Biswajit Jana · Independent research

Seven-sector TESS timing study · 2026

WASP-19 b One day in the life of an irradiated giant

A reproducible transit and ephemeris analysis built from seven unmodified TESS SPOC light curves. The purpose is not to announce orbital decay, but to measure what TESS alone can—and cannot—resolve.

Timed transits204
Timing baseline2,528 d
Conditional Ṗ−1.38 ± 4.02 ms yr⁻¹
Model preferenceΔBIC = −5.20
Artistic concept of the hot Jupiter WASP-19 b near its active host star
Artistic concept—not an observation and not to scale. The strongly heated hemisphere, dark circulation bands, active late-G host and thin blue-gray scattering limb translate measured system properties and the cross-instrument atmospheric literature into a visual hypothesis. The detailed colors and cloud map are interpretive; disputed TiO claims are deliberately not visualized as fact. Visual basis and provenance.
01 · Principal result

The simpler ephemeris remains preferred.

The seven-sector dataset independently times 204 transits across 2,528 days. A quadratic ephemeris gives a conditional period derivative of −1.38 ± 4.02 ms yr⁻¹, but its additional curvature parameter is not justified by the likelihood gain: ΔBIC(linear − quadratic) = −5.20.

Under the stated Gaussian and equilibrium-tide assumptions, the 95% negative bound is Ṗ > −9.26 ms yr⁻¹, corresponding to Q′★ > 3.0 × 105. This is a sensitivity limit—not a tidal-decay detection.

Observed-minus-calculated timing measurements and residuals for 204 WASP-19 b transits
Figure 1. Top: observed-minus-archive timings with beta- and sector-jitter-inflated uncertainties; linear and quadratic ephemerides are overlaid. Bottom: residuals from the preferred linear model. Machine-readable values are available in the event-timing and model-statistics CSV files.
02 · Robustness

No single sector determines the timing conclusion.

Seven leave-one-sector-out refits probe sensitivity to individual observing campaigns. Every deletion prefers the linear model: ΔBIC ranges from −4.84 to −5.17, while Ṗ ranges from −2.66 to −0.17 ms yr⁻¹. This test addresses sector leverage; it does not capture every shared systematic.

Leave-one-sector-out period derivative estimates for seven TESS sectors
Figure 2. Conditional Ṗ estimates after deleting each sector. Horizontal bars show one-standard-deviation uncertainties; annotations give ΔBIC(linear − quadratic). The vertical rust line and band show the full seven-sector estimate.
03 · Measurement chain

From archive light curves to an auditable timing bound.

01

Homogeneous input

Seven target-specific, 120-second SPOC light curves are kept unmodified. Mixed-cadence and FFI products are excluded to avoid duplicate or heterogeneous measurements.

02

Sector-level transit

Each sector receives a bounded circular, quadratic-limb-darkened fit with timing freedom and a local baseline. Residual time-averaging beta inflates formal errors.

03

Event timing

The sector shape is fixed while midpoint, intercept, and slope are fitted for every transit. Empirical sector jitter prevents a few formally precise events from dominating.

04

Ephemeris comparison

Conditioned weighted least squares compares linear and quadratic O−C models. BIC penalizes the quadratic model's extra coefficient.

05

Deletion analysis

Both ephemerides are refitted seven times, omitting one sector per run. The full inference is retained alongside the deletion ensemble.

06

Fail-closed checks

SHA-256 tests verify every FITS file. Regression tests rerun the real-data analysis and require all declared figures and tables.

Seven independently fitted TESS transits of WASP-19 b
Figure 3. Independent sector fits after timing adjustment.
Residual root mean square versus bin size for seven TESS sectors
Figure 4. Time-averaging residual diagnostic used to estimate correlated-noise inflation.
Seven TESS sector depth estimates and combined depth
Figure 5. Sector depths are compatible after error inflation: combined model depth 22,557.2 ± 154.2 ppm; Cochran Q = 5.42 for 6 degrees of freedom (p = 0.491).
04 · Literature context

The atmospheric evidence is not one settled detection.

WASP-19 b illustrates why cross-instrument reproducibility and stellar activity matter. The studies below use different instruments, epochs, wavelength ranges, reduction choices, and model spaces. Broadband TESS photometry cannot adjudicate molecular abundances, so this repository does not turn those claims into a new retrieval.

StudyReported resultInterpretive scope
Huitson et al. (2013)H₂O at 4σ; low or absent TiOHST optical-to-near-IR transmission spectrum; active-star treatment is central.
Sedaghati et al. (2017)TiO, haze, Na and H₂O reportedCombined VLT/FORS2 and HST evidence; historically influential but not reproduced uniformly.
Espinoza et al. (2019)Featureless optical spectrum; subsolar TiO/NaMagellan/IMACS analysis and joint interpretation that challenges a simple strong-TiO account.
ESPRESSO study (2021)TiO constrained to roughly 100× subsolar; blue scattering retainedHigh-resolution optical evidence with a different sensitivity and model space.
05 · Open materials

Every public number traces to a file or script.

Observations and provenance

Seven target-specific TESS SPOC light curves are stored as received. Exact archive identifiers, retrieval date, product-selection rule, and SHA-256 digests are recorded in data/SOURCE.md. System values come from the committed NASA Exoplanet Archive row.

pip install -r requirements.txt
python scripts/analyze_transit.py
python scripts/analyze_multisector.py
python scripts/analyze_timing_limits.py
pytest tests/ -v
06 · Scope

What this analysis does not establish.

  • The orbit is circular and quadratic limb-darkening coefficients are fixed representative values.
  • Composite semi-major axis and stellar-radius uncertainties are not propagated.
  • Midpoint freedom introduces a bounded timing search; BIC is used as the support gate.
  • PDCSAP processing, dilution and stellar variability can bias inferred geometry.
  • Sector-level detrending creates shared systematics among event timings.
  • Empirical jitter reduces formal overconfidence but does not make timings independent.
  • The Ṗ interval is conditional on a quadratic ephemeris and diagonal timing errors.
  • The analysis cannot reject apsidal precession or replace longer heterogeneous timing baselines.
07 · Sources

Primary references.

  1. Hebb et al. (2010), discovery paper for WASP-19 b.
  2. Ricker et al. (2015), Transiting Exoplanet Survey Satellite, doi:10.1117/1.JATIS.1.1.014003.
  3. TESS Team, TESS Light Curves — All Sectors, MAST, doi:10.17909/t9-nmc8-f686.
  4. Petrucci et al. (2020), Discarding orbital decay in WASP-19b after one decade, arXiv:1910.11930.
  5. Rajkumar et al. (2026), Long-term monitoring of WASP-19 b, doi:10.1051/0004-6361/202556822.
  6. Huitson et al. (2013), HST optical-to-near-IR transmission spectrum of WASP-19 b, doi:10.1093/mnras/stt1243.
  7. Sedaghati et al. (2017), Detection of titanium oxide in the atmosphere of a hot Jupiter, doi:10.1038/nature23651.
  8. Espinoza et al. (2019), ACCESS: a featureless optical transmission spectrum for WASP-19 b, doi:10.1093/mnras/sty2691.