Abstract
Scale setting is of central importance in lattice QCD. It is required to predict dimensional quantities in physical units. Moreover, it determines the relative lattice spacings of computations performed at different values of the bare coupling, and this is needed for extrapolating results into the continuum. Thus, we calculate a new quantity, $w_0$, for setting the scale in lattice QCD, which is based on the Wilson flow like the scale $t_0$. It is cheap and straightforward to implement and compute. In particular, it does not involve the delicate fitting of correlation functions at asymptotic times. It typically can be determined on the few per-mil level. We compute its continuum extrapolated value in 2 + 1-flavor QCD for physical and non-physical pion and kaon masses, to allow for mass-independent scale setting even away from the physical mass point. We demonstrate its robustness by computing it with two very different actions (one of them with staggered, the other with Wilson fermions) and by showing that the results agree for physical quark masses in the continuum limit.
Introduction
In Lattice Quantum Chromodynamics (QCD), physical quantities are calculated on a discrete spacetime grid. However, the lattice spacing $a$—the distance between grid points—is a dimensionless input parameter. To convert simulation results into physical units (like GeV or fm), researchers must "set the scale" by calculating a dimensionful quantity and comparing it to experimental data. Traditionally, the Sommer scale ($r_0$) has been used, but it suffers from high statistical noise and requires complex fitting procedures. The Budapest-Marseille-Wuppertal (BMW) collaboration introduces $w_0$, a new high-precision scale based on the Wilson flow, designed to overcome these limitations.
Methodology: The Wilson Flow and $w_0$
The proposed scale utilizes the Wilson flow (or gradient flow), a process that smooths gauge fields by evolving them according to a diffusion-like equation. As flow time $t$ increases, high-frequency fluctuations (noise) are suppressed. The primary observable is the action density $\langle E(t) \rangle$.
While the previously defined scale $t_0$ uses the value of $t^2 \langle E(t) \rangle$ directly, the BMW collaboration found that using the derivative $W(t) = t \frac{d}{dt} \{t^2 \langle E(t) \rangle\}$ provides better control over discretization errors. The scale $w_0$ is defined as the square root of the flow time $t$ where $W(t) = 0.3$. This method is computationally efficient and avoids the systematic uncertainties associated with long-distance fits.
Key Results and Universality
To demonstrate the robustness of $w_0$, the authors performed simulations using two distinct lattice formulations: staggered fermions with 2-stout smearing and Wilson-clover fermions with 2-HEX smearing. These simulations covered a range of quark masses, including the physical point.
The study yielded a physical value of $w_0 = 0.1755(18)(04) \text{ fm}$. A critical finding is the universality of the scale: despite the different discretization schemes, both actions produce consistent results in the continuum limit. Furthermore, $w_0$ exhibits weak dependence on quark masses, making it an ideal "anchor" for various lattice ensembles.
Conclusion and Limitations
The $w_0$ scale represents a significant advancement in lattice QCD precision, offering a "ruler" that is both easy to compute and highly precise (at the per-mil level). However, researchers must remain mindful of certain factors. The current value is specific to $N_f = 2+1$ QCD; the inclusion of charm quarks ($N_f = 2+1+1$) requires further study. Additionally, finite-volume effects can become significant if the simulation box is too small ($L < 2$ fm), and careful monitoring of autocorrelations is necessary to ensure statistical independence in the flow measurements.
Summary updated