---
title: "The reactor anomaly and the spectrum bump ask different questions | Sandbox Physics"
description: "A reactor spectrum can fall short in total and still show a local excess. Both are comparisons with predictions; fuel and distance tests help find the cause."
canonical_url: "https://sandboxphysics.com/stories/neutrinos/reactor-anomaly-and-spectrum-bump/"
language: "en"
page_type: "story"
alternate_languages:
  zh-CN: "https://sandboxphysics.com/zh/stories/neutrinos/reactor-anomaly-and-spectrum-bump/"
last_modified: "2026-10-10"
---

[2026 Nobel Prize in Physics](https://sandboxphysics.com/nobel/2026/physics/)

# The reactor anomaly and the spectrum bump ask different questions

One discrepancy is about the total number of antineutrinos; the other is about where events fall in energy. Each has its own reference prediction, tests and possible causes, and they should be sorted out separately before either is called new physics.

Sandbox Physics · 2026-10-10 · Interpreting evidence

## The anomaly appeared when the prediction went up

The reactor antineutrino anomaly took shape in 2011, when revised emission calculations raised the predicted yields. Earlier short-distance measurements then looked low against the new reference. The observed counts had not suddenly fallen; the denominator of the comparison had changed.

Mention and colleagues’ historical reanalysis gave the observed-to-predicted ratio shown below. That number belongs to their selected set of experiments and their calculations. Later flux models and experimental corrections change the comparison, so it should not be quoted as a timeless global deficit.

The ratio depends on more than the emission prediction. Exposure, the number of target protons, the interaction cross section, efficiency and backgrounds all sit between the predicted source and the counted events. An absolute discrepancy can arise at any of these steps. A surprising ratio calls for checks, but on its own it does not say which step is wrong.

$$
R_{\mathrm{2011}}=\frac{N_{\mathrm{observed}}}{N_{\mathrm{predicted}}}=0.943\pm0.023
$$

The combined short-baseline ratio from Mention et al. (2011), using their revised reference and uncertainty treatment. It is a historical value, not an updated ratio for all reactors.

Sources: [1] · [2] · [3]

## The bump is an excess in one part of the spectrum

Daya Bay, RENO and Double Chooz each found more events than their adopted predictions in a limited part of the prompt-energy spectrum. The familiar “five MeV bump” names roughly where it sits in prompt energy. That is not the exact energy of the incoming antineutrinos, and the bump is not a sharp line from a new particle.

A local excess and a deficit in the total do not contradict each other. A spectrum can sit above its reference in one energy range and below it elsewhere. Scaling two spectra to the same area hides a difference in rate but leaves a difference in shape visible, so the plotting convention matters.

RENO reported that the excess rose and fell with reactor power, and several detectors saw related structure. That points to an origin in the reactors, not a background that stays constant while the cores change. It does not single out a particular nuclear branch or show that every detector-response uncertainty is gone.

Sources: [4] · [5] · [6]

1. 01 **Total yield**

   Compare the total with a named prediction

2. 02 **Local spectrum**

   Check the energy axis and normalization

3. 03 **Telling causes apart**

   Vary the fuel mixture or the distance

Total rate, spectral shape, fuel dependence and distance dependence each test something different. A spectrum can show a local excess and still have a lower total.

## Changing fuel tests whether one scale factor is enough

In 2017 Daya Bay followed the yield as the fraction of fissions from plutonium changed. If every isotope fell short by the same fraction, the yield would change with fuel in one way; if the error lay mainly in one isotope’s predicted yield, it would change differently. A running reactor could therefore test more than a single average.

The analysis found that the isotopes did not all fall short of their reference by the same fraction. Under the tested assumptions, that is in tension with explaining the whole rate anomaly by sterile neutrinos alone. It does not show that no sterile state can exist in any parameter range.

