Brightness variations of the FUor-type eruptive star V346 Nor
aa r X i v : . [ a s t r o - ph . S R ] J a n Astronomy & Astrophysicsmanuscript no. paper c (cid:13)
ESO 2018July 4, 2018
Brightness variations of the FUor-type eruptive star V346 Nor ⋆ Á. Kóspál , P. Ábrahám , Ch. Westhues , and M. Haas Konkoly Observatory, Research Centre for Astronomy and Earth Sciences, Hungarian Academy of Sciences, PO Box 67, 1525Budapest, Hungary e-mail: [email protected] Astronomisches Institut, Ruhr-Universität Bochum, Universitätsstraße 150, D-44801 Bochum, GermanyReceived date; accepted date
ABSTRACT
Decades after the beginning of its FU Orionis-type outburst, V346 Nor unexpectedly underwent a fading event of ∆ K = / VVV survey to outline the brightnessevolution. In our VLT / NaCO images, we discovered a halo of scattered light around V346 Nor with a size of about 0 ′′ .
04 (30 au). TheVISTA data outlined a well-defined minimum in the light curve at late 2010 / early 2011, and tentatively revealed a small-amplitudeperiodic modulation of 58 days. Our latest data points from 2016 demonstrate that the source is still brightening but has not reachedthe 2008 level yet. We used a simple accretion disk model with varying accretion rate and line-of-sight extinction to reproduce theobserved near-infrared magnitudes and colors. We found that before 2008, the flux changes of V346 Nor were caused by a correlatedchange of extinction and accretion rate, while the minimum around 2010 was mostly due to decreasing accretion. The source reacheda maximal accretion rate of ≈ − M ⊙ yr − in 1992. A combination of accretion and extinction changes was already invoked in theliterature to interpret the flux variations of certain embedded young eruptive stars. Key words. stars: formation – stars: circumstellar matter – infrared: stars – stars: individual: V346 Nor
1. Introduction
FU Orionis-type stars (FUors) are low-mass pre-main sequenceobjects characterized by 4-6 magnitude optical outbursts due totemporarily enhanced accretion from the circumstellar disk tothe star (Hartmann & Kenyon 1996). Following the outburst ofthe first such object, FU Ori in 1937, now more than two dozenFUors and FUor candidates are known (Audard et al. 2014).V346 Nor was discovered by Elias (1980) as a source withina few arcseconds of the HH 57 nebulosity, the latter being afaint, compact H α emitting knot (Schwartz 1977), located in theSa 187 molecular cloud within the Norma 1 association, at adistance of 700 pc (Reipurth 1981, see also Fig. 1). A few yearslater Graham (1983) reported the appearance of a star-like sourceat the northeastern tip of HH 57, probably coinciding with thesource in Elias (1980). They mentioned that the star was notvisible in 1976 (Schwartz 1977), but a di ff use patch is clearlydiscernible in the blue plates of the ESO / SERC Sky Survey,obtained in April–June 1975 (Holmberg et al. 1974; Reipurth1981). Therefore, V346 Nor transformed from a faint di ff usenebula to a bright point-source some time between 1976 and1980. Based on this and on the spectroscopic properties of thestar, Reipurth & Krautter (1983) suggested that V346 Nor wasundergoing a FUor-type outburst. Frogel & Graham (1983) pre-sented photometry from 2.2 to 20 µ m and remarked that the col-ors of the object are similar to those of FU Ori and V1057 Cyg.Subsequent near-infrared (near-IR) photometry indi-cated that V346 Nor was gradually brightening in the K band, reaching a broad maximum between about 1990 and2000 (Frogel & Graham 1983; Reipurth & Wamsteker 1983;Reipurth 1985; Kenyon & Hartmann 1991; Molinari et al. 1993;Prusti et al. 1993; Reipurth et al. 1997; Ábrahám et al. 2004,see also Fig. 3). Kenyon & Hartmann (1991) presented thebroad-band optical-IR spectral energy distribution (SED) of offset Fig. 1.
