Systematic errors in partially-quenched QCD plus QED lattice simulations

Antonin Portelli, Stephan Dürr, Z. Fodor, Julien Frison, Christian Hoelbling, S. D. Katz, Stefan Krieg, Thorsten Kurth, Laurent Lellouch, Thomas Lippert, Alberto Ramos, Kálman Szabó

Abstract: At the precision reached in current lattice QCD calculations, electromagnetic effects are becoming numerically relevant. Here, electromagnetic effects are included by superimposing $U(1)$ degrees of freedom on $N_f = 2+1$ QCD configurations from the Budapest-Marseille-Wuppertal Collaboration. We present preliminary results for the electromagnetic corrections to light pseudoscalar meson masses and discuss some of the associated systematic errors.

Introduction

In the study of subatomic physics, isospin symmetry—the assumption that up ($u$) and down ($d$) quarks are identical—is a useful but imperfect approximation. This symmetry is broken by the intrinsic mass difference between the quarks and their differing electric charges ($+2/3e$ and $-1/3e$, respectively). As Lattice QCD simulations reach sub-percent precision, these previously negligible electromagnetic (QED) effects have become larger than statistical uncertainties. To achieve higher accuracy, QED must be integrated into the strong force models. This paper explores the calculation of electromagnetic corrections for light mesons (pions and kaons) and quantifies the systematic errors inherent in simulating these forces on a finite computational grid.

Methodology and Datasets

The researchers utilized a "partially-quenched" approach, superimposing a non-compact $U(1)$ electromagnetic field onto existing $N_f = 2+1$ QCD gauge configurations provided by the Budapest-Marseille-Wuppertal (BMW) Collaboration. In this framework, "valence" quarks (which constitute the measured mesons) are charged, while "sea" quarks (virtual quarks in the vacuum) remain neutral. This computational shortcut introduces a specific systematic error known as quenching.

The study analyzed a subset of 2010 SU(3) gauge configurations with the following parameters:

  • Lattice Spacing: Approximately $0.12\text{ fm}$.
  • Pion Masses: 12 distinct masses ranging from $135\text{ MeV}$ to $422\text{ MeV}$.
  • Volumes: Four spatial volumes ranging from $(2\text{ fm})^3$ to $(5.8\text{ fm})^3$.

Finite-Volume Effects

Electromagnetism is a long-range force, making simulations highly sensitive to the finite size of the computational "box" (lattice). A significant finding of this work is the inadequacy of standard SU(3) Chiral Perturbation Theory (ChPT) in fitting the finite-volume data. Instead, a simpler dimensional analysis suggesting that corrections scale as $1/L^2$ (where $L$ is the box size) provided a much more robust fit to the lattice results.

Finite-volume corrections for the kaon mass squared splitting compared to theoretical models.

Key Results and Dashen's Theorem

The analysis yielded preliminary values for electromagnetic squared mass splittings and corrections to Dashen's theorem. Dashen's theorem predicts how electromagnetic effects vanish in certain limits; however, this study observed significant violations, which are critical for accurately determining individual quark masses.

The reported preliminary values include:

  • Pion Splitting: $\Delta_{EM}M_\pi^2 = 1283(30)_{\text{stat}} \text{ MeV}^2$
  • Kaon Splitting: $\Delta_{EM}M_K^2 = 2179(34)_{\text{stat}}(100)_{\text{qu}} \text{ MeV}^2$
  • Dashen Violation (Absolute): $\Delta_A D = 896(37)_{\text{stat}}(100)_{\text{qu}} \text{ MeV}^2$
  • Dashen Violation (Relative): $\Delta_R D = 0.70(4)_{\text{stat}}(8)_{\text{qu}}$
Extrapolation of the pion mass squared splitting to the physical point.

These results are consistent with previous phenomenological estimates but provide a more direct calculation from first principles.

Limitations and Future Outlook

While these results mark significant progress, several limitations remain. The current study relies on a single lattice spacing ($a \approx 0.12\text{ fm}$); future work will require multiple spacings to perform a proper continuum extrapolation. Additionally, the "partial quenching" of sea quarks introduces an estimated $5\%$ error in kaon mass splittings. Moving toward "fully dynamical" QED, where sea quarks are also charged, and incorporating non-degenerate quark masses are the next steps toward fully disentangling isospin-breaking effects.

Summary updated

Abstract

At the precision reached in current lattice QCD calculations, electromagnetic effects are becoming numerically relevant. Here, electromagnetic effects are included by superimposing U(1) degrees of freedom on N f = 2 + 1 QCD configurations from the Budapest-Marseille-Wuppertal Collaboration. We present preliminary results for the electromagnetic corrections to light pseudoscalars mesons masses and discuss some of the associated systematic errors.
BibTeX
@inproceedings{portelli2012systematic,
  author = {Antonin Portelli and Stephan Dürr and Z. Fodor and Julien Frison and Christian Hoelbling and S. D. Katz and Stefan Krieg and Thorsten Kurth and Laurent Lellouch and Thomas Lippert and Alberto Ramos and Kálman Szabó},
  title = {{Systematic errors in partially-quenched QCD plus QED lattice simulations}},
  booktitle = {Proceedings of XXIX International Symposium on Lattice Field Theory — PoS(Lattice 2011)},
  year = {2012},
  doi = {10.22323/1.139.0136},
}
Figure