⚛️ lattice

Disentangling Scheme Dependence in Quasi-PDFs with a Transverse-Momentum Cutoff

This paper employs a transverse-momentum cutoff to systematically disentangle scheme-dependent contributions from infrared divergences in the one-loop nonsinglet quark quasi-PDF, thereby clarifying the renormalization-group behavior and providing a benchmark for matching to lightcone parton distributions.

Original authors: Junegone Chay

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Junegone Chay

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Trying to See the Invisible

Imagine you are trying to understand the internal structure of a proton (a tiny particle inside an atom). Physicists call the map of how the smaller parts (quarks and gluons) move inside the proton a Parton Distribution Function (PDF). Think of this PDF as a "speed limit map" for the particles inside the proton.

The problem is that this map is drawn on a "lightcone"—a special, fast-moving perspective that is impossible to capture directly in a computer simulation. Our computers (using a method called Lattice QCD) can only take snapshots of the proton while it is sitting still or moving slowly in "Euclidean time" (a static, 3D-like view).

To get the "speed limit map" (the real PDF) from the "static snapshot" (the computer simulation), scientists use a bridge called Quasi-PDFs. This bridge connects the static snapshot to the real map.

The Problem: The "Dirty" Bridge

Building this bridge involves doing some very complex math. However, this math comes with two types of "noise" or "dirt" that messes up the calculation:

The Real Physics (Infrared Divergence): This is the actual signal we need. It represents the real, physical behavior of the particles. We want to keep this.
The Math Artifacts (Scheme Dependence): This is the "dirt" caused by the specific rules (renormalization schemes) we use to make the math work. Different rules create different types of dirt. If we don't separate the dirt from the signal, our final map will be wrong.

The paper argues that while scientists have been building this bridge, they haven't always been very clear about which part of the math is the "real signal" and which part is just "mathematical cleanup."

The Solution: The "Transverse-Momentum Cutoff" (TMC)

The author, Junegone Chay, decides to test a specific set of cleaning rules called the Transverse-Momentum Cutoff (TMC) scheme.

Imagine you are cleaning a muddy floor.

The Old Way: You might just sweep everything into a pile and hope the dirt stays there.
The TMC Way: You put a specific barrier (a cutoff) on the floor. Anything that tries to go beyond that barrier is immediately identified as "dirt" and swept away.

In this paper, the author uses this barrier to systematically separate the "dirt" (scheme-dependent parts) from the "floor" (the real physics).

The Two Types of "Dirt" Found

When the author applied this TMC cleaning method, they found two distinct types of mathematical dirt that needed to be removed before the bridge could be used:

The Linear "Sticky" Dirt (Linear Divergence):
- The Analogy: Imagine the bridge is made of a sticky tape. As you pull the tape, it gets longer and longer, creating a "linear" mess that grows with the size of the tape.
- The Paper's Finding: In the TMC scheme, there is a "linear divergence" (a mess that grows linearly) caused by the way the bridge is constructed. The author shows that this specific mess must be cut off and thrown away (assigned to a "counterterm") because it doesn't exist in the real world map we are trying to build.
The "Edge" Dirt (Boundary Divergence):
- The Analogy: Imagine the bridge is a long road. Most of the road is fine, but at the very far ends (infinity), the road starts to crumble or fade out.
- The Paper's Finding: Because the Quasi-PDF exists on a "full line" (it can go from negative infinity to positive infinity), there is special "dirt" that appears only at the very edges ( $x = \pm \infty$ ). The author discovered that this edge dirt is different from the "sticky tape" dirt. It needs its own special cleanup bucket.

The Result: A Cleaner Bridge

By using this method, the author successfully separated the math into three clear piles:

The Counterterm (The Trash Bin): This contains all the "dirt" that depends on the specific rules used (the cutoff $\Lambda$ ). This includes the sticky linear mess and the edge crumbling. This is thrown away.
The Remainder (The Clean Signal): This is what is left after the trash is removed. It contains the "real physics" (the infrared divergence) that matches the real world.
The Matching Coefficient (The Blueprint): This is the final instruction manual that tells us how to turn the clean signal from the computer into the real-world map.

Why This Matters

The paper doesn't just do the math; it clarifies how the math works.

It shows that you can't just say "remove the cutoff dependence." You have to be careful about where that dependence comes from (the middle of the road vs. the edges).
It proves that even though different cleaning rules (schemes) create different types of dirt, if you clean them correctly, the final "speed limit map" (the matching coefficient) ends up being the same as if you had used a different cleaning method (like the standard MS scheme).

Summary in One Sentence

This paper acts like a master plumber who installs a new filter (the TMC scheme) to clearly separate the "gunk" caused by the filter itself from the "clean water" of real physics, ensuring that the final map of the proton's interior is accurate and free of mathematical artifacts.