V346 Nor (black plus sign in the center) and its surroundingsin a
JHK S color composit image. The observations were taken withinthe VISTA Variables in The Via Lactea (VVV) Survey on March 15,2010. North is up and east is to the left. The displayed area is 2 ′ . × ′ . V346 Nor, while Weintraub et al. (1991) published submillime-ter and millimeter photometry for it. Both groups concluded thatthe object is surrounded by a significant amount of circumstellarmaterial, in the form of an actively accreting disk and a flattenedenvelope. Recently, Kraus et al. (2016) reported a dramaticbrightness decrease of V346 Nor and a subsequent brightening.
Article number, page 1 of 7 & Aproofs: manuscript no. paper -2-1012 D EC o ff s e t ( a rc s e c ) J H K S N o r m a li z e d i n t e n s i t y J H K S Fig. 2.
Top:
VLT / NaCo J , H , and K S images of V346 Nor from 2008.The color scale is logarithmic. Bottom:
Radial intensity profiles ofV346 Nor (solid curves) and of another, fainter star visible in the fieldof view (dashed curves) measured in our NaCo J , H , and K S images.The uncertainty of the brightness profile of the fainter star is indicatedby error bars, while the uncertainty of V346 Nor’s profile is less thanthe curve thickness. They interpret these results as a 2-3 orders of magnitude dropin the accretion rate, followed by the onset of a new outburst.In order to better understand this spectacular event, and tofollow up the evolution of the system, we present new near-IRobservations of V346 Nor, and re-evaluate the data taken bythe VISTA telescope. We analyze the brightness and colorvariations observed in V346 Nor and compare the results withsimilar fading events of highly accreting young stellar objectsfrom the literature.
2. Observations, data reduction, and photometry
We observed V346 Nor with the NaCo adaptive optics instru-ment on the UT4 of European Southern Observatory’s VeryLarge Telescope (VLT) at Cerro Paranal, Chile, on April 10 / ′′ . We obtained J , H , and K S -band images with the N20C80dichroic and the 13 mas pixel − scale camera. We observed2MASS J16323308 − ′ away from V346 Nor and has similar 2MASS magnitudes( J = H = K S = ′′ and sky annulus between 2 ′′ . ′′ .
9. Thislarge aperture was chosen to include all the flux of V346 Nor,which appears slightly extended in our NaCo images (see be-low), therefore, our photometry can be compared to earlier un-resolved photometry from the literature. The obtained instru-mental magnitudes were converted to standard magnitudes usingthe 2MASS values for the photometric standard. We also down-loaded archival J and K S -band NaCo observations from June 12,2003, and reduced and extracted photometry from them in a sim-ilar way. The obtained brightnesses of V346 Nor are presentedin Tab. B.1.Our group observed the area around V346 Nor with theInfraRed Imaging System (IRIS) at Bochum Observatory near Cerro Armazones. IRIS is a 80 cm telescope equipped with a1k ×
1k infrared camera. The system provides a resolution of0.74 ′′ / pixel and a field-of-view of 13 ′ × ′ . Data were taken be-tween June 26, 2010 and July 1, 2010, as well as between August22 and 25, 2016, in the J , H and K S bands. Individual frameswere combined to eliminate the sky signal and correct for flat-field di ff erences. V346 Nor was not visible in the J band, but wasdetected in all H and K S frames. All J -band images obtained onthe same night were combined into mosaics, and 3 σ upper lim-its were determined. For the other filters, we performed aperturephotometry using the same aperture and sky annulus sizes as forthe NaCo images. For the photometric calibration, we used a setof about 50 2MASS stars with quality flag ‘A’ to determine theo ff set between the instrumental and the 2MASS magnitudes. Wefound that no color term was needed for this transformation. Theuncertainty of the final photometry is the quadratic sum of theformal uncertainty of the aperture photometry and the photomet-ric calibration. The resulting J upper limits and HK S magnitudesare presented in Tab. B.1.We observed V346 Nor with the SMARTS 1.3 m telescopeat Cerro Tololo on June 7 and August 9, 2016. The telescopeis equipped with the ANDICAM instrument, which provides si-multaneous optical and IR images. The CCD for the ANDICAMis a Fairchild 447 2k ×
