A search of public NASA TESS data for transiting planets, eclipsing multiple stars and unusual dimming events that had not been reported, run end to end from the MAST open-data archive. Its principal result is the planet candidate described below. Nothing in this repository is a confirmed planet.
R. Vora (2026), A 36.34-day Sub-Neptune Candidate Transiting the Nearby K Dwarf TIC 402710280, Identified in TESS Full-Frame Images. Preprint: doi:10.5281/zenodo.23390317 (
paper/tic402710280.pdf). Data and code: doi:10.5281/zenodo.23390274. Pre-registered test of the 2027 transits: doi:10.5281/zenodo.23390325.
| Host | TIC 402710280 (TYC 9041-03677-1), K dwarf, T = 9.24, 66.8 pc |
| Period | 36.33986 ± 0.00051 d, from four transits (TESS Sectors 39, 93, 100, 102) |
| Radius | 3.75 ± 0.22 R⊕ (Rp/R★ = 0.0421) |
| Equilibrium temperature | 498 ± 21 K |
| False-positive probability | 0.003 ± 0.004 (TRICERATOPS); no high-resolution imaging yet |
| Next transits | 2027 Feb 19 (TESS Sector 114); 2027 May 3, 03:41 UTC (full night at La Silla) |
The preprint also reports a 407 ± 26 ppm signal with a period of 3.78 days whose source (the target or a background star 3.5″ away) the present data cannot identify.
| Path | Content |
|---|---|
paper/ |
Manuscript (AASTeX): tic402710280.template.tex is the source; pipeline/paper_numbers.py fills every number from the result files into tic402710280.tex. Figures in paper/figures/. |
release/tic402710280_zenodo.zip |
The data-and-code record for Zenodo: light curves of every transit window, our photometry of Sectors 100–103, transit times, posterior summaries and chains, all vetting results, the pipeline snapshot and the manuscript. |
release/prereg_tic402710280/ |
Pre-registered test of the 2027 transits: protocol, frozen test code, predictions and validation. |
release/ZENODO_UPLOAD.md |
Upload order and metadata for the three Zenodo records. |
Reproducing the paper from pipeline/ (after the Sector 100–103 light curves exist, ffi_new.py):
python audit_402710280.py # preliminary times, Sector 66 diagnostics, combined centroid
python fit_new.py # transit shape, re-timing, ephemeris, predictions
python shape_chains.py free # free limb darkening
python timing_diagnostics_402710280.py # baseline/window/flux-series timing tests, variability
python secondary_scan_402710280.py # secondary eclipse over phases 0.3-0.7
python blend_shape_402710280.py # diluted-eclipse test of the 3.5" neighbour
python fpp_triceratops.py 10 # TRICERATOPS
python second_signal_402710280.py # the 3.78-day signal; then fpp_second_signal.py
python paper_figures.py && python paper_numbers.py && (cd ../paper && latexmk -pdf tic402710280.tex)
python build_release.py && python build_prereg.py
| Track | Looking for | Why it might still be hidden |
|---|---|---|
| A | Small planets (≈1–2 R⊕) around nearby K and M dwarfs | Stars with only full-frame data, or few two-minute sectors, were never searched across all their sectors |
| B | Long-period planets seen as single or "duo" transits | Standard searches need several transits per search window |
| C | Eclipsing triples and 2+2 quadruples | Extra eclipse sets appear only after the main binary is subtracted |
| D | Deep aperiodic or asymmetric dips | No standard pipeline looks for them |
The search covered 10,324,827 public light curves (QLP full-frame light curves of Sectors 1–98 and SPOC two-minute
light curves of nine sectors), plus our own photometry of the 2026 full-frame images (Sectors 99–108). Every search
is repeated on light curves flipped upside down, so that the false-alarm rate is measured rather than assumed. The
methods are described in docs/METHODS.md; the survey report is
results/report/index.html.
All inputs are public and are read directly from the MAST open-data bucket on AWS (s3://stpubdata): SPOC
two-minute, TESS-SPOC and QLP light curves, the calibrated full-frame image cubes (via astrocut), the TESS Input
Catalog v8.2 and Gaia DR3. Known objects come from ExoFOP (TOI and community-TOI tables), the NASA Exoplanet Archive,
SIMBAD, VizieR and NASA ADS.
| Path | Content |
|---|---|
pipeline/ |
All analysis code; pipeline/README.md groups the scripts by role. |
results/ |
Compact, committed outputs: per-sector survey tables, candidate lists, vetting results, dossiers and the report. |
docs/ |
Methods, and the brief used to run the survey in parallel worker sessions. |
Bulk data (light curves, catalogue extracts, caches) are kept outside the repository, in $ADASTRA_WORK
(default: ../work next to the repository).
python3.13 -m venv .venv && .venv/bin/pip install -r requirements.txt
export ADASTRA_WORK=/path/to/scratch # optional
Code: MIT (LICENSE). Data products, figures and text: CC BY 4.0 (LICENSE-DATA).
To cite this work, see CITATION.cff.
The analysis code and the manuscript were developed with the assistance of an AI system (Claude, Anthropic); the author directed the work and is responsible for its content.