RNO 54: A Previously Unappreciated FU Ori Star

We present evidence in support of the hypothesis that the young stellar object RNO 54 is a mature-stage FU Ori type source. The star was first cataloged as a ``red nebulous object"in the 1980s but appears to have undergone its outburst prior to the 1890s. Present-day optical and near-infrared spectra are consistent with those of other FU Ori type stars, both in the details of spectral line presence and shape, and in the overall change in spectral type from an FGK-type in the optical, to the M-type presented in the near-infrared. In addition, the spectral energy distribution of RNO 54 is well-fit by a pure-accretion disk model with parameters: $\dot{M} = 10^{-3.45\pm0.06}$ $M_\odot$ yr$^{-1}$, $M_* = 0.23\pm0.06 \ M_\odot$, and $R_\mathrm{inner} = 3.68\pm0.76 \ R_\odot$, though we believe $R_\mathrm{inner}$ is likely close to its upper range of $4.5 R_\odot$ in order to produce a $T_\mathrm{max} = 7000$ K that is consistent with the optical to near-infrared spectra. The resulting $L_\mathrm{acc}$ is $\sim 265 \ L_\odot$. To find these values, we adopted a source distance $d=1400$ pc and extinction $A_V=3.9$ mag, along with disk inclination $i=50$ deg based on consideration of confidence intervals from our initial disk model, and in agreement with observational constraints. The new appreciation of a well-known source as an FU Ori type object suggests that other such examples may be lurking in extant samples.


INTRODUCTION
FU Ori stars are a rare breed of young stellar object (YSO); see Herbig (1977); Hartmann & Kenyon (1996) and Fischer et al. (2022) for reviews.Like other YSOs, FU Ori stars have significant luminosity in the infrared to millimeter portion of their spectral energy distributions, and they can be closely associated with optical and/or near-infrared nebulae.Such observational features are generally consistent with a young star that is surrounded by a substantial amount of circumstellar dust, distributed over of tens to thousands of astronomical units, that often heavily obscures the central proto-or pre-main sequence star.
However, beyond the standard YSO features, members of the FU Ori class have certain additional observational characteristics that distinguish them.Specifically, FU Ori optical spectral types are those of FGK-type giants to supergiants, while near-infrared spectra exhibit the features of M-type giants to supergiants.The distinctive wavelengthdependent temperature pattern, as well as the broad and "flat-bottomed" absorption line profiles in FU Ori photospheres are such that explanations such as starspots or a multiple star system, can be discarded.In addition, all FU Ori type stars show strong wind/outflow signatures as evidenced by asymmetrically blue-shifted absorption or P Cygni profiles in certain lines.
The interpretion that FU Ori objects are YSOs in a state of prolonged accretion outburst is commonly accepted.During initiation of the outburst, a protoplanetary disk surrounding a young star increases its rate of mass transfer from the disk onto the central star, transitioning from low-state accretion within the broad range of ∼ 10 −8 to 10 −11 M ⊙ yr −1 , to enhanced levels in the high-state approaching 10 −4 to 10 −5 M ⊙ yr −1 .FU Ori spectral energy distributions (SEDs) and even more convincingly, their high-dispersion spectra (Kenyon et al. 1988;Welty et al. 1992;Rodriguez & Hillenbrand 2022;Carvalho et al. 2023b;Liu et al. 2022), can both be reproduced by pure-accretion disk models having a temperature and a velocity gradient with radius.
The prototype, FU Ori, and several other members of the FU Ori class (e.g.HBC 722) are located in or near well-known star forming regions.Other examples (e.g.Gaia 17bpi) are more isolated and associated with small dark clouds but not significant star forming regions.
In this paper we present evidence regarding a relatively isolated YSO that was documented as such in the literature decades ago, but appears to have been overlooked all of these years as a bona fide FU Ori type star.Although no accretion outburst was observed in RNO 54, the object is empirically similar to classical members of the class, and both its spectral energy distribution and its high-dispersion spectrum can be well-fit by an accretion disk model.RNO 54 is therefore a neglected, but likely FU Ori star.This same conclusion was independently reached in a recent paper by Magakian et al. (2023).

