFigure 6 requires confirming that the regions have sim-ilar M (cid:63) populations, due to the known correlation be-tween disk dust mass and stellar mass (e.g., Andrewset al. 2013; Ansdell et al. 2016; Barenfeld et al. 2016;Pascucci et al. 2016). Two-sample tests have previouslydemonstrated that the stellar mass populations in Lu-pus, σ Orionis, and Upper Sco are likely drawn from thesame parent population (Barenfeld et al. 2016; Ansdellet al. 2016, 2017). To confirm that the M dust popula-tions in λ Orionis and Upper Sco are indeed statisticallyindistinguishable, we again use two-sample tests to com-
LMA Survey of λ Orionis Disks M dust [M ] P M d u s t OMC-2 (1 Myr)Lupus (2 Myr) Orionis (3 Myr) Orionis (5 Myr)Upper Sco (5 Myr)
Figure 6.
The disk dust mass ( M dust ) cumulative distribu-tion for λ Orionis compared to several other regions surveyedby ALMA, calculated with the Kaplan-Meier Estimator toaccount for upper limits (see Section 5.2). Approximate agesfor each region are provided for reference. pare their stellar mass populations. Using the M (cid:63) val-ues from Table 1 for λ Orionis and those from Barenfeldet al. (2016) for Upper Sco, we find p -values of 0.77 and0.11 for the T-test and Wilcoxon rank-sum test, respec-tively (calculated with scipy.stats in Python). Thusthe M (cid:63) populations are likely drawn from the same par-ent population and so the M dust distributions can bereliably compared.Figure 5 also illustrates the decline in M dust distribu-tions with age, but highlights differences at the high-mass end, which are not readily apparent from the cu-mulative distributions in Figure 6. In Figure 5 , the λ Orionis detections are plotted against the upper 32%of the M dust populations in the comparison regions (32%was chosen to match the detection rate in λ Orionis).Interestingly, the highest-mass disks in all the regionsappear to converge around ∼
100 M ⊕ regardless of ageor environment. Although the number of disks at suchhigh masses falls steeply after a few Myr, one or more“outlier” disks appear to persist even at ∼ HD 245185
HD 245185 hosts a clear outlier disk in λ Orionis, hav-ing over an order of magnitude higher millimeter con-tinuum and CO flux than the rest of the sample (Fig-ure 4). HD 245185 is a well-studied Herbig Ae/Be star inthe literature and is the only early-type (A0) star in oursample (Table 1) with a mass of ∼ M (cid:12) (e.g., Folsomet al. 2012). Figure 3 shows the 1.25 mm continuum mapand the CO zero-moment and first-moment maps fromour ALMA observations. We note that the proper mo-tions ( µ α = 0 . µ δ = − .
93) and distance (427 +21 − pc)from Gaia
DR2, as well as the systemic velocity derivedfrom our ALMA CO observations ( v LSR ≈
13 km s − ),are all consistent with cluster membership.Equation 1 yields M dust ≈ M ⊕ , an order of magni-tude higher than the next most massive disk in λ Orionis(Figure 5). One concern is that this is due to HD 245185being the only hot star in our sample, given that the sim-plified relation in Equation 1 assumes T dust = 20 K forall disks. However, only some of the difference (a factorof ∼
3) may be accounted for if, rather than assumingan isothermal disk, we scale the dust temperature withstellar luminosity using T dust = 25 K × ( L (cid:63) /L (cid:12) ) . assuggested by the radiative transfer model grid of An-drews et al. (2013). We do not use this scaling, however,as it remains uncertain whether such a clear relationshipholds in the real disk population. For example, Tazzariet al. (2017) found no relation between T dust and stel-lar properties when modeling the ALMA visibilities ofresolved Lupus disks.Another reason why a massive disk around HD 245185is unusual at the evolved age of λ Orionis is that disksaround intermediate-mass stars typically dissipate twiceas fast as those around late-type stars (at least basedon infrared emission, which traces the warm inner disk;e.g., Ribas et al. 2015). It is unlikely that HD 245185 ismuch younger than the average disk in the cluster: sev-eral authors have estimated the age of HD 245185, e.g.,6 . ± . . ± . λ Orionis to rec-oncile this system. We discuss possible explanations forsuch long-lived primordial disks in Section 6.3. DISCUSSION6.1.
