Determination of $\alpha_S$ in the $SU(3)$ Yang-Mills theory

This paper presents a strategy to determine the running of the strong coupling in $SU(3)$ Yang-Mills theory using a finite-volume scheme with twisted boundary conditions and a gradient-flow coupling, demonstrating preliminary continuum extrapolation results that promise reduced statistical errors and the elimination of linear cutoff effects.

The Gist

The Big Picture: Measuring the "Glue" of the Universe

Imagine the universe is held together by a super-strong, invisible glue. In physics, this is called the Strong Force, and it's what keeps the tiny particles inside an atom (protons and neutrons) stuck together.

Scientists have a number they call $\alpha_S$ (alpha-S) that measures exactly how strong this glue is. Knowing this number precisely is crucial. If you get it wrong, your calculations for how particles behave in giant particle smashers (like the Large Hadron Collider) will be off, and you might miss discovering new physics.

Currently, we know this number pretty well, but there's still a tiny bit of uncertainty. This paper is about a team of physicists trying to sharpen that number even further, aiming for sub-percent precision (meaning they want to be right within less than 1% of the true value).

The Problem: The "Zoom" Issue

To measure this glue strength, physicists use a technique called Lattice QCD. Imagine the universe isn't a smooth sheet, but a giant 3D grid (like a chessboard made of tiny cubes).

To measure the glue, they have to look at the grid at different "zoom levels":

1. Zoomed in: Looking at the tiny cubes (high energy).
2. Zoomed out: Looking at the whole board (low energy).

The challenge is that the grid itself is an approximation. If the cubes are too big, your measurement is blurry (this is called a "cutoff effect"). To get the real answer, you have to simulate the grid with smaller and smaller cubes and then mathematically "extrapolate" to see what the answer would be if the cubes were infinitely small (zero size).

The Analogy: Imagine trying to measure the circumference of a circle by drawing it on graph paper.

• If you use thick markers and big squares, your circle looks jagged and your measurement is wrong.
• If you use fine pens and tiny squares, it looks smoother.
• The "continuum extrapolation" is the math you do to guess what the perfect, smooth circle would look like if you had infinite resolution.

The New Strategy: The "Split-Step" Method

In the past, scientists tried to jump from a small grid to a big grid in one giant leap to measure how the glue strength changes. This is like trying to walk up a steep, rocky mountain in one giant stride. It's hard, you might slip, and your measurement of the slope (the "systematic error") is messy.

The authors propose a new way: The "Split-Step" Method.

Instead of one giant leap, they break the journey into two smaller, easier steps:

1. Step 1 (Change the Scale): Keep the grid size the same, but change how you measure the glue (the "renormalization scale"). This is like looking at the same mountain but through a different telescope.
2. Step 2 (Change the Volume): Keep the measurement method the same, but actually make the grid bigger. This is like physically walking up the mountain.

Why is this better?

• Step 1 is the tricky part where the "grid artifacts" (the jaggedness of the graph paper) usually mess things up. By isolating this, they can focus all their computing power on fixing just this one part.
• Step 2 is much cleaner and easier to calculate.
• By doing them separately, they get a much smoother, more accurate path to the top of the mountain.

The Tools of the Trade

To make this work, they used some fancy tools:

• Twisted Boundary Conditions: Imagine a video game world where if you walk off the right edge, you appear on the left. Usually, this creates weird "ghost" effects at the edges. The authors use a special "twist" (like twisting a rubber band) so that the edges don't mess up the math. This removes a whole class of errors that other teams have to struggle with.
• Gradient Flow: Imagine the grid is a muddy field. If you pour water on it, the mud flows and settles into a smooth, clean surface. The "Gradient Flow" is a mathematical way of "smoothing out" the noisy, jagged data on the grid so they can see the true signal underneath.
• Topology Freezing: Sometimes, the grid gets "stuck" in a specific shape (like a knot) and can't change. The authors use a special filter to ignore these stuck shapes so they only measure the smooth, flowing parts.

The Results: A Clearer Picture

The paper presents "preliminary results," meaning they have run the simulations and the math looks promising.

They compared their new "Split-Step" method against the old "One Giant Leap" method. The results showed:

• Less Error: The "Split-Step" method had much smaller statistical errors.
• Better Control: The math used to smooth out the grid (the extrapolation) was much more stable.
• More Data: Because they split the problem, they could use more of their simulation data to make the final calculation.

The Bottom Line

This paper is a blueprint for a more precise way to measure the strength of the universe's strongest glue. By breaking a difficult math problem into two smaller, manageable steps and using a special "twisted" grid setup, the team has shown a path to reducing the uncertainty in our understanding of the fundamental forces of nature.

If you think of the current measurement as a blurry photo of a distant star, this new method is like switching to a high-definition camera with a better lens, promising a crystal-clear image of the universe's building blocks.

Technical Summary

1. Problem Statement

The strong coupling constant, $\alpha_S$, is a fundamental parameter in Quantum Chromodynamics (QCD) essential for perturbative calculations of LHC cross-sections. Current uncertainties in $\alpha_S$ contribute significantly to theoretical errors in these calculations.

• Context: The most precise determinations to date utilize a "decoupling strategy," which relates the QCD $\Lambda$ parameter to that of pure $SU(3)$ Yang-Mills theory (the heavy quark mass limit).
• Bottleneck: In previous high-precision studies, approximately 20% of the total squared error in the QCD coupling originated from the determination of the pure gauge $\Lambda$ parameter.
• Goal: This work aims to improve the precision of the pure gauge $\Lambda$ parameter determination to the sub-percent level by reducing systematic errors associated with the continuum extrapolation of the step-scaling function.