BACKGROUND ON THE YOUNG STELLAR OBJECT RNO 54
RNO 54, also known as PDS 120, IRAS 05393+2235 and GN 05.39.2, is located at 05:42:21.24+22:36:47.1 (J2000.).The parallax, proper motions, and radial velocity of RNO 54 seem to make it kinematically isolated from nearby stars on the sky.No kinematically similar sources could be found in Gaia DR3 data (Gaia Collaboration et al. 2023).Furthermore, no young stellar objects could be identified in nearby projection in SIMBAD (Wenger et al. 2000).RNO 54 thus appears to be a relatively isolated YSO.
RNO 54 has a flat-spectrum type Class I spectral energy distribution.It also has a cometary or ring-like nebula similar to those associated with the YSOs V1515 Cyg, FU Ori, and Parsamian 21, among others, but differently shaped than the conical nebulae exhibited by e.g.PV Cep, R Mon and V1647 Ori.Although its sky location is just north of the greater Orion large star-forming complex, towards the galactic plane, RNO 54 has a Gaia parallax that corresponds to a much further (3.5 times) distance of 1.40 ± 0.05 kpc.While the error bar is only a few percent, the somewhat high RUWE value of 1.4 could make the kinematic parameters untrustworthy, but is also explainable by the nebulous nature of the source.Other Gaia-derived parameters include log T ef f = 6700 K and log g = 2.32, and A o (similar to A V ) = 4.21 mag.Besides the temperature and gravity expected for an FU Ori type object, RNO 54 also has a high luminosity reported by Gaia.The LU M − F LAM E value is 780 L ⊙ , which is derived from the observed Gaia G magnitude, the implied A G extinction, and a bolometric correction appropriate to the spectral type.
In the literature, RNO 54 was first noted by Cohen (1980) as a Red Nebulous Object with cometary morphology, that was bright in the near-infrared and had Hα emission.Cohen (1980) assigned a spectral type of F5 II and an extinction A V = 3.8 ± 0.3 mag.Goodrich (1987) suggested that nebulae, which had already been associated with FU Ori objects by Herbig (1977), could be used as signposts to identify additional sources that might be "either postor pre-outburst FU Orionis stars."Indeed, Goodrich (1987) specifically mentioned RNO 54 as an example, and also provided spectra typed as early G Ib or II.Later, Torres et al. (1995) included RNO 54 in a table of "probable post-FU Ori stars", calling it an F8 II and noting its Hα and Li I 6707 Å profiles.
No actual outburst has been detected for RNO 54, but if indeed an FU Ori star, the outburst is constrained to have occurred prior to its detection on the epoch 1951.85Palomar sky survey plate used by Cohen (1980).The outburst was also likely prior to the Astrographic Catalog, for which the position is reported by Fresneau (1983) at epoch 1893.08 with photographic (blue) magnitude 13.0, close to the current value.
Since these early papers, RNO 54 has been included only in survey studies, namely the Magakian (2003) catalog of reflection nebulae, the Lumsden et al. (2013) Red MSX Source Survey, an OH maser catalog by Engels & Bunzel (2015), and the López et al. (2021) survey of large-scale Herbig Haro flows associated with IRAS-identified young stellar objects.As this manuscript was being prepared, a single-object study by Magakian et al. (2023) appeared.These authors come to the same conclusion we do about RNO 54: that it is very likely an FU Ori type star.Both papers reinforce the original speculation of Goodrich (1987).