Do Supernovae Impact Planet Formation? λ Orionis may provide a rare snapshot of an evolvedSFR that is ∼ λ Orionis to those in other SFRs2
Ansdell et al. may therefore provide an opportunity to study how su-pernovae affect planet formation.Supernovae are theorized to strip significant amountsof mass from disks around nearby pre-main sequencestars. Close & Pittard (2017) ran three-dimensional hy-drodynamic simulations of protoplanetary disks, with arange of masses and inclinations, subject to a super-nova occurring 0.3 pc away. They reported an “in-stantaneous stripping” phase with mass-loss rates of10 − M (cid:12) yr − lasting 10–100 years, followed by moremoderate but extended ablation with mass-loss rates of10 − to 10 − M (cid:12) yr − . For the low-mass (0.1 M Jup )and moderate-mass (1.0 M Jup ) disks in their simula-tions, up to 90% and 30% of the disk mass, respec-tively, was removed during the instantaneous strippingphase; these disk masses are typical in σ Orionis, a pos-sible example of a pre-supernova OB cluster. High-mass(10 M Jup ) disks, however—similar to the outlier aroundHD 245185 in λ Orionis (Section 5.3)—were largely un-affected. Since the peak ram pressure in the simulationsof Close & Pittard (2017) strongly depends on distancefrom the supernova (dropping off as d − ), we would ex-pect most disks within 0.3 pc to be significantly depletedin mass, with those further out relatively unaffected.Indeed, as shown in Figure 7, we observe but do notdetect four disks with projected separations < . λ Orionisappears to follow the general decline in disk dust masswith age seen in other SFRs, and is statistically indis-tinguishable from that of the similarly aged Upper Scoassociation (Figure 6; Section 5.2), implies that a su-pernova occurring several Myr into disk evolution doesnot significantly reduce the amount of planet-formingsolid material that would otherwise be available at thisage, except potentially for disks that were within a smallfraction of a parsec from the supernova event.Additionally, pre-supernova feedback may mute theeffects of the actual supernova event on the surroundingenvironment. Recent simulations that combine stellarwinds and photoionization with supernova events (Lucaset al. 2020) found that pre-supernova feedback sculptslow-density channels in the gas, through which super-nova energy can more freely escape into the wider inter-stellar medium. As a result, supernova explosions mayhave only moderate, though more widespread, effectson the surrounding natal molecular clouds and their
Projected Separation from Ori [pc] M d i s k = × M d u s t [ M J u p ] Figure 7.
Disk mass ( M disk ), assuming a gas-to-dust ratioof 100, as a function of projected separation from the centralOB star ( λ Ori) in λ Orionis. Circles are ALMA Band 6continuum detections and downward-facing triangles are 3 σ upper limits; stars indicate CO detections. The verticaldotted line denotes the 0.3 pc distance of the supernova inthe simulations of Close & Pittard (2017) (see Section 6.1). star/disk populations. When occurring late in disk evo-lution, it is possible that these moderate effects from su-pernovae are simply negligible when compared to otherdisk evolutionary processes.One caveat to this interpretation, however, is thatwe cannot rule out that a supernova also occurred inthe SFRs against which we are comparing the λ Ori-onis disk population. In particular, a supernova mayhave also occurred ∼ ζ Oph and pulsarPSR 1932+11059, which may have once been close bina-ries in Upper Sco before the pulsar progenitor explodedas a supernova (Hoogerwerf et al. 2000, 2001). However,the large uncertainties on the present-day kinematics ofthe runaway objects, including an unknown radial ve-locity for the pulsar, make the location of the presumedsupernova, and thus its potential impact on the UpperSco disk population, unclear.Another caveat is that our ALMA sample is selectedfrom targeted
Spitzer observations (Section 3), thus onlyincludes disks that are currently within ∼ λ Orionis members outsidethe
Spitzer survey region are radially expanding outward(Figure 1; Kounkel et al. 2018), the concern is that ourALMA survey missed some disks that were once muchcloser to the cluster core and thus potentially most af-fected by the supernova. However, the faster-moving
LMA Survey of λ Orionis Disks ∼ − (relative to thecluster median), which translates to ∼ (cid:38)
