The Day-Night Asymmetry: When the Earth Helps Solar Neutrinos Oscillate

The Mikheyev–Smirnov–Wolfenstein effect — coherent forward scattering of neutrinos on electrons — is the standard explanation for the observed flavor pattern of solar neutrinos. It explains why the high-energy solar 8B neutrinos arrive at Earth as approximately one-third ${\nu }_e$ and two-thirds ${\nu }_{\mu }/{\nu }_{\tau }$, while the low-energy pp neutrinos are almost entirely ${\nu }_e$. The effect operates inside the Sun, where the central density is high enough to drive an adiabatic flavor conversion as the neutrino propagates outward.

But the same physics applies, in muted form, when solar neutrinos pass through the Earth on their way to a detector. The Earth’s electron density is much lower than the Sun’s core, but the path length can be 12,800 kilometres (the diameter), which gives the matter potential time to act. The result is a small modification of the oscillation pattern: at night, when neutrinos have just transited the Earth, the electron-flavor survival probability is slightly different than during the day, when they arrived from above.

This is the day-night asymmetry. It was predicted in the 1980s, after the MSW framework was established, as a clean test of the matter-effect interpretation of the solar neutrino problem. It is small — a few percent at best — and detecting it requires very large statistics. But it has now been measured, primarily by Super-Kamiokande, and the result is consistent with the predicted matter-enhanced oscillation pattern.

This post is about the physics of the day-night asymmetry, why it is so small, how it has been measured, and what it confirms.

The matter potential

A neutrino propagating through ordinary matter experiences a potential due to coherent forward scattering on the electrons. For an electron neutrino, the potential is: $V_e=\sqrt{2}{G}_F{n}_e$ where $G_F$ is the Fermi constant and $n_e$ is the electron number density. The numerical value, for the average density of the Earth’s mantle ($n_e\sim 2$ mol/cm³), is roughly $10^{-13}$ eV. For the Sun’s core ($n_e\sim 100$ mol/cm³), it is $\sim 5\times 10^{-12}$ eV.

For comparison, the typical neutrino energy difference $\Delta E=\Delta m^2/(2E)$ at solar energies (a few MeV) and $\Delta m^2=7.5\times 10^{-5}$ eV² is approximately $10^{-11}$ eV. So the solar matter potential is comparable to the vacuum oscillation energy, while the Earth potential is about an order of magnitude smaller.

The condition for matter effects to be important is that $V_e$ be comparable to or larger than the vacuum oscillation frequency. In the Sun, this is satisfied for the higher-energy 8B neutrinos. In the Earth, the matter potential is just barely large enough to produce a noticeable effect — at the percent level.

Day-night regeneration

When a solar neutrino exits the Sun, it has been converted by the matter effect into a state that is dominantly the heavier mass eigenstate ${\nu }_2$ (in the standard normal-ordering picture). At the Sun’s surface, the matter potential drops to zero, and the neutrino propagates as a vacuum state — a coherent superposition of mass eigenstates with definite mass.

Through eight light-minutes of vacuum to Earth, the relative phases between the mass eigenstates accumulate. The arrival state at Earth is, in flavor basis, a specific mixture: roughly $|U_{e2}{|}^2={\sin }^2{\theta }_{12}$ probability of being ${\nu }_e$ and the complement of being ${\nu }_{\mu }/{\nu }_{\tau }$. This corresponds to the well-known solar electron-neutrino survival probability of about 1/3 for high-energy 8B neutrinos.

When the neutrino traverses the Earth, the matter potential along the path partially regenerates the electron-flavor content. The mass eigenstates pick up additional phases that depend on the matter potential, and the resulting flavor composition at the detector is shifted slightly toward higher ${\nu }_e$ probability.

Quantitatively, the regenerated electron-survival probability is: $P_{ee}^{\text{night}}=P_{ee}^{\text{day}}+f_{\text{reg}}$ where the regeneration factor $f_{\text{reg}}$ depends on the path length through the Earth, the energy of the neutrino, and the mixing parameters. For 8B neutrinos at Super-Kamiokande, $f_{\text{reg}}$ is approximately $0.02$ — a 2-3% increase in the electron-neutrino fraction at night compared to day.

How it gets measured

The asymmetry observable is: $A_{DN}=2\times \frac{R_{\text{day}}-R_{\text{night}}}{R_{\text{day}}+R_{\text{night}}}$ where $R_{\text{day}}$ and $R_{\text{night}}$ are the elastic-scattering rates measured during day and night respectively. (For Super-K, “elastic scattering” means $\nu +e^{-}\to \nu +e^{-}$, sensitive to all flavors but weighted toward ${\nu }_e$.)

The expected sign is negative: the night rate is slightly higher because more electron neutrinos have been regenerated in transit through the Earth. The expected magnitude for Super-K’s 8B-energy analysis is approximately $A_{DN}\approx -3\%$, with significant dependence on the analysis threshold and the assumed mixing parameters.

To measure a 3% asymmetry against a substantial background and statistical fluctuations, you need very large solar-neutrino samples — tens of thousands of events at minimum. Super-Kamiokande accumulated over 50,000 solar 8B events through Phase IV (2008-2018), which made the measurement statistically possible.

The systematic uncertainties are dominated by detector response variations as a function of source angular position (the solar zenith angle changes throughout the year and across day/night) and by the energy calibration. Super-K’s solar analysis has a long history of careful systematic study to extract the small asymmetry signal.

