SUMMARY:
A source in the constellation of Puppis,
located at RA,Dec = 07 59 15.26, -31 08 44.6 (J2000.),
and coincident with WISEA J075915.26-310844.6, likely the same as IRAS 07572-3100,
appears to have undergone a large-amplitude outburst in 2015
that has lasted until the present (early 2021).
The full amplitude appears to be 8 mag in the mid-infrared, though seemingly much smaller in the optical.
A new nebula also appeared, as discussed in announcements
from the citizen scientists and amateur astronomers referenced above.
One group was examining imaging data from DECaPS,
and the other ASAS-SN lightcurves.
An assessment of the source environment, spectral energy distribution,
multi-wavelength lightcurve information, and post-outburst optical spectrum follows.
At present, it is not obvious that the object is a young star,
though there is some compelling evidence in that direction.
The spectrum seems consistent with an evolved, AGB type star,
yet that interpretation is not entirely satisfactory either.
CDSportal color images from DECaPS (left) and Spitzer (right):
The optical nebula recorded by the DECam Plane Survey (DECaPS) is about 1.5'
in extent,
and the structure is reported by Neagu2021 as variable.
The infrared images also indicate an extended source,
with a smaller size at 3.6um and a larger size at 4.5um.
ENVIRONMENT:
The known IRAS source is projected on a small dark cloud region
that is evident in dust maps from Herschel, but not well-studied.
It was designated as ``cold clump" PGCC G247.99-00.79 in 2016A&A...594A..28P
The broader views below from Spitzer, Herschel/PACS, and Herschel/SPIRE
indicate that the nearest major star-forming region, RCW 19, is
about 5 degrees away in projection.
The source is essentially in the galactic plane, at l = 247.98, b = -0.78.
The catalog of Winston2020
shows that there are several groups of young stellar objects
corresponding to the bright-rimmed dark patch to the northeast,
its adjacent bright dust patch,
and to the brightest dust region due north of the source position.
Examination of the Herschel HiGal dust temperature maps of Marsh2017,
provided at
http://www.astro.cardiff.ac.uk/research/ViaLactea/PPMAP_Results/l248_results/,
shows that the dust temperature towards the dark cloud is only 12.4 K.
An empirical correlation between column density and temperature
suggests a line-of-sight extinction in this direction of Av=5-20 mag.
SOURCE SED AND COLORS:
The spectral energy distribution that can be assembled from PanSTARRS, 2MASS,
Spitzer, WISE, AKARI, IRAS, and Herschel/PACS
is shown in solid points below.
Larger-beam measurements (IRAS, Herschel/SPIRE) are indicated by open points.
There is some question about the PanSTARRS-i measurement as, although the source
appears in PS-1 (Chambers2016), this position is very close to the edge of the survey
and it is not possible to examine the image at the usual image servers.
In blue are measurements that occcurred subsequent to the 2015 outburst,
from Gaia, ATLAS, and NEOWISE.
For the recent ATLAS measurements we converted the two measured "c" and "o" bands
into equivalants in PanSTARRS "gri" filters using the equations given by Tonry2018.
The model photosphere that has been reddened and normalized
to the short wavelength photometry is consistent with
the spectroscopy presented below, but somewhat arbitrary.
The overall SED is red, and can be modelled as having dust at a range
of temperatures,
indicating a range of distances from the illuminating source.
In young star nomenclature this would be a Class I type SED,
but the slope would be unusually steep, alpha = +1.24.
Objects such as GY92 197 have slopes this steep, but they are rare,
and for the most part, invisible shortward of K-band.
Available 2MASS and WISE/NEOWISE measurements in the pre-outburst stage produce the following colors:
As noted above, these infrared colors would be unusually red for a young stellar object,
but they are consistent with an AGB star.
If the object were an AGB star,
the colors would align more with an oxygen-rich rather than a carbon-rich source
(https://www.aanda.org/articles/aa/full_html/2014/04/aa22818-13/F5.html).