10 pc from the cluster core, it isunlikely that they were particularly close to the super-nova when it occurred, especially if the outward accel-eration was aided by gravitational feedback of the dis-persed gas (Zamora-Avil´es et al. 2019). Moreover, al-though Kounkel et al. (2018) explain the observed radialmotions as due to the “single-trigger expansion” causedby the supernova, alternative explanations staged pre-supernova are also viable: such kinematics can be a con-sequence of any mechanism that disperses the intraclus-ter gas (e.g., stellar winds or radiation; see Winter et al.2019), and pre-supernova feedback may preclude the ac-tual supernova event as the main mechanism driving theremoval of the gas potential (e.g., Lucas et al. 2020).Indeed, an alternative explanation is that a super-nova has not yet actually occurred in λ Orionis, andthat the observed features described in Section 2 orig-inate from other aspects of the cluster history, such aspre-supernova feedback. In fact, the expected chemicaleffects from a supernova are not readily apparent in ourcurrent ALMA data. Supernovae are production sitesof cosmic rays (see review in Grenier et al. 2015), thusthe cosmic ray ionization rate of H ( ζ CR ) should be en-hanced in λ Orionis. Typical ionization rates in molec-ular clouds are ζ CR ∼ − s − and may be even lowerin disk midplanes (Cleeves et al. 2014), however after asupernova the levels can be enhanced to ζ CR ∼ − or10 − s − (e.g., Indriolo et al. 2010; Le Petit et al. 2016).At these levels, the transformation of CO into methanoland hydrocarbons proceeds much faster in the ice andgas, on scales of < λ Orionis disks, and at levels expected from theirmillimeter continuum emission (Figure 4). However, thesample size of gas detections is small; deeper observa-tions and additional molecular lines will help determineif our current CO non-detections are due to dispersal ofthe gas or chemical transformation of the CO.Nevertheless, the massive stars in λ Orionis, throughpre-supernova feedback and/or a recent supernova eventitself, appear to have sculpted many of the observationalfeatures of the region, yet have not significantly reducedthe available planet-forming material in the overall diskpopulation beyond what is expected at this evolved age.6.2.
External Photoevaporation in λ Orionis
OB stars can also impact the disk population throughexternal photoevaporation (e.g., Johnstone et al. 1998; St¨orzer & Hollenbach 1999), a process that is nowthought to be one of the main environmental factors de-pleting disk material (e.g., Scally & Clarke 2001; Selleket al. 2020), even in typical galactic UV environments(e.g., Facchini et al. 2016; Haworth et al. 2016; Win-ter et al. 2018). While direct detection of externallydriven photoevaporative winds is observationally chal-lenging (Henney & O’Dell 1999; Rigliaco et al. 2009;Haworth & Owen 2020), indirect evidence based on theexpected impacts on more easily observable disk proper-ties is growing (e.g., Fang et al. 2012; Mann et al. 2014;Kim et al. 2016; Guarcello et al. 2016; Haworth et al.2017; van Terwisga et al. 2019b). In particular, usingALMA to estimate disk masses, Ansdell et al. (2017)found in σ Orionis a dearth of massive ( (cid:38) M Jup ) diskswithin ∼ λ Orionis, there is a lackof even low-mass ( (cid:38) M Jup ) disks within ∼ µ m, at which point they are toolarge to be affected by external photoevaporation andinstead experience rapid inward radial drift (Haworthet al. 2018; Sellek et al. 2020). Thus any detectabledistance-dependent trend in disk mass may have beenwashed out due to the dynamical evolution of the clus-ter and the evolution of the disk dust population probedby ALMA for a couple Myr beyond the age of σ Orionis.6.3.
Outlier Disks in Evolved Regions
The handful of relatively nearby SFRs (at (cid:46)
400 pc;the distance of Orion or closer) with evolved ages ( ∼ λ Ori-onis, Upper Sco, and the TW Hya Association (TWA).Each of these evolved regions contains at least one long-lived primordial disk that remains much more massivethan the rest of the surviving disk population. Youngerregions, in contrast, exhibit a smooth distribution indisk masses, regardless of whether they are in low-mass4