Technical Summary: Disentangling Scheme Dependence in Quasi-PDFs with a Transverse-Momentum Cutoff

Problem Statement
Quasi-parton distribution functions (quasi-PDFs) serve as a bridge between Euclidean spatial correlations in lattice QCD and lightcone parton distribution functions (PDFs) via Large-Momentum Effective Theory (LaMET). While the matching between quasi-PDFs and lightcone PDFs relies on the cancellation of identical collinear infrared (IR) divergences, the treatment of ultraviolet (UV) divergences and renormalization prescriptions remains a source of ambiguity. Unlike the standard $\overline{\text{MS}}$ scheme for lightcone PDFs, there is no consensus on the renormalization scheme for quasi-PDFs. Various schemes (e.g., RI/MOM, ratio, hybrid) have been proposed, often motivated by lattice regularization needs. A critical challenge is disentangling the scheme-dependent contributions (associated with UV cutoffs and renormalization prescriptions) from the scheme-independent physical content required for matching. Specifically, the separation of linear divergences (inherent to spatial Wilson lines in lattice or cutoff schemes) from the logarithmic structures and finite terms necessary for the matching coefficient is not always transparent.

Methodology
The authors utilize a transverse-momentum cutoff (TMC) scheme as a diagnostic setting to systematically decompose the one-loop nonsinglet quark quasi-PDF. The methodology involves:

Asymptotic Separation: Instead of naively subtracting cutoff-dependent tails at fixed momentum fraction $x$ , the authors isolate the scheme-dependent part by analyzing the asymptotic behavior of the integrand as the transverse momentum $k_\perp \to \infty$ . They identify the scheme-dependent sector as the integral of the asymptotic form up to the cutoff $\Lambda$ , while the remainder is defined as the convergent integral of the difference between the full integrand and its asymptotic form.
Distributional Analysis: The calculation is performed in momentum space, and results are expressed as distributions on the full real line ( $-\infty < x < \infty$ ), acknowledging that quasi-PDFs support regions where $x > 1$ and $x < 0$ due to large negative gluon momentum fractions.
Component Decomposition: The one-loop calculation is broken down into tadpole, sail, vertex, and wave-function renormalization diagrams. The authors explicitly separate terms with explicit dependence on the cutoff $\Lambda$ (assigned to the counterterm) from the finite and IR-divergent remnants.
Boundary Treatment: Special attention is paid to the behavior at $x = \pm \infty$ . The authors demonstrate that expanding the large- $|x|$ tails as distributions generates UV boundary poles (regulated via dimensional regularization in the expansion) which must be absorbed into the counterterm, distinct from the explicit $\Lambda$ -dependent terms.

Key Contributions and Results

Systematic Decomposition: The paper establishes a clear decomposition of the bare quasi-PDF $\tilde{q}^{(1)}_{\text{bare}}$ $\tilde{q}_{bare}^{(1)}$ into a TMC counterterm $\tilde{q}^{(1)}_{\text{TMC,ct}}$ $\tilde{q}_{TMC,ct}^{(1)}$ and a renormalized quasi-PDF $\tilde{q}^{(1)}_{\text{TMC,ren}}$ $\tilde{q}_{TMC,ren}^{(1)}$ .
- The counterterm contains:
  - The linearly divergent tadpole contribution, expressed as a double plus distribution $[1/(1-x)^2]^{[-\infty, \infty]}_{++(1)}$ proportional to $\Lambda/p_z$ .
  - The logarithmic wave-function renormalization term proportional to $\ln(\Lambda^2/\mu^2)$ .
  - UV boundary poles at $x = \pm \infty$ arising from the distributional expansion of large- $|x|$ tails.
- The renormalized quasi-PDF retains the full collinear IR pole (required to cancel the pole in the lightcone PDF) and the finite terms relevant for matching.
Matching Coefficient: The one-loop matching coefficient $C^{(1)}_{\text{TMC}}$ is derived by subtracting the renormalized lightcone PDF (in $\overline{\text{MS}}$ ) from the renormalized TMC quasi-PDF. The result shows that after removing the explicit TMC-dependent sector, the nonlocal $x$ -dependent terms agree with the standard $\overline{\text{MS}}$ full-line matching result.
Renormalization Group (RG) Behavior: The authors derive the RG equation for the renormalized quasi-PDF in the TMC scheme. They distinguish between:
- The finite- $x$ anomalous dimension, driven by the logarithmic wave-function counterterm.
- The boundary anomalous dimension, supported at $x = \pm \infty$ , arising from the UV boundary poles.
  The paper clarifies that while the boundary terms do not affect the matching coefficient at finite $x$ , they are essential for the RG evolution of the quasi-PDF as a distribution on the full line.
Scheme Dependence of Finite Terms: The analysis reveals that the finite constant in the $\delta(1-x)$ term of the matching coefficient depends on the specific renormalization prescription for the wave-function contribution (e.g., the finite part remaining after subtracting the $\Lambda$ -dependent or $1/\epsilon$ pole).

Significance
The paper claims to provide a "benchmark" for understanding scheme dependence in quasi-PDFs. By using the TMC scheme, which makes cutoff dependence explicit, the authors demonstrate how to systematically separate the counterterm (scheme-dependent) from the physical matching remnant. The primary significance lies in clarifying that:

The counterterm for a cutoff-regulated quasi-PDF must be defined as a distribution on the full line, incorporating both explicit cutoff dependence and boundary UV terms.
This separation preserves the collinear IR structure required for LaMET matching without disturbing the finite- $x$ matching kernel.
The approach offers a framework to examine analogous separations in other renormalization schemes (such as RI/MOM), potentially aiding in the construction of scheme-independent matching procedures.

The work does not propose new experimental applications but focuses on the theoretical consistency and organization of perturbative calculations in LaMET, ensuring that the transition from lattice correlations to physical PDFs is theoretically well-defined.