BAT data for 2017–2026 do not confirm the 20.058-day cycle in IGR J16493−4348 with four block noise tests

We tested the known 20.058-day X-ray cycle in the binary IGR J16493−4348 with later public Swift/BAT data. We used 2,543 daily rates from 2017-06-20 to 2026-10-20. We fit a stable sine wave, time-interval offsets, and four harmonics of the binary orbit. Random-sign residual tests used 20-, 40-, 80-, and 160-day blocks. The fitted semiamplitude is 1.208 × 10^-4 count cm^-2 s^-1. The four p-values range from 0.122 to 0.211. Conservative approximate 95% amplitude bounds range from 2.609 to 2.863 × 10^-4 count cm^-2 s^-1. These bounds depend on the stated block noise assumptions. Recovery tests use the actual late observation times and eight phases. They recover a stable signal with the fitted early amplitude in 85.9% to 96.5% of trials. The bounds still include that early amplitude. Thus the result does not show that the physical cycle stopped. A 19–21 day scan also gives no significant peak under its search-wide tests. Pointing-data fits support the daily-data result. A source-specific date audit finds 4,503 integer MJD labels one day later than YEAR/DAY. The same fields occur in both frozen ASCII and FITS tables. Mission seconds support YEAR/DAY for these rows. This extends a prior SCI audit to this source. The package includes raw inputs, provenance, code, retained simulations, figures, row ledgers, and offline reproduction instructions.

IGR J16493-4348 · Swift/BAT · superorbital modulation · X-ray binary · correlated noise · injection recovery · data quality · time labels · reproducible astronomy