Compact research reactors with highly enriched fuel give a different mixture, dominated by uranium-235 fission. STEREO’s measured spectrum provides a direct reference for that isotope. Comparing fuel mixtures tests the nuclear predictions; comparing distances searches for an oscillation phase. They are different tests.

Sources: [7] · [8] · [9]

## A nuclear-model error and a new neutrino would show up differently

The nuclear calculations depend on fragment yields, decay branches and corrections for transitions beyond the simplest allowed approximation. Hayes and colleagues showed that the uncertain treatment of these forbidden transitions can matter at the level of the historical discrepancy. A better error model changes how much an apparent mismatch can be made to mean.

A simple extra sterile state would instead add a new oscillation length. The clearest reactor test is a spectrum that changes with distance in the predicted way once geometry and detector response are included. A feature that looks the same at nearby distances points more naturally to the source spectrum than to that oscillation.

Later flux-model studies, including Giunti and colleagues’ 2021 comparison, found that the size of the rate anomaly depends on which prediction is used. So “the anomaly has disappeared” also needs qualifying: agreement under one model is a result about that model and dataset, and it does not settle every neutrino anomaly.

Sources: [10] · [11] · [8] · [9]

## How to read a reactor anomaly

A useful report names the dataset, the prediction, the energy definition and how the uncertainties are correlated. It then says whether the comparison tests the total rate, the shape, the fuel dependence or the distance dependence. Without that, two papers can seem to disagree when they are asking different questions of overlapping data.

Oscillation fits that compare near and far detectors can stay robust even when the shared spectrum prediction is imperfect. Absolute flux measurements feel that imperfection more, because testing the prediction is their purpose. One experiment can therefore give strong evidence on mixing while its source-spectrum question stays open.

The next article follows the distance test directly. That is what an anomaly is good for: it suggests explanations that make different predictions and prompts new checks. Treating the word anomaly as evidence for a new particle skips the step that could tell those explanations apart.

Sources: [12] · [7] · [8] · [13]

## What we learned

A spectrum can fall short in total and still show a local excess; both are comparisons with predictions. Watching how the signal changes with fuel and with distance helps tell errors in the nuclear model from a particular sterile-neutrino oscillation explanation. Calling something an anomaly does not decide its cause.

## Try it in the Lab

[Reactor coincidence](https://sandboxphysics.com/labs/reactor-coincidence/)

## Primary sources & revision

1. [Mention et al. · The Reactor Antineutrino Anomaly (2011)](https://arxiv.org/abs/1101.2755)
2. [Mueller et al. · Improved Predictions of Reactor Antineutrino Spectra (2011)](https://arxiv.org/abs/1101.2663)
3. [Patrick Huber · On the determination of anti-neutrino spectra from nuclear reactors (2011)](https://arxiv.org/abs/1106.0687)
4. [Daya Bay Collaboration · Measurement of the Reactor Antineutrino Flux and Spectrum (2015/2016)](https://arxiv.org/abs/1508.04233)
5. [Seon-Hee Seo for RENO · New Results from RENO and The 5 MeV Excess (2014)](https://arxiv.org/abs/1410.7987)
6. [Double Chooz Collaboration · Improved measurements of the neutrino mixing angle (2014)](https://arxiv.org/abs/1406.7763)
7. [Daya Bay Collaboration · Evolution of the Reactor Antineutrino Flux and Spectrum (2017)](https://arxiv.org/abs/1704.01082)
8. [STEREO Collaboration · Uranium-235 spectrum and sterile-neutrino test (Nature, 2023; arXiv v3, 2024)](https://arxiv.org/abs/2210.07664v3)
9. [PROSPECT Collaboration · Final PROSPECT-I Short-Baseline Oscillation Search (2024 submission; 2025 publication)](https://arxiv.org/abs/2406.10408v2)
10. [Hayes et al. · Systematic Uncertainties in the Analysis of the Reactor Neutrino Anomaly (2013/2014)](https://arxiv.org/abs/1309.4146)
11. [Giunti, Li, Ternes & Xin · Reactor antineutrino anomaly in light of recent flux model refinements (2021/2022)](https://arxiv.org/abs/2110.06820)
12. [Daya Bay Collaboration · Observation of electron-antineutrino disappearance (2012)](https://arxiv.org/abs/1203.1669)
13. [Particle Data Group · Statistics (2025)](https://pdg.lbl.gov/2025/reviews/rpp2025-rev-statistics.pdf)