In terms of mass loss, guided by Riebel2012, the K-W3 color
would correspond to a relatively low mass loss rate of 10^-8 Msun/year
(in the pre-outburst stage).
SPECTRUM:
An optical spectrum was acquired on 2021, February 3 UT (20210203)
at the W.M. Keck Observatory with ESI, the Echellette Spectrograph and Imager (Shennis2002).
Two 1200 second exposures were taken with the 0.5" slit to achieve spectral resolution R=8000.
The spectral range 5,000-10,000 is covered with adequate signal-to-noise in the reddest 7 spectral orders.
The source is clearly a late-type star, with a few weak emission lines.
Spectral typing suggests an M6-M7 III.
Low gravity is indicated by the shape of the CaH region,
as well as in the shapes of the prominent TiO and VO bands, and MgH 5200.
Low gravity signatures are also shown by the relatively weak
NaI and KI lines.
Weak emission lines in the spectrum have the following properties:
The Halpha line is resolved, with velocity width 2x the sky lines
(10.5 pixels FWHM vs 3.5 for most sky lines).
The Halpha velocity width is ~140 km/s (FWHM), perhaps with enhanced red-side emission.
This is quite narrow for a young stellar object, in which velocities
of several hundred km/s are characteristic of rapidly accreting stars,
but it is also rather large compared to the 10-15 km/s that is typical of AGB star winds.
The measured value is, however, essentially identical to the expected
escape velocity of a (n evolved) star with M = 5 Msun and R = 100 Rsun.
This suggests that the kinetic energy of the (outflowing?) gas measured in Halpha
could indeed come from the gravitational potential energy of the star.
The forbidden-line emission strengths suggest shocked gas.
Such lines are common in young accreting stars which drive outflows.
They seem rare for lower AGB stars, but can be seen in Miras.
For example, Sibbons2015 reported about 15% of their survey sample
to show [SII] and [NII] lines (with 25% showing Halpha emission).
These objects were nearly all C-type AGB stars, however
whereas IRAS 07572-3100, if an AGB star, would definitely still be an O-type.
Finally, our spectrum shows LiI 6707 with W_lambda = 625 mA.
This is a fairly typical value for late type young stellar objects.
However, red giants and red supergiants can also show LiI from dredge-up processes.
The measured ratios of e.g. LiI 6708 / CaI 6673 and LiI 6708 / KI 7699 are also in the range of these objects.
No strong conclusions can be drawn from the LiI data.
SOURCE PROPERTIES:
There is no distance available for the source itself.
However, looking at Gaia parallaxes for a few tens of objects
in the catalog of Winston2020, leads to an estimate of about 1.5 kpc.
We take this as the distance to the dust structures that are prominent
in the Hershel PACS and SPIRE maps shown above.
The nearest identified candidate YSO
(SSTGLMA G247.9941-00.7828 at 07 59 19.140 -31 09 40.83)
is ~1.5' away, and has a parallax suggesting distance in the range 900 - 1700pc.
If the outbursting source is associated with these clouds, we should be able to come up with a combination of intrinsic source luminosity, distance, and foreground extinction that is consistent with the observations.
Taking the W1 = 12.0 mag pre-outburst brightness and assuming no change in spectral type, i.e. assuming the progenitor is also a late-type M giant, the bolometric correction of such a source is approximately BC_W1 = 3.25 mag (based on assumed colors for an M6 star of BC_V = 4.36 and V-W1 = 7.6). The resulting Mbol = -2.55 and logL = 2.9 Lsun are consistent with the properties of a 2 Msun AGB star. The evolutionary models of Fadeyev2017 suggest that this HRD location could correspond to the first helium shell flash in such a star, thus providing some possible reason for the outburst. Reducing the assumed bolometric correction will increase the luminosity, and thus the inferred mass of the AGB track. The implied absolute magnitude M_W1 = -5.8 mag is consistent with values published at http://stev.oapd.inaf.it/YBC/ for a star with Teff = 3500 and log g = 0.5.
