⚛️ phenomenology

A quantitative study of two-loop splitting in double parton distributions

This paper demonstrates that incorporating two-loop corrections and approximate heavy quark mass effects into the perturbative splitting kernels significantly improves the quantitative accuracy and predictive stability of double parton distributions at small inter-parton distances.

Original authors: Markus Diehl, Peter Ploessl

Published 2026-02-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Markus Diehl, Peter Ploessl

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

Imagine you are watching a high-energy collision between two protons, like two tiny, incredibly dense bags of marbles smashing into each other. Usually, when they collide, a single "marble" (a parton, like a quark or gluon) from one bag hits a single marble from the other, creating a burst of new particles. This is the standard, expected event.

But sometimes, something more complex happens: two marbles from the first bag hit two marbles from the second bag at the exact same time. This is called Double Parton Scattering (DPS). It's like two separate billiard balls from one rack hitting two separate balls from another rack simultaneously.

This paper is a deep dive into the math that predicts how often this "double hit" happens, specifically focusing on a tricky mechanism where the two marbles in one bag didn't start as two separate marbles, but were actually split from a single marble just before the collision.

Here is a breakdown of the paper's key ideas using everyday analogies:

1. The "Parent" Problem: Splitting from One

The authors focus on a specific scenario: What if the two particles involved in the collision came from the same "parent" particle that split in two just moments before?

The Analogy: Imagine a parent (a single parton) giving birth to twins (two partons) right before a race. If you try to count how many twin-races happen, you have to account for the fact that these twins came from the same source.
The Issue: In the past, scientists used a rough, "first-approximation" math (called Leading Order) to calculate this. It was like using a sketchy map. The paper shows that this sketchy map was very unstable. Depending on how you tweaked the map, the predicted number of events could change by a factor of 10 or more! That's useless for making precise predictions.
The Solution: The authors upgraded the math to a "second-order" precision (Next-to-Leading Order or NLO). Think of this as swapping the sketchy map for a high-definition GPS. The result? The predictions became much more stable. The wild swings disappeared, and the scientists could finally trust the numbers.

2. The "Double Counting" Dilemma

There is a major headache in this physics: Double Counting.

The Analogy: Imagine you are counting people at a party.
- Scenario A (DPS): You count people who arrived in pairs (two separate cars).
- Scenario B (SPS): You count people who arrived in a limo (a single complex vehicle).
- The Problem: Sometimes, a limo drops off two people who look exactly like a pair that arrived in two separate cars. If you aren't careful, you count the same two people twice—once as a "pair" and once as part of the "limo."
The Fix: The paper refines the "Subtraction Term." This is a mathematical tool used to subtract the "limo" people from the "pair" count so you don't double-count them.
The Innovation: The authors created a new, smoother way to draw the line between "pair" and "limo." Instead of a jagged, confusing border, they built a smooth ramp. This makes the calculation much more flexible and less sensitive to arbitrary choices, ensuring the final count is accurate.

3. The "Heavy" Quarks (The Heavyweights)

Protons contain light quarks (up, down) and heavy quarks (charm, bottom, top).

The Analogy: Imagine the light quarks are like ping-pong balls and the heavy quarks are like bowling balls.
The Issue: When the "splitting" happens, the math changes depending on whether the particle is light or heavy. If you treat a bowling ball like a ping-pong ball, your math breaks. Previous methods often had "jumps" or "glitches" in the math when switching from treating a quark as heavy to treating it as light.
The Fix: The authors developed a "massive scheme." Instead of abruptly switching rules, they created a smooth transition zone.
- The Result: In the old method, the math would jump up and down wildly (like a bumpy road). In their new method, the road is smooth. Even though they are using an approximation (since the full math for heavy quarks at this level of precision doesn't exist yet), it is far more realistic and removes those ugly, unphysical jumps.

4. Why This Matters

Why do we care about these double hits?

The "Signal" vs. "Noise": In particle physics, we often look for rare, new particles (the "signal"). But DPS is a background "noise" that can look like the signal. To find the new physics, we need to know exactly how much "noise" there is.
Stability is Key: Before this paper, the predictions for this noise were shaky. If you changed a tiny setting in the computer, the answer changed wildly. This paper shows that by using better math (NLO), the predictions are now solid. This allows physicists to say, "We are 95% sure this background noise is this big," which is crucial for claiming a discovery of new physics.

Summary

Think of this paper as the team that fixed the GPS and the Traffic Rules for a very complex highway (the proton collision).

They upgraded the map from a sketch to a high-def GPS (NLO corrections), making predictions stable.
They smoothed out the traffic rules to stop people from being counted twice (better subtraction).
They made sure the heavy trucks (heavy quarks) are handled smoothly without causing traffic jams (discontinuities).

The result? A much more reliable way to predict what happens when protons smash together, helping us understand the universe at its smallest scales.

1. Problem Statement

Double Parton Scattering (DPS) is a process where two partons from each proton undergo separate hard scatterings. The theoretical description relies on Double Parton Distributions (DPDs), which quantify the joint probability of finding two partons with momentum fractions $x_1, x_2$ and transverse separation $y$ .

The paper addresses two critical theoretical challenges:

Perturbative Stability at Small Distances: At small transverse distances ( $y$ ), DPDs are dominated by the perturbative splitting of a single parton into two. While Leading Order (LO) calculations exist, they suffer from severe dependence on the initial scale ( $\mu_{init}$ ) and the cutoff scale ( $y_{cut}$ ), often leading to uncertainties of an order of magnitude or more. The impact of Next-to-Leading Order (NLO) corrections (two-loop splitting kernels) on this stability was previously unquantified.
Double Counting: There is an overlap between the DPS mechanism (specifically the "1v1" contribution where both partons split from one parent) and higher-loop corrections to Single Parton Scattering (SPS). A subtraction scheme is required to avoid double counting, but the construction of this subtraction term and its sensitivity to scale choices needed revisiting, particularly when including NLO effects and heavy quark masses.

