Comment on "Testing claims of the GW170817 binary neutron star inspiral affecting β -decay rates''
CComment on “Testing claims of the GW170817 binary neutron starinspiral affecting β -decay rates” E. Fischbach a, ∗ , D. E. Krause b,a , M. Pattermann a a Department of Physics and Astronomy, Purdue University, West Lafayette, IN 47907, USA b Physics Department, Wabash College, Crawfordsville, IN 47933, USA
Abstract
We show that recently reported radioactive decay data support an earlier observation of anunexpected signal associated with the GW170817 binary neutron star inspiral.
Keywords:
Neutron stars, radioactivity, neutrinosIn Ref. [1] experimental evidence was reported for an unexpected signal associated withthe GW170817 binary neutron star inspiral [2, 3]. As noted there, this evidence derivedfrom a laboratory experiment in which the decay rates of the isotopes Si-32 and Cl-36 weremeasured quasi-simultaneously in alternating time periods in a common detector. The unex-pected signal described in Ref. [1] was a surprising correlation at the 2 . σ level between theSi-32 and Cl-36 decay rates during the 5-hour period following the inspiral, which contrastswith the absence of any significant correlation either before or after this period. In whatfollows we show that, when properly analyzed, data reported recently in Ref. [4] exhibit acorrelation similar to what was reported in Ref. [1] during the same time interval. Giventhat the data reported in Refs. [1] and [4] were acquired from two experiments that werespatially separated by several thousand kilometers, the new data of Refs. [4] lend additionalsupport to the indications in Ref. [1] of an unexpected signal associated with the GW170817binary neutron star inspiral.To place Ref. [4] in proper context, there are by now numerous published reports oftime-varying nuclear decay rates. These are manifested, for example, by annual variations ofdecay rates attributable to the annually varying Earth-Sun distance, in a significant numberof different nuclei in experiments using different techniques. (See Ref. [1] for references.)On the other hand, there have also been numerous claims that such effects have not beenseen [5]. What is important in evaluating the various different claims is recognizing thatsince there is at present no theory to explain such effects, there is also no framework thatwould allow for a direct comparison of one experiment to another especially when usingdifferent nuclei. In particular, this implies that experiments utilizing different radionuclidesmay have very different intrinsic sensitivities to external perturbations, and this observationis a central issue in combining the data in Refs. [1] and [4]. ∗ Corresponding author
Email address: [email protected] (E. Fischbach) a r X i v : . [ nu c l - e x ] F e b t follows from the previous discussion that in planning a new experiment to searchfor possible variations in decay rates associated with astrophysical phenomena, one shouldcarefully choose radionuclides where time-varying decay rates have previously been observed.In Ref. [1], the radionuclides under study were Si-32 and Cl-36, both of which had previouslybeen shown to manifest time-varying decay rates [6, 7].In contrast, the same cannot be said for all the radionuclides studied in the Ref. [4]: Tostart with, Cs-137 has been studied in great detail in Ref. [8] based on long-term measure-ments carried out at the Physikalisch-Technische-Bundesanstalt (PTB) in Germany over the9 year period 1999–2008. The authors find that “. . . the PTB measurements of the decayrate of Cs-137 show no evidence of an annual oscillation. . . ”. The same conclusion for Cs-137 arises from Ref. [9] which analyzed 5.4 years of data from a sample of Cs-137 which wason board the recent MESSENGER mission to Mercury. Their conclusion for the relevantparameter ξ was consistent ξ = 0, i.e. no effect for Cs-137.Turning next to Ti-44, the only directly relevant data are those of O’Keefe, et al. [7].This reference reports evidence in their Fig. 1 of an annual oscillation in the ratio of decayrates Na-22/Ti-44. Although we cannot disentangle the respective contributions of Na-22and Ti-44 to the ratio, we cannot rule out that Ti-44 contributes at least part of the signalfor the observed annual oscillation, in which case the Ti-44 data in Ref. [4] are relevant inthe present context.Finally we consider the data reported in Ref. [4] on Co-60. This nuclide, which has beenextensively studied by Parkhomov, et al. [10], exhibits a clear annual variation, and hencethe Co-60 data are also directly relevant for the present discussion.Turning now to Fig. 1 and Table 1 of Ref. [4], which summarized their results, we