Peak frequency has a negative fitted phase slope in 104 CHIME bursts of FRB 20180916B with a linear time term

We tested whether fitted burst peak frequency is associated with activity phase inside the CHIME 400–800 MHz band. CHIME/FRB Catalog 2 release 20251231 gives 113 events for FRB 20180916B. Quality flags leave 104 events in 54 activity cycles from 2018 to 2023. Each event has one weight and uses its median component peak. With a 16.34-day reference period, fixed cycle counts, and a linear calendar-time term, the fitted slope is -1.142 days per frequency doubling. The nominal 95% cycle-cluster interval is -1.712 to -0.571 days. A model with separate cycle intercepts gives -1.637 days, with an interval of -2.847 to -0.427, from 77 events in 27 cycles. Whole-cycle bootstrap intervals are also negative. All 54 fits that remove one cycle have a negative slope. The raw rank correlation is weak, and the 20 events in 2022–2023 do not establish the trend alone. We prove that a linear time term makes the phase slope invariant to a trial period when reference cycle counts are kept. This property does not confirm the clock. The result extends a known association across observing bands to a measured-peak catalog test inside one band. It is exploratory and conditional on detection, the spectral fits, and the selected model. Beam response, exposure, and selection effects prevent an intrinsic activity-window or physical-mechanism claim. We preserve the event audit, provenance, code, results, figures, and reproduction instructions.

FRB 20180916B · CHIME/FRB Catalog 2 · fast radio bursts · activity phase · spectral peak · cycle clusters · period sensitivity · reproducible astronomy