⚛️ nuclear experiments

Reconstructing jet anisotropies with cumulants

Original authors: Tanner Mengel, Niseem Magdy, Ron Belmont, Anthony Timmins, Christine Nattrass

Published 2026-06-25

📖 5 min read🧠 Deep dive

Original authors: Tanner Mengel, Niseem Magdy, Ron Belmont, Anthony Timmins, Christine Nattrass

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 smashing two heavy atoms together at nearly the speed of light. When they collide, they create a tiny, super-hot drop of "primordial soup" called the quark-gluon plasma (QGP). This soup is so dense and energetic that it behaves like a perfect fluid, swirling and flowing.

Scientists have long known that the smaller particles (like pions) flying out of this soup swirl in a specific pattern, like water going down a drain. But they've been puzzled by the jets—high-speed streams of particles that shoot out from the collision. Do these jets also swirl in a pattern? If so, why? And how do we measure it without getting confused by the noise of the swirling soup?

This paper is like a detective story where the authors build a "fake crime scene" in a computer to test a new way of solving the mystery.

The Problem: The Foggy Window

Imagine you are trying to watch a specific dancer (a jet) spin in a crowded, foggy ballroom (the QGP). The crowd is moving in waves, and the fog makes it hard to see exactly where the dancer is or how fast they are spinning.

The Old Way: Scientists used to try to guess the dancer's spin by looking at the general direction the whole crowd was leaning (the "event plane"). But this method can be tricky; sometimes the crowd's movement looks different depending on how you count them, leading to errors of up to 10%.
The New Idea: The authors wanted to see if they could use a more sophisticated counting method called "cumulants" to isolate the dancer's spin from the crowd's noise.

The Experiment: Building a Fake Ballroom

Since they can't perfectly control a real atomic collision, the authors built a computer simulation (a "fake ballroom") to test their method.

The Background (TennGen): They created a simulation of the swirling crowd (the QGP) that acts like a hydrodynamic fluid.
The Dancer (Pythia-8): They added high-speed jets into this crowd.
The Twist: They programmed the jets to spin in specific, known patterns. Sometimes the spin was constant, sometimes it got stronger as the jet got faster, and sometimes they even used a weird, unrealistic "wavy" pattern just to see if their method could handle extreme cases.

The Solution: The "Group Hug" Technique

To measure the jet's spin, the authors used a technique called multi-particle correlations.

The 2-Particle Method: Imagine asking two people in the crowd, "Are you looking at the dancer?" and comparing their answers. This gives a rough idea.
The 4-Particle Method: Now, imagine asking four people at once. This is like a "group hug" of data. It's much harder for random noise or local glitches (like one person bumping into another) to fake a signal when you need four people to agree. This helps filter out the "fog" and the "noise" to see the true spin of the jet.

The Magic Trick: Unfolding the Truth

Even with the best cameras, the simulation has "blur" (measurement errors). The authors used a mathematical trick called Bayesian unfolding.

The Analogy: Imagine you have a blurry photo of a spinning top. You know exactly how your camera blurs things. The "unfolding" process is like using a computer program to reverse-engineer the blur, sharpening the image until you see the top spinning exactly as it was in reality.
They tested this by taking their "blurry" simulated data and running it through the unfolding process. The result? The computer perfectly recovered the original spin patterns they had programmed in, even the weird, wavy ones.

What They Found

The paper claims that:

It Works: Their new method (using 2- and 4-particle cumulants) can accurately reconstruct the spinning patterns of jets, even when the data is messy.
It's Robust: Whether the jet's spin was constant or changed wildly with speed, the method found the truth.
It's Better: The 4-particle method (the "group hug") gave results that matched the 2-particle method, suggesting that in their simulated environment, the extra noise wasn't messing up the measurement.

Why This Matters (According to the Paper)

The authors say this is a crucial step because:

It helps us understand how jets lose energy in the plasma (jet quenching).
It offers a way to study these patterns in smaller collision systems (like smashing a proton into a gold nucleus), where the old "event plane" method doesn't work well because the crowd is too small to define a clear direction.

In short: The authors built a virtual lab to prove that a new, smarter way of counting particles can cut through the noise of atomic collisions to reveal exactly how high-speed jets behave inside the quark-gluon plasma. They didn't discover a new particle or change physics laws; they proved their measuring tape is accurate and ready for real-world use.

Technical Summary: Reconstructing Jet Azimuthal Anisotropies with Cumulants

Problem Statement
In relativistic heavy-ion collisions, the formation of quark-gluon plasma (QGP) leads to azimuthal anisotropies in hadron production. While low transverse momentum ( $p_T < 3$ GeV) anisotropies are well-described by hydrodynamic flow, high- $p_T$ ( $> 10$ GeV) hadrons and jets also exhibit non-zero anisotropy, likely driven by path-length dependent jet quenching within the QGP. A significant challenge in this field is the interpretation of these high- $p_T$ anisotropies, particularly in smaller collision systems (e.g., $p$ +A) where jet quenching signatures are ambiguous.

