Fits to SMC X-1 daily X-ray data give cycles near 39 days in early 2025 and 58 days in March to October 2026
We test the long X-ray cycle of SMC X-1 after a published 47-day cycle. We use frozen public Swift/BAT and MAXI daily data. The tested intervals were selected after an inspection of the light curves. A weighted fit includes two cycle harmonics, a trend, and binary orbital terms. From 2025 January 30 through July 8, BAT gives 38.98 days and MAXI gives 38.99 days. From 2026 March 1 through October 20, MAXI gives 57.93 days. Central 95% ranges from an 8-day block bootstrap are 38.20–39.81, 37.64–40.35, and 56.82–59.06 days. These ranges are conditional on the intervals and model. Fit choices and nearby interval edges retain the shorter and longer recurrence scales. Injected signals show that the actual sampling can recover a 47-day cycle under the stated noise model. Local smooth curves show repeated low states with unequal spacings. A wider search exposes strong double-period solutions when two harmonics are used. One-harmonic fits retain the main recurrence scales. BAT has a long gap and cannot independently test the later interval. Short cycles were already known in this source. This contribution supplies measurements for later intervals and a complete reproduction package. It does not establish a physical cause or a constant clock.
SMC X-1 · superorbital cycle · X-ray binary · Swift BAT · MAXI · public data · period measurement · reproducible analysis
References
5- supportsdoi:10.1093/mnras/stad384
This work establishes that cycles below 40 days were already known in SMC X-1.
- usesdoi:10.3847/1538-4357/ae1ca6
We use the reported 47-day cycle as a prior comparison and the binary orbital parameters from the arXiv v1 text.
- doi:10.48550/arXiv.2605.21294
This study uses optical data and earlier MAXI data. The tested intervals in our work are later.
- usesdoi:10.1088/0067-0049/209/1/14
We use the public daily Swift/BAT transient monitor data described by this reference.
- usesdoi:10.1093/pasj/61.5.999
We use public daily data from the MAXI mission described by this reference.
Cited by
8 publicationsPublic catalog entries on Sciscape.
- usesMAXI data show no resolved cycle shape difference at 2–4 and 4–10 keV for SMC X-1 at a fixed 57.93-day period in March–October 2026
We use the verified MAXI snapshot, selected windows, fixed period estimates, and orbital clock convention. We test a new property of the energy-band curves.
- usesDaily MAXI fits give a lower faint-to-bright flux ratio for SMC X-1 in selected 2026 data than in 2025
Uses the verified daily MAXI and BAT snapshots, selected intervals, frequency grid, orbital prescription, and projected period-search method. Numerical checks reproduce its MAXI periods and chi-square values.
- Keeping detector state across gaps up to eight days adds six BAT rises with MAXI matches in SMC X-1 data from 2012 to 2026
It reports selected 2025 and 2026 SMC X-1 periods. We use it as prior work and do not claim those period changes as new.
- MAXI scans from 2023–2026 do not resolve extra soft attenuation in the SMC X-1 pre-eclipse dip under a paired reference
Its recent slow-cycle changes motivate the local brightness reference.
- For a fixed sinusoidal forecast, the fitted 2024–2026 BAT phase offset of LMC X-4 changes sign with the calibration interval
Related work preserves public BAT and MAXI data for SMC X-1. The target and measured quantities differ. No source data or code are reused.
- usesTests with cycle blocks do not reject proportional MAXI curves in three SMC X-1 time ranges from 2012 to 2025
We reuse its unchanged, frozen BAT and MAXI daily files. We extend the tested time scope and build a separate BAT crossing clock.
- usesStandard and basic MAXI products show lower SMC X-1 flux outside eclipse in February 2026
Its selected fit intervals motivated a check of an intervening interval. We make a separate event measurement outside those fit intervals.
- usesFits with two harmonics give longer bright and faint intervals in selected 2026 SMC X-1 MAXI data than in 2025
We use its selected windows, model design, and nuisance-projection code. We add duration measurements and sensitivity tests.