2. Methodology

The authors propose a novel strategy combining finite-volume scaling, twisted boundary conditions, and gradient flow techniques.

A. Finite-Size Scaling and Step-Scaling

Instead of inverting a perturbative series directly, the authors use the step-scaling function $\sigma_s(u)$, which quantifies the change in the running coupling $\bar{g}^2$ when the renormalization scale $\mu$ is changed by a factor $s$ (typically $s=2$).
$$\sigma_s(u) = \bar{g}^2(\mu/s) \big|_{\bar{g}^2(\mu)=u}$$
This allows the integration of the $\beta$-function (required for $\Lambda$) to be performed in discrete steps.

B. Scheme Selection

• Boundary Conditions: The authors employ Twisted Boundary Conditions (TBC) instead of the standard Schrödinger Functional (Dirichlet) or periodic conditions.
  • Advantage: TBCs preserve translational invariance, eliminating linear cutoff effects ($O(a)$) and reducing statistical errors compared to other finite-volume schemes.
• Observable: They utilize Gradient Flow (GF) couplings. The flow equation smears the gauge field, automatically renormalizing composite operators. The coupling is defined via the dimensionless energy density $t^2 \langle E(t) \rangle$ at a flow time $t$ related to the box size $L$ by $\sqrt{8t} = cL$.
• Topology: To avoid topology freezing in small volumes, a projector $\delta_Q$ is used to restrict measurements to the zero-topology sector.

C. The "Split" Strategy (Key Innovation)

Traditionally, the step-scaling function is determined by doubling the lattice volume ($L \to 2L$) at a fixed bare coupling. The authors adopt a two-step decomposition proposed in Ref. [10]:

1. $\mathcal{J}_1(u)$: Change the renormalization scale by factor $s$ (change $c$) at fixed volume $L$.
2. $\mathcal{J}_2(u)$: Change the volume by factor $s$ ($L \to sL$) at fixed renormalization scale.
   The total step-scaling function is reconstructed as $\sigma(u) = \mathcal{J}_2(\mathcal{J}_1(u))$.

• Rationale: Cutoff effects are predominantly associated with the change in scale ($\mathcal{J}_1$). The volume change ($\mathcal{J}_2$) is less sensitive to lattice artifacts. This separation allows for better control over systematics and utilizes more data points for the volume extrapolation.

3. Numerical Setup and Simulation

• Action: Standard Wilson plaquette action with twisted boundary conditions implemented via specific plaquette weights ($w(p) = e^{2\pi i/3}$).
• Flow Equation: $O(a^2)$ improved Zeuthen flow equation to eliminate $O(a^2)$ discretization errors.
• Energy Density: $O(a^2)$ improved combination of plaquette and clover definitions.
• Lattice Parameters: Simulations performed on volumes $L/a \in [8, 48]$ with bare couplings $\beta \in [5.9, 15.0]$. Flow parameters $c \in \{0.2, 0.3, 0.4\}$ were used.
• Software: LatticeGPU.jl for simulations and ADerrors.jl for error analysis (using the $\Gamma$-method and automatic differentiation).

4. Key Results

The authors compared the direct determination of $\sigma(u)$ against the split determination ($\mathcal{J}_1$ and $\mathcal{J}_2$) using preliminary data. They evaluated systematics using FLAG review criteria:

• $\eta$ (Level Arm): Ratio $[a_{max}/a_{min}]^2$. Higher is better.
• $\delta$ (Extrapolation Size): The deviation of the extrapolated value from the finest lattice value, normalized by uncertainty. Lower is better ($\delta < 3$ is considered good).

Comparative Findings (Table 1):

• Direct Method ($\Sigma$): $\eta = 2.25$, $\delta = 2.27$.
• Split Method ($\mathcal{J}_1$): $\eta$ ranges from 4.00 to 5.76; $\delta$ ranges from 1.33 to 1.80.
• Split Method ($\mathcal{J}_2$): $\eta = 4.00$; $\delta = 0.35$.

Observations:

1. Reduced Systematics: The split method yields significantly larger level arms ($\eta$) and smaller extrapolation sizes ($\delta$), indicating superior control over continuum extrapolation.
2. Linear Scaling: The volume change step ($\mathcal{J}_2$) exhibits very small scaling violations, allowing for robust linear extrapolations in $(a/L)^2$.
3. Precision: Despite the decomposition, the relative statistical error of the final extrapolated values remains comparable to the direct method (approx. $4.8 \times 10^{-4}$ to $8.6 \times 10^{-4}$).

5. Significance and Conclusion

This work presents a robust strategy to achieve sub-percent precision in the determination of the $SU(3)$ Yang-Mills $\Lambda$ parameter, a critical input for the decoupling method used to determine $\alpha_S$ in full QCD.

• Technical Advancement: By decoupling the scale change from the volume change, the authors isolate the dominant source of cutoff effects, allowing for more reliable continuum extrapolations.
• Methodological Improvement: The use of twisted boundary conditions combined with gradient flow eliminates linear cutoff effects and reduces statistical noise.
• Impact: The preliminary results demonstrate that the split strategy offers better quality extrapolations (higher $\eta$, lower $\delta$) without sacrificing statistical precision. This paves the way for a more precise determination of the strong coupling constant, directly benefiting high-energy physics phenomenology at the LHC.

The authors conclude that this approach is a promising path forward for precision lattice QCD, with full results expected in future publications.