betaFeedback

WorksPhysics and Astronomy

Towards an unbiased jet energy loss measurement

Abstract

The modifications imprinted on jets due to their interaction with Quark Gluon Plasma (QGP) are assessed by comparing samples of jets produced in nucleus-nucleus collisions and proton-proton collisions. The standard procedure ignores the effect of bin migration by comparing specific observables for jet populations at the same reconstructed jet transverse momentum (pTp_T). Since jet pTp_T is itself modified by interaction with QGP, all such comparisons confound QGP induced modifications with changes that are simply a consequence of comparing jets that started out differently. The quantile matching procedure introduced by Brewer et al. directly estimates average fractional jet energy loss (QAAQ_{AA}) and can thus mitigate this pTp_T migration effect. In this work, we validate the procedure in more realistic scenarios that include medium response. We study the evolution of QAAQ_{AA} with jet radius, its sensitivity to minimum particle pTp_T and medium response as implemented in two different models for jet evolution in heavy-ion collisions. Further, we use this procedure to establish that the difference between inclusive jet and γ+γ+jet nuclear modification factors (RAAR_{AA}) is dominated by differences in the spectral shape, leaving the colour charge of the jet initiating parton with a lesser role to play. Additionally, we compare QAAQ_{AA} to an experimentally proposed proxy for fractional jet energy loss, SlossS_{loss}, showing that both quantities are similar, although the former provides a more clear physical interpretation. Finally, we show the size of the pTp_T migration correction for four different substructure observables and how to reliably use the quantile procedure experimentally to improve existing measurements.

Prepared for submission to JHEP

Introduction

From the structure of Quantum Chromodynamics (QCD), the fundamental theory of strong interactions, a particularly interesting system emerges: the quark-gluon plasma (QGP), a hot and dense state of matter where quarks and gluons are not confined inside hadrons. This system is believed to have been produced in the early epochs of the Universe and, experimentally, it has been recreated in the aftermath of ultrarelativistic heavy ion collisions at the Relativistic Heavy Ion Collider (RHIC) [1, 2, 3] and the Large Hadron Collider (LHC) [4, 5, 6]. These experimental programmes have allowed the exploration of the many-body properties of QCD in such extreme regimes, and led to the remarkable discovery that the QGP is a nearly ideal liquid. About 1 fm/c1\ \mathrm{fm}/c (∼3 ys\sim3\ \mathrm{ys}) after a heavy ion collision, QGP production is signaled by an abrupt increase in both particle and energy densities, with energy densities well above those expected inside typical hadrons. After its production, the QGP acquires hydrodynamical behaviour and expands, losing its collective properties on a timescale of roughly 10 fm/c10\ \mathrm{fm}/c (see e.g. [7] for a comprehensive review of heavy-ion collisions).

To be able to quantitatively describe the properties of the QGP and their time evolution, a probe which is produced concurrently is needed. For this it is common to use highly energetic probes [8] produced in the rare high transverse momentum scatterings happening between the quark and gluon constituents of the nucleons in the nucleus. The most powerful ones are hadronic jets, the result of the fragmentation of highly boosted quarks and gluons. Jets are extended objects in space and time, since their fragmentation evolution develops through multiple energy scales from the initial hard scattering (≳0.1\gtrsim0.1 TeV) down to soft non-perturbative scales of hadronization (≲0.1\lesssim0.1 GeV). For this reason, they are sensitive to the different stages of the QGP’s evolution. The vacuum description of jet evolution, as it happens in proton-proton (pp) collisions, is well established perturbatively through precise resummations of large logarithms in the form of parton showers, but jets in nuclei collisions are different because they interact with the produced background medium in a non-trivial way, stimulating processes which compete with the vacuum-like development of the ensuing cascade [9, 10]. The imprints of these interactions are carried through to the final particle distribution inside jets, with part of the energy thermalizing to the medium. This implies that jets reconstructed in nucleus-nucleus (AA) collisions are quenched versions of pp jets plus additional correlated background stemming from the medium’s response to the jet’s passage. The study of these medium induced effects is referred to as jet quenching [8, 11, 12, 13, 14] and it provides the framework through which one can compare jets produced in pp collisions with jets in AA collisions with the final aim of inferring the properties of the QGP.

