Eur. Phys. J. C (2020) 80:96
https://doi.org/10.1140/epjc/s10052-020-7683-6
Regular Article - Experimental Physics
Generalized Lomb–Scargle analysis of
36
Cl decay rate
measurements at PTB and BNL
Akanksha Dhaygude
a
, Shantanu Desai
b
Department of Physics, IIT Hyderabad, Kandi, Telangana 502285, India
Received: 18 December 2019 / Accepted: 26 January 2020
© The Author(s) 2020
Abstract Recently Pomme et al. (Solar Phys 292:162,
2017) did an analysis of
36
Cl radioactive decay data from
measurements at the Physikalisch-Technische Bundesanstalt
(PTB), in order to verify the claims by Sturrock and col-
laborators of an influence on beta-decay rates measured at
Brookhaven National Lab (BNL) due to the rotation-induced
modulation of the solar neutrino flux. Their analysis excluded
any sinusoidal modulations in the frequency range from 0.2
to 20/year. We carry out an independent analysis of the same
PTB and BNL data, using the generalized Lomb–Scargle
periodogram to look for any statistically significant peaks in
the range from 0 to 14 per year, and by evaluating the sig-
nificance of every peak using multiple methods. Our results
for the PTB data are in agreement with those by Pomme
et al. For BNL data, we do find peaks at some of the same
frequencies as Sturrock et al., but the significance is much
lower. All our analysis codes and datasets have been made
publicly available.
1 Introduction
Sturrock and collaborators have argued in a number of works
over more than a decade (eg. Refs. [2–4] and references
therein) that beta decay rates for a large number of nuclei
exhibit variability and show periodicities at multiple frequen-
cies, some of which have been associated with solar rotation
as well as other processes in the solar core. They have also
found similar peaks at 12.7 per year in the Super-K solar
neutrino flux (from the first 5 years of data) [5], which they
have argued to be due effects of solar rotation. Furthermore,
they have correlated the two sets of findings, and argued for
an influence of solar neutrinos on beta-decay rates.
However, many other groups have failed to reproduce the
periodicities in the beta-decay results, while analyzing the
a
e-mail: ep15btech11005@iith.ac.in
b
e-mail: shntn05@gmail.com
same data as well as decays of the same elements from other
experiments. A review of some of these claims and rejoinders
can be found in Refs. [3,6,7]. In our previous works, we have
also carried out an independent analysis of some of these
claims and found evidence of periodicities at some of the
same frequencies as found by Sturrock et al., albeit with a
lower significance [8,9].
In this work, we focus on addressing the claimed period-
icities in the beta decay rates of
36
Cl. Sturrock et al. [10]have
argued for periodicities with periods at 1/year and 12.7/year
(or 28.7 days) in the
36
Cl decay rates of the Brookhaven
National Lab [11] counting experiment. They have argued
that the peak at 12.7/year is indicative of the synodic rotation
rate of the radiative zone of the Sun, since it matches the value
of 28.7 days determined using helioseismology [12,13].
These results were rebutted by Pomme et al. [1], who
found no evidence for periodicities in the decay rates of
36
Cl using more accurate measurements at Physikalisch-
Technische Bundesanstalt Braunschweig (PTB), obtained
using the triple-to-double coincidence ratio measurement
techniques [6]. Pomme [14,15] has also raised concerns
about the detectorstabilityand control of experimentaluncer-
tainties in the BNL measurements, which are now more than
three decades old. Furthermore, the invariability of the decay
constants for
36
Cl was also demonstrated using triple-to-
double coincidence ratio measurements [16], to refute claims
of oscillations ascribed to the changes in Earth–Sun distance.
In this work, we independently try to adjudicate the con-
flicting results between these two works by doing an inde-
pendent analysis of the beta decay residual data from both the
BNL and PTB measurements (which were kindly provided to
us by S. Pomme) using the Generalized Lomb–Scargle peri-
odogram [17–19]. We search the frequency range between 0
and 14/year (or up to 26 days), since this covers the frequency
range associated with solar rotation [10,20]. We calculate
the significance of the peaks using all the available methods
provided in the astropy [21] library used to calculate the
Lomb–Scargle periodogram.
0123456789().: V,-vol
123