TIC 237226943
An uncatalogued eclipsing-binary candidate from TESS.
Eclipsing-binary candidate: 1.9983-day period, a 1.76% primary eclipse, no significant secondary eclipse. Automated cross-checks question this period - see below.
TIC 237226943 is a candidate eclipsing binary star with an orbital period of 1.9983 days, detected through automated analysis of TESS photometry. The primary eclipse produces a 1.02% dimming signal with a detection significance of 168.5 in the BLS periodogram, while no secondary eclipse was detected. Pixel-level analysis yielded a difference-imaging signal-to-noise of 39.2, though the centroid offset of 1.0 arcsec from the target raises an important consideration: TESS's resolution cannot confidently separate this star from a nearby bright neighbor, with the centroid separation of only 0.051 pixels falling within the instrument's stated systematic uncertainty. The candidate does not appear in six major variable-star catalogs, and the pipeline's machine-learning model assigns it a 0.62 screening score. Further observations would be needed to definitively confirm whether the signal originates from the target star itself.
- Surface brightness ratio
- companion faint - no secondary eclipse
- Companion temperature
- below ~2,860 K
- Companion / star radius
- ~0.13
- Face hidden at primary eclipse
- ~2%
- Separation
- ~8.6 x the star's radius
- Eccentricity
- consistent with circular
Zoomed: both stars drawn about 2x larger than their real size relative to the orbit, so they overlap for longer than the real eclipse lasts. A star's surface never gets dimmer during an eclipse - part of it is simply hidden. What drops is the total light we receive, shown exactly by the readout and bar below from the real light curve; the glow dims with it, exaggerated so small dips show. Colours follow temperature on a stellar colour scale, saturated so the differences show; the star's is its catalogue value (6,041 K). The companion's colour, sizes and orbit shape are estimates from the eclipse depths, duration and timing, assuming central eclipses and spherical stars - good to perhaps tens of percent. Only the relative orbit is shown; both stars really circle their common centre of mass. Animated about 7,194x faster than real on average, slowing down through each eclipse, and starting from where the system is in its orbit right now. Hover to pause.
- Secondary2026-10-19 08:46 UTC
- Primary2026-10-20 08:45 UTC
- Secondary2026-10-21 08:43 UTC
Computed from the fitted ephemeris. Approximate - not corrected for your location or for light-travel time across Earth's orbit.
The signal
A repeating, consistent dip in brightness.
Residual scatter 1.169% (observed minus binned median).
Automated cross-checks question this period - see Caveats below.
- Period
- 1.99834 d
- Primary depth
- 1.76%
- Epoch (T0)
- 1765.997 BTJD
- Duration
- 2.02 h
The catalog check
Cross-matched against known variable-star catalogs to rule out anything already listed.
- VSXNot found
- ASAS-SNNot found
- Gaia eclipsing-binary tableNot found
- Gaia variable classificationNot found
- TESS EB catalog (Prsa+2022)Not found
- ExoFOP TOI listNot found
0/6 catalogs list a known variable at this position.
No known variable star or catalogued match at this position.
- Catalogs checked
- 6
- Matches
- 0
The pixel check
Confirms the dimming is on the target star, not a nearby source.
Out/in share one sqrt colour scale; in-eclipse = out-of-eclipse minus difference. TESS sector 85.
- 39.2
- 0.05 px1"
- -0.0150 phase
- TIC 18845740480.10 px, dT 4.4
- 0.91
- 85
The dimming centroid sits almost as close to a bright neighbour as to the target.
- Centroid offset
- 0.05 px
- Difference SNR
- 39.2
- Capable neighbour
- TIC 1884574048
- TESS sector
- 85
The Gaia view
A second spacecraft's measurements: how far away the star is, how bright and what colour, and whether its position fits a single star.
The latest Gaia context file does not include this candidate yet, so there is no Gaia view to show.
The ground-survey check
An independent instrument, different years: does real ZTF photometry show the eclipse at the period TESS measured?
- ZTF g (723 pts, 1 outlier)Too few outliers to test
- ZTF r (732 pts, 2 outliers)Confirmed (p=0.0100)
Folded at the TESS-measured period of 1.99834 d, at least one band shows individual ZTF epochs - real observations, years apart - landing far below the baseline brightness and clustered in phase far tighter than chance would produce given how many such epochs that band has (p < 0.01, chance-corrected for the outlier count).
Source: ZTF public light-curve archive (IRSA), independent of the TESS data used to detect this candidate.
Real ZTF ground-survey photometry backs this period: one band shows epochs, years apart, landing more than 5 sigma below the baseline brightness and clustered in phase far tighter than chance would produce given how many such epochs that band has (p < 0.01) - a different instrument finding the same eclipse.
- Bands confirmed
- ZTF r
- Deepest clustered dip
- 77.0 sigma (ZTF r)
- Chance p-value
- 0.0100 (ZTF r)
- Period tested
- 1.99834 d
Fold it yourself
Adjust the period and watch the eclipse appear.
Shown at the true period. The raw light curve needed to re-fold at other periods has not been exported yet - this slider unlocks automatically once it is.
The period the pipeline measured. Drag the slider away from it and the eclipse smears out.
- True period
- 1.99834 d
- BLS period
- 1.99834 d
- Odd/even z
- 1.94
Caveats
Known limitations and things to check.
Cross-check:No eclipse at 16 of 229 covered epochs where the 1.99834 d period predicts one (each under half the typical depth, > 5 sigma below it). The period is probably wrong - the dips seen may be separate eclipses of a longer-period system.
- Thin margin:The ON_TARGET call rests on a neighbour-competition margin of just 0.051 px (dimming centroid 0.049 px from target vs 0.101 px from the nearest bright-enough neighbour) - inside diag_pixel.py's own stated centroid systematic of 0.10-0.3 px. TESS's resolution cannot confidently separate this star from that neighbour.
VIIIInstrument
Raw values, fits and technical details.
Instrument
Raw values, fits and technical details.
Full statistics
- 1.99834 d
- 1.99834 dsame
- 1.76% +/- 0.03%seen 198/229
- ML screening score
- 0.625logistic_regression
- < 0.06%none at 5 sigma
- 1.94< 3
- 0.3996 dpower 0.01
- 2.0 h
- Distance (TIC, Gaia-based)
- 1106 pc+/- 24
- 13.34 / 12.90
- 6041 K / 2.06 Rsun
- RA / Dec (J2000)
- 19h 47m 40.1s +77d 56m 44s
- 168.5 / 39.2
- 162,767 ptsover 1898 d
Eclipse close-ups
Same flux scale on both panels. Line: binned median. Shown in place of an orbit animation - this pipeline measures eclipses, it does not fit an orbit.
Period alias analysis
Odd/even depth z = 1.94 (gate 3.0): alternate eclipses match, so the BLS period stands.
O - C timing
181 eclipses timed, rms 15.71 min, reduced chi-square 325.34 (179 dof): more scatter than the timing errors explain - worth a closer look.
Each eclipse timed by fitting the star's own mean eclipse profile; fitted P = 1.998183 +/- 0.000000 d.