With no extinction, such a source would have to be at 36 kpc, but if the foreground extinction is Av = 7 mag, then the source would have distance 1.45 kpc, plausibly just behind the dust clump towards which it is projected. We note that this extinction value is consistent with the temperature-extinction correlation referred to above. It is somewhat higher than the Av=3 mag illustrated in the SED above.
The catalog of Marton2019 classifies the source as having 52% chance of being extragalactic, and 47% a young stellar object. The Vioque2020 gives it an 81% chance of being a pre-main sequence star.
OUTBURST LIGHTCURVE:
A major brightening of IRAS 07572-3100 / WISEA J075915.26-310844.6
clearly occurred in 2015,
as first reported by Théveno2021.
However, the NEOWISE photometry is well in the saturation regime,
and care is needed when interpreting the lightcurve rise and its apparent plateau.
This is especially the case when there is evidence for source extension
in the Spitzer images,
as well as from the measured PACS/SPIRE source sizes.
Below is a combined Spitzer, WISE, and NEOWISE lightcurve.
The NEOWISE measurements are the profile-fitting magnitudes,
which attempt to account for the full extent of the source.
The saturated photometry (brighter than the horizontal lines of the same color),
means that the true brightness of the plateau is uncertain,
but W1 approximately 4th mag and W2 approximately 2nd mag can be inferred.
Below is the aperture photometry version of just the NEOWISE portion of
the lightcurve.
Using increasing aperture sizes, from the smallest (5.5")
value to larger values, steadily decreases the measured magnitudes,
supporting the previous arguments for the extended nature of the source.
The aperture magnitudes are fainter than above, due to the saturation, but also because
the aperture (24.75" in this case) may not capture the full extent of the source.
For these reasons, it is hard to know whether the apparent bluer-when-brighter behavior is believable.
Because the NEOWISE cadence is around 6 months, the lightcurve peak may be undersampled.
There is indeed evidence for this in the optical data shown below.
Optical time-domain surveys have also captured portions of the brightening event.
Shown below are the ASAS-SN (Shappee2014) and ATLAS (Tonry2018,Heinze2018) lightcurves,
which seem to indicate a sharp peak with steep evolution on either side.
ASAS-SN may have captured the peak, but is not sensitive enough
to measure the bright-state plateau in the lightcurve.
ATLAS apparently missed the true peak, but samples well the bright-state plateau.
Summarizing the lightcurve behavior, the NEOWISE data indicate
an 8 mag outburst initiated at the beginning of 2015,
with the object rising >7 mag in the mid-infrared over 6-12 months.
Because of the saturation, it is unclear how to interpret the approach
to the plateau, and whether that does not occur until 2017, or closer to late 2015.
ASAS-SN and ATLAS data both indicate that the optical peak was in late 2016,
a full year after the major infrared brightening had occurred.
The optical lightcurve has steep evolution on either side of its peak
with a seemingly steeper slope on the decline side;
a the optical plateau level is about 3 mag below the inferred lightcurve peak.
The total optical brightening is unclear, and could be as little as
1.5 mag at g, 2.5 mag at r, and 3.5 mag at i -- according to comparison of
ATLAS measurements with quiescent-state gri data from VPHAS and PanSTARRS.
Below are all the lightcurves together, with the infrared shifted downward/fainter by a significant amount.
The current source brightness is estimated at V approximately 19.5 mag (AB),
and K approximately 5-6 mag (regular Vega).
Colors estimated from the ATLAS data are g-r = 2.8 and r-i = 1.8.
THOUGHTS:
This interesting object was first highlighted and publicized
by amateur astronomers, to whom we are certainly very grateful.
However, early interpretation of the brightening of IRAS 07572-3100
as a young star undergoing an FU Ori or an EX Lup type outburst,
does not appear to be supported by spectroscopic and other evidence.
This object seems to be something more puzzling!
