BMW/DMZ calculation of the hadronic vacuum polarisation for the muon magnetic moment

This paper presents the latest hybrid determination of the hadronic vacuum polarisation contribution to the muon magnetic moment by the BMW and DMZ collaborations, achieving 0.45% precision and overturning the theoretical consensus to resolve the long-standing discrepancy with experimental measurements.

The Gist

Here is an explanation of the paper, translated from complex physics jargon into a story about detective work, building blocks, and a very specific kind of cosmic "glitch."

The Mystery of the Wobbly Muon

Imagine the universe is a giant,精密 (precision) clockwork machine. For decades, scientists have been trying to predict exactly how fast a specific gear in this machine—the muon—should wobble as it spins.

The muon is a tiny, unstable particle, like a heavy cousin of the electron. When you put it in a magnetic field, it wobbles (precesses) at a specific rate. Scientists call this the "magnetic moment."

The Problem:
For twenty years, there was a nagging discrepancy.

• The Experiment: Real-world scientists (at Fermilab and Brookhaven) measured the wobble. It was wobbling faster than the theory predicted.
• The Theory: The "Standard Model" (our best rulebook for how the universe works) said it should wobble slightly slower.
• The Gap: The difference was small, but statistically significant. It was like a GPS saying you are in New York, but your car is clearly in New Jersey. This gap hinted at "New Physics"—perhaps invisible particles or forces we haven't discovered yet.

The Two Ways to Solve the Puzzle

To fix the math, scientists needed to calculate a specific, messy part of the equation called Hadronic Vacuum Polarization (HVP). Think of the vacuum not as empty space, but as a bubbling soup of virtual particles popping in and out of existence. The muon interacts with this soup, which changes its wobble.

There were two ways to calculate how much the soup affects the muon:

1. The "Recipe Book" Method (Data-Driven):
   Scientists looked at past experiments where they smashed electrons and positrons together to create particles. They used these real-world measurements as ingredients to cook up the answer. This was the "consensus" method for a long time.

2. The "From Scratch" Method (Lattice QCD):
   Instead of using a recipe book, scientists tried to build the universe from the ground up using supercomputers. They created a 3D grid (a lattice) and simulated the laws of physics point-by-point to see what the soup does. This is the BMW/DMZ method described in this paper.

The Plot Twist: The "From Scratch" Team Wins

In 2020, the team behind this paper (the BMW collaboration) finished their "From Scratch" calculation.

• The Shock: Their result was different from the "Recipe Book" method.
• The Result: Their calculation matched the experimental measurement perfectly.
• The Implication: If the "From Scratch" math is right, the "Recipe Book" was wrong. And if the math is right, there is no mystery. The Standard Model is correct, and we don't need "New Physics" to explain the muon's wobble.

However, this created a new problem: Why did the "Recipe Book" (experimental data) disagree with the "From Scratch" (theory) calculation?

The New Solution: The "Hybrid" Approach

In this new paper (presented in 2025), the team refined their "From Scratch" calculation and introduced a clever Hybrid Strategy.

Imagine you are trying to measure the length of a very long rope.

• The Middle Section: The middle of the rope is easy to measure with a ruler (Lattice QCD).
• The Ends: The very ends of the rope are frayed and hard to measure precisely with a ruler because they wiggle too much (this is the "long-distance" part of the calculation).

The team realized that while the "ruler" (Lattice) struggles with the frayed ends, the "Recipe Book" (experimental data) is actually very good at measuring those specific ends because they are dominated by simple, low-energy physics.

The Hybrid Fix:

1. They used their super-precise "From Scratch" ruler for the middle of the rope (where the physics is complex).
2. They borrowed the "Recipe Book" measurement for the frayed ends (where the data is reliable).
3. They stitched these two together.

The Result: A Perfect Fit

By using this hybrid approach, the team achieved a result with 0.45% precision.

• The new calculation matches the experimental measurement of the muon's wobble almost perfectly (within 0.5 sigma, which is essentially a statistical tie).
• The Conclusion: The "glitch" in the universe was likely an error in how the old "Recipe Book" data was combined, not a sign of new particles.

