⚛️ phenomenology

A resonance-aware MC@NLO QCD+EW-matched calculation of lepton-pair production

This paper presents the first automated, resonance-aware MC@NLO matching of NLO QCD+EW corrections to an interleaved QCD+QED parton shower, specifically developed to eliminate spurious higher-order terms from recoil assignments and validated through predictions for Drell-Yan lepton pair production.

Original authors: Lois Flower, Joanne Roper, Marek Schönherr

Published 2026-07-08

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Lois Flower, Joanne Roper, Marek Schönherr

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: Tuning the Particle Physics Microscope

Imagine particle physicists are like astronomers trying to take a crystal-clear photo of a distant galaxy. For a long time, the "blur" in their photos came from the telescope (experimental error). But now, with the upcoming High-Luminosity LHC (a massive particle collider), the telescope is so sharp that the blur is coming from the theory used to interpret the photo.

To get a perfect picture, the theoretical predictions need to be incredibly precise. This paper is about building a better "lens" for one specific type of particle event: Drell-Yan production, where two protons smash together and create a pair of leptons (like electrons or muons), often passing through a heavy, unstable particle called a Z boson (the "resonance").

The authors have created a new computer simulation method that combines two different types of physics forces:

QCD (Strong Force): The glue that holds atomic nuclei together. It's loud, messy, and happens all the time.
EW/QED (Electroweak/EM Force): The force behind electricity and magnetism. It's quieter but becomes very important when you need extreme precision.

The Problem: The "Recoil" Mess

Think of a particle collision like a game of billiards. When a ball (a particle) hits another, it bounces off, and the table (the other particles) has to absorb the "recoil" to keep everything balanced.

In standard computer simulations, the program has to decide which ball on the table takes the recoil. Usually, it picks a neighbor. This works fine for simple collisions.

But what happens when there is a "Resonance" (the Z boson)?
Imagine the Z boson is a fragile, spinning glass vase sitting on the table. If the billiard ball hits the vase, the vase wobbles. In standard simulations, the program might accidentally assign the recoil to a ball on the other side of the vase. This makes the vase wobble in a way that doesn't make physical sense, creating "ghost" errors in the calculation. These errors are small in theory but can become huge and messy in the computer numbers, ruining the precision.

The Solution: A "Resonance-Aware" Algorithm

The authors developed a new rule for their simulation software (called MC@NLO) to handle this. They call it "Resonance-Aware."

The Old Way: The program blindly assigns recoil to any nearby particle, even if it means shaking the fragile vase.
The New Way: The program looks at the vase first. It asks, "Is this particle close enough to the vase to feel its wobble?"
- If yes (Soft emission): It treats the vase as a single unit and assigns the recoil carefully so the vase doesn't break.
- If no (Hard emission): It realizes the vase has been hit so hard it's effectively split into "production" (making the vase) and "decay" (the vase breaking apart). It then treats these as two separate events, ensuring the recoil doesn't cross the gap between them in a way that breaks the physics.

The "Interleaved" Dance

Another challenge is that QCD (strong force) and QED (electromagnetic force) happen at the same time.

QCD is like a rowdy crowd at a concert, constantly shouting and pushing (gluons splitting).
QED is like a few people whispering in the crowd (photons).

In the past, simulations often handled the rowdy crowd first and the whisperers later, or vice versa. This paper introduces an "Interleaved" approach. It's like a dance where the rowdy crowd and the whisperers take turns stepping forward in a perfectly synchronized rhythm. This ensures that when a whisper (photon) happens, it's calculated with the exact same precision as the shout (gluon).

Fixing the "Negative" Glitches

When they combined these two forces, they found a new glitch: sometimes the math produced "negative probabilities" (which is impossible in the real world) in certain regions of the data.

To fix this, they added a "Sudakov Factor."

Analogy: Imagine you are driving a car. If you drive too fast, your brakes might overheat and fail. The Sudakov factor is like a smart cruise control that automatically slows you down before you reach the point where the brakes would fail. It suppresses the unphysical "negative" results by applying a safety brake to the simulation, ensuring the numbers stay positive and realistic.