canimmediately set aside the Cs-137 data for the reasons discussed above. We are then left witha figure in which the Ti-44 and Co-60 data in the relevant inspiral shaded region in Ref. [4]exhibit correlations which appear to be strikingly similar to the Si-32 and Cl-36 correlationsin the corresponding inspiral shaded region in Fig. 2 of Ref. [1].In Fig. 1 below we exhibit both Fig. 2 of Ref. [1] and Fig. 1 of Ref. [4]. The data forTi-44 and Co-60 have been obtained from the arXiv version of Ref. [4], after setting aside theCs-137 data, as discussed above. Specifically, there is now an evident correlation among thedata for the four relevant isotopes (Si-32, Cl-36, Ti-44, and Co-60) which have previouslybeen shown to evidence periodic variations in their respective decay rates. (This correlationis already clearly evident in Ref. [4] itself, once the Cs-137 data are set aside.) Althoughthe timing of the peaks and troughs in Ti-44 and Co-60 data do not correspond precisely tothose in Ref. [1], this is most likely due binning effects, and to the fact that the Si-32 andCl-36 data were acquired in the same detector but in slightly shifted time intervals.Although Ref. [4] suggests that the authors may be disputing the effect observed inRef. [1], the preceding analysis suggests that their data actually support the observationsreported previously in Ref. [1]. Intuitively it appears unlikely that two sets of observations ofa correlation in decay data in the very same ∼ igure 1: Shown in this figure are the data from Ref. [1] for Si-32 and Cl-36 (top), and the data for Ti-44 andCo-60 obtained from the arXiv version of Ref. [4] (bottom). The gray shaded region denotes the ∼ ∼ . σ , while Ti-44 and Co-60 are correlated at ∼ . σ as discussed in text. that we are considering are confined to data acquired from 2 disparate experiments overexactly the same time interval.In summary, it appears that the data in Ref. [4], when properly analyzed, may actuallysupport the earlier observation published in Ref. [1] of an unexpected correlation in the decayrates of two radionuclides in the experiment during the same time interval encompassingGW170817. If we set aside the Cs-137 data from Table 1 of Ref. [4] for the reasons discussedabove, the statistical significance of the Ti-44/Co-60 correlation appears to be in the rangeof σ = (1 . σ , as reported in Ref. [4]. Although the authors of Ref [4] reduce thesignificance of their own correlation in the post-inspiral region from an initial value of 2.4 σ to 1.9 σ because of the “. . . arbitrary choice of the time interval. . . ”, this reduction seemsinappropriate given that this interval is, in fact, dictated by Ref. [1].Although the difference in the experimental setups in Refs. [1] and [4] preclude formallycombining their respective data sets in the inspiral region to infer a combined effectivesignificance of σ e , we can presume that the joint probability of detecting these two inde-3endent sets of correlations in the same time interval has a statistical significance that isat least greater than the σ = 2 . σ significance quoted in Ref. [1]. (A heuristic estimate, σ e ∼ (cid:112) σ + σ , would yield σ e ∼ . σ using the data of Refs. [1] and σ = 2 . σ. )We conclude by noting that since other groups may have also acquired relevant data dur-ing the GW170817 inspiral period from nuclei that have previously exhibited decay anoma-lies, we encourage members of the community to revisit their data for possible signals duringthis period. Acknowledgements
We are indebted to Virgil Barnes, Laura Cayon, Tim Meese, and Jeff Scargle for helpfuldiscussions.
References [1] E. Fischbach et al., Indications of an unexpected signal associated with the GW170817 binary neutronstar inspiral, Astropart. Phys. 103 (2018) 1.[2] B. P. Abbott, et al., GW170817: Observation of Gravitational Waves from a Binary Neutron StarInspiral, Phys. Rev. Lett. 119 (2017) 161101.[3] B. P. Abbott, et al., Multi-messenger Observations of a Binary Neutron Star Merger, ApJL 848 (2017)L12.[4] P. A. Breur, et al., Testing claims of the GW170817 binary neutron star inspiral affecting β -decay rates,Astropart. Phys. 119 (2020) 102431; arXiv:1909.06990 [nucl-ex].[5] S. Pomm´e, G. Lutter, M. Marouli, K. Kossert, O. N¨ahle, On the claim of modulations in radon decayand their association with solar rotation, Astropart. Phys. 97 (2018) 38.[6] D. E. Alburger, G. Harbottle, and E. F. Norton, Half-life of Si, Earth Planet. Sci. Lett. 78 (1986)168.[7] D. O’Keefe, et al., Spectral content of Na/ Ti decay data: implications for a solar influence, Astro-phys. Space Sci. 344 (2013) 297.[8] J. H. Jenkins, et al., Concerning the time dependence of the decay rate of