NEW SPECTRAL DATA AND COMPARISON TO BENCHMARK FU ORI STARS
Based on a re-read of the Goodrich (1987) and Cohen (1980) papers undertaken as part of general scholarship, we had suspicions about the nature of RNO 54, and sought to investigate whether modern data could confirm an FU Ori interpretation.During fall of 2020, we thus obtained optical and near-infrared spectra in order to investigate this hypothesis in detail.
Figure 1.Optical spectrum of RNO 54 (magenta) compared to the FU Ori objects HBC 722 (PTF 10qpf), V1057 Cyg, and V1515 Cyg, and to the continuum-plus-emission objects V1331 Cyg and V2492 Cyg (PTF 10nvg).RNO 54 has an absorption spectrum similar to the broad-line FU Ori stars V1057 Cyg and HBC 722 (PTF 10qpf); V1515 Cyg has narrower and deeper lines than the others, due to its nearly face-on inclination.RNO 54 thus appears to be an extreme accretor, and does not have the emission line spectrum of mere atypically-high accretion rate sources.

Data Acquisition and Reduction
Our optical spectrum was obtained using the W.M. Keck Observatory and Echellette Spectrograph and Imager (ESI Sheinis et al. 2002) by J. van Roestel on 2020-09-12 UT in 1.2 ′′ seeing.The 1.0 ′′ slit was used with 2-pixel spatial binning, yielding an effective spectral resolution of R ≈ 7, 500 in a single 300 sec integration.Data reduction was kindly performed by T. Kupfer using the MAKEE package 1 written by T. Barlow.The one-dimensional spectra cover the range ≈ 3, 900 − 10, 000 Å with some overlap between orders.
Our infrared spectrum was obtained using the NASA's IRTF and SpeX instrument Rayner et al. (2003) by K. De on 2020-09-22 UT.The 0.5 ′′ slit was used yielding an effective spectral resolution of R ≈ 1, 200 in the SXD mode, covering ∼ 0.7 − 2.5 µm.Dithered exposures were used to obtain a total integration time of 180 sec on source.The spectral images were reduced using the Spextool package Cushing et al. (2004) with flux calibration performed using the XT ellCor package Vacca et al. (2003).

Basic Spectral Analysis
A segment of the optical spectrum appears in Figure 1  Our optical and near-infrared spectra of RNO 54 enable investigation of properties of the "photosphere", which appears to be dominated by absorption lines produced in an accretion disk.We can also study the spectral signatures of outflowing material in winds.
The Keck/ESI and the IRTF/SpeX spectra were taken ten days apart, and are essentially identical in the wavelength range of their overlap when convolved to the same spectral resolution.There is no significant variability in either absorption or emission line strengths and profiles.Comparing our data with spectral standards and with model atmospheres implies a heliocentric radial velocity of about +15 km/s for RNO 54.
Source extinction can also be probed spectroscopically, given that the optical spectrum of RNO 54 displays clear DIB absorption at 5850, 6270, 6379, and 6614 Å. Relations established in Carvalho & Hillenbrand (2022) can be applied to the 6614 Å DIB strength of 332 ± 2 m Å, to infer a reddening value of E(B-V) = 1.35 mag and thus a visual extinction estimate of A V = 4.2 mag.We note that this is identical to the Gaia DR3 value quoted above.

Signatures of an FU Ori Type Spectrum
The spectrum of RNO 54 shows both strong wind and outflow signatures evident as P Cygni type lines, and symmetrically broadened disk-like absorption lines.Both features are associated with bona fide members of the FU Ori class of YSOs.

