Revisiting the role of saturation in diffractive vector meson production

This paper presents a global Bayesian analysis using a Color Glass Condensate framework and Gaussian-process emulators to demonstrate that correcting for electromagnetic dissociation effects in LHC data resolves previous tensions between proton and nuclear datasets, enabling a consistent simultaneous description of coherent and incoherent diffractive J/ψ photoproduction in both γ+p and γ+Pb collisions.

The Gist

Imagine the inside of a proton (a tiny particle in an atom) as a bustling city filled with invisible messengers called gluons. These gluons carry the force that holds the proton together. When you zoom in really close, especially when the gluons are moving very slowly, they start to crowd together so densely that they begin to overlap and interact in complex ways. Physicists call this crowded, saturated state the "Color Glass Condensate" (CGC). It's like a traffic jam where the cars (gluons) are so packed they can't move freely anymore.

To understand this traffic jam, scientists smash particles together at huge speeds, like in the Large Hadron Collider (LHC). They look at a specific event called diffractive vector meson production. Think of this as shining a high-energy flashlight (a photon) at a proton or a heavy lead nucleus and watching how a specific type of particle (a J/ψ meson) bounces off. The way it bounces tells us about the density and arrangement of the gluon traffic jam inside.

The Problem: A Mismatch in the Data

For a while, physicists had a puzzle. When they used their best mathematical models (the CGC framework) to predict how these particles would bounce off a single proton, the predictions matched the data perfectly. However, when they tried to use the same model to predict what would happen with a heavy lead nucleus (which is like a giant city of protons), the model failed.

The model predicted that the lead nucleus would behave in a certain way at high energies, but the actual experiments showed something different. It was as if the model was saying, "The traffic jam should be this heavy," but the experiments were saying, "No, it's actually lighter." This created a "tension" or a disagreement between the proton data and the lead data. To make the numbers match, scientists had to artificially shrink their predictions by a factor (called a "K factor"), which felt like cheating to fix a broken model.

The Solution: Cleaning Up the Mess

The authors of this paper realized there might be a hidden variable they hadn't accounted for: Electromagnetic Dissociation (EMD).

Here is a simple analogy: Imagine you are trying to count how many people enter a building through a specific door. But, every time someone enters, the wind (electromagnetic forces) sometimes blows a few extra people in through a side window, or knocks some people out the back. If you don't account for this wind, your count of people entering through the main door will be wrong.

In the LHC experiments, the "wind" (EMD) was causing some events to be miscounted or missed entirely. The experimental data they had been using was slightly "dirty" because it didn't fully correct for this effect.

The Discovery

The researchers took the latest experimental data and applied a "cleaning filter" to correct for this electromagnetic dissociation. They then ran their global analysis again, comparing the proton and lead data side-by-side.

The result was a breakthrough:

1. The Tension Disappeared: Once the data was corrected for the "wind," the proton and lead data finally agreed with each other. The model no longer needed to be "cheated" with an artificial shrinking factor.
2. One Model Fits All: The same set of rules (the CGC framework) that worked for protons now worked perfectly for lead nuclei without any extra adjustments.
3. Better Understanding of Gluons: The corrected data showed that the "traffic jam" of gluons in the lead nucleus behaves exactly as the theory predicted, just without the noise of the experimental errors.

Why This Matters

This paper doesn't invent a new theory or build a new machine. Instead, it acts like a detective who realizes the crime scene photos were slightly blurry. By sharpening the photos (correcting the data), the detective found that the suspect (the CGC theory) was innocent all along.

The authors conclude that the "Color Glass Condensate" theory is a consistent and accurate way to describe how gluons behave in both small protons and large lead nuclei, provided we look at the experimental data with the right corrections. It resolves a long-standing disagreement in the physics community and gives us a clearer picture of the fundamental building blocks of matter.