The Big Picture Metaphor

Think of the Standard Model as a masterpiece painting.

• For years, art critics (experimentalists) pointed at a small smudge on the canvas and said, "This doesn't look right! The artist must have used a secret, unknown color (New Physics)!"
• The "Recipe Book" (old data analysis) agreed with the critics, saying, "Yes, the smudge is real."
• The "From Scratch" team (BMW) looked at the painting and said, "No, we can recreate this painting pixel-by-pixel on a computer. When we do, the smudge disappears. The painting is perfect."
• This new paper is the team saying, "We've double-checked our pixel-by-pixel recreation, and we've even borrowed a brushstroke from the critics to fix the edges. The painting is perfect. The artist (Nature) followed the rules exactly. No secret colors needed."

Why This Matters

This is a massive victory for Quantum Field Theory. It shows that our current understanding of the universe is incredibly robust. It validates the Standard Model to a precision of 0.31 parts per million.

While some physicists might be disappointed that they didn't find "New Physics" (which would have been exciting), this result is a triumph of human ingenuity. It proves that with enough computing power, better algorithms, and clever hybrid thinking, we can solve the universe's most complex puzzles using the rules we already know.

Technical Summary

Here is a detailed technical summary of the paper "BMW/DMZ calculation of the hadronic vacuum polarisation for the muon magnetic moment" by F.M. Stokes et al. for the BMW and DMZ collaborations.

1. Problem Statement

The paper addresses the long-standing discrepancy between experimental measurements and Standard Model (SM) theoretical predictions of the muon anomalous magnetic moment ($a_\mu$).

• The Discrepancy: For two decades, experiments (Brookhaven and Fermilab) have shown a deviation of approximately $4.2\sigma$ from the SM prediction based on "data-driven" evaluations of the Hadronic Vacuum Polarisation (HVP). This suggested the potential existence of new physics.
• The Conflict: In 2020, the Budapest-Marseille-Wuppertal (BMW) collaboration performed a lattice QCD calculation of the HVP that was significantly larger than the data-driven value, aligning with the experimental measurement and seemingly resolving the tension. However, this created a new tension with the data-driven approach.
• The Challenge: Recent updates to experimental data-driven inputs (specifically $e^+e^- \to \pi^+\pi^-$ cross-sections) have shown significant spread and inconsistencies (e.g., between BaBar, KLOE, and CMD-3), making the data-driven determination unreliable. Conversely, lattice calculations must overcome systematic uncertainties related to finite volume effects, continuum extrapolation, and scale setting.

2. Methodology

The authors present a hybrid calculation that combines the strengths of high-precision lattice QCD with data-driven inputs to minimize uncertainties.

A. Lattice QCD Setup

• Action: Tree-level Symanzik gauge action and one-link staggered fermion action for $2+1+1$ dynamical quark flavors.
• Smearing: Stout smearing ($\rho=0.125$) for gauge links in the fermion action; HEX smearing for large-volume ensembles.
• Ensembles: Simulations performed at seven different lattice spacings (finest $a \approx 0.048$ fm) and varying volumes.
  • Standard volumes: $\approx 6$ fm spatial extent.
  • Dedicated large volumes: $\approx 11$ fm spatial extent to suppress finite-size effects exponentially.
• Blinding: The 2024 analysis was performed completely blind (data multiplied by random factors) to prevent unconscious bias.

B. Analysis Strategy

• Global Fits: The HVP contribution is extracted by fitting the lattice data to a function of lattice spacing ($a$) and quark masses ($m_l, m_s$).
  • Continuum Extrapolation: Uses a form $A_0 + A_2 a^2 \alpha_s^\gamma + A_4 a^4 + A_6 a^6$. Variations in $\gamma$ and the inclusion of higher-order terms are used to estimate systematic errors.
  • Model Averaging: Different fit forms are combined using weights derived from a modified Akaike Information Criterion (AIC) to account for the number of parameters and data points.
• Euclidean-Time Windows: The integral for the HVP is split into three distinct windows to isolate specific sources of uncertainty:
  1. Short-distance ($t < 0.4$ fm): Dominated by discretization effects. Calculated using lattice QCD with perturbative theory corrections.
  2. **Intermediate-distance ($0.4 < t < 1.0$fm):** The region of the$\rho$-resonance where data-driven inputs show the most tension. Calculated entirely via lattice QCD.
  3. Long-distance ($t > 1.0$ fm): Sensitive to finite-volume effects and statistical noise.