First published 2026-10-10; last revised 2026-10-10. Original explanatory text and diagrams by Sandbox Physics. Illustrations are schematic; no experimental event records are reproduced here. This is an independent educational publication, not an official Nobel or experiment collaboration publication.

[← Previous: JUNO: why must a huge detector resolve tiny spectral features?](https://sandboxphysics.com/stories/neutrinos/juno-fine-spectrum-and-mass-ordering/) [Next: Short reactor baselines: look for a change with distance →](https://sandboxphysics.com/stories/neutrinos/short-baseline-reactor-sterile-tests/)

## Keep reading

Neutrinos · Reading path

### Read reactors and the Earth

1. [Article **Where does a reactor’s antineutrino spectrum come from?**](https://sandboxphysics.com/stories/neutrinos/reactor-antineutrino-spectrum-origins/)
2. [Article **KamLAND: testing a solar mystery with reactors on Earth**](https://sandboxphysics.com/stories/neutrinos/kamland-reactor-oscillations/)
3. [Article **Daya Bay: why compare a near detector with a far one?**](https://sandboxphysics.com/stories/neutrinos/daya-bay-near-far/)
4. [Article **RENO and Double Chooz: why repeat the mixing-angle measurement?**](https://sandboxphysics.com/stories/neutrinos/reno-double-chooz-independent-mixing-tests/)
5. [Article **JUNO: why must a huge detector resolve tiny spectral features?**](https://sandboxphysics.com/stories/neutrinos/juno-fine-spectrum-and-mass-ordering/)
6. [You are here **The reactor anomaly and the spectrum bump ask different questions**](https://sandboxphysics.com/stories/neutrinos/reactor-anomaly-and-spectrum-bump/)
7. [Article **Short reactor baselines: look for a change with distance**](https://sandboxphysics.com/stories/neutrinos/short-baseline-reactor-sterile-tests/)
8. [Article **Geoneutrinos: what can particles tell us about the Earth’s heat?**](https://sandboxphysics.com/stories/neutrinos/geoneutrinos-earth-radiogenic-heat/)

Neutrinos · Topic

### Unexpected results and sterile neutrinos

Gallium, reactor and timing experiments have all reported surprises. Which ones held up, and could a fourth kind of neutrino explain the rest?

1. [Article **How does a faint signal become a neutrino discovery?**](https://sandboxphysics.com/stories/neutrinos/neutrino-discovery-evidence-chain/)
2. [Article **Gallium experiments: opening the Sun’s low-energy window**](https://sandboxphysics.com/stories/neutrinos/gallium-low-energy-solar-neutrinos/)
3. [Article **Why can a neutrino change flavor?**](https://sandboxphysics.com/stories/neutrinos/neutrino-flavor-oscillations/)
4. [You are here **The reactor anomaly and the spectrum bump ask different questions**](https://sandboxphysics.com/stories/neutrinos/reactor-anomaly-and-spectrum-bump/)
5. [Article **Short reactor baselines: look for a change with distance**](https://sandboxphysics.com/stories/neutrinos/short-baseline-reactor-sterile-tests/)
6. [Article **OPERA: tau appearance and a corrected clock anomaly**](https://sandboxphysics.com/stories/neutrinos/opera-tau-appearance-and-timing-correction/)
7. [Article **Sterile neutrinos: an explanation must pass several tests**](https://sandboxphysics.com/stories/neutrinos/sterile-neutrinos-anomalies-and-tests/)