A constant-period clock gives higher forecast scores for 2024 TESS and 2025 JWST times of TOI-2109 b under stated priors and errors
We test whether earlier sinusoidal timing models predict later transit times of TOI-2109 b better than a constant-period clock. We use fixed public measurements and two observation-date splits. The first test trains on 85 times through 2023 and scores 21 TESS times from 2024. The second trains on 106 times through 2024 and scores one JWST time from 2025. The forecast calculation includes coefficient uncertainty and alternative signal periods. With the stated Gaussian coefficient priors and reported independent timing errors, the log predictive density ratios for sinusoid over constant period are -0.661 and -3.371. The constant-period forecast thus gives 1.94 and 29.1 times higher predictive density. Earlier fitted BIC gains still favor the sinusoid by 32.53 and 31.74. The full-data evidence also remains in favor of a sinusoid. Changes to the tested period and amplitude priors keep the forecast ordering. Shared facility and year offsets weaken the differences. We provide a source-checked timing table, an audit of seven inconsistent orbit labels, exact forecasts, simulations, code, figures, and reproduction instructions. This result tests the prediction limits of a fixed sinusoidal timing model. It does not exclude a companion or establish a cause for the timing changes.
TOI-2109 b · transit timing · forecast validation · sinusoidal timing model · posterior prediction · public astronomical data
References
6- usesdoi:10.3847/1538-4357/ade057
The public timing CSV at a fixed repository commit supplies the 2020 to 2024 measurement facts. We assign consistent orbit coordinates and test chronological forecasts.
- usesdoi:10.48550/arXiv.2609.26894
Version 1 supplies the factual 2025 JWST transit time and error used as the second test observation. We do not reproduce the paper's full-data fit or its figures.
- doi:10.1051/0004-6361/202451068
This work reported possible sinusoidal timing changes and a possible outer companion. It supplied the CHEOPS measurements used in the later compilation.
- doi:10.3847/1538-3881/ac26bd
This work reported the planet and supplied early observations in the timing compilation.
- A quadratic model with a sinusoid gives a period decrease of 5.14 milliseconds per year for 109 TOI-2109 b transit times sci:sha256:91ce245bd258d9c50b1272ef17f9852159a3c093ae9d3bf2cebbef5a58c3c0e5
This prior SCI audit checked rate conversion and alternative sinusoidal periods in the full 2026 table. The present study adds earlier-to-later forecast tests and integrates over alternative periods.
- Nine of twelve quadratic clocks selected from a 2022 catalog have larger later TESS errors than linear clocks under fixed timing errors sci:sha256:0758e443088192734141f2eaa439d3fc2bd935d58de99112a70788d5677437f3
This prior SCI study tested external transit-clock forecasts across a planet sample. The present study tests sinusoidal models for one system with a joint probability score. No code is reused.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- Local fits to 45 TESS transits preserve the 115-day wave preference in TOI-2109 b timing data with a constant-period clock and independent errors
This prior study tests predictions from the timing table. Our fit preference is conditional and does not replace its forecast tests.
- Single TOI-2109 b times fail a decay test without WASP at two fixed means (2027–2031 weekly grid; five-second errors)
This earlier study tested historical forecasts for later TESS and JWST times under stated priors. We test future observations. Both fitted clocks include a sinusoid with a searched period.