2k chip, which we used with 2 × ′′ / pixel, and afield of view of about 6 ′ × ′ . The IR Array for the ANDICAMis a Rockwell 1k ×
1k HgCdTe “Hawaii” Array, also used with2 × ′′ / pixel binned scale and 2 ′ . × ′ . VRI optical andCIT / CTIO
JHK
IR filters. A 5-point dithering was done to en-able bad pixel removal and sky subtraction in the IR images.Bias and flat correction for the optical images were done bythe Yale SMARTS team. Although HH 57 is faintly visible inour V and R images, V346 Nor itself is not detected in theoptical. We used magnitude values from the UCAC4 catalog(Zacharias et al. 2013) to calibrate the images and determined3 σ upper limits of V > R > I > JHK magnitudesfor V346 Nor are presented in Tab. B.1.V346 Nor was covered as part of the VISTA Variables inthe Via Lactea Survey, an ESO public survey using the VISTA4.1 m telescope and the VIRCAM near-IR camera (Minniti et al.2010). We downloaded all VIRCAM images from this surveycovering V346 Nor. To obtain photometry that can be com-pared with our NaCO, IRIS, and SMARTS data, we performedour own flux extraction with the same aperture as describedabove. At some epochs, our values di ff er from those publishedby Kraus et al. (2016) for the same measurements. The reason isprobably a di ff erent treatment of the known non-linearity of theVIRCAM detectors for bright sources (e.g., Saito et al. 2012).The correction we applied is described in details in Appendix A.The J and H photometry, as well as the K s -band results after thenon-linearity correction are also given in Tab. B.1.
3. Results
The top part of Figure 2 shows our NaCo
JHK S images from2008 of V346 Nor, while the bottom panel displays the nor-malized, azimuthally averaged radial brightness distributions ofV346 Nor, and another, fainter star visible in the field of viewat a distance of 5 ′′ .
7, position angle of 13 ◦ east of north. As-suming that this nearby faint star is a point source, the compari- Article number, page 2 of 7óspál et al.: Brightness variations of V346 Nor M a g n i t u d e M a g n i t u d e Fig. 3.
Near-IR light curves of V346 Nor. Data points are fromElias (1980); Frogel & Graham (1983); Graham & Frogel (1985);Reipurth & Krautter (1983); Reipurth & Wamsteker (1983); Reipurth(1985); Kenyon & Hartmann (1991); Molinari et al. (1993); Prusti et al.(1993); Reipurth et al. (1997); Quanz et al. (2007); Connelley et al.(2008), the 2MASS, DENIS, AllWISE and NEOWISE catalogs(Cutri et al. 2003; Cutri 2013; Cutri et al. 2015), and this work. Down-ward arrows indicate upper limits. son of the brightness profiles show that V346 Nor is extended inthe J and H bands, while it is consistent with a point source inthe K S band. Deconvolved sizes using Gaussian deconvolutionare 0 ′′ . ± ′′ .
016 (29 ±
11 au) in J , 0 ′′ . ± ′′ .
010 (25 ± H , and we can give an upper limit of 0 ′′ .
02 (13 au) for the K S -band size. The size of the near-IR emitting region in circumstel-lar disks is typically only a few au, therefore, the emission seenin the NaCo images has to be scattered light. The J and H im-ages are slightly asymmetric with the northern part slightly moreextended than the southern part. No such asymmetry is evidentin the K S image.Figure 3 shows the near-IR light curves of V346 Nor. Thefirst few data points indicate a brightening between 1979 and1983, followed by a relatively constant period until about 1988.Afterwards, the K and L -band data show a gradual brighteninguntil 1992, already reported in Ábrahám et al. (2004). The J and H light curves were rather flat, with small, < JHK photome-try in 1979.Some time around 2008, V346 Nor started a dramatic fading,and reached a minimum around late 2010-early 2011 (Fig. 3,bottom). Afterwards, the source quickly brightened by ∆ K = ∆ K = J A V = 10 mag 19791983-88/2003 1989-1999 2008 20102016 Fig. 4.