Current measurements of jet differential azimuthal anisotropies ( $v_n^{jet}$ ) often rely on the event plane method, which defines the event plane using soft particles. However, discrepancies between event plane methods and multi-particle cumulant methods can reach up to 10%, complicating the interpretation of experimental data. Furthermore, factors such as long-range non-flow effects and initial/final state fluctuations can influence flow correlations measured via 2-particle correlations. There is a need to determine if multi-particle cumulant methods can robustly reconstruct jet anisotropies, disentangle short- and long-range correlations, and provide a consistent framework for analyzing high- $p_T$ behavior across different collision systems.

Methodology
The authors propose a simulation-based framework to test the feasibility of measuring jet azimuthal anisotropies using 2- and 4-particle cumulants. The methodology involves the following steps:

Event Simulation:
- Background: A hydro-like background is generated using TennGen, a data-driven generator tuned to Au+Au collision data at $\sqrt{s_{NN}} = 200$ GeV (20–30% centrality). TennGen reproduces flow correlations and matches experimental multiplicities and $p_T$ spectra.
- Jet Signal: Jets are generated using Pythia-8 (Monash Tune 13) with a minimum hard scattering $p_T$ cutoff of 5 GeV.
- Merging: Pythia-8 jets are merged with TennGen background particles. The leading jets are realigned to simulate specific differential azimuthal anisotropies ( $v_n^{jet}$ ) with various functional forms, including constant values, linear increases with $p_T$ , and sinusoidal dependencies.
- Clustering: Merged constituents are clustered into anti- $k_t$ jets ( $R=0.4$ ) using FastJet. Jet momentum is corrected for background using a jet-multiplicity subtraction method.
Cumulant Construction:
- The study utilizes the traditional cumulant method to construct $v_n^{jet}\{2k\}$ (where $k=1$ for 2-particle and $k=2$ for 4-particle correlations).
- Correlations are calculated between the jet of interest (Region A, $|\eta| < 1.1-R$ ) and reference flow particles (Region C, $3.1 < |\eta| < 5.1$ ) to minimize non-flow effects.
- Equations for 2-particle ( $v_n^{jet}\{2\}$ ) and 4-particle ( $v_n^{jet}\{4\}$ ) cumulants are defined using event-averaged correlations of the form $\langle e^{in(\psi - \phi)} \rangle$ .
Unfolding Procedure:
- To address bin-by-bin migration caused by finite resolution in uncorrected jet momentum measurements, a Bayesian unfolding technique (RooUnfold 2.0.0) is employed.
- A four-dimensional response matrix is constructed mapping reconstructed-level correlations and $p_T$ to truth-level correlations and $p_T$ .
- The unfolding is iterated (typically 3–5 times) until the $\chi^2$ between unfolded and truth values stabilizes.

Key Contributions

Feasibility Demonstration: The paper demonstrates the feasibility of measuring $v_n^{jet}$ using 2- and 4-particle cumulants in a simulated Au+Au environment.
Robustness Testing: The authors test the reconstruction method against a wide range of input $v_n^{jet}$ functional forms, including physically unrealistic sinusoidal dependencies, to ensure the method is not biased by specific momentum dependencies.
Unfolding Validation: A critical contribution is the validation of the Bayesian unfolding procedure. The authors show that the method can accurately recover the true functional form of $v_n^{jet}(p_T)$ even when the response matrix is constructed from models with different functional forms than the input truth.
Comparison of Cumulant Orders: The study compares 2-particle and 4-particle cumulant results, highlighting their agreement in the absence of flow fluctuations, which serves as a baseline for future studies involving fluctuating systems.

Results

Reconstruction Accuracy: The unfolded results for $v_n^{jet}\{2\}$ and $v_n^{jet}\{4\}$ accurately recover the input truth values for constant, linearly increasing, and sinusoidal $v_n^{jet}(p_T)$ models.
Independence from Response Matrix: As shown in Figure 2, the corrected $v_n^{jet}\{2\}$ values recover the true constant input ( $4\%$ ) regardless of whether the response matrix was built using a constant, scaled constant, or linearly dependent model. This confirms the robustness of the unfolding against model assumptions.
Cumulant Agreement: In the simulated scenario (which lacks flow fluctuations), the 2-particle ( $k=1$ ) and 4-particle ( $k=2$ ) results agree, reflecting the underlying model nature.
Harmonic Sensitivity: The method successfully captures anisotropies for harmonic orders $n=2, 3,$ and $4$, with the results reflecting the input sensitivity to flow harmonic order attenuation.

Significance and Claims
The paper claims that multi-particle cumulant methods are a robust tool for reconstructing jet azimuthal anisotropies. The primary significance lies in the ability of these methods to:

Provide a consistent technique for analyzing high- $p_T$ jet anisotropies, potentially resolving discrepancies seen with event plane methods.
Offer a pathway to disentangle long-range and short-range correlations between jets, which is crucial for understanding jet-medium interactions.
Enable the study of jet interactions and anisotropies in smaller collision systems (e.g., $p$ +A, $d$ +A) where traditional event plane reconstruction is not possible.

The authors note that while the current simulation assumes a static background without event-by-event flow fluctuations, future applications to real data must account for effects such as uncorrelated jets, resonance decays, and significant flow fluctuations in small systems. The methodology presented serves as a foundational step toward applying these techniques to experimental data to better understand the mechanisms driving high- $p_T$ hadron anisotropies.