Ansdell et al. (e.g., Lupus) or high-mass (e.g., OMC-2) SFRs. This isillustrated in Figure 5, which shows the upper 32% of thedisk dust mass distributions in different SFRs (32% waschosen to match the detection rate in our λ Orionis sur-vey). In Figure 5, all of the SFRs exhibit similar slopesuntil the highest-mass disks, at which point the distribu-tions of the evolved Upper Sco and λ Orionis regions risesteeply due to their outlier disks (we do not show TWAin Figure 5, as ALMA surveys of its disk population arenot yet published). Additionally, the most massive disksin each SFR, regardless of age or environment, all have M dust ∼ M ⊕ , which suggests that the outlier disksin the evolved regions were not distinct from birth, butrather that some mechanism stopped the otherwise nat-ural removal of disk material (e.g., by accretion onto thestar and/or winds) as the disks evolved.A potential explanation is that one or more(sub-)stellar companions formed in these disks, on orbitsthat are inhibiting the dispersal of outer disk materialwhile also avoiding disk truncation/disruption, result-ing in particularly long-lived primordial disks. A single,massive super-Jupiter on an initially circular orbit mayexcite its own eccentricity (e.g., D’Angelo et al. 2006),enabling the clearing of a large inner cavity while pre-serving the outer disk; this has been demonstrated forthe ∼ (cid:46) λ Orionis, UpperSco, and TWA show evidence of such companions thatcould explain their extended primordial disk lifetimes. For HD 245185 in λ Orionis, Kama et al. (2015) placedit in the population of Herbig Ae/Be stars whose stel-lar abundances are depleted in refractory elements whilehosting warm/flared transition disks, as opposed to thepopulation with solar abundances hosting cold/flat fulldisks. They explained the chemical peculiarity of theformer population with giant planets filtering out dustymaterial as it flows through the disk, resulting in the ac-cretion of high gas-to-dust ratio material onto the star.These massive planets would then also be responsiblefor clearing out the large gap or cavity in the disk duststructure that defines the transition disk classification(Espaillat et al. 2014). Although our ALMA observa-tions do not resolve any gaps or cavities in the diskaround HD 245185 (Figure 3), the resolution is poor( ∼
60 au in radius). As the SED of HD 245185 exhibitsa mid-infrared dip (Ansdell et al. 2015), future ALMAobservations may still resolve structure in the disk.For Upper Sco, one of the most massive disks isaround 2MASS J16042165-2130284 (also known simplyas J1604), a negligible accretor (Manara et al. 2020) ofK2 spectral type that hosts a face-on transition disk witha large inner cavity seen in millimeter emission (Math-ews et al. 2012; Ansdell et al. 2020). Evidence pointsto one or more high-mass planetary companions clear-ing out the cavity as well as misaligning an unseen innerdisk component that casts variable shadows on the outerdisk detected in scattered light (e.g., Takami et al. 2014;Pinilla et al. 2018). This misaligned inner disk is alsothought to cause the “dipper” variability observed inspace-based light curves (e.g., Cody & Hillenbrand 2018;Ansdell et al. 2020). Another outlier disk in Upper Scois around the G-type star 2MASS J15583692-2257153(or HD 143006), which also hosts a face-on transitiondisk exhibiting “dipper” variability (Ansdell et al. 2020).This disk has a large gap, several narrow rings/gaps,and a bright asymmetry resolved by ALMA (Andrewset al. 2018b) suggesting a warped inner disk driven by alow-mass stellar or high-mass planetary companion, po-tentially orbiting at a few au (P´erez et al. 2018). Finally,the most massive disk in Upper Sco is around 2MASSJ16113134-1838259 (or AS 205); this is not a transitiondisk, but displays clear spirals arms detected by ALMA,indicative of strong dynamical interactions induced bya known external companion (Kurtovic et al. 2018) thatmay be dominating its disk structure and evolution.The outlier disk in TWA is around TW Hya itself.It also has an inner disk warp, this time detected in thegas kinematics (Rosenfeld et al. 2012), which could orig-inate from a massive planetary companion orbiting thisK-type star (e.g., Facchini et al. 2014). Extremely high-resolution ALMA observations have detected a gap at
LMA Survey of λ Orionis Disks (cid:46) ∼ in situ , preserving the outerregions and resulting in an outlier disk. SUMMARYWe present an ALMA Band 6 survey of the proto-planetary disk population in the λ Orionis OB cluster.This region is important for studying disk evolution andplanet formation due to its evolved age and the recentsupernova that may have occurred in its core. Our keyfindings are as follows: • The millimeter emission from λ Orionis disks isweak, but not particularly unusual given the clus-ter’s evolved age of ∼ µ Jy median rms corresponding to 3 σ dust mass upper limits of ∼ M ⊕ . Stacking the 30non-detections gives a 4 σ mean signal of 20 µ Jy( ∼ M ⊕ ), indicating that deeper observationsshould produce