An Exomoon Survey of 70 Cool Giant Exoplanets and the New Candidate Kepler-1708 b-i

Not checked yet

• All115 supplied v1 pages have reading evidence; this closes the supplied manuscript cursor. The stellar-table caption now identifies isochrone medians and68.3% intervals for69 hosts/70 candidate IDs. Physical/journal identity, later-version collation, raw data, code, posteriors, cited originals, official award records and independent numerical or fine-glyph checks remain unreviewed.
• The tentative moon, dependent Kepler treatments and injections, selection/noise/priors and ordinary alternatives remain. New methods show overlapping catalogues and a false-positive versus false-negative tradeoff. Different TTV forecasts have different stated bases, with no measured follow-up. Native radius units, KIC identifiers, table labels, repeated bibliography and ambiguous J.S. contribution roles remain unresolved; frozen name-encoding variants do not replace established authors.
• Public acknowledgments name D.K. for NASA80NSSC21K0960 and J.Sz. for ERC/Horizon2020 948467; PLEIADES analysis uses HEC-SMD-17- 1386. Mission and Gaia/DPAC funding, computing support and26 thanked donors are separate roles. Legal award recipients, PI, amounts and dates remain unspecified or unverified. These acknowledgments and a correspondence address establish no private approach, achieved programme access or custody. No whole-paper or ET/UAP proof follows.
• January12,2022 is the submission/native v1 margin date; PDF creation January14,2022 and HTTP modification January23,2023 remain separate layers. Earlier native FPP wording conflicts with its spurious-signal discussion, and the opaque HTTP ETag differs from the measured byte hash. These unresolved issues, original journal/physical fidelity and fine-glyph limitations persist; no silent semantic or identifier repair is claimed.
• New methods retain dependent detrendings, selection and noise assumptions,69 stars/70 candidates, differing stellar inputs, model priors, density and stability cuts, transit-biased Beta assumptions and numerical evidence estimates. False-positive and false-negative exclusions do not establish moon absence. Ordinary blend alternatives can favor marginal evidence despite a better moon maximum likelihood. Computation was reported, not reproduced.
• A blend Bayes factor6.8 and native51/72 versus70 denominator wording remain qualified. Density equations, normal/log/uniform priors and KDE boundary estimates are model-dependent. Annotation glyphs, identifiers, dimensional wording and raw-data/posterior/high-resolution originals remain unreviewed or unresolved. No new funding, private approach, achieved programme access or ET/UAP proof is established.
• Conservative candidate rejection does not exclude a moon or establish a physical spot model. Eight reductions share Kepler data and fitted templates; pixel nulls and known-star PRF localization do not exclude all noise or unknown blends. Planet false-positive probability0.024% is separate from moon validation. Occultation limits330/350ppm have distinct stated methods. Page64 two-planet wording and signs -16.2/+7.0 versus earlier moon+23.2 remain unresolved; the manuscript tail is now read, while raw/code/cited-apparatus originals remain unread. No new funding or approach claim occurs.
• The1% false-positive rate from200 selected same-star null injections is not a99% moon posterior. Rejected realizations were retried and Gaussian-process detrending omitted. Occurrence priors and target-specific true/false-positive rates remain unknown or heterogeneous; finite outside-transit Gaussian-noise nulls do not prove every likelihood assumption. The authors cannot calculate accurate candidate odds. March24,2023 is a dated forecast, not an observed transit or current schedule.
• Native page65 repeats the same evidence subscript on both sides of a stated difference; prior page64 wording/sign conflict persists. Page68 prior wording, page71 start-time wording and Equation7’s printed sum remain literal, without silent correction. Native page79 folio is uncertain despite extracted footer79. Bibliography, cited figures/tables, code/data and physical/journal-version fidelity remain distinct gaps; no new award recipient or private approach appears.
• Extended figures compare shared Kepler reductions and fitted models. KDE resampling does not count new planets; eight detrendings are dependent. Blend marginal evidence/BIC and spot AIC can favor ordinary alternatives despite better moon chi-square. Six target-pixel panels locate a template signal without proving its reality or excluding all noise. Data-on-request and software links are offers, with no request, receipt or execution.
• SupplementaryTable1 retains70 candidate rows and distinct tick, cross, blank and N/A states;11 numerical secondary candidates are not confirmed moons. The previously partial stellar table is now completed through its caption on98, with69 hosts for70 candidate IDs; KIC9079767 continuation was read. Stellar medians and intervals are model inputs, not a new object census. Earlier extracted graph/table controls are nonprose artifacts.
• Terminal figures reuse Kepler/Gaia/models and DR25 selection. Planet FPP1in4237 and rounded posterior1.000 are separate from moon existence. Two injected false positives103/161 have positive fitted moon radii and strong Bayes factors. Selected Gaussian/null compatibility does not prove universal noise assumptions. Small fitted centroid offsets and blend probabilities are model outputs, not removal of every alternative.
• Page109 repeats top-left in its caption although the likelihood panel is top-right; page115 retains a visible[ht] artifact. Earlier sign, prior, glyph and date conflicts remain. Text character counts describe LF-normalized returned text, with unchanged raw CRLF file bytes independently hashed. Earlier raw-substring wording is explicitly qualified in the new private receipt; no files were altered or pages recredited.