Building a census of Slow-Deep Occultations with Time-Domain Surveys

collaborators & Team

My awesome advisors!

Time domain | stellar populations | variable stars

(he/him)

variability tree

Only a few classes of variable sources have been extensively characterized on the decade-long timescales. This means that large baseline surveys (i.e LSST, Gaia) will discover new/anomalous classes of variable stars that are comparable to the timescales of the survey duration, speed, and depth

why anomalous study stars?

Anomalous stellar systems
can teach us new variability mechanisms and different stellar environments (i.e YSO, AGB, RcB) and their surroundings

Case study: E-aur

• First case study of an unusual variable
  star that was known since the 1820s
   
• Initial theories of stellar occultation challenged our view on stellar theory. Until the hypothesis of Huang et al. 1965 of a disk occultation can cause a flat-bottomed dimming event
   
• Until the early 21st century, E-Aur was the record holder for the longest orbital period eclipsing binary. 

Case study: E-aur

mid-eclipse hole

Summary

1. Large distance error 0.4-4kpc (...now with Gaia, low mass-model is supported)
    
2. Challenge with the classification of primary, and tension with the mass of the system
    
3. Central star not visible (likely obscured)

high mass

low mass

• ​Normal FI-supergiant 15 solar masses with companion star type BV at 13 solar masses
   
• A high mass scenario suggests a young age solution.
  This would indicate the occulter to be a YSO disk (that usually lasts on the order of a few Myr)
   

• Normal FI-supergiant with ~1 solar mass with a secondary star at 5 solar masses
   
• A low mass scenario suggests that the primary has recently entered the post-AGB phase and the disk has formed due to mass exchange. However, spectroscopic follow-up does not suggest post-AGB

primary

Kloppenborg et al. 2010

Interferometric images suggest
the disk is an optically thick disk
is likely a debris disk rather than a YSO-disk as it would be required by the short evolutionary timescales in the 15 Mdot scenario!

Case study: E-aur

imaged disk with the CHARA via optical interferometry

conclusions

ongoing challenges

• Disk accretion during 3rd contact (Gibson et al. 20018)
• The formation channel of such a system is unclear
• Presence of rare-Earth elements in optical spectra

Gaia photometric
science alerts

• Gaia Science Alerts (GSA) daily alert release of interesting transients identified by Gaia. Alerts are designed to pick up on both fast & slow evolving transients
   
• Limiting magnitude of ~20.5 Gaia G band
   
• ~45% of the 21,800 events have been classified. Majority of them still remain unknown and unclassified
   
• Almost all stellar events contain astrometric solutions

optical light curve: Gaia17bpp

• Longest & deepest dimming duration
• No previous reporting groups?!!
• Odd location on the color-magnitude diagram

Archival photometry: IR-Optical

No obvious dimming event within the last 66 years!

Digitize Access to a Sky Century @ Harvard (DASCH)

constraining the primary star

Conclusions:

• Photometric SED classification suggest M-giant (consistent with E-Aur previous analogs) with Teff~4100 K and ~55 solar radii
• Preliminary SED modeling reveals a ~400 K drop from the eclipse

astroARIADNE (Vines & Jenkins 22)

Giant or Dwarf?

Photometric Classification

• Naive Bayes Classifier on LAMOST M dwarf &
  giants spectroscopically classified (Zhong et al. 2015)
• 17bpp classified as giant based on 2MASS/WISE color

Spectroscopic Classification

dwarf

giant

dwarf

giant

• Absence of Na I D and strength of Ca II line indicates
  small surface gravity values
• No reported peculiarities or ongoing accretion
  from the optical spectrum

APO 3.5 m ARC

Further clues

high order super-Gaussian function

• Optical-IR colors suggest a reddening profile during egress
• IR (W1-W2) show a blue trend from the quiescent phase?
  • Maybe a central hot source or Rayleigh scattering...? Extinction alone cannot contribute to this effect

Assuming b=0, circular orbit,

and R/R*~1 using Kepler's law:

orbital period

duration (FWHM)

orbital period >1000 years assuming the secondary has a small mass

Long duration event

A long-duration (FWHM~3.1 years) and deep flat dimming event suggest: 

extended object or large semi-major axis 

Preliminary light curve

modeling

Toy model assuming an opaque

transiting disk (ignoring chromatic effects)
(van de Kamp et al. 2021)

Orbital velocity is suspiciously too

small & disk size is small for a YSO

Compared to the literature:

Gaia17bpp sets a new record!

future prospects with gaia & LSST

conclusions

1.

We report the discovery of the longest and deepest single dimming event recovered in the literature. The progenitor of the dimming event is not constrained

1.

2.

Both photometric and spectral classification indicates the presence of the primary star being a likely M-giant at Teff~4100 K with a radius of 55.6

3.

Using optical to IR colors, we find that the

color evolution near the optical bandpasses is flat near the photometric minimum. The IR color evolution during the eclipse has a blue color excess from the quiescent star

4.

A systematic search for such dipping events will further shed light on the dimming mechanism. The Gaia Alerts System is already full of a small sample of such rare stellar transients. LSST will likely find more!

extra slides

GSA compared to other

surveys

gsa alert identification

A new source previously not seen and rising to G<19 mag will be considered an alert candidate

New source

Detection

Old Source

detection

Old Source

mean-rms

detection

skewness threshold

Detection

• Egress/ingress -> red
• eclipse -> bluer (than star?!

Case study: E-aur

Several tensions: 

• Mass model 
• Constraints on distance
• Classification 

Summary

Caroll et al. 1991

• 27-year orbital period
• First discovered in the early 1820's if Fritch (german astronomer)
• spectrum does not change during the eclipse
• First theories suggested a stellar eclipse with a semi-transparent star
• E-aur eclipse through the 20th century was considered to be the largest star
• Huang et al. 1965 suggest the disk model theory tilted -- but issues with that came about the whole in the mid-eclipse
• Central star not visible... and some mass transfer (Gibson 2018) using the He I 10 730 Å EW that reveals the disk structure
• Recent data from Gaia DR2 p~ 2.4 mas... d~350-525 parsec (supports small mass ratio (q<1) model
   

aavso.org

Torsten Bronger CC BY-SA 3.1