Comparisons of annual modulations in MINOS with the event rate modulation in CoGeNT

The CoGeNT Collaboration has recently published results from a fifteen month data set which indicate an annual modulation in the event rate similar to what is expected from weakly interacting massive particle interactions. It has been suggested that the CoGeNT modulation may actually be caused by other annually modulating phenomena, specifically the flux of atmospheric muons underground or the radon level in the laboratory. We have compared the phase of the CoGeNT data modulation to that of the concurrent atmospheric muon and radon data collected by the MINOS experiment which occupies an adjacent experimental hall in the Soudan Underground Laboratory. The results presented are obtained by performing a shape-free ${\chi }^2$ data-to-data comparison and from a simultaneous fit of the MINOS and CoGeNT data to phase-shifted sinusoidal functions. Both tests indicate that the phase of the CoGeNT modulation is inconsistent with the phases of the MINOS muon and radon modulations at the $3.0\sigma$ level.

Figure 1

The residuals of the MINOS Far Detector muon rates, radon levels and CoGeNT event rates as a function of time. The MINOS muon and radon data have been scaled by factors of 10 and one-half, respectively, to fit on the same graph and, for this figure, use the same binning as the CoGeNT results. The vertical dashed lines indicate the start of a new calendar year. The arrow marks the date where a dark matter signal is expected to peak.

Figure 2

Confidence limit contours for the period and phase as determined by fitting the CoGeNT event rate and MINOS Far Detector muon rate and radon level data to Eq. (1). The best fit values are given in Table 1. The vertical and horizontal black lines mark the expected period and phase for a dark matter signal, respectively.

Figure 3

The $\Delta {\chi }^2$ distributions comparing the phases of the MINOS muon rate and radon level data to the phase of the CoGeNT event rate data using Eq. (2). The $\Delta {\chi }^2$ curves are calculated with respect to their ${\chi }^2$ minima. The flattening of the $\Delta {\chi }^2$ curves indicates that these exclusions are limited by the confidence with which the CoGeNT data can exclude the null modulation hypothesis.

Figure 4

The $\Delta {\chi }^2$ distributions comparing the phases of the MINOS muon rate and radon level data to the phase of the CoGeNT event rate data using Eq. (3). The $\Delta {\chi }^2$ curves are calculated with respect to their ${\chi }^2$ minima.