Revival of the Reactor Antineutrino Anomaly

This paper reports that the 2023 summation model for reactor antineutrino fluxes revives the Reactor Antineutrino Anomaly at a $2.2\sigma$ significance level, creating a $3.8\sigma$ tension with gallium anomaly results and other neutrino data, which is reduced to $1.3\sigma$ when gallium systematic uncertainties are enlarged.

The Gist

Imagine a giant, high-tech bakery (a nuclear reactor) that bakes four specific types of cookies every day: Uranium-235, Uranium-238, Plutonium-239, and Plutonium-241. According to the master recipe books (theoretical models), we know exactly how many "cookie crumbs" (antineutrinos) each type of cookie should drop as it cools down.

For years, scientists have been standing outside the bakery with very sensitive crumb-counters. They noticed a mystery: the counters were catching fewer crumbs than the recipe books predicted. This missing crumb problem was called the Reactor Antineutrino Anomaly.

Here is the story of the paper, broken down simply:

1. The Mystery Comes and Goes

• The First Clue (2011): Scientists updated the recipe books and realized, "Wait, we were expecting way too many crumbs!" The missing crumb count was significant (a 2.5σ anomaly). It was a big deal.
• The False Hope (2021): New recipe updates came out. These new books said, "Actually, we overestimated the crumbs even more." When scientists used these new numbers, the missing crumb count almost vanished. The mystery seemed solved; the gap between prediction and reality shrank to almost nothing.
• The Twist (Now): The authors of this paper looked at the very latest recipe book published in 2023 by a French team (called the CEA model). This book is special because it includes a very detailed "uncertainty budget"—a checklist of every possible mistake the bakers could have made.
• The Result: When they used this new 2023 book, the missing crumb count came back. The gap is now significant again (2.2σ). The anomaly has been "revived."

2. The "Ghost Cookie" Explanation

If the recipe books are right, but we are still missing crumbs, where did they go?
The most popular theory is that the crumbs aren't missing; they are changing shape.

• Imagine the crumbs are "active" particles that we can see.
• The theory suggests that some of them turn into "sterile" ghosts (particles that don't interact with anything and are invisible to our counters) as they travel from the oven to the counter.
• This is called 3+1 oscillation: three normal types of particles plus one invisible "ghost" type.

3. The Great Tug-of-War

The authors tried to fit this "ghost cookie" theory to all the data they have. They ran into a massive problem: The data sets don't agree with each other.

• Team A (The Reactors): The reactor data (including the new CEA model) says, "Yes, ghosts exist, and here is how many."
• Team B (The Gallium Experiments): These are experiments using radioactive sources (like a different kind of cookie) to test for ghosts. They say, "Yes, ghosts exist, but the numbers are totally different from Team A."
• Team C (The Sun & KATRIN): These are experiments looking at the sun or measuring particle mass. They say, "We don't see any ghosts."

When you try to combine all these teams into one big group hug, it's a disaster. The math shows a 3.8σ tension. In plain English, it's like trying to force a square peg into a round hole while the peg is screaming, "I don't fit!" The data from the Gallium experiments is fighting the data from the Reactors and the Sun so hard that the whole theory looks shaky.

4. The "Stretchy Ruler" Solution

Since the math is screaming that something is wrong, the authors asked: Who is holding the ruler wrong?

They suspected that the Gallium experiments (Team B) might have underestimated their own measurement errors. Maybe their "uncertainty" was too tight, making the disagreement look worse than it really is.

So, they did something clever, inspired by how the Particle Data Group (the referees of physics) handles conflicting measurements:

• They took the Gallium data and stretched their uncertainty ruler. They made the "error bars" (the margin of doubt) 3.8 times wider.
• The Result: Suddenly, the tension dropped from a screaming 3.8σ down to a calm 1.3σ.

The Conclusion

By stretching the uncertainty of the Gallium experiments, the authors managed to make all the different data sets (Reactors, Sun, KATRIN, and Gallium) agree with each other again.

In summary:

1. The mystery of missing reactor crumbs is back thanks to a new 2023 recipe book.
2. The "ghost particle" explanation is still the best guess, but the data is messy.
3. The biggest conflict is between different types of experiments.
4. If we assume the Gallium experiments were a bit too confident in their precision, the whole picture becomes consistent, and the "ghost" theory survives.

The paper doesn't claim this proves ghosts exist; it just says, "If we fix the way we measure the errors in one specific experiment, the math finally adds up."

Technical Summary

Technical Summary: Revival of the Reactor Antineutrino Anomaly

Problem Statement
The Reactor Antineutrino Anomaly (RAA) refers to a persistent deficit between the measured event rate of reactor antineutrinos ($\bar{\nu}_e$) and theoretical predictions. Originally identified in 2011 with a significance of $2.5\sigma$ based on the Huber-Muller (HM) flux calculations, the anomaly was thought to be resolved in 2021. New flux calculations by the Estienne-Fallot (EF) group (2019) and the Kopeikin-Skorokhvatov-Titov (KI) group (2021) reduced the deficit to approximately $1\sigma$, suggesting the anomaly was an artifact of flux modeling rather than new physics. This paper investigates whether the anomaly persists when considering the latest 2023 summation model flux calculation (CEA) by a French research group, which includes a comprehensive uncertainty budget.