Super-Kamiokande’s measurement

The first significant Super-K day-night measurement was published in 2014 (Abe et al., Physical Review D 91, 052012, based on Phase I+II+III+IV data through 2014). The reported asymmetry was: $A_{DN}=(-3.3\pm 1.0\text{(stat)}\pm 0.5\text{(syst)})\%$

This is a 2.7$\sigma$ result. The sign and magnitude are consistent with the predicted matter-effect regeneration through the Earth. With Phase IV data extended through 2018 and additional Phase V data, the significance grew but remained at the 3-3.5σ level.

A 3-sigma effect in solar physics is not a discovery in the strict 5-sigma sense, but in this context the result is considered convincing for several reasons:

• The sign is correct (negative, as predicted)
• The magnitude is consistent with the global oscillation parameters from other measurements
• The energy dependence (slightly larger asymmetry at higher energies, as predicted) is observed
• The result is reproduced by SNO’s independent analysis

SNO’s 2013 combined analysis (Aharmim et al., Physical Review C 88, 025501) reported an asymmetry consistent with Super-K’s value, providing independent confirmation through a different detector technology and analysis chain.

Why the asymmetry matters

The day-night asymmetry serves as an independent test of the MSW oscillation framework. The framework is well-established from the original solar neutrino flavor measurements at SNO and from the energy dependence of the survival probability across the gallium, chlorine, and water-Cherenkov experiments. But the day-night asymmetry tests the matter-effect physics through the Earth — a different medium, a different path length, a different geometric configuration.

The fact that the asymmetry is observed at the predicted magnitude provides a clean cross-check. If solar neutrino oscillation were due to some non-MSW mechanism — for example, decoherence, decay, or a non-standard interaction — the day-night asymmetry might not match the standard prediction. The agreement with the standard MSW expectation rules out such alternatives at the level of the measurement precision.

The energy dependence

The day-night asymmetry varies with neutrino energy. The predicted behavior:

• Below ~3 MeV: Asymmetry decreases as the matter effect through the Earth becomes weak.
• Around 5-10 MeV: Asymmetry is at its maximum (~3-4%).
• Above ~10 MeV: Asymmetry decreases as the propagation becomes more vacuum-like.

Super-K has reported the asymmetry in two energy bins (above and below 7 MeV) and the energy dependence is consistent with the prediction. SNO’s analysis is in different energy windows and is also consistent.

The energy-dependent pattern is an additional consistency check beyond the absolute magnitude. The matter effect predicts both the average asymmetry and how it varies with energy; both are in agreement with the data.

Earth path-length dependence

In principle, the asymmetry should also depend on the path length through the Earth — neutrinos passing through the core (longer path, higher density) should show a slightly different regeneration than those passing through only the mantle. Super-K has reported the asymmetry as a function of the cosine of the zenith angle of the Sun, with finer binning probing the path-length dependence.

The current Super-K analysis is statistics-limited for this path-length dependence; the data are consistent with the predicted shape but the error bars are large enough that alternative models are not strongly excluded.

Future experiments — particularly Hyper-Kamiokande with its much larger 260-kt fiducial mass — will reduce statistical uncertainties on the path-length dependence. JUNO’s ~20 kt liquid scintillator with low energy threshold will also contribute to lower-energy day-night studies.

What the result confirms

The day-night asymmetry confirms three things:

1. The MSW matter effect operates as predicted, both inside the Sun (where it converts most of the high-energy 8B neutrinos to non-electron flavors) and through the Earth (where it partially regenerates the electron flavor).

2. Solar neutrino oscillation is genuinely a coherent quantum-mechanical effect, not decoherence or particle decay. The small day-night regeneration requires that the mass eigenstates retain phase coherence over astronomical distances.

3. The standard three-flavor mixing picture is sufficient — no additional flavor or non-standard interaction is needed to explain the data. The MSW model with the measured ${\theta }_{12}$ and $\Delta m_{21}^2$ values predicts the day-night asymmetry, and the prediction is correct.

These confirmations are subtle. They are not as headline-grabbing as the original SNO neutral-current measurement that established neutrino flavor change. But they are quantitatively important — they constrain alternatives that would otherwise survive the original flavor measurements.

Looking ahead

The day-night asymmetry will become a more precise observable as next-generation experiments accumulate data. Hyper-Kamiokande’s larger fiducial mass and improved energy resolution will tighten the asymmetry measurement to 1% precision or better, allowing tests of the predicted energy and zenith-angle dependence at high statistical significance.

JUNO, with its 20-kt liquid scintillator and low-energy threshold, will measure 7Be solar neutrinos with high statistics. The day-night asymmetry for 7Be is small (the energy is below the regeneration peak) but the corresponding precision will allow new tests of the framework at lower energies.

DUNE’s ${\nu }_e$ charged-current sensitivity could also contribute, though its solar program is not as central to its physics goals as the long-baseline beam.

By the end of the decade, the day-night asymmetry will be a well-measured observable across multiple experiments and energy ranges, providing a precision test of MSW physics that is fully independent of the original flavor measurements. The framework will be constrained at the percent level — and any deviation from the expected pattern would be a signal of new physics, ranging from non-standard interactions to coupled fields beyond the Standard Model.

For now, the result stands as a quiet confirmation: the Earth, like the Sun, modifies neutrino flavors through coherent forward scattering, exactly as the textbook MSW formalism predicts.