Which star varies? 0.248 d signal blended with a catalogued red giant (TIC 427154884)

Hi all!

I’m an independent researcher in Romania working on TESS light-curve analysis. I’ve been investigating an anomaly at TIC 427154884 (Tmag 12.1, catalogued cool giant, Teff ~4017 K, R ~30 R☉, ~4 kpc, crowded/reddened field), and I’d appreciate advice on whether it’s worth a spectrum.

QLP light curves across sectors 16, 17, 24, 56, 76, 77, 83 (2019–2024) show a clean ~14% peak-to-peak variation at P = 0.2479071 d (orbital 0.4958142 d if binary). A single ephemeris fits all seven sectors to ~0.001 cycle across both the 1800 s and 200 s cadences, and a per-pixel Lomb–Scargle map on the raw TESScut FFI localizes the signal to the target pixel, so it’s astrophysical, not a pipeline artifact. The period is far too short for the catalogued giant. Independent optical photometry does not detect the signal at the target: ZTF (~2″, injection-validated in g) and ATLAS forced photometry (o-band, 3023 epochs, the red band matched to TESS) both show no power at the period, and a 2MASS / AllWISE / Gaia search finds no resolved hidden source.
The evidence points to an unresolved (within ~0.7″), very-red, short-period variable blended with the giant, plausibly a foreground red-dwarf eclipsing/contact binary, though a genuinely single anomalous star can’t be excluded from photometry alone.

What would settle it is a single spectrum (composite / two-lined signature, or RV variation on the 0.2479 / 0.4958 d cycle), or sub-arcsecond near-IR imaging. Full analysis, figures, and methods are in the attached report and on ExoFOP-TESS (data tag 477311,(https://exofop.ipac.caltech.edu/tess/target.php?id=427154884).
Happy to answer questions or share the light curves. Any thoughts on feasibility for a small-telescope spectrograph, or pointers to who might be a good match

Thanks,
Munteanu Sabin.

The actual variable is the EW-type binary NSVS 3278320, located 54" away:

When working with TESS you have to be careful of blending at distances up to 2 or 3 arc minutes.
If you had not detected the signal in ZTF data, that would already discard the unsolved close binary scenario because the ZTF data should also be blended and the signal should be seen there too.

Best wishes,
Sebastian

1 Like

Dear Sebastian,

Thank you for sharing this with me. This information will be incredibly helpful for evaluating other potential targets in the future.

Best regards,
Sabin.