The Results: A Sharper Photo

The authors tested this new method on the production of electron-positron pairs (the "lepton pairs").

Validation: They checked their math against known fixed calculations and found it matched perfectly.
The Difference: When they compared their new "Resonance-Aware" method against the old "Blind" method, they saw a difference.
- The Mass Distribution (how heavy the particles look): The old method distorted the shape of the Z boson peak by about 0.2% to 0.5%. While this sounds small, in the world of high-precision physics, it's like measuring a human hair's width and being off by a fraction of a millimeter.
- The Photon Energy: The new method changed the prediction for high-energy photons by up to 15% in the extreme tails of the data.

Conclusion

This paper presents the first fully automated way to simulate these complex particle collisions where the "Strong Force" and "Electromagnetic Force" dance together, while carefully protecting fragile "Resonance" particles from being shaken apart by the simulation's math.

By making the simulation "aware" of these resonances, they have removed artificial distortions, providing a much cleaner, more precise theoretical prediction. This is essential for the future of particle physics, where scientists need to spot tiny deviations from the Standard Model that could reveal new, undiscovered physics.

Technical Summary: A Resonance-Aware MC@NLO QCD+EW-Matched Calculation of Lepton-Pair Production

Problem Statement
As the field of collider physics transitions toward the High-Luminosity LHC (HL-LHC) era, experimental uncertainties are shrinking, demanding commensurate improvements in theoretical precision. While Next-to-Leading Order (NLO) Quantum Chromodynamics (QCD) calculations are standard, Electroweak (EW) effects are becoming increasingly critical, particularly for observables involving color-neutral particles or large weak isospin. Furthermore, QED radiation at hadron colliders can induce large corrections to specific observables.

A significant challenge arises when matching fixed-order NLO QCD+EW calculations to parton showers (PS) in the presence of internal resonances (e.g., $Z$ or $W$ bosons). Standard parton shower algorithms, such as those based on the Catani-Seymour dipole formalism, rely on recoil assignment to spectator partons to maintain momentum conservation. In resonant processes, this recoil assignment can alter the virtuality of the resonant propagator, introducing spurious higher-order terms that are numerically large but theoretically uncontrolled. These artifacts distort the shape of the resonance and compromise the precision required for processes like Drell-Yan lepton-pair production ( $pp \to \ell^+\ell^-$ ).

Methodology
The authors present the first automated matching of NLO QCD+EW to an interleaved QCD+QED parton shower using the MC@NLO method within the Catani-Seymour dipole formalism, implemented in the SHERPA event generator. The methodology comprises three main components:

Interleaved QCD+QED Dipole Shower:
The authors extend the CSSHOWER framework to evolve QCD and QED emissions simultaneously. The shower evolution is defined by a generating functional where the ordering parameter $t$ is the transverse momentum for emissions and virtuality for splittings. The Sudakov form factor is constructed as a product of QCD and QED components, ensuring an interleaved evolution rather than a simple sum. To handle the non-positive definite splitting functions inherent in QED dipoles (due to charge correlations), a weighted shower approach is employed, with specific strategies to minimize negative weights by restricting dipole spectators.
MC@NLO Matching with Sudakov Modification:
The standard MC@NLO master equation is adapted to match the interleaved shower to NLO QCD+EW matrix elements. This involves generating "S-events" (Born + virtual + integrated dipoles) and "H-events" (real corrections).
- H-event Modification: The authors identify that in interleaved showers, the QCD dynamics can induce large Sudakov suppressions ( $\bar{\Delta} \ll 1$ ) that, when combined with QED real emissions in H-events, lead to unphysical negative contributions in the differential cross-section. To mitigate this, they introduce an additional Sudakov factor, $\Delta_n(t_n, t_{n+1})$ , computed via the standard interleaved shower, applied directly to the H-events. This "MC@NLO $_{sud}$ " approach suppresses unphysical higher-order terms ( $O(\alpha_s \alpha)$ ) while maintaining formal NLO accuracy.
Resonance-Aware Subtraction and Evolution:
To address the resonance problem, the authors develop a resonance-aware modification to the subtraction and shower algorithms. The phase space is divided into "resonant" and "non-resonant" regimes based on two criteria:
- Virtuality ( $\Delta_r$ ): The distance of the resonant propagator's virtuality from its nominal mass, normalized by the width $\Gamma_r$ .
- Emission Scale ( $t$ ): The hardness of the emission relative to the resonance width squared ( $\Gamma_r^2$ ).
In the resonant regime (soft/collinear emissions relative to the width), standard dipoles spanning the resonance are active. In the non-resonant regime (hard emissions resolving the resonance), the process factorizes into production and decay subprocesses. Dipoles connecting the production and decay sectors are suppressed or switched off using $\Theta$ -functions. This ensures that recoil assignments do not artificially distort the resonant propagator's virtuality distribution. For neutral resonances (like the $Z$ boson), this involves modifying final-initial and initial-final dipoles; for charged resonances, additional resonant dipoles would be required (noted as future work).

