⚛️ phenomenology

Investigating the transverse-momentum- and pseudorapidity-dependent flow vector decorrelation in p--Pb collisions with a Multi-Phase Transport model

This study utilizes the Multi-Phase Transport (AMPT) model to systematically investigate transverse-momentum and pseudorapidity-dependent flow vector decorrelations in p--Pb collisions at 5.02 TeV, demonstrating that the string-melting version successfully describes experimental data while highlighting the critical roles of initial conditions and partonic interactions, and the necessity of subtracting nonflow effects for accurate analysis.

Original authors: Siyu Tang, Zuman Zhang, Chao Zhang, Liang Zheng, Renzhuo Wan

Published 2026-01-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Siyu Tang, Zuman Zhang, Chao Zhang, Liang Zheng, Renzhuo Wan

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 the Large Hadron Collider (LHC) as a massive, high-speed billiard table. Usually, physicists smash heavy balls (lead nuclei) together to create a super-hot, super-dense soup of particles called "quark-gluon plasma." This soup behaves like a perfect, frictionless fluid.

But recently, scientists started smashing smaller balls together—like a proton (a tiny particle) hitting a lead nucleus. They were surprised to find that even in these tiny collisions, this "perfect fluid" behavior still appeared. This paper investigates a specific, weird quirk of that fluid: decorrelation.

Here is a simple breakdown of what the authors did and found, using everyday analogies.

The Big Idea: The "Wave" That Gets Out of Sync

When the fluid forms, it doesn't just sit there; it ripples. Imagine throwing a stone into a calm pond. You get perfect, circular waves moving outward. In a perfect world, if you look at the wave at the edge of the pond and the wave in the middle, they should move in perfect harmony.

However, in these tiny collisions, the "ripples" (called flow vectors) get messy.

The Problem: The direction and strength of the wave change depending on where you look (how far forward or backward along the collision path) and how fast the particles are moving.
The Term: This messiness is called decorrelation. It's like a choir where everyone starts singing the same song, but as you move from the front row to the back row, or from the slow singers to the fast singers, everyone starts singing slightly different notes or at different times. They lose their synchronization.

The Experiment: A Virtual Laboratory

The authors didn't just look at real data; they built a sophisticated computer simulation called the AMPT model. Think of this model as a video game engine that simulates the entire life cycle of a collision, from the split-second impact to the final spray of particles.

They ran the simulation with different "settings" to see what causes the choir to lose its tune:

The Starting Line (Initial Conditions): How the collision begins. Did the proton hit the lead nucleus dead-center or off to the side?
The Parton Phase: The moment the particles break apart into their smallest components (quarks and gluons) and bounce off each other like billiard balls.
The Hadron Phase: The moment those components recombine into larger particles and bounce around again before flying off.
The "Noise" (Non-flow): Sometimes particles are linked not because of the fluid, but because they came from the same jet (like two shrapnel pieces flying from a single explosion). This is "noise" that can fake a signal.

What They Discovered

1. The "String Melting" Works
The version of their simulation that treats the initial collision as "melting strings" of energy matched the real-world data from the LHC best. It successfully recreated the "out-of-sync" waves.

2. The Final Bounce Doesn't Matter Much
They found that the final stage, where the big particles bounce off each other (hadronic scatterings), has almost no effect on the decorrelation.

Analogy: Imagine a chaotic mosh pit. The authors found that whether the people in the pit bump into each other or not doesn't change the fact that the music (the flow) was already out of sync before they even started dancing. The "messiness" was baked in from the start.

3. The Start and the Middle Are Key
The main reasons the waves get out of sync are:

How the collision starts: The initial unevenness of the proton and lead nucleus.
The Parton Phase: How the tiny particles bounce around in the middle.
Analogy: If you want to fix the choir's timing, you need to fix how they start singing and how they interact in the middle of the song. Changing what happens at the very end won't help.

4. The "Jet" Noise is a Big Deal (Especially at the Edges)
This is a crucial finding. In the middle of the collision, the "noise" from jets (shrapnel) doesn't mess up the measurement much. But at the edges (forward and backward directions), this noise is huge.

Analogy: Imagine trying to hear a whisper in a quiet room (the middle of the collision). It's easy. But if you stand near a loudspeaker blasting music (the edges), the whisper gets drowned out. The authors found that in small collisions, the "loudspeaker" (long-range jet correlations) is actually responsible for making the waves look even more out of sync than they really are. If you don't subtract this noise, you get a distorted picture.

The Conclusion

This paper tells us that the "out-of-sync" behavior in tiny proton-lead collisions is real and is driven by the initial chaos of the crash and the interactions of the tiny particles inside.

However, it also warns us: Don't trust the edges. In these small systems, the "noise" from jets plays a massive role in making the data look messy. To understand the true nature of this tiny fluid, scientists must carefully filter out that noise, especially when looking at the forward and backward directions.

In short: The fluid is real, the messiness is real, but you have to be very careful not to confuse the messiness of the fluid with the noise of the shrapnel.

