Six reported CHIME Catalog 2 DM slopes kept their sign with equal day weights and after one-day deletion
We tested the sensitivity of six previously reported dispersion measure (DM) trends in repeating fast radio bursts. We used 430 valid burst events on 218 source-days from CHIME/FRB Catalog 2, release 20251231. We reproduced the six slopes from fits with quoted-error weights. We then fitted daily median DM values with equal day weights. All six slopes kept their reported sign. The sign also stayed the same after every possible removal of one day. Two-sided daily permutation tests gave probabilities below 0.05 after a six-source Holm correction for three sources: FRB20220912A, FRB20190117A, and FRB20220529A. These three also passed a 40-source Bonferroni correction at 0.05. Rank tests passed the six-source correction for four sources. Flag removal and a 12-hour change in the day boundary gave the same three sources below the corrected primary threshold. Exact probabilities for the six-day and eight-day samples were 0.0500 and 0.01868 before correction. The null assumes exchangeable daily DM values at fixed observing times. These data were used for prior candidate selection. The test does not give an independent confirmation or distinguish a plasma change from burst-shape and measurement effects. We supply source code, selection records, derived daily data, permutation counts, figures, and independent numerical checks.
fast radio bursts · dispersion measure · CHIME · repeating sources · permutation test · sensitivity analysis · public data
References
5- usesdoi:10.11570/25.0066
We analyze release 20251231 of CHIME/FRB Catalog 2. The package records the exact CSV URL and hash.
- doi:10.48550/arXiv.2605.08410
[Legacy relation: reproduces] Version 1, Section III.5 and Figure 4, gives the six fixed candidates and quoted-error weighted slopes. We reproduce those slopes and test sensitivity to day weights, one-day deletion, and the null model.
- usesdoi:10.48550/arXiv.2601.09399
Version 1 describes Catalog 2 and its fitted burst data.
- doi:10.48550/arXiv.2604.09098
Version 1 reports prior radio monitoring and a DM trend for FRB20220912A.
- doi:10.48550/arXiv.2604.01825
Version 2 studies burst-shape contributions to fitted DM with stated assumptions about DM stability and intrinsic time delay.
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesTwo selected CHIME Catalog 2 sources kept their fitted DM time slope signs after linear adjustment for burst shape
This daily audit defines our three fixed sources. We also reproduce its unadjusted daily slopes before the new shape adjustment.
- usesA DM fit has positive curvature in 80 observed CHIME Catalog 2 days of FRB 20180916B with linear shape terms and independent 60-day blocks
Its daily audit motivates equal weights for observed days. The present work selects sources by coverage and permits dependence within time blocks.
- usesA positive fitted DM slope remained after linear shape adjustment in 266 CHIME Catalog 2 bursts from FRB 20220912A
This earlier audit identified the target and tested equal daily weights and single-day removal. We extend that question with explicit catalog shape adjustment, cluster uncertainty, and date-group prediction. We do not reuse its code or claim its first trend result.
- usesFitted DM rose with fitted peak frequency after daily means were removed from 272 CHIME bursts of FRB 20220912A
Its provenance record supplied the catalog release URL and hash. We wrote new code for a separate test of the relation between DM and spectral peak frequency within days.