Jet wake from linearized hydrodynamics

Abstract

We explore how to improve the hybrid model description of the particles originating from the wake that a jet produced in a heavy ion collision leaves in the droplet of quark-gluon plasma (QGP) through which it propagates, using linearized hydrodynamics on a background Bjorken flow. Jet energy and momentum loss described by the hybrid model become currents sourcing linearized hydrodynamics. By solving the linearized hydrodynamic equations numerically, we investigate the development of the wake in the dynamically evolving droplet of QGP, study the effect of viscosity, scrutinize energy-momentum conservation, and check the validity of the linear approximation. We find that linearized hydrodynamics works better in the viscous case because diffusive modes damp the energy-momentum perturbation produced by the jet. We calculate the distribution of particles produced from the jet wake by using the Cooper-Frye prescription and find that both the transverse momentum spectrum and the distribution of particles in azimuthal angle are similar in shape in linearized hydrodynamics and in the hybrid model. Their normalizations are different because the momentum-rapidity distribution in the linearized hydrodynamics analysis is more spread out, due to sound modes. Since the Bjorken flow has no transverse expansion, we explore the effect of transverse flow by using local boosts to add it into the Cooper-Frye formula. After including the effects of transverse flow in this way, the transverse momentum spectrum becomes harder: more particles with transverse momenta bigger than 2 GeV are produced than in the hybrid model. Although we defer implementing this analysis in a jet Monte Carlo, as would be needed to make quantitative comparisons to data, we gain a qualitative sense of how the jet wake may modify jet observables by computing proxies for two example observables: the lost energy recovered in a cone of varying open angle, and the fragmentation function. We find that linearized hydrodynamics with transverse flow effects added improves the description of the jet wake in the hybrid model in just the way that comparison to data indicates is needed. Our study illuminates a path to improving the description of the wake in the hybrid model, highlighting the need to take into account the effects of both transverse flow and the broadening of the energy-momentum perturbation in spacetime rapidity on particle production.

Jet tomography in hot QCD medium with deep learning

“…In this proceedings, we review the power of our novel deep learning techniques to locate with precision the jet creation point in the transverse plane, by selecting jets according to jet energy loss 𝜒, width 𝑅 𝑔 and orientation, which constitutes a significant step towards the exploitation of tomographic power of energetic jets. In future work, the tomographic power is expected to be further improved by considering the interplay between the jet and the local properties of the medium, e.g., the local hydrodynamic flow [23,24,[37][38][39] or spatial-temporal gradients [21,22,24], which determine preferred directions and deformed radiation spectra for the soft emissions from the jet. A direct extraction of the traversed length 𝐿 of energetic jets in the QGP is highly desirable to push forward this series of studies.…”

Deep Learning for the classification of quenched jets

“…The ability to do so will result in a subsample of jets produced in PbPb that only includes jets that have experienced significant modification making the assessment of modifications, at observable level, clearer. A recent study [16] showed that a convolutional neural network (CNN) trained on jet images for jets modified using the strong/weak Hybrid model [17] including effects of medium response [18] allowed for the extraction, on a jetby-jet basis, of the p T the jet would have had if no QGP was present. These important results rely, at least in significant part, on the presence of a medium response component which is the leading feature identified by the CNN as signalling the strength of quenching effects.…”

“…These important results rely, at least in significant part, on the presence of a medium response component which is the leading feature identified by the CNN as signalling the strength of quenching effects. However, the medium response remains the most model-dependent component of state-of-the-art [18][19][20] jet quenching simulations and is not inconceivable that features highlighted by the CNN may be model-specific and absent in real data.…”

Disentangling Jet Modification in Jet Simulations and in Z+Jet Data

“…Jets that develop while immersed in a hot background medium have modified structure relative to those formed in vacuum, for example due to medium-induced emission [13][14][15][16] or drag [17][18][19][20][21][22]. In addition, partons in the jet excite a "wake" in the medium they pass through [23][24][25][26][27][28][29]. This wake carries momentum in the jet direction "lost" by the jet partons and after hadronization therefore yields lower-p T particles from the medium that are correlated with the jet direction.…”