more detections. Only 5 disks arealso detected in the CO line (Figure 3); how-ever, the lack of gas detections is consistent withthe weak continuum emission, based on the cor-relation between millimeter continuum and COemission seen in younger regions (Figure 4). • The effects of massive stars, in the form of pre-supernova feedback and/or a supernova event it-self, do not appear to significantly reduce the over-all planet-forming capacity of a population of pro-toplanetary disks that is already a few Myr intoevolution. This is based on comparing the diskmass distribution in λ Orionis to that of otherSFRs, in particular the similarly aged Upper Scoassociation (Figures 5 and 6). One explanationis that supernovae are only effective at strippingmass from nearby disks that are within a smallfraction of a parsec. Additionally, pre-supernovafeedback may sculpt low-density channels in theintercluster gas, through which energy can moreeasily escape, significantly muting the impact ofsupernovae events on the surrounding disk popu-lation. However, more work is needed to confirmthe occurrence of the supernova event in λ Orionisand/or determine whether a recent supernova alsooccurred in Upper Sco. • Massive “outlier” disks lingering in evolved ( ∼ λ Orionis, Upper Sco, andTWA show evidence for one or more (sub-)stellarcompanions. Because these massive disks wouldnot be considered outliers in younger ( ∼ λ Orionis willbuild up larger numbers of continuum and line detec-tions to improve our constraints on population-levelstatistics. Understanding the disk population of thecluster members outside of the
Spitzer survey area willalso ensure that the ALMA survey is complete. Ob-taining better constraints on the stellar and accretion6
Ansdell et al. properties of the disk-hosting stars in λ Orionis withwide-band and/or high-resolution spectra will allow usto search for trends seen in other regions, as well asany evidence for external photoevapoaration and viscousevolution. Deeper observations with ALMA of multiplemolecular lines tracing the gas content and chemistrywill help determine whether our observations reflect diskgas dispersal or the transformation of CO due to en-hanced cosmic ray ionization from the supernova. De-tailed theoretical studies may also provide insight intothe peculiar kinematics of λ Orionis and its links to thestar-formation history of the region; similar studies inUpper Sco are also important, in particular to assess thepossibility of a recent supernova having also occurred inthat region. Indeed, λ Orionis still has much to teach usabout perhaps one of the most common types of planet-forming environments in the nearby Galaxy.ACKNOWLEDGMENTSMA and EC acknowledge support from NASA grantNNH18ZDA001N/EW. TJH is funded by a Royal So-ciety Dorothy Hodgkin Fellowship. This project hasreceived funding from the European Union’s Horizon2020 research and innovation programme under theMarie Sklodowska-Curie grant agreement No 823823(DUSTBUSTERS). This work was partly supportedby the Deutsche Forschungs-Gemeinschaft (DFG, Ger-man Research Foundation) - Ref no. FOR 2634/1TE 1024/1-1. This work makes use of the followingALMA data: ADS/JAO.ALMA
Facilities:
ALMA,
Gaia
Software: astropy (Astropy Collaboration et al.2013, 2018),
ASURV (Lavalley et al. 1992),
CASA (Mc-Mullin et al. 2007), matplotlib (Hunter 2007)
LMA Survey of λ Orionis Disks A. ALMA OBSERVING LOGTable 3 summarizes the 13 execution blocks that collected the ALMA observations of our λ Orionis disk sample(Section 4.1). It provides the date and time of the observations in Coordinated Universal Time (UTC Date), number ofantennae used ( N ant ) and their baseline range ( L base ), precipitable water vapor (PWV) at the time of the observations,and names of the bandpass, flux, and gain calibrators applied in the pipeline calibration. Table 3.
ALMA Observing Log
UTC Date N ant L base PWV Calibrators(End Time) (m) (mm) (Bandpass/Flux, Phase)2018-09-09 44 15–1213 2.1 J0423 − Ansdell et al. B. ALMA CO SPECTRAFigure 8 presents the ALMA Band 6 CO spectra for all 44 sources in our λ Orionis disk sample. Extraction of thespectra is described in Sections 4.3.
LO 65
LO 1079
LO 1152
LO 1359
LO 1589
LO 1624
LO 1840
LO 2088
LO 7957
LO 2712
LO 2989
LO 2993
LO 3360
LO 3506
LO 3597
LO 3746
LO 7951
LO 3785
HD 245185
LO 3887
LO 3942
LO 4021
LO 4111
LO 4126
LO 4155
LO 4163
LO 4187
LO 4255
LO 4363
LO 4407
LO 4520
LO 4531
LO 4817
LO 4916
LO 5267
LO 5447
LO 5679
LO 5916
LO 6191
LO 6866
Velocity (km/s) -505 F l u x ( m J y ) LO 6886
LO 7402
LO 7490
LO 7528
Figure 8.