Despite the simplicity of this reasoning, fully describing heavy-ion jets is not an easy task since it requires formulating a parton shower in the medium which encapsulates all the relevant medium-induced effects while providing an accurate evolution model for the background medium. Not only this, but before following through with this reasoning, a simple yet very relevant question should first be answered: how exactly should one compare samples of vacuum and medium jets? The standard procedure to draw this comparison consists on selecting a window of reconstructed jet transverse momentum (pTp_{\mathrm{T}}) and calculating the value of a given observable for both populations of jets. However, if our aim is to quantify jet modifications with respect to a vacuum baseline and if those modifications depend either explicitly or implicitly on pTp_{\mathrm{T}}, this approach is biased. The key problem is that it ignores that medium jets migrate to lower pTp_{\mathrm{T}}, i.e., jets lose energy as a result of interaction with QGP. Thus, comparisons drawn with this procedure confound QGP induced modifications with changes that are simply a consequence of comparing jets that were originated by hard partons with different pTp_{\mathrm{T}}. To perform an unbiased comparison of jet populations, one thus needs to have an estimate of jet energy loss.

In order to quantify how much energy jets in AA collisions lose with respect to their pp counterparts, it is useful to understand the physical mechanisms contributing to this effect. Interactions between the jet and QGP can generally be understood as momentum exchanges of the jet cascade with the medium. Among other effects, these exchanges result in elastic energy loss [15], part of which thermalizes to the medium. The deposited energy, being correlated with the jet’s direction of propagation, can end up being reconstructed along with the final state jet constituents. One usually refers to this contribution to a heavy-ion jet as medium response. This way, depending on the jet reconstruction radius, part of the lost energy in elastic processes is recovered. The accumulation of successive scatterings of the partons in the evolving shower with the QGP can resolve fluctuations in their wavefunction into real radiation. This medium-induced radiation has been tracted analytically within several perturbative schemes [16, 17], [18, 19, 20, 21, 22], [23, 24] and [25, 26]. Regardless of the treatment, successive emissions stimulated by the medium are typically not angular ordered like what is predicted in a vacuum cascade [27]. Hence, part of these emissions have large angles with respect to the jet, falling outside of the reconstruction radius and significantly changing the jet energy distribution.

We emphasize that in this work we study jet energy loss and not parton energy loss and that this quantity has an important interplay with the jet radius [28]. This means that the energy lost by jets reconstructed in AA collisions is with respect to the baseline of jets reconstructed in pp collisions, the energy of which is already a fraction of the initiating parton’s energy due to part of the vacuum shower falling outside of the jet reconstruction radius [29]. Recent works [30, 31, 32] have tried to quantify jet energy loss as well as correlate it with path length and partonic flavour in a model-independent way, by parameterizing an energy loss distribution and extracting the relevant parameters through Bayesian inference.

Experimentally, the nuclear modification factor RAAR_{AA}, which is useful to quantify jet suppression, is often also used to qualitatively describe jet energy loss. It cannot, however, directly quantify jet energy loss, given that it takes the ratio of AA over pp cross sections at the same reconstructed pTp_T, ignoring pTp_T migration. This bias is exacerbated by the steeply falling nature of the spectrum, to which the RAAR_{AA} is sensitive. In fact, if we consider two jet spectra with significantly different steepnesses, e.g. inclusive jet and boson+jet samples, and impose a fixed energy loss for all jets, then the less steeper spectrum will present a smaller suppression, i.e., RAAR_{AA} closer to 1. Hence, for the same energy loss one obtains different RAAR_{AA} because the shape of the spectrum was varied. Another problem caused by pTp_T migration is selection bias (see e.g., [33]), whereby samples of in-medium jets selected in a given pTp_T window are biased towards those jets that were less modified due to the spectrum steepness. Both these problems can be partially evaded in boson+jet events [34], because there we have a reference for the pTp_T of the hard parton that originated the jet. This way, one can compare pp and AA jets that started out with the same pTp_T. However, the size of the statistical sample of boson+jet events is significantly smaller than that of inclusive jet events, which one would like to use1. We therefore need another solution.

