Bayesian Inference analysis of jet quenching using inclusive jet and hadron suppression measurements

The JETSCAPE Collaboration reports a new determination of the jet transport parameter $\hat{q}$ in the Quark-Gluon Plasma (QGP) using Bayesian Inference, incorporating all available inclusive hadron and jet yield suppression data measured in heavy-ion collisions at RHIC and the LHC. This multi-observable analysis extends the previously published JETSCAPE Bayesian Inference determination of $\hat{q}$, which was based solely on a selection of inclusive hadron suppression data. JETSCAPE is a modular framework incorporating detailed dynamical models of QGP formation and evolution, and jet propagation and interaction in the QGP. Virtuality-dependent partonic energy loss in the QGP is modeled as a thermalized weakly-coupled plasma, with parameters determined from Bayesian calibration using soft-sector observables. This Bayesian calibration of $\hat{q}$ utilizes Active Learning, a machine–learning approach, for efficient exploitation of computing resources. The experimental data included in this analysis span a broad range in collision energy and centrality, and in transverse momentum. In order to explore the systematic dependence of the extracted parameter posterior distributions, several different calibrations are reported, based on combined jet and hadron data; on jet or hadron data separately; and on restricted kinematic or centrality ranges of the jet and hadron data. Tension is observed in comparison of these variations, providing new insights into the physics of jet transport in the QGP and its theoretical formulation.

💡 Research Summary

The JETSCAPE Collaboration presents a comprehensive Bayesian inference study of the jet transport coefficient (\hat{q}) in the quark‑gluon plasma (QGP), incorporating all available inclusive hadron and jet nuclear‑modification factor ((R_{AA})) measurements from RHIC and the LHC. Building on a previously published analysis that used only a subset of hadron data, this work expands the observable set to include reconstructed jet suppression, thereby providing a multi‑observable calibration of (\hat{q}/T^{3}).

The theoretical framework is the modular JETSCAPE event generator. Bulk evolution is modeled with Trento initial conditions, a short free‑streaming stage, followed by 2+1‑D viscous hydrodynamics (VISHNU) and Cooper‑Frye particlization, with UrQMD handling the hadronic afterburner. Hard processes are generated with Pythia8 (PP19 tune) for the initial‑state radiation and hard scattering; vacuum final‑state radiation is simulated with the MATTER module. In‑medium jet evolution combines the virtuality‑dependent MATTER model with the Linear Boltzmann Transport (LBT) model, allowing a temperature‑dependent transport coefficient (\hat{q}) to be extracted. The definition of (\hat{q}) follows the standard mean transverse‑momentum‑squared per unit path length, with its field‑theoretic expression derived from Hard Thermal Loop (HTL) effective theory.

For the statistical analysis, the authors employ Bayesian inference with priors taken from a previous soft‑sector calibration (shear and bulk viscosities, initial‑condition parameters). The likelihood compares simulated (R_{AA}) for both jets and hadrons to experimental data across a wide range of centralities and transverse momenta. Because direct evaluation of the full model is computationally prohibitive, an active‑learning workflow is used: a Gaussian‑process emulator is trained on an initial design, and the emulator’s uncertainty guides the selection of new simulation points, dramatically reducing the number of required full JETSCAPE runs (of order 10⁴).

Four calibration scenarios are explored: (i) combined jet + hadron data, (ii) jet data only, (iii) hadron data only, and (iv) restricted kinematic or centrality subsets. The posterior distributions for (\hat{q}/T^{3}) are broadly consistent with a scaling proportional to (T^{3}), yielding typical values of 2–4 GeV²/fm at a reference temperature of 0.4 GeV. However, systematic differences emerge. Jet‑only calibrations prefer a larger (\hat{q}) at high (p_{T}) (> 30 GeV) than hadron‑only fits, reflecting the sensitivity of reconstructed jets to wide‑angle radiation and energy flow outside the jet cone. Centrality dependence shows a clear trend: the most central (0–10 %) collisions require the highest (\hat{q}), while peripheral (30–50 %) collisions yield lower values, mirroring the expected temperature and density gradients in the medium.

Importantly, the study identifies tensions between the posterior distributions obtained from different observable sets and kinematic windows. These tensions suggest that the current implementation of virtuality‑dependent energy loss (MATTER + LBT) does not fully capture all aspects of jet quenching, especially the interplay between collinear and large‑angle emissions. Model uncertainties—such as the MATTER–LBT matching scale, the free‑streaming duration, and the treatment of the medium’s correlation length—contribute to the observed discrepancies, as do experimental systematic uncertainties (centrality determination, pp reference).

The authors discuss pathways to resolve these issues: incorporating photon‑jet and Z‑jet correlations, measuring intra‑jet substructure (splitting functions, radial energy profiles), and performing Bayesian model averaging across different energy‑loss formalisms (BDMPS, GLV, Higher‑Twist). They also advocate for extending the active‑learning framework to include additional theory parameters, thereby enabling a more exhaustive exploration of the QGP’s transport properties.

In summary, this work demonstrates that a Bayesian, multi‑observable approach combined with machine‑learning‑driven emulator techniques can provide a robust, quantitatively precise extraction of the jet transport coefficient (\hat{q}). By explicitly quantifying the tensions between jet and hadron data, the study highlights where current theoretical models fall short and outlines a clear roadmap for future experimental and theoretical efforts to achieve a universal, observable‑independent determination of QGP jet transport properties.