ALMA Band 6 CO spectra for λ Orionis disks. Horizontal dashed red lines indicate 3 × the median channelrms, while horizontal dotted black lines indicate the zero flux level. Vertical dashed lines are the source v LSR , when available(Table 1). Clear detections ( > σ ) are highlighted in purple, while marginal detections ( ∼ σ ) are indicated with orange. LMA Survey of λ Orionis Disks
Adams, F. C. 2010, ARA&A, 48, 47,doi: 10.1146/annurev-astro-081309-130830Alecian, E., Wade, G. A., Catala, C., et al. 2013, MNRAS,429, 1027, doi: 10.1093/mnras/sts384Alexander, R. 2012, ApJL, 757, L29,doi: 10.1088/2041-8205/757/2/L29Andrews, S. M. 2015, PASP, 127, 961, doi: 10.1086/683178—. 2020, arXiv e-prints, arXiv:2001.05007.https://arxiv.org/abs/2001.05007Andrews, S. M., Rosenfeld, K. A., Kraus, A. L., & Wilner,D. J. 2013, ApJ, 771, 129,doi: 10.1088/0004-637X/771/2/129Andrews, S. M., Terrell, M., Tripathi, A., et al. 2018a, ApJ,865, 157, doi: 10.3847/1538-4357/aadd9fAndrews, S. M., & Williams, J. P. 2005, ApJ, 631, 1134,doi: 10.1086/432712Andrews, S. M., Wilner, D. J., Zhu, Z., et al. 2016, ApJL,820, L40, doi: 10.3847/2041-8205/820/2/L40Andrews, S. M., Huang, J., P´erez, L. M., et al. 2018b,ApJL, 869, L41, doi: 10.3847/2041-8213/aaf741Ansdell, M., Williams, J. P., & Cieza, L. A. 2015, ApJ, 806,221, doi: 10.1088/0004-637X/806/2/221Ansdell, M., Williams, J. P., Manara, C. F., et al. 2017, AJ,153, 240, doi: 10.3847/1538-3881/aa69c0Ansdell, M., Williams, J. P., van der Marel, N., et al. 2016,ApJ, 828, 46, doi: 10.3847/0004-637X/828/1/46Ansdell, M., Williams, J. P., Trapman, L., et al. 2018, ApJ,859, 21, doi: 10.3847/1538-4357/aab890Ansdell, M., Gaidos, E., Hedges, C., et al. 2020, MNRAS,492, 572, doi: 10.1093/mnras/stz3361Astropy Collaboration, Robitaille, T. P., Tollerud, E. J.,et al. 2013, A&A, 558, A33,doi: 10.1051/0004-6361/201322068Astropy Collaboration, Price-Whelan, A. M., Sip˝ocz, B. M.,et al. 2018, AJ, 156, 123, doi: 10.3847/1538-3881/aabc4fBailer-Jones, C. A. L., Rybizki, J., Fouesneau, M.,Mantelet, G., & Andrae, R. 2018, AJ, 156, 58,doi: 10.3847/1538-3881/aacb21Baraffe, I., Homeier, D., Allard, F., & Chabrier, G. 2015,A&A, 577, A42, doi: 10.1051/0004-6361/201425481Barenfeld, S. A., Carpenter, J. M., Ricci, L., & Isella, A.2016, ApJ, 827, 142, doi: 10.3847/0004-637X/827/2/142Bayo, A., Barrado, D., Stauffer, J., et al. 2011, A&A, 536,A63, doi: 10.1051/0004-6361/201116617Beckwith, S. V. W., Sargent, A. I., Chini, R. S., & Guesten,R. 1990, AJ, 99, 924, doi: 10.1086/115385Bell, C. P. M., Naylor, T., Mayne, N. J., Jeffries, R. D., &Littlefair, S. P. 2013, MNRAS, 434, 806,doi: 10.1093/mnras/stt1075 Booth, A. S., & Ilee, J. D. 2020, MNRAS, 493, L108,doi: 10.1093/mnrasl/slaa014Bosman, A. D., Walsh, C., & van Dishoeck, E. F. 2018,A&A, 618, A182, doi: 10.1051/0004-6361/201833497Bowler, B. P. 2016, PASP, 128, 102001,doi: 10.1088/1538-3873/128/968/102001Cazzoletti, P., Manara, C. F., Baobab Liu, H., et al. 2019,A&A, 626, A11, doi: 10.1051/0004-6361/201935273Cieza, L. A., Ru´ız-Rodr´ıguez, D., Hales, A., et al. 2019,MNRAS, 482, 698, doi: 10.1093/mnras/sty2653Clarke, C. J. 