Disk Lines
Figure 3 demonstrates that the optical spectrum of RNO 54 is close in appearance to the recent FU Ori outburst sources Gaia 17bpi (Hillenbrand et al. 2018) and Gaia 18dvy (Szegedi-Elek et al. 2020), and unlike the low-state accreting classical T Tauri star BP Tau.The latter has more low-excitation absorption lines from species like Ti I and Fe I visible, whereas the three FU Ori stars (counting RNO 54) have low-excitation lines as well as strong absorption from high-excitation, hotter lines such as Fe II, seen especially towards the bluer range of spectrum (4000-5500 Å).BP Tau has some of these same lines in emission, e.g.Fe II 4924, 5018, 5169 Å.In addition, BP Tau shows considerably less broadening than the FU Ori sources, with a v sin i ≈ 13 km/s that is typical for a T Tauri star.For RNO 54, absorption widths are around 85 km/s (HWHD) for the disk lines.
The optical photosphere of RNO 54 (Figure 1, top panel of Figure 2, and Figure 3) shows lines like Ca I, Mg I, and Fe I that are typically seen in cool photospheres.However, the lines are shallow (few percent depths) and broadened in a non-gaussian way, more like line-splitting, which can be seen even in the only moderate spectral dispersion of our ESI spectrum.Furthermore, there are many intermediate excitation lines from Ni I, Fe I, and Fe II (e.g.5316, 6516 Å), and clear indications of an even hotter spectral contributions as well.Absorption is obvious in many lines with excitation potential (EP) of 5-10 eV, for example Mg II (e.g.4481, 7877, 7896 Å), Si I (e.g.6518, 8752 Å) with 5.0 eV, the Ca II 8912, 8927 Å doublet2 as well as 9890 Å (7.0 eV) and C I lines in the 9061-9112 Å range (7.5 eV).At the same time, there is a cool contribution as evidenced by weak TiO bandhead absorption at 8859 and 9209 Å.Neither the hotter atomic lines nor the cooler molecules are seen in typical T Tauri stars, such as BP Tau.While Figures 1 and 3 demonstrate many of these lines; not all are illustrated.
The infrared spectrum of RNO 54 (Figure 2) shows clear molecular band absorption in TiO (Y band), CN (Y band), VO (Y and J bands), H 2 O (J, H, and K bands), and CO (H and K bands).There are also atomic lines such as Na I, Ca I, Al I, Mg I and Si I. Notable are the higher-excitation atomic lines from Ca II and C I (around 9000 Å).
Consistent with other FU Ori stars displaying a mixed spectrum with both cooler and hotter opacity contributors, there is also strong evidence for low gravity.In the optical, this is indicated by enhanced Ba II 6141 and 6496 Å absorption, lines with positive luminosity effect (Andrievsky 1998), as well as the strengths of several Ti I lines and the ratio of Fe II 5316 Å to Fe I 5328 Å that are highlighted by Carvalho et al. (2023a) as gravity-sensitive.In the infrared, important indicators of low surface gravity are the Sr II lines in Y-band (Sharon et al. 2010), CN in J-band (Wallace et al. 2000), and the strong Si I line in the H-band (Meyer et al. 1998).
Finally, RNO 54 clearly shows Li I 6707 Å (see Figure 4), a line which indicates stellar youth and is detected in most young stellar objects (excepting those with very high veiling due to rapid accretion).Li I 6707 Å is ubiquitously present in the extreme accretor FU Ori stars, and usually has a disk-like broadening often with additional kinematic signatures, perhaps from outflow. Figure 3. Portions of the optical spectrum of RNO 54 (magenta) compared to the recent FU Ori outburst sources Gaia 18dvy (blue) and Gaia 17bpi (orange), as well as to a non-outbursting but heavily accreting classical T Tauri star, BP Tau (green).Spectra have been normalized and offset, for clarity.The lower three objects are similar in terms of their line presence, depth, and width, exhibiting absorption from a mix of low-excitation and higher-excitation neutral and ionized species.The top spectrum, by contrast, is less complicated and while similar to the FU Ori objects over small spectral ranges, has fewer high-excitation lines, somewhat deeper lines, and much narrower lines.

