A Neutrino Mass Puzzle

The differences between the three neutrino masses is known to considerable precision and is most easily communicated in meV units. But the absolutely masses are not very meaningfully constrained from direct measurements of the neutrino masses. Cosmology based constraints on the sum of the three neutrino masses are much more restrictive. 

But a new preprint based upon newly released DESI data shows that we are at the brink of being over constrained, as the sum of the three neutrino masses estimated from cosmology is likely to fall below the minimum sum of the three neutrino masses from neutrino oscillation data. 

This suggests that the cosmology based bounds may be unreliable. This isn't all that stunning given that myriad problems with the LambdaCDM model, even though the sum of the neutrino masses estimates from cosmology are quite robust over a wide range of assumptions.

On the other hand, cosmology based bounds aren't the only factors that favor a normal neutrino mass hierarchy over an inverted neutrino mass hierarchy. A normal neutrino mass hierarchy leaves two possibilities: either a near minimal sum of the three neutrino masses, independent of cosmology considerations, or a nearly degenerate set of neutrino masses, as the main possibilities. The large size of the PMNS matrix mixing angles relative to those in the CKM matrix for quarks suggests that the hierarchy of neutrino masses may indeed be less steep than that of the quarks. But there is no established theory to quantify that intuition.

The baryon acoustic oscillation (BAO) analysis from the first year of data from the Dark Energy Spectroscopic Instrument (DESI), when combined with data from the cosmic microwave background (CMB), has placed an upper-limit on the sum of neutrino masses, ∑mν<70 meV (95%). In addition to excluding the minimum sum associated with the inverted hierarchy, the posterior is peaked at ∑mν=0 and is close to excluding even the minimum sum, 58 meV at 2σ. 

In this paper, we explore the implications of this data for cosmology and particle physics. The sum of neutrino mass is determined in cosmology from the suppression of clustering in the late universe. 

Allowing the clustering to be enhanced, we extended the DESI analysis to ∑mν<0 and find ∑mν=−160±90 meV (68%), and that the suppression of power from the minimum sum of neutrino masses is excluded at 99% confidence. 

We show this preference for negative masses makes it challenging to explain the result by a shift of cosmic parameters, such as the optical depth or matter density. 

We then show how a result of ∑mν=0 could arise from new physics in the neutrino sector, including decay, cooling, and/or time-dependent masses. These models are consistent with current observations but imply new physics that is accessible in a wide range of experiments. 

In addition, we discuss how an apparent signal with ∑mν<0 can arise from new long range forces in the dark sector or from a primordial trispectrum that resembles the signal of CMB lensing.

Nathaniel Craig, Daniel Green, Joel Meyers, Surjeet Rajendran, "No νs is Good News" arXiv:2405.00836 (May 1, 2024).

Another preprint today greatly loosens the bound on the muon neutrino mass from more direct measurements to about 150 keV, which is a huge step backwards from the previous direct detection bound of 420 eV proposed in the year 2014.

The tightest bound on the lightest neutrino mass eigenstate is about 0.8 eV (and is likely to drop to 0.2 eV as experimental methods improve). Combining the current bound and the mass differences permitted from neutrino oscillation data keeps all three neutrino mass eigenstates at under 1 eV in either a normal or an inverted neutrino mass hierarchy (and the sum of the three neutrino masses under 2.6 eV). So the neutrino mass bounds from the new preprint and from the 2014 study are meaningless.

If the lightest neutrino mass were constrained to be 0.2 eV or less, the sum of the three neutrino masses in either hierarchy, could be constrained to be no more than about 0.71 eV (0.66 eV in a normal hierarchy), which while not competitive with the potentially discredited cosmology bounds, would still be far less of a setback.

The background from the muon neutrino mass preprint states the following:

As a result of the discovery of neutrino flavor oscillations, neutrinos are thought to have a non-zero mass, as opposed to the standard model (SM). The detection of neutrino oscillations in the atmospheric Super-Kamiokande and solar Sudbury Neutrino Observatory (SNO) experiments provided initial evidence supporting the existence of nonzero neutrino masses. In the 21st century, several neutrino oscillation experiments were conducted, providing precise measurements for the phenomenon of neutrino oscillation. However, these experiments revealed that two out of the three flavors are heavy, and the massive flavor possesses a mass of at least 0.05 eV. Nonetheless, it should be noted that these experiments could only determine mass-squared differences between the flavors and were unable to directly measure the individual mass of each flavor. 

It is crucial to highlight that the theoretical concept of neutrino oscillation was initially proposed by the Russian scientist Bruno Pontecorvo to elucidate the absence of detected atmospheric and solar neutrinos. 

Measuring neutrino mass is of a great importance due to its implications regarding not only refining our understanding about the nature of the universe and dark matter, but also conceivably providing insights into some new physics beyond SM. Therefore, one of the main objectives of particle physicists has been to measure neutrino masses for several years. Consequently, numerous experiments have been conducted since 1991 to measure the mass of neutrinos based on tritium beta decay, from Los Alamos to the Karlsruhe Tritium Neutrino experiment in 2022. 

At Los Alamos, researchers established an upper limit of 11 eV at a 95% confidence level for the mass of the electron anti-neutrino, m(¯ νe). In the first run in 2019, the KATRIN experiment significantly improved the sensitivity of m(¯ νe), setting a new upper bound of 1.1 eV at 90% confidence level (CL), which represents an improvement by a factor of about two compared to the previous limit. Furthermore, in the second run in 2022, they achieved a more precise upper bound of 0.9 eV at 90% CL. The results from the KATRIN 2019 (first run) were then combined with those from KATRIN 2022 (second run), resulting in a more accurate upper limit of 0.8 eV at a 90% CL for m(¯ νe). 

On the other hand, the most successful attempts to measure the mass of muon neutrinos, as the second flavor, were in 1982, involving the measurement of muon neutrinos from pion decay in flight and achieving an upper bound of less than 500 keV, and in 1996, measuring muon neutrinos from the decay of pions at rest, resulting in an upper limit of less than 170-190 keV.