Technical Summary: Probing Transverse-Momentum- and Pseudorapidity-Dependent Flow Vector Decorrelation in p–Pb Collisions

Problem Statement
High-energy heavy-ion collisions generate a quark-gluon plasma (QGP) characterized by collective flow, typically quantified by Fourier coefficients ( $v_n$ ) of the azimuthal particle distribution. Recent observations indicate that event-by-event fluctuations in the initial energy density lead to the decorrelation of flow vectors as a function of transverse momentum ( $p_T$ ) and pseudorapidity ( $\eta$ ). While these phenomena are well-documented in large collision systems (Pb–Pb), their systematic theoretical understanding in small collision systems (p–Pb and pp) remains incomplete. Specifically, the relative contributions of initial conditions, partonic interactions, hadronic rescatterings, and nonflow correlations (such as jets) to $p_T$ - and $\eta$ -dependent flow vector decorrelations in p–Pb collisions require further investigation.

Methodology
This study employs the A Multi-Phase Transport (AMPT) model, specifically the string-melting version, to simulate p–Pb collisions at $\sqrt{s_{NN}} = 5.02$ TeV. The model encompasses four stages: initial conditions (generated via HIJING with Lund fragmentation), partonic scattering (via the Zhang's Parton Cascade, ZPC), hadronization (via quark coalescence), and hadronic rescattering (via the Relativistic Transport, ART, model).

To disentangle the physical mechanisms driving decorrelation, the authors systematically varied key model parameters across six configurations (Par1–Par6):

Initial Conditions: Varied Lund string parameters ( $a$ and $b$ ) to alter string tension.
Partonic Interactions: Varied the partonic scattering cross-section ( $\sigma$ ) from 0 mb (no partonic scattering) to 0.5 mb and 3 mb.
Hadronic Effects: Adjusted the cutoff time ( $t_{max}$ ) for hadronic rescattering (30 fm/c, 0.6 fm/c, and 0 fm/c) to isolate the impact of hadronic scatterings and resonance decays.

The study utilized two-particle correlation methods to define observables:

$p_T$ -dependent decorrelation: Analyzed via the ratio $v_2\{2\}(p_T) / v_2[2](p_T)$ and the factorization ratio $r_2(p_T^a, p_T^b)$ .
$\eta$ -dependent decorrelation: Analyzed via the product of factorization ratios $\sqrt{r_2(\eta_a, \eta_b)r_2(-\eta_a, -\eta_b)}$ , a method adapted for asymmetric systems like p–Pb to isolate symmetry-plane fluctuations.

A template fit method, developed by the ATLAS collaboration, was applied to subtract nonflow contributions (specifically long-range jet correlations) from the correlation functions.

Key Results

$p_T$ -Dependent Decorrelation: The AMPT model with default parameters (Par1) reasonably reproduces ALICE measurements of $v_2\{2\}$ $v_{2} {2}$ and the ratio $v_2\{2\}/v_2[2]$ $v_{2} {2} / v_{2} [2]$ . The ratio deviates from unity at $p_T \gtrsim 2$ $p_{T} ≳ 2$ GeV/c, indicating $p_T$ $p_{T}$ -dependent flow vector fluctuations.
- Nonflow: After subtracting long-range jet correlations, the ratio remains consistent with un-subtracted results, suggesting that $p_T$ -dependent decorrelation in the mid-rapidity region is driven primarily by real flow signals rather than nonflow.
- Physics Drivers: The magnitude of decorrelation is sensitive to initial conditions (Lund parameters) and partonic scattering cross-sections. Hadronic scatterings were found to have a negligible impact on $p_T$ -dependent decorrelation.
$\eta$ -Dependent Decorrelation: The model describes CMS data for the $\eta$ $η$ -dependent factorization ratio, showing a linear decrease (decorrelation) as $|\eta|$ $∣ η ∣$ increases.
- Nonflow: In contrast to the $p_T$ case, nonflow subtraction significantly enhances the observed decorrelation. The contribution from long-range jet correlations is substantial in the $\eta$ -dependent channel, particularly in peripheral collisions (40–60%), where the trend of decorrelation can even reverse after subtraction.
- Physics Drivers: Similar to the $p_T$ case, initial conditions and partonic interactions influence the magnitude of $\eta$ -decorrelation, while hadronic scatterings show weak sensitivity.
Model Limitations: While the model captures the general trends, it systematically underestimates the data in certain kinematic regions (e.g., forward/backward rapidity) and requires further tuning of initial conditions and parton scattering processes to fully match experimental data across all centralities and rapidities.

Significance and Claims
The paper claims to present the first systematic transport model study of $p_T$ - and $\eta$ -dependent flow vector decorrelations in small collision systems. The primary findings are:

Mechanism Identification: The decorrelation of the flow vector in p–Pb collisions is driven by event-by-event fluctuations in the initial state, which are propagated to the final state through partonic interactions. Hadronic rescatterings play a minimal role in these specific decorrelation observables.
Role of Nonflow: The study highlights a critical distinction between $p_T$ and $\eta$ dependencies: while $p_T$ -dependent decorrelation at mid-rapidity is robust against nonflow, longitudinal ( $\eta$ ) decorrelation is heavily influenced by long-range jet correlations (nonflow).
Future Utility: The results suggest that flow vector decorrelation can serve as a probe for the transport properties of the hot, dense medium formed in small collision systems. The work provides a referential baseline for future experimental measurements and theoretical developments, emphasizing the need to consider nonflow contributions when analyzing longitudinal fluctuations in small systems.

The Big Idea: The "Wave" That Gets Out of Sync

The Experiment: A Virtual Laboratory

What They Discovered

The Conclusion

More like this