Near-IR color-magnitude diagram for V346 Nor. The reddeningpath is marked with dashed line (Cardelli et al. 1989). The solid graycurves indicate our reddened accretion disk model fits (see details intext). lowing 3 years, indicating a slower brightening rate. As of 2016August, V346 Nor has not yet reached the brightness level it dis-played in 1980–2000. This is well visible in the near-IR lightcurves, but our optical upper limits also supports this, as the starwould have been visible in our
VRI images had it been as brightas between 1980–2000. The deep minimum was also visible inthe WISE 3.4 µ m photometry, although with smaller amplitude.The lower panel of Fig. 3 shows that the minimum in the K -band has a parabola-like light curve shape. By fitting andsubtracting a second-order polinomial from the photometry be-tween 2010 and 2014, the obtained residuals are on the orderof 0.2 mag, and suggest a possible periodic modulation. Wecalculated a Lomb-Scargle periodogram for the residuals, andfound a tentative 58 ± × − (see also Fig. B.1 in Appendix B). Simi-lar periodicities in the light curves were already found in, e.g.,V1647 Ori, where it was explained by an orbiting dust cloud(Acosta-Pulido et al. 2007), and in V960 Mon, where is was ex-plained by a putative close companion (Hackstein et al. 2015).
4. Discussion and Conclusions
The color-magnitude diagram in Fig. 4 shows that all data pointsuntil 2008 form an approximately linear strip, while the latermeasurements deviate from this trend, suggesting di ff erent phys-ical mechanisms for the brightness and color changes beforeand after 2008. In order to reproduce the observations, in eachepoch we fitted the JHK S data points by a steady, opticallythick, geometrically thin, viscous accretion disk, with radiallyconstant mass accretion rate. Such disk models were success-fully proposed and used to reproduce the SEDs of FUors byHartmann & Kenyon (1996), Zhu et al. (2007), and Kóspál et al.(2016). We calculated the disk’s SED by integrating the fluxes ofconcentric annuli between the stellar radius and R out , assumingblackbody emission. We reddened the model fluxes by di ff er-ent A V values, using the extinction law from Savage & Mathis(1979) with R V = out = ff ect on thenear-IR fluxes), we have only two free parameters, the productof the stellar mass and the accretion rate M ˙ M , and the extinc-tion A V . We fixed the stellar mass and radius to typical low-massYSO values of 1 M ⊙ and 3.0 R ⊙ (the resulting accretion rate is in-versely proportional to the adopted stellar mass). The inclination Article number, page 3 of 7 & Aproofs: manuscript no. paper of the V346 Nor disk is not known, thus we adopted 2 / π ≈ ◦ ,the mean expected value if the disk is randomly oriented. Thefitting procedure was performed with χ minimization. More ex-treme extinction laws (up to R V = A V and ˙ M values by less than 15%, which iswithin the formal uncertainty of our fitting procedure.The bluest points in Fig. 4 correspond to measurementsobtained in 1983-88 and 2003, when the system exhibitedapproximately the same brightness and color. We found thatthese points can be well fitted by our disk model with ˙ M = × − M ⊙ yr − and line-of-sight reddening of A V = M = × − M ⊙ yr − and ˙ M = × − M ⊙ yr − , respectively.The reddening, however, was significantly higher, 16.7 mag in1979 and 21.5 mag in 2008. We simulated the time evolutionof the system by computing a sequence of models of gradu-ally changing A V from 21.5 to 6.7 mag and ˙ M from 4.5 × − to 1.0 × − M ⊙ yr − . The resulting line is plotted in Fig. 4. Formost epochs, our disk model fits the SED shape well with typicalformal uncertainties of 1-2 mag in A V and 10-30% in ˙ M . Afterthe minimum in 2010, the shape of the SEDs can be reproducedless well with the accretion disk model, resulting in formal un-certainties of 6-8 mag in A V , and up to a factor of 6 in ˙ M .The data points obtained between 1989 and 