Key Results
The authors validate their framework using $pp \to \nu_e \bar{\nu}_e$ (non-resonant) and $pp \to e^+ e^-$ (resonant) at $\sqrt{s} = 13.6$ TeV.

Validation against Fixed Order: For non-resonant processes, the MC@NLO QCD+EW results match fixed-order NLO QCD+EW calculations closely for hard emissions, with S-events dominating soft regions and exhibiting correct Sudakov suppression.
Correction of Artifacts: In the $pp \to e^+ e^-$ channel, the standard MC@NLO QCD+EW calculation (without the H-event Sudakov modification) produces unphysical negative cross-sections and kinks in the photon transverse momentum ( $p_{\perp, \gamma}$ ) and dilepton mass ( $m_{\ell\ell}$ ) spectra. The inclusion of the additional Sudakov factor in MC@NLO $_{sud}$ removes these artifacts, restoring physical behavior and agreement with YFS-resummed QED corrections.
Resonance-Aware Impact: Comparing resonance-aware and standard (resonance-unaware) calculations for $pp \to e^+ e^-$ $pp \to e^{+} e^{-}$ reveals:
- Dilepton Mass: A distortion of up to 0.5% in the invariant mass distribution below the $Z$ peak ( $m_{\ell\ell} \approx 60$ GeV) when using standard dipoles. The resonance-aware treatment corrects this, ensuring the shape is independent of subtraction scheme parameters.
- Photon $p_\perp$ : Significant differences in the hard photon tail. The resonance-aware calculation enhances the hardest photon emissions by up to 15% at $p_{\perp, \gamma} \approx 1$ TeV compared to the standard approach, attributed to the proper handling of ISR-FSR interference effects via resonance-spanning dipoles.
Statistical Enhancement: The paper investigates artificially enhancing QED splittings to improve statistics. While a factor of $c=5$ improves photon sampling without artifacts, $c=10$ introduces unphysical sub-percent deviations due to weight fluctuations, leading the authors to recommend against aggressive enhancement in the matched first emission.

Significance and Claims
The paper claims to present the first automated matching of NLO QCD+EW to an interleaved QCD+QED parton shower using the MC@NLO method in the Catani-Seymour dipole formalism. Its primary significance lies in:

Precision for HL-LHC: Providing a theoretical tool capable of matching the increasing precision of experimental data by consistently including both QCD and EW higher-order corrections.
Resonance Handling: Solving the specific problem of spurious higher-order terms introduced by recoil assignments in resonant processes. The proposed resonance-aware modification ensures that the virtuality distribution of resonant propagators is not distorted by the parton shower, which is vital for precision measurements of the $Z$ boson and PDF extraction.
Artifact Removal: Demonstrating that unphysical negative weights and spectral distortions in interleaved QCD+EW matching can be cured by a physically motivated Sudakov modification to H-events.

The authors note that while the resonance-aware algorithm is general, their current implementation is process-specific (focusing on neutral-current Drell-Yan). They acknowledge that extending this to color- or charge-charged resonances (like top quarks or $W$ bosons) requires additional splitting functions where the resonance acts as both emitter and spectator, a task reserved for future work.