Among the confirmed young star outbursts that have occurred
over the past few years, there is significant diversity
in the shape of the outburst lightcurves,
as well as in their early spectral appearance.
Each object has been its own case study!
Thus, it is worth considering if the spectrum of the outburst state of IRAS 07572-3100
could suggest an entirely new phenomenon in young stellar objects.
If it really is a young star, as indicated by the apparent association with a dust clump,
albeit one with little evidence of ongoing star formation aside from two dusty mid-infrared sources,
why does it have the very late-type spectrum and the steeply rising SED that it does?
And isn't 8 mag a rather large outburst for a YSO?
The cuspy nature of the lightcurve near its optical peak also seems inconsistent
with previous examples of accretion-driven outbursts in young stars.
Could the object possibly be the initial emergence of a newborn star from its molecular core,
where we are observing only the heated inner parts of an infalling dust envelope and not yet the central star itself?
Neither the relative infrared and optical outburst amplitudes,
nor the cuspy optical lightcurve shape near peak, suggest an extinction clearing event however.
What about binary effects, perhaps a protostellar merger?
If IRAS 07572-3100 is an old/evolved star, as suggested by the spectrum, why does it have the location it does?
As we have argued above, there is a combination of distance and extinction that
produces the approximately correct position in the HR diagram for this scenario.
But shouldn't there be evidence for pulsation if the progenitor was an uncataloged dusty AGB?
One can argue around this if the limited number of infrared measurements (four from
Spitzer, WISE, and the first two NEOWISE epochs), all lined up in phase.
What exactly does the 8 mag burst correspond to,
given that it has lasted so long (5-6 years at present),
and that the spectrum in the outburst state is so innocuous looking?
Speculation about the consequences of a helium (shell) flash or later-stage thermal pulsations are just that,
since the empirical evidence seems more impressive, and not entirely consistent with
the expectations at the stellar surface were one of these dramatic scenarios occurring deep in the interior.
Further afield would be that we are witnessing the actual birth of a planetary nebula.
Again, the lack of pulsation and the O-type rather than C-type spectrum argue against this.
The lightcurve shape approaching peak and then during the initial decline to plateau
does seem consistent with eruptive events in evolved objects, possibly driven by binary interactions.
Similar to the above suggestion, we could be witnessing the origin of a pre-planetary nebula stage,
with enhanced accretion on to either a primary or secondary driving the outburst.
Although the >100 km/s wind speed seems consistent with observations of pre-planetaries (Hofner2018),
again the issue is the lack of evidence for a late-stage AGB progenitor.
Considering even more exotic systems, there is some resemblance to
e.g. the early decline (2009) spectrum of the stellar merger V1309 Sco (Kaminski2015);
however, IRAS 07572-3100 lacks the forbidden emission line strength
and the heavy element signatures (e.g. RbI, CsI) of this source.
Furthermore, by the time V1309 Sco was at a similar post-outburst age,
its spectrum had evolved from one of strong molecular absorption to pure continuum+emission.
These non-YSO explanations also require a coincidence argument
for the projected location of the dust-enshrouded source precisely on a cold dust clump.
The presence of the nebula, which is demonstrated to have appeared
following the 2015 outburst (Théveno2021) and to have faded from 2017 to 2018 (Neagu2021),
suggests that the illuminating source is within or just behind the dust clump.
This decrease in nebular brightness seems consistent with the timing of the post-peak optical decline and ``light echo" effects.
We note that the guider camera images from our Keck/ESI spectrum did not indicate much nebulosity as of early 2021.
At present, the nature of IRAS 07572-3100 is ambiguous.
The current lengthy plateau in the outburst lightcurve should be monitored for changes,
with new spectroscopic investigation in the event of a re-brightening or dimming.
I thank colleagues Michael Kuhn (who also produced the dust map figure above),
Viraj Karambelkar, Roc Cutri, Philip Massey, Tom Greene, and Jim Fuller for discussions about this source.