Wind/Outflow Lines
As shown in Figure 4, RNO 54 exhibits Hα emission with a P Cygni type profile, indicative of wind.The 10% width of the emission part of the Hα line is 170 km/s.Blueshifted absorption dips are seen at about -90 and -175 km/s, which also appear at a more muted level in the Ca II triplet line profiles.The "double-dip" structure on the blue side is not atypical for FU Ori type sources.The higher Balmer series lines are in absorption, with asymmetric profiles having extra blueshifted absorption.
Wind signature is also seen in the absorption profiles of certain metal lines, including the deep Na I D doublet at 55% of continuum, the K I 7669,7699 Å doublet, and the O I 7773 Å triplet.Furthermore, there is weak evidence in RNO 54 for hot wind lines, such as Si II 6347,6371 Å (8.1 eV) that are prominent in early-stage FU Ori outbursts (see e.g.Carvalho et al. 2023a on V960 Mon).The O I 8446 Å (9.5 eV) line is also present, but also weaker than in other examples of FU Ori stars.
In the infrared spectrum, there is a strong pure-absorption blue-asymmetric profile in the He I 10830 line, with absorption out to -650 km/s.Absorption is present in the Paschen series from Paβ up through the optical lines, and in Brγ, but there is no signature in upper level Brackett series lines.The H I absorption line widths are approximately 150 km/s (HWHD), and thus broader than the disk lines.
Overall, the presence of outflowing hot gas in the RNO 54 system is clear.The wind profiles are not as deep nor as broadly and asymmetrically blueshifted as they appear in some other FU Ori stars, which we suspect is due to the relatively old age of the outburst.
In terms of jet activity, our spectrum does not show low-excitation forbidden lines such as [O I] or [S II] that are often observed in "low-state" accreting young stellar objects, and indicative of shocked material.While not typically seen in the "high-state accretor" FU Ori stars, such lines have been observed in V1057 Cyg and V960 Mon, FU Ori outburst sources that have faded substantially.However, Magakian et al. (2023) did identify optical forbidden-line emission at spatially offset positions from the RNO 54 point source, suggesting possible past ejections.

ACCRETION DISK MODEL MATCHED TO RNO 54 SPECTRA AND SED
In this section, we demonstrate that in addition to displaying the spectral signatures of an FU Ori star, RNO 54 is well-fit by a standard pure-accretion disk model.
We fit the SED following a procedure similar to that described in Carvalho et al. (2023b).The priors we impose on the MCMC fit are: a Kroupa IMF constraining the central star mass, and a requirement imposed by the line broadening in the ESI spectrum, that v max × sin i should be drawn from a Gaussian distribution centered on 85 km s −1 with a standard deviation of 10 km s −1 .The v max is calculated at each iteration as the Keplerian velocity of the innermost annulus of the disk: v max = GM * /R inner .Our fitting process initially considered the parameters: disk accretion rate Ṁ , stellar mass M * , stellar radius R * , disk inclination i, extinction A V , and distance d.The fit converged with decently well-behaved posteriors, but in order to better constrain the most important disk parameters, we chose to adopt fixed values for d, A V , and i.The initial distance posterior was 1490 ± 670 pc and we thus adopted d = 1400 pc, the Gaia DR3 value.The initial extinction posterior was 3.93 ± 0.67 mag and we thus adopted 3.93 mag, which is consistent within the errors of the value of 4.2 mag resulting from the DIBs measurement as well as Gaia DR3.The initial inclination posterior was 47 ± 14 deg and we thus adopted 50 deg, also consistent with the morphology of optical and infrared images of the source that show a cometary morphology in its scattered light nebula.In particular, it is the inclination constraint that helps narrow the parameters of the accretion disk.
The posteriors for the remaining parameters are shown in the corner plot in Figure 5 for our final fit.The system parameters are: Ṁ = 10 −3.45±0.06M ⊙ yr −1 , M * = 0.23 ± 0.06 M ⊙ , and R inner = 3.68 ± 0.76 R ⊙ .The resulting best-fit SED is a good match to the optical/NIR photometry and the SpeX spectrophotometry and is also shown in Figure 5.
We can derive additional parameters of the RNO 54 system, namely the maximum temperature of the disk, T max and the accretion luminosity L acc , via comparison with the V960 Mon system that was similarly modelled in detail by Carvalho et al. (2023b).Using our infrared spectrum for RNO 54 and that for V960 Mon from Carvalho et al. (2023b) at the January 2016 epoch, when the object had cooled somewhat from the earlier outburst epoch, we find that the objects closely match one another, especially in many of the molecular features.We assume, therefore, that the T max of the two objects should be approximately similar, and adopt for RNO 54 the T max ∼ 7000 K estimated in Carvalho et al. (2023b) for V960 Mon at this late epoch.This temperature is somewhat cooler than the formal temperature following from the most likely system parameters given above, but is within the 1σ values, notably reached by making R inner larger.
Continuing this line of argument, we find that dereddening the RNO 54 spectrum by A V = 3.5 mag gives a good match to the dereddened V960 Mon spectrum, which confirms our best-fit A V from the SED fitting.Assuming then a distance of 1.4 kpc to RNO 54 and an inclination of 50 deg, we find an integrated luminosity of L SpeX ∼ 137 L ⊙ .In V960 Mon, the L acc ∼ 2 L SpeX , which would give L acc ∼ 265 L ⊙ for RNO 54.Fixing T max and assuming T max ∝ L acc /R 2 inner 1/4 , we can then compute R inner ∼ 4.5 R ⊙ and that Ṁ M * ∼ 7.72 × 10 −5 M 2 ⊙ yr −1 .These values are in good agreement with the system parameters we derived above from the SED fit.