2007, MNRAS, 376, 1350,doi: 10.1111/j.1365-2966.2007.11547.xCleeves, L. I., Bergin, E. A., & Adams, F. C. 2014, ApJ,794, 123, doi: 10.1088/0004-637X/794/2/123Close, J. L., & Pittard, J. M. 2017, MNRAS, 469, 1117,doi: 10.1093/mnras/stx897Cody, A. M., & Hillenbrand, L. A. 2018, AJ, 156, 71,doi: 10.3847/1538-3881/aaceadConcha-Ram´ırez, F., Wilhelm, M. J. C., Portegies Zwart,S., & Haworth, T. J. 2019, MNRAS, 490, 5678,doi: 10.1093/mnras/stz2973Cumming, A., Butler, R. P., Marcy, G. W., et al. 2008,PASP, 120, 531, doi: 10.1086/588487Cunha, K., & Smith, V. V. 1996, A&A, 309, 892Dale, J. E., Ngoumou, J., Ercolano, B., & Bonnell, I. A.2014, MNRAS, 442, 694, doi: 10.1093/mnras/stu816D’Angelo, G., Lubow, S. H., & Bate, M. R. 2006, ApJ, 652,1698, doi: 10.1086/508451Davidson-Pilon, C., Kalderstam, J., Jacobson, N., et al.2020, CamDavidsonPilon/lifelines: v0.24.15, v0.24.15,Zenodo, doi: 10.5281/zenodo.3934629de Geus, E. J. 1992, A&A, 262, 258Debes, J. H., Poteet, C. A., Jang-Condell, H., et al. 2017,ApJ, 835, 205, doi: 10.3847/1538-4357/835/2/205Diplas, A., & Savage, B. D. 1994, ApJS, 93, 211,doi: 10.1086/192052Dolan, C. J., & Mathieu, R. D. 1999, AJ, 118, 2409,doi: 10.1086/301075—. 2001, AJ, 121, 2124, doi: 10.1086/319946Eistrup, C., Walsh, C., & van Dishoeck, E. F. 2016, A&A,595, A83, doi: 10.1051/0004-6361/201628509Espaillat, C., Muzerolle, J., Najita, J., et al. 2014, inProtostars and Planets VI, ed. H. Beuther, R. S. Klessen,C. P. Dullemond, & T. Henning, 497,doi: 10.2458/azu uapress 9780816531240-ch022Facchini, S., Clarke, C. J., & Bisbas, T. G. 2016, MNRAS,457, 3593, doi: 10.1093/mnras/stw240Facchini, S., Ricci, L., & Lodato, G. 2014, MNRAS, 442,3700, doi: 10.1093/mnras/stu1149 Ansdell et al.
Fang, M., van Boekel, R., King, R. R., et al. 2012, A&A,539, A119, doi: 10.1051/0004-6361/201015914Fazio, G. G., Hora, J. L., Allen, L. E., et al. 2004, ApJS,154, 10, doi: 10.1086/422843Folsom, C. P., Bagnulo, S., Wade, G. A., et al. 2012,MNRAS, 422, 2072,doi: 10.1111/j.1365-2966.2012.20718.xFung, J., & Chiang, E. 2017, ApJ, 839, 100,doi: 10.3847/1538-4357/aa6934Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al.2016, A&A, 595, A1, doi: 10.1051/0004-6361/201629272Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al.2018, A&A, 616, A1, doi: 10.1051/0004-6361/201833051Galli, P. A. B., Bouy, H., Olivares, J., et al. 2020, arXive-prints, arXiv:2001.05190.https://arxiv.org/abs/2001.05190Greaves, J. S., & Rice, W. K. M. 2010, MNRAS, 407, 1981,doi: 10.1111/j.1365-2966.2010.17043.xGrenier, I. A., Black, J. H., & Strong, A. W. 2015, ARA&A,53, 199, doi: 10.1146/annurev-astro-082214-122457Guarcello, M. G., Drake, J. J., Wright, N. J., et al. 2016,arXiv e-prints, arXiv:1605.01773.https://arxiv.org/abs/1605.01773Haworth, T. J., Boubert, D., Facchini, S., Bisbas, T. G., &Clarke, C. J. 2016, MNRAS, 463, 3616,doi: 10.1093/mnras/stw2280Haworth, T. J., Facchini, S., Clarke, C. J., & Cleeves, L. I.2017, MNRAS, 468, L108, doi: 10.1093/mnrasl/slx037Haworth, T. J., Facchini, S., Clarke, C. J., & Mohanty, S.2018, MNRAS, 475, 5460, doi: 10.1093/mnras/sty168Haworth, T. J., & Owen, J. E. 