C. The Hybrid Approach

The core innovation of this work is the hybrid method:

• The short and intermediate windows are calculated using the high-precision BMW lattice results.
• The long-distance tail ($t > 2.8$ fm), which contributes only $\sim 5\%$ to the total value but dominates the lattice uncertainty due to finite-size effects, is replaced by a data-driven determination.
• Rationale: The long-distance tail is dominated by low-energy states well below the $\rho$-resonance. In this region, experimental data from different sources (BaBar, KLOE, CMD-3, $\tau$-decays) are in good agreement, unlike the $\rho$-peak region.
• Systematics: The authors include additional uncertainties to account for variations in how the data-driven tail is constructed (e.g., averaging order, inclusion of specific experiments like CMD-3).

D. Scale Setting

• The 2020 result used the $\Omega$ baryon mass ($M_\Omega$) for scale setting.
• The 2024 result switched to the pion decay constant ($f_\pi$) to avoid potential contamination from unquantified $K\Xi$ scattering states in the GEVP analysis of $M_\Omega$. The result using $f_\pi$ is consistent with the $M_\Omega$ result.

3. Key Contributions

• Precision Improvement: The 2024 calculation achieves a precision of 0.45% on the HVP contribution, surpassing the precision of data-driven evaluations.
• Resolution of Tensions: By using a hybrid approach, the authors bypass the inconsistencies in the $\rho$-resonance region of experimental data while avoiding the large finite-volume uncertainties of pure lattice calculations for the long tail.
• Validation of Standard Model: The new result brings the theoretical prediction into excellent agreement with the experimental measurement, reducing the discrepancy to 0.5$\sigma$.
• Methodological Rigor: The paper details the transition from linear/quadratic continuum extrapolations to cubic forms, the use of large-volume ensembles (11 fm), and the implementation of a fully blinded analysis.

4. Results

• HVP Contribution: The calculated leading-order HVP contribution is:
  $$a_\mu^{\text{LO-HVP}} = 704.5(6.1) \times 10^{-10}$$
  (Note: The uncertainty is dominated by the hybrid tail and physical point determination).
• Total Muon g-2: Combining this HVP with other SM contributions (QED, Electroweak, HLbL), the total theoretical prediction is:
  $$a_\mu^{\text{SM}} = 11659203.3(6.2) \times 10^{-10}$$
• Comparison with Experiment: The experimental value (from Fermilab/Brookhaven) is approximately $11659204.0(4.3) \times 10^{-10}$ (based on the paper's context of 0.31 ppm precision).
• Discrepancy: The difference between the new SM prediction and the experimental measurement is 0.5$\sigma$.
• Consistency: The result is consistent with independent lattice calculations from Mainz/CLS and RBC/UKQCD, as well as the "White Paper" consensus updated with lattice inputs.

5. Significance

• Standard Model Validation: The result provides a remarkable validation of the Standard Model to a precision of 0.31 ppm. It suggests that the previously observed $4.2\sigma$discrepancy was likely due to issues in the data-driven evaluation of the HVP (specifically the treatment of$e^+e^- \to \pi^+\pi^-$ data) rather than new physics.
• Paradigm Shift: This work supports a shift in the field where lattice QCD is becoming the primary tool for determining the HVP contribution, potentially replacing data-driven methods for the most critical energy regions.
• Future Direction: The paper highlights that while the "tension" is resolved, the field must now focus on understanding the remaining small discrepancies in experimental cross-section data (CMD-3 vs. others) and further refining lattice techniques to reduce the 0.45% uncertainty even further.

In summary, the BMW/DMZ collaboration has delivered a high-precision, hybrid lattice QCD calculation that reconciles theory and experiment for the muon $g-2$, effectively closing the door on the "new physics" hypothesis that was previously thought to be necessary to explain the anomaly.