1999 are situatedabove the model line. These SEDs can also be fitted with ourdisk model, but with higher ˙ M values than at any time before. A V was between 12.1 and 19.2 mag in this period. In particular,we found that the accretion rate showed a maximum in 1992 Jan-uary, with ˙ M = × − and A V = A V from19.2 to 12.1 mag and ˙ M from 9.8 × − to 3.5 × − M ⊙ yr − ,also plotted in Fig. 4.Kraus et al. (2016) suggested that the minimum around 2010is related to a large drop in the accretion rate. Using our accretiondisk model, we found that by keeping a constant A V = M to the SED measured in 2010 we canreach the 2008 data point (Fig. 4). In the minimum, the measuredfluxes constrain the model accretion rate below 4 × − M ⊙ yr − ,thus the change in accretion rate was at least a factor of 100 ormore, in agreement with the findings of Kraus et al. (2016). Inour modeling, the scattered light component, indicated by ourNaCo observations in 2008, was not included, since its consis-tent treatment would be beyond the scope of this Letter.Our results demonstrated that while the rapid fading in 2010was an accretion event, the flux evolution beforehand was dueto a correlated change in extinction and accretion rate together.That is, increasing accretion rate is accompanied by growingextinction towards the source. V346 Nor is similar to a groupof highly variable young stellar objects whose flux changes aredue to a combined e ff ect of changing accretion rate and variablecircumstellar extinction. Such objects are, e.g., H α
11, PV Cep,V1647 Ori, and V899 Mon (Kun et al. 2011a,b; Mosoni et al.2013; Ninan et al. 2015). V346 Nor also resembles the youngeruptive star V2492 Cyg in several aspects: the pre-outburst po-sition of V2492 Cyg in the near-IR color-color diagram is closeto the point of the 2010 minimum of V346 Nor, it also underwenta large accretion change at the beginning of its outburst, and inthe high state, the line-of-sight extinction is continuously vary-ing (Kóspál et al. 2011, 2013; Hillenbrand et al. 2013). There-fore, the observed variability in both V346 Nor and V2492 Cygare governed by a combination of changing accretion and extinc-tion. The minimum of V346 Nor in 2010 was immediately fol-lowed by the re-brightening of the source. As of 2016, the sourceis on its way back to its 2008 state in the color-magnitude dia-gram (Fig. 4). This suggests that the brightening is governed bythe same process that was responsible for the fading, namelychanging accretion at a constant high extinction. Indeed, oursimple accretion disk model can reproduce the
JHK S fluxesmeasured in 2016 by assuming ˙ M = × − M ⊙ yr − , and A V = Acknowledgements.
The authors thank the referee for his / her useful comments.This work was supported by the Momentum grant of the MTA CSFK LendületDisk Research Group, and by the NKFIH research fund OTKA 101393. References
Ábrahám, P., Kóspál, Á., Csizmadia, S., et al. 2004, A&A, 428, 89Acosta-Pulido, J. A., Kun, M., Ábrahám, P., et al. 2007, AJ, 133, 2020Audard, M., Ábrahám, P., Dunham, M. M., et al. 2014, Protostars and PlanetsVI, 387Bell, K. R. & Lin, D. N. C. 1994, ApJ, 427, 987Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245Connelley, M. S., Reipurth, B., & Tokunaga, A. T. 2008, AJ, 135, 2496Cutri, R. M. 2013, VizieR Online Data Catalog, 2328Cutri, R. M., Mainzer, A., Conrow, T., et al. 2015, Explanatory Supplement tothe NEOWISE Data Release Products, Tech. rep.Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online DataCatalog, 2246Elias, J. H. 1980, ApJ, 241, 728Frogel, J. A. & Graham, J. A. 1983, IAU Circ., 3792Graham, J. A. 1983, IAU Circ., 3785Graham, J. A. & Frogel, J. A. 1985, ApJ, 289, 331Hackstein, M., Haas, M., Kóspál, Á., et al. 2015, A&A, 582, L12Hartmann, L. & Kenyon, S. J. 1996, ARA&A, 34, 207Hillenbrand, L. A., Miller, A. A., Covey, K. R., et al. 2013, AJ, 145, 59Holmberg, E. B., Lauberts, A., Schuster, H.