2. Methodology

The authors employ a rigorous perturbative QCD framework based on the factorization formalism developed in previous works [8, 17, 18].

Formalism:
- DPDs are decomposed into a splitting part ( $F^{spl}$ ), calculable via perturbative kernels, and an intrinsic part ( $F^{intr}$ ), modeled non-perturbatively.
- The total cross-section is formulated as $\sigma = \sigma_{DPS} - \sigma_{sub} + \sigma_{SPS}$ , where $\sigma_{sub}$ removes the double-counted region.
- The authors introduce a new subtraction scheme that interpolates between the DPS regime (small $y$ ) and the SPS regime (large $y$ ) using a smooth interpolation function $\rho(y)$ , rather than relying solely on scale variations at the initial condition.
Calculations:
- NLO Kernels: They utilize the recently computed NLO splitting kernels for massless unpolarized partons.
- Heavy Quark Masses: They develop an approximate "massive scheme" for NLO. Since full NLO massive kernels are not yet available, they construct kernels that satisfy the correct renormalization group equations (RGE) and possess the correct limiting behaviors for $y \ll 1/m_Q$ and $y \gg 1/m_Q$ , interpolating smoothly between them.
- Evolution: DPDs are evolved from an initial scale $\mu_{init}(y) \sim 1/y$ to the hard scales of the process ( $\mu_1, \mu_2$ ) using DGLAP and Collins-Soper equations.
Numerical Setup:
- Kinematics relevant for LHC (Run 2 and future 100 TeV colliders), including $W^\pm W^\pm$ , $W^\pm J/\Psi$ , and dijet production.
- Use of MSHT20 PDF sets.
- Analysis of scale variations ( $\mu_{init}$ and $\nu = b_0/y_{cut}$ ) to estimate theoretical uncertainties.

3. Key Contributions

Quantification of NLO Impact: The paper provides the first comprehensive quantitative study of how two-loop (NLO) corrections to splitting kernels affect DPDs and double parton luminosities.
Refined Subtraction Scheme: A new, flexible construction for the double-counting subtraction term ( $I_{sub}$ ) is presented. It separates the subtraction into a DPS-like component (initialized at the cutoff scale) and an SPS-like component (evaluated at the hard scale), smoothly interpolated.
Approximate Massive NLO Kernels: The authors propose a practical method to include heavy quark mass effects (charm, bottom, top) in NLO splitting calculations, addressing the discontinuities inherent in fixed-flavor number schemes.
Color Non-Singlet Analysis: The study extends the analysis to color octet channels, which are crucial for understanding the full color structure of DPS but are often suppressed by rapidity evolution.

4. Key Results

A. Stability of Predictions (Massless Case)

Scale Dependence: At LO, the double parton luminosities exhibit huge dependence on the initial scale $\mu_{init}$ (variations of an order of magnitude). At NLO, this dependence is significantly reduced, improving the stability of predictions.
Subtraction Term: The NLO corrections to the subtraction term are generally small but non-negligible. For specific channels (e.g., $u\bar{u}$ at central rapidities), the subtraction term nearly cancels the raw DPS contribution, leaving a small residual that is highly sensitive to scale choices.
Intrinsic vs. Splitting: For processes like $W^+W^-$ , the splitting contribution (which starts at NLO) is comparable in magnitude to the intrinsic contribution, highlighting the necessity of NLO calculations even for processes where LO splitting vanishes.

B. Heavy Quark Mass Effects

Discontinuities: In the massless scheme, switching active flavors at specific scales creates large, unphysical discontinuities in the DPDs, especially for heavy quarks ( $c, b, t$ ).
Massive Scheme Improvement: The approximate NLO massive scheme significantly reduces these discontinuities compared to LO.
Scheme Dependence: While the massive scheme yields different central values than the massless scheme, the dependence on the transition parameters ( $\alpha, \beta$ ) is reduced at NLO. However, for top quarks, the massless scheme fails to reproduce the massive scheme even at NLO, with deviations reaching 40%.

C. Color Structure

Color Singlet vs. Octet: Color octet DPDs are generally suppressed at large $y$ due to Collins-Soper evolution (rapidity evolution). However, in the four-gluon channel, octet contributions can be comparable to singlet ones.
Cancellation: In channels with $q\bar{q}$ pairs (e.g., $W^+W^-$ ), the cancellation between the DPS term and the subtraction term is very strong at central rapidities, making the final result highly sensitive to the cutoff scale $y_{cut}$ .

5. Significance

Theoretical Precision: The study demonstrates that including two-loop corrections is not merely a formal improvement but a practical necessity for obtaining reliable DPS predictions. The reduction in scale uncertainty transforms DPS from a qualitative estimate into a quantitative tool.
Phenomenology: The results are directly applicable to LHC analyses involving same-sign $W$ pairs, $J/\Psi$ production, and multi-jet events. The improved stability allows for better constraints on the effective cross-section $\sigma_{eff}$ and the underlying parton correlations.
Future Directions: The paper identifies the need for NLO splitting kernels for polarized partons and more accurate treatments of heavy quark masses to further reduce theoretical uncertainties. It lays the groundwork for precision DPS phenomenology at the High-Luminosity LHC and future colliders.

In conclusion, Diehl and Plößl establish that NLO corrections are essential for stabilizing the theoretical description of double parton scattering, particularly regarding scale dependence and the treatment of heavy quark masses, thereby enabling more robust comparisons with experimental data.