Technical Summary

Technical Summary: Revisiting the Role of Saturation in Diffractive Vector Meson Production

Problem Statement
Understanding the behavior of gluons at small parton momentum fractions ($x$) is essential for characterizing hadronic and nuclear properties. In this regime, the rapid growth of gluon density is expected to be moderated by non-linear recombination effects, leading to a saturated state described by the Color Glass Condensate (CGC) effective theory. While CGC-based calculations have successfully described coherent and incoherent diffractive $J/\psi$ photoproduction in $\gamma + p$ collisions at HERA and the LHC, a persistent tension has been observed when applying the same framework to $\gamma + \text{Pb}$ collisions. Previous global Bayesian analyses indicated that proton and nuclear datasets could not be simultaneously described without introducing an overall normalization factor ($K \ll 1$) and adjusting evolution parameters to artificially enhance non-linear dynamics. This discrepancy suggested a fundamental inconsistency between the CGC predictions for protons and nuclei or a systematic issue in the experimental data interpretation.

Methodology
The authors perform a new global Bayesian analysis of coherent and incoherent diffractive $J/\psi$ photoproduction in $\gamma + p$ and $\gamma + \text{Pb}$ collisions. The study utilizes a CGC-based framework where cross sections are derived from the average and variance of the diffractive scattering amplitude over target configurations. Key theoretical components include:

• Wave Functions: A Boosted Gaussian parametrization for the $J/\psi$ wave function.
• Dipole Amplitude: Constructed from Wilson lines using the McLerran–Venugopalan model at $x_0 = 0.01$, evolved to smaller $x$ via event-by-event JIMWLK evolution.
• Geometry: Proton shape fluctuations are modeled using a constituent hot-spot picture, while nuclear configurations utilize Woods–Saxon sampled nucleon positions.
• Statistical Framework: A multivariate Gaussian likelihood is employed, replacing full CGC calculations with Gaussian-process emulators trained on a large set of simulations (reusing validated emulators from previous work).
• Data Correction: The primary methodological update involves replacing the original ultraperipheral collision data with measurements corrected for the expected effect of electromagnetic dissociation (EMD). This correction addresses the systematic underestimation of extracted $\gamma + \text{Pb}$ cross sections caused by exclusive events failing experimental exclusivity requirements.

The analysis constrains a parameter set characterizing proton geometry, fluctuations, color charge density, and infrared regulators. The study compares fits with and without an overall normalization factor $K$ and performs a Bayesian model comparison using the Bayes factor to assess the necessity of additional model freedom.

Key Contributions and Results

1. Resolution of Tension: The inclusion of EMD-corrected $\gamma + \text{Pb}$ data substantially reduces the previously observed tension between proton and nuclear datasets. The corrected data allow for a consistent simultaneous description of diffractive $J/\psi$ production in both $\gamma + p$ and $\gamma + \text{Pb}$ collisions within the standard CGC framework.
2. Elimination of Ad-Hoc Normalization: Unlike previous analyses that required a normalization factor $K \sim 0.3$ to reconcile the datasets, the new analysis with EMD-corrected data prefers $K \approx 1$. The posterior distributions for the evolution parameters ($m_{\text{JIMWLK}}$ and $\Lambda_{\text{QCD}}$) become well-constrained and consistent between proton and nuclear fits, removing the multi-modality and boundary preferences seen in earlier work.
3. Bayesian Model Comparison: A Bayes factor analysis ($\ln B_{A/B} = 2.57 \pm 0.01$) provides moderate evidence in favor of the baseline model (fixed $K=1$) over the extended model (free $K$). This indicates that the improved fit quality does not justify the added complexity of an additional normalization parameter once the EMD correction is applied.
4. Parameter Constraints: Separate fits reveal that while $\gamma + p$ data tightly constrain geometric parameters, $\gamma + \text{Pb}$ data alone cannot fully constrain all features of the proton structure. However, the EMD correction shifts the preferred JIMWLK evolution parameters in the nuclear fit toward values compatible with the proton fit, suggesting the previous discrepancy originated from the treatment of the nuclear data rather than a fundamental flaw in the CGC framework.

Significance
The paper demonstrates that electromagnetic-dissociation veto corrections in ultraperipheral collisions can substantially impact the extraction of small-$x$ nuclear dynamics. By applying these corrections, the authors show that the CGC framework can consistently describe diffractive vector meson production off both protons and heavy nuclei without invoking artificial suppression factors. The results suggest that the previously observed tension between proton and nuclear measurements was largely an artifact of uncorrected experimental data rather than a failure of the saturation physics model. The authors note that while the EMD corrections are theory-driven and not yet official releases from experimental collaborations, their inclusion resolves the inconsistencies, supporting the validity of gluon saturation effects predicted by the CGC.