RNO 54 in Context
Compared to Gaia 18dvy and Gaia 17bpi, both recently outbursting FU Ori stars that are also optically visible sources, RNO 54 seems to have broader and shallower absorption lines.This is probably due to its higher inclination, which we found in our disk modelling to be close to 50 deg.The line ratios, in particular the relative strengths of Fe I and Fe II lines, also suggests that RNO 54 is hotter than both Gaia 18dvy and Gaia 17bpi.For the same reasons, RNO 54 is potentially cooler than V960 Mon, even at its later epochs, as evidenced by its stronger TiO and VO bands, as well as its stronger Fe I, and relatively weaker Fe II lines.
V960 Mon is a recent outburst erupting in late 2014, but one that has faded and cooled quickly.older FU Ori ouburst than V960 Mon, and observed spectroscopically only many decades after its assumed outburst, these lines along with H I Paschen and Bracket absorption are still present, demonstrating wind, but very weak.An exception is He I 10830 Å which is exceptionally strong in RNO 54.
In terms of other information on RNO 54, there is low level variability in the optical lightcurves that are publically available from e.g.ASAS-SN, ZTF and ATLAS.The variations are quasiperiodic on long timescales (months) with amplitudes of about 0.2 mag, which is nothing unusual for a YSO.The mid-infrared lightcurve and colorcurve from NEOWISE are both relatively flat, and again unremarkable.Magakian et al. (2023) Several parameters of RNO 54 are found to be similar between the present study and that of Magakian et al. (2023).The extinction was reported as A V = 2.5 − 3.0 mag in Magakian et al. (2023) whereas we find a value closer to 4 mag.The luminosity was L = 300 − 400 L ⊙ in Magakian et al. (2023), whereas we find a lower 265 L ⊙ for the inner accretion disk, though closer to 580 L ⊙ in the total SED including the mid-and far-infrared.

Other Remarks
It is not unreasonable to assume that RNO 54 is only one of at least a few tens of recognized YSOs that is actually an object experiencing an accretion outburst.YSO selection techniques based on infrared excess as a disk indicator, or activity diagnostitics such as x-ray emission, Hα emission, or photometric variability, have identified several hundred thousand YSOs within 1-2 kpc of the Sun.However, a relatively small fraction of these have been studied spectroscopically.
The tell-tale FU Ori signature at low-dispersion is a wavelength-dependent spectral type from optical FGK to near-infrared (M), and at high-dispersion, a mixed-temperature low-gravity spectrum over short spectral ranges.In spectroscopic studies that consider only optical data, or only near-infrared data (but not both), this could render bona fide FU Ori stars removed from samples due to similarity to background giant contaminants (FGK types in optical spectra, and M types in infrared spectra).
Given the recent discovery rate of approximately one new FU Ori outburst every 1-2 years, and the lower rates in the past (one discovery only every ∼10 years), there are likely to be a few tens of unrecognized "FUOr-like" stars in which no outburst was detected/noticed, but the object currently does exhibit the spectral signatures of being in an outburst state.