2020, MNRAS, 492, 5030,doi: 10.1093/mnras/staa151Henney, W. J., & O’Dell, C. R. 1999, AJ, 118, 2350,doi: 10.1086/301087Herczeg, G. J., & Hillenbrand, L. A. 2015, ApJ, 808, 23,doi: 10.1088/0004-637X/808/1/23Hern´andez, J., Calvet, N., Hartmann, L., et al. 2009, ApJ,707, 705, doi: 10.1088/0004-637X/707/1/705Hern´andez, J., Morales-Calderon, M., Calvet, N., et al.2010, ApJ, 722, 1226,doi: 10.1088/0004-637X/722/2/1226Hern´andez, J., Calvet, N., Brice˜no, C., et al. 2007, ApJ,671, 1784, doi: 10.1086/522882Hildebrand, R. H. 1983, QJRAS, 24, 267Hollenbach, D., Johnstone, D., Lizano, S., & Shu, F. 1994,ApJ, 428, 654, doi: 10.1086/174276Hoogerwerf, R., de Bruijne, J. H. J., & de Zeeuw, P. T.2000, ApJL, 544, L133, doi: 10.1086/317315—. 2001, A&A, 365, 49, doi: 10.1051/0004-6361:20000014 Hunter, J. D. 2007, Computing In Science & Engineering,9, 90, doi: 10.1109/MCSE.2007.55Indriolo, N., Blake, G. A., Goto, M., et al. 2010, ApJ, 724,1357, doi: 10.1088/0004-637X/724/2/1357Johnstone, D., Hollenbach, D., & Bally, J. 1998, ApJ, 499,758, doi: 10.1086/305658Kama, M., Folsom, C. P., & Pinilla, P. 2015, A&A, 582,L10, doi: 10.1051/0004-6361/201527094Kenyon, S. J., & Hartmann, L. 1995, ApJS, 101, 117,doi: 10.1086/192235Kim, J. S., Clarke, C. J., Fang, M., & Facchini, S. 2016,ApJL, 826, L15, doi: 10.3847/2041-8205/826/1/L15Kounkel, M., Covey, K., Su´arez, G., et al. 2018, AJ, 156,84, doi: 10.3847/1538-3881/aad1f1Kurtovic, N. T., P´erez, L. M., Benisty, M., et al. 2018,ApJL, 869, L44, doi: 10.3847/2041-8213/aaf746Lang, W. J., Masheder, M. R. W., Dame, T. M., &Thaddeus, P. 2000, A&A, 357, 1001Lavalley, M., Isobe, T., & Feigelson, E. 1992, AstronomicalSociety of the Pacific Conference Series, Vol. 25, ASURV:Astronomy Survival Analysis Package, ed. D. M. Worrall,C. Biemesderfer, & J. Barnes, 245Le Petit, F., Ruaud, M., Bron, E., et al. 2016, A&A, 585,A105, doi: 10.1051/0004-6361/201526658Lee, D., Seon, K.-I., & Jo, Y.-S. 2015, ApJ, 806, 274,doi: 10.1088/0004-637X/806/2/274Lucas, W. E., Bonnell, I. A., & Dale, J. E. 2020, MNRAS,493, 4700, doi: 10.1093/mnras/staa451Luhman, K. L., & Esplin, T. L. 2020, AJ, 160, 44,doi: 10.3847/1538-3881/ab9599Lynden-Bell, D., & Pringle, J. E. 1974, MNRAS, 168, 603,doi: 10.1093/mnras/168.3.603Maddalena, R. J., & Morris, M. 1987, ApJ, 323, 179,doi: 10.1086/165818Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al.2017, AJ, 154, 94, doi: 10.3847/1538-3881/aa784dManara, C. F., Morbidelli, A., & Guillot, T. 2018a, A&A,618, L3, doi: 10.1051/0004-6361/201834076Manara, C. F., Prusti, T., Comeron, F., et al. 2018b, A&A,615, L1, doi: 10.1051/0004-6361/201833383Manara, C. F., Natta, A., Rosotti, G. P., et al. 2020, arXive-prints, arXiv:2004.14232.https://arxiv.org/abs/2004.14232Mann, R. K., Di Francesco, J., Johnstone, D., et al. 2014,ApJ, 784, 82, doi: 10.1088/0004-637X/784/1/82Mathews, G. S., Williams, J. P., & M´enard, F. 2012, ApJ,753, 59, doi: 10.1088/0004-637X/753/1/59Mathieu, R. 2015, λ Ori: A Case Study in Star Formation,147–163, doi: 10.1007/978-3-662-47290-3 11
LMA Survey of λ Orionis Disks Mathieu, R. D. 2008, The λλ