-E., & West, R. M. 1974, A&AS, 18,463Kenyon, S. J. & Hartmann, L. W. 1991, ApJ, 383, 664Kóspál, Á., Ábrahám, P., Acosta-Pulido, J. A., et al. 2013, A&A, 551, A62Kóspál, Á., Ábrahám, P., Acosta-Pulido, J. A., et al. 2011, A&A, 527, A133Kóspál, Á., Ábrahám, P., Acosta-Pulido, J. A., et al. 2016, ArXiv e-printsKraus, S., Caratti o Garatti, A., Garcia-Lopez, R., et al. 2016, MNRAS, 462, L61Kun, M., Szegedi-Elek, E., Moór, A., et al. 2011a, ApJ, 733, L8Kun, M., Szegedi-Elek, E., Moór, A., et al. 2011b, MNRAS, 413, 2689Minniti, D., Lucas, P. W., Emerson, J. P., et al. 2010, New A, 15, 433Molinari, S., Liseau, R., Lorenzetti, D., & Graham, J. 1993, IAU Circ., 5727Mosoni, L., Sipos, N., Ábrahám, P., et al. 2013, A&A, 552, A62Ninan, J. P., Ojha, D. K., Baug, T., et al. 2015, ApJ, 815, 4Prusti, T., Bontekoe, T. R., Chiar, J. E., Kester, D. J. M., & Whittet, D. C. B.1993, A&A, 279, 163Quanz, S. P., Henning, T., Bouwman, J., et al. 2007, ApJ, 668, 359
Article number, page 4 of 7óspál et al.: Brightness variations of V346 Nor
Reipurth, B. 1981, A&AS, 44, 379Reipurth, B. 1985, A&A, 143, 435Reipurth, B. & Krautter, J. 1983, IAU Circ., 3823Reipurth, B., Olberg, M., Gredel, R., & Booth, R. S. 1997, A&A, 327, 1164Reipurth, B. & Wamsteker, W. 1983, A&A, 119, 14Saito, R. K., Hempel, M., Minniti, D., et al. 2012, A&A, 537, A107Savage, B. D. & Mathis, J. S. 1979, ARA&A, 17, 73Schwartz, R. D. 1977, ApJS, 35, 161Weintraub, D. A., Sandell, G., & Duncan, W. D. 1991, ApJ, 382, 270Zacharias, N., Finch, C. T., Girard, T. M., et al. 2013, AJ, 145, 44Zhu, Z., Hartmann, L., Calvet, N., et al. 2007, ApJ, 669, 483
Article number, page 5 of 7 & Aproofs: manuscript no. paper
VISTA K V I S T A - K M A SS V346 Nor
Fig. A.1.
Demonstration of the non-linearity correction applied for thephotometry of V346 Nor. K s m a g n i t u d e Fig. A.2.
VISTA / VIRCAM photometry of V346 Nor without (greenplus signs) and with (red circles) correction for non-linearity. Asterisksindicate our IRIS photometry. The black solid curve is a parabola fit-ted to the data points to remove the long-term trend before the periodanalysis (see Sec. B).
Appendix A: Near-IR photometry of V346 Nor
The comparison of our aperture photometry with the 2MASScatalog revealed the known detector issue that stars brighter thanabout 11 mag in the K S band enter the non-linear regime of theVIRCAM detector (Saito et al. 2012). While close to its min-imum V346 Nor was below this limit, after about May 2013 itbecame brighter than 11 mag. In order to correct for the underes-timation of the signal, we plotted the o ff sets between the instru-mental and the 2MASS magnitudes for all stars with quality flag‘A’ in the image as a function of the instrumental magnitude.We fitted the distribution of points with a first or second orderpolynomial, and determined the o ff set valid for V346 Nor fromthis fit (for an example, see Fig. A.1). The necessary correctiondue to the non-linearity was typically in the 0.1–0.2 mag range,with a few higher values up to 0.4–0.6 mag. Fig. A.2 shows the K S -band light curve without and with the non-linearity correc-tion, demonstrating that our correction significantly reduced thescatter of the data points obtained close in time. Appendix B: Period analysis
10 100 1000Period (d)02468101214 L o m b - S c a r g l e P o w e r -2 -3 ∆ K m a g Fig. B.1.
Top:
Lomb-Scargle periodogram of the light curve ofV346 Nor after removing a parabolic trend, as illustrated in Fig. A.2.The highest peak corresponds to a period of 58 days. The dashed linesshow the powers corresponding to false alarm probabilities of 10 − and10 − . Bottom:
Phase-folded light curve showing the data points afterremoving the parabolic trend and folded with a period of 58 days.Article number, page 6 of 7óspál et al.: Brightness variations of V346 Nor
Table B.1.
Near-IR photometry of V346 Nor.
Date JD − J H K S Telescope2003-06-12 52 802.86 9.81 ± ± ± ± ± ± ± ± ± > ± > ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±0.02 IRIS