Methodology
The authors perform a global analysis of short-baseline reactor antineutrino data using the updated CEA flux model.

1. Flux Models: The study compares four theoretical models for the inverse beta decay (IBD) yield: HM (2011), EF (2019), KI (2021), and the new CEA (2023). The CEA model is a summation method calculation that provides a detailed covariance matrix for the uncertainties of the four fissionable isotopes ($^{235}\text{U}$, $^{238}\text{U}$, $^{239}\text{Pu}$, $^{241}\text{Pu}$).
2. Data Set: The analysis utilizes experimental data from 28 reactor experiments (including Bugey, Rovno, ILL, Krasnoyarsk, STEREO, DANSS, RENO, Double Chooz, and Daya Bay). The dataset includes both absolute rate measurements and spectral ratio measurements.
3. Statistical Framework: The authors employ a $\chi^2$ minimization technique (Eq. 3) to determine the ratio of measured to predicted yields ($R_M$) and the corresponding significance of the deficit (RAA). They account for experimental correlations and the covariance matrices of the theoretical flux models.
4. Oscillation Analysis: The paper tests the "3+1" neutrino oscillation hypothesis, where the electron neutrino mixes with a light sterile neutrino ($\nu_s$). This involves fitting the data to the survival probability $P_{ee}$ dependent on the mixing angle $\sin^2 2\vartheta_{ee}$ and the squared-mass splitting $\Delta m^2_{41}$.
5. Global Consistency: The compatibility of the RAA with other anomalies and constraints is evaluated using the Parameter Goodness-of-Fit (PG) test. The datasets considered include the Gallium Anomaly (GA), solar neutrino bounds (SOL), KATRIN mass bounds, combined reactor spectral ratio measurements (SPE), and the Daya Bay search for sub-eV sterile neutrinos.
6. Handling Tensions: To address significant tensions between datasets, the authors apply a procedure inspired by the Particle Data Group (PDG) treatment of incompatible measurements. They specifically enlarge the uncertainties of the Gallium Anomaly data to reduce global tension, as the primary conflict lies between Gallium data and the reactor/solar/KATRIN datasets.

Key Contributions and Results

• Revival of the RAA: The analysis of the 2023 CEA flux model reveals a revival of the Reactor Antineutrino Anomaly. The deficit is measured at $2.2\sigma$, which is significantly larger than the $1.1\sigma$–$1.2\sigma$ found with the EF and KI models and comparable to the original $2.5\sigma$–$2.6\sigma$ HM result.
• Role of Uncertainties: The authors demonstrate that the magnitude of the RAA is highly sensitive to the covariance structure of the flux models. While the CEA model predicts a lower $^{235}\text{U}$ yield (reducing the expected flux), the specific correlations in the CEA uncertainty budget result in a larger anomaly compared to EF and KI. Replacing CEA covariances with those of EF or KI models reduces the RAA to $1.3\sigma$ and $0.8\sigma$ respectively, highlighting the critical importance of validating uncertainty correlations.
• 3+1 Oscillation Constraints: In the 3+1 oscillation framework, the CEA-RAA data allows for specific regions in the $(\sin^2 2\vartheta_{ee}, \Delta m^2_{41})$ plane. The reactor spectral ratio measurements (SPE) impose tight bounds on $\Delta m^2_{41}$ in the range $0.1 \text{ eV}^2 \lesssim \Delta m^2_{41} \lesssim 2 \text{ eV}^2$.
• Global Tension: A combined fit of all datasets (CEA-RAA, GA, SOL, KATRIN, SPE, and Daya Bay) reveals a significant tension of $3.8\sigma$. The primary source of this tension is the incompatibility between the Gallium Anomaly data and the reactor/solar/KATRIN data sets. The CEA-RAA itself shows only moderate tension ($1.6\sigma$) with the reactor spectral ratio measurements.
• Reconciled Global Fit: By enlarging the Gallium Anomaly uncertainties by a factor derived from the PG test (effectively scaling them to reduce the $3.8\sigma$ tension), the authors achieve a global fit with a reduced tension of $1.3\sigma$. This procedure yields allowed regions for the oscillation parameters that are consistent across all datasets under the 3+1 hypothesis.

Significance and Claims
The paper claims to demonstrate that the Reactor Antineutrino Anomaly has not been resolved but has instead re-emerged with the adoption of the 2023 CEA flux calculation. The authors assert that the anomaly is now at a level ($2.2\sigma$) that warrants serious consideration, nearly matching the original discovery significance.

Regarding the explanation via sterile neutrinos, the paper concludes that while the 3+1 model can accommodate the data, the global picture is complicated by the Gallium Anomaly. The authors argue that the most plausible resolution to the current global tension is not the rejection of the 3+1 model, but rather the recognition that the uncertainties in the Gallium source experiments may have been underestimated. By adjusting these uncertainties, the paper presents a "reconciled" global analysis that provides reliable allowed regions for sterile neutrino parameters, assuming the validity of the CEA flux model and the 3+1 oscillation framework. The work emphasizes that accurate validation of flux model uncertainties and correlations is essential for determining the true status of the RAA.