SUMMARY
In this paper, we have presented evidence for the young stellar object RNO 54 as satisfying all of the commonly accepted criteria for a post-outburst FU Ori type star.No actual photometric outburst was detected in this source, but the hypothesized burst is constrained to have occurred prior to the 1950s, and likely prior to the 1890s.Evidence in favor of FU Ori status -beyond the common YSO features of having a Class I SED showing broad infrared excess, a nebular environment, and the Li I 6707 Å spectral signatures of stellar youth -includes: • An optical absorption spectrum featuring metal line absorption from a range of hotter and cooler lines that are broadened to 85 km/s, consistent with the rotational broadening of a disk; • An infrared spectrum featuring a similar mix of higher and lower excitation potential atomic lines, as well as prominent molecular bands.
• Low surface-gravity indicators such as atomic Sr II and Si I lines, as well as molecular bands due to CN, H 2 O absorption sufficiently deep to produce a "triangular H-band" spectral shape, and strong CO absorption.
• Wind/outflow evident in the P Cygni type profile of Hα, and a hint of such in Ca II triplet lines, as well as blueshifted asymmetry in other atomic absorption line profiles, prominently He I 10830 Å.
• The match of observed spectrophotometry to an accretion disk model with system parameters: Ṁ = 10 −3.45±0.06M ⊙ yr −1 , M * = 0.23 ± 0.06 M ⊙ , and R inner = 3.68 ± 0.76 R ⊙ ; we note, however, that R inner is likely to be close to its upper range of 4.5R ⊙ .The corresponding radiative parameters of the accretion disk are thus T max ≈ 7000 K and L acc ≈ 265L ⊙ .
We speculate that RNO 54 may be only one of at least a few tens of known YSOs that is an object in an unrecognized state of accretion outburst.

Figure 2 .
Figure2.Infrared spectrum of RNO 54 compared to that of V960 Mon.The two sources do not have similar continuum shapes, with V960 Mon artifically adjusted to attempt to match the reddening of RNO 54.However, both sources have absorption features from a mix of molecules (TiO, VO, H2O, CO; red), low-excitation (Mg I, Ca I, Al I), and higher-excitation (Si I, S I, C I, Ca II) neutral and ionized species (black), as well as the Sr II, CN, and Si I indicators of low surface gravity (orange).In addition, both sources show broad blueshifted absorption in the He I 10830 line (see inset for RNO 54), indicating a strong wind.

Figure 4 .
Figure 4. Segments of the optical spectrum of RNO 54 emphasizing lines showing wind signatures, including the K I 7669,7699 Å doublet, the O I 7773 Å triplet, the deep Na I D doublet, the Mg Ib doublet, Hα, and the Ca II triplet.The Li I 6707 Å signature of stellar youth is also highlighted, showing line splitting consistent with a disk profile, similar to Ca I 6717 Å.

Figure 5 .
Figure 5. Left: assembled photometric SED (blue points), spectrophotometry from SpeX (black line), and accretion disk model fit (red line).Right: "corner plot" showing the physical parameters in the model; red lines mark median values while blue lines are modal values.The gas disk model is appropriate over the optical and near-infrared wavelength range, where accretion dominates the emission.Beginning in the mid-infrared, dust contributions from an inactive or passive disk become important, while in the far-infrared dust emission from an envelope dominates; these cooler contributions to the SED are not part of our modelling effort.