This paper is organized as follows. In § 2 we introduce the quantile procedure as it was defined in the original paper [36], further motivating it as a way to calculate an energy loss proxy and clarifying the ideal scenarios where it works exactly. We also relate it to other existing energy loss proxies. Then, in § 3, we detail the physics implemented on the generators used to obtained the event samples analysed in this work, as well as the the relevant parameters for both the event generation setup and the analysis underlying jet reconstruction. In § 4, the quantile procedure is validated in a more realistic scenario than in [36], i.e., including medium response and for multiple jet radii. The following results in § 5 consist on studying, for two different event generators, how the energy loss proxy obtained via quantile matching varies with jet radius and with minimum particle pTp_T cut. Additionally, a study of the colour charge dependence of energy loss is carried out, focusing on disentangling the spectrum steepness influence (which is large in RAAR_{AA}) from the actual energy lost by jets. This is done for γ\gamma+jet, γ+q\gamma+q, γ+g\gamma+g and inclusive jet samples. In § 6, an obvious obstacle to the experimental implementation of the quantile procedure, namely the pTp_T cutoff of the spectrum, is addressed and solved. A comparison with another energy loss proxy used in [37] is made. § 7 shows the degree to which some jet substructure observables are changed when applying the quantile procedure to reduce the pTp_T migration bias. Conclusions are presented in § 8.

Quantile matching

Quantile matching as a means to estimate the average energy lost by jets through interaction with a QGP was introduced in [36]. The probability of a jet having transverse momentum within a given (pT,1,pT,2)(p_{T,1}, p_{T,2}) interval is given by the cumulative jet cross section

Σ(pT,1,pT,2)=∫pT,1pT,2dpTdσdpT.(1)\Sigma(p_{T,1},p_{T,2}) = \int_{p_{T,1}}^{p_{T,2}} dp_T \frac{d\sigma}{dp_T}. \tag*{(1)}

The matching procedure aims at establishing a correspondence between a given transverse momentum (pTv)(p_T^v) in the pp jet spectrum and a transverse momentum (pTq)(p_T^q) in the AA jet spectrum by equating appropriately normalized upper cumulative cross sections

Σpp(pTv,+∞)=ΣAA(pTq(pTv),+∞),(2)\Sigma^{pp}(p_T^v,+\infty) = \Sigma^{AA}(p_T^q(p_T^v),+\infty), \tag*{(2)}

i.e., by calculating a quantile function pTqp_T^q which depends on the value of pTvp_T^v. This translates to finding out pTqp_T^q for a given pTvp_T^v such that AA jets with pT>pTq(pTv)p_T > p_T^q(p_T^v) are produced with the same probability as pp jets with pT>pTvp_T > p_T^v. As argued in the original paper [36], in a scenario where a jet with a given pTp_T always loses the same energy ϵ(pT)\epsilon(p_T) and where that loss of energy either decreases or grows sufficiently slow with pTp_T (i.e., dϵ/dpT<1d\epsilon/dp_T < 1), the difference pTv−pTq(pTv)p_T^v-p_T^q(p_T^v) is equal to ϵ(pTv)\epsilon(p_T^v). This is simply a consequence of the conservation of number of jets in the case where jet transverse momenta are exactly and identically ordered in pp and AA. In reality, the energy lost by jets originating from hard partons with the same pTp_T can vary considerably [38, 39, 40, 41]. The probabilistic nature of QCD branching and of the interaction of partons with the QGP, together with the varying amount of traversed QGP, all contribute to disperse the energy loss around some mean value [30] and away from the ideally ordered scenario. Thus, in [36] the authors explored the accuracy of

1−QAA(pTv)=1−pTq(pTv)pTv,(3)1-Q_{AA}(p_T^v) = 1-\frac{p_T^q(p_T^v)}{p_T^v}, \tag*{(3)}

as a proxy for the average fractional jet energy loss of jets.

A crude analytic estimate of QAAQ_{AA} can be obtained as follows.2 Assume that the jet spectrum in pp is described by a single power law

dσppdpT(pT)=ApT−n,(4)\frac{d\sigma^{pp}}{dp_T}(p_T) = A p_T^{-n}, \tag*{(4)}

where AA is a normalization constant, and that the quenched jet spectrum can be obtained by taking an energy loss ϵ\epsilon for each jet in the pp spectrum with probability P(ϵ,pT)P(\epsilon,p_T) [43]:

dσAAdpT(pT)=∫0+∞dϵ P(ϵ,pT)dσppdpT(pT+ϵ)≈A(pT+⟨ϵ⟩(pT))−n.(5)\frac{\mathrm{d}\sigma^{AA}}{\mathrm{d}p_T}(p_T)=\int_{0}^{+\infty}\mathrm{d}\epsilon\,P(\epsilon,p_T)\frac{\mathrm{d}\sigma^{pp}}{\mathrm{d}p_T}(p_T+\epsilon)\approx A\left(p_T+\langle\epsilon\rangle(p_T)\right)^{-n}. \tag*{(5)}

In the last equality, we assumed that the fractional energy loss ϵ/pT\epsilon/p_T is small, i.e., that P(ϵ,pT)P(\epsilon,p_T) is negligible unless ϵ/pT≪1\epsilon/p_T\ll1, and that nn is sufficiently large (n∼5n\sim5). The result is then written to first order in ⟨ϵ⟩(pT)/pT\langle\epsilon\rangle(p_T)/p_T, with ⟨ϵ⟩(pT)=∫0+∞dϵ P(ϵ,pT)ϵ\langle\epsilon\rangle(p_T)=\int_{0}^{+\infty}\mathrm{d}\epsilon\,P(\epsilon,p_T)\epsilon.

The upper cumulative jet cross sections in Eq. (2) read

Σpp(pTv,+∞)=∫pTv+∞dpTdσppdpT=An−1(pTv)1−n,(6)\Sigma^{pp}(p_T^v,+\infty)=\int_{p_T^v}^{+\infty}\mathrm{d}p_T\frac{\mathrm{d}\sigma^{pp}}{\mathrm{d}p_T}=\frac{A}{n-1}(p_T^v)^{1-n}, \tag*{(6)}
ΣAA(pTq,+∞)=∫pTq+∞dpTdσAAdpT≈An−1(pTq+⟨ϵ⟩(pTq))1−n,(7)\Sigma^{AA}(p_T^q,+\infty)=\int_{p_T^q}^{+\infty}\mathrm{d}p_T\frac{\mathrm{d}\sigma^{AA}}{\mathrm{d}p_T}\approx\frac{A}{n-1}\left(p_T^q+\langle\epsilon\rangle(p_T^q)\right)^{1-n}, \tag*{(7)}

where we further assumed that ⟨ϵ⟩(pT)\langle\epsilon\rangle(p_T) varies slowly with pTp_T compared to pT−np_T^{-n} and that nn is constant. Inserting Eqs. (6) and (7) into Eq. (2), we have

1−pTq(pTv)pTv=1−QAA(pTv)≈⟨ϵ⟩(pTv)pTv,(8)1-\frac{p_T^q(p_T^v)}{p_T^v}=1-Q_{AA}(p_T^v)\approx\frac{\langle\epsilon\rangle(p_T^v)}{p_T^v}, \tag*{(8)}

where we took ⟨ϵ⟩(pTq)≈⟨ϵ⟩(pTv)\langle\epsilon\rangle(p_T^q)\approx\langle\epsilon\rangle(p_T^v), i.e., the zeroth order term in the expansion of ⟨ϵ⟩\langle\epsilon\rangle in powers of (pTv−pTq)(p_T^v-p_T^q). This confirms that 1−QAA1-Q_{AA} is a proxy for the average fractional energy loss as a function of pTp_T. This proxy is independent of the spectrum slope when assuming a pT−np_T^{-n} shape with constant nn and assumes a small relative energy loss with a mild pTp_T dependence.

One can also formulate a local version of the integral equation in Eq. (2) by differentiating it with respect to pTvp_T^v

dpTqdpTvdσAAdpTq(pTq(pTv))=dσppdpTv(pTv),(9)\frac{\mathrm{d}p_T^q}{\mathrm{d}p_T^v}\frac{\mathrm{d}\sigma^{AA}}{\mathrm{d}p_T^q}(p_T^q(p_T^v))=\frac{\mathrm{d}\sigma^{pp}}{\mathrm{d}p_T^v}(p_T^v), \tag*{(9)}

which results in a first order differential equation for pTq(pTv)p_T^q(p_T^v). A calculation of pTq(pTv)p_T^q(p_T^v) based on Eq. (9) is potentially more sensitive to statistical uncertainties in measured spectra. Additionally, one should keep in mind that, although Eq. (9) does not need an integration up to arbitrarily large pTp_T in contrast with Eq. (2) (see § 6.1), it is, in principle, still sensitive to a choice of pTqp_T^q evaluated at some momentum. This is because Eq. (9) is a first order differential equation in pTqp_T^q for which an initial condition needs to be specified.

Interestingly, the left-hand side of Eq. (9) was first introduced in [44] in the context of obtaining a simple energy loss estimate by looking at the suppression of the π0\pi^0 spectrum.3 It was used to quantify the number of particles in a bin of pTp_T after suppression. The jacobian factor dpTq/dpTvdp_T^q / dp_T^v accounts for the change in the relative particle density per pTp_T bin as a consequence of induced energy loss. Calculating pTqp_T^q, whether via Eq. (9) or Eq. (2), is a natural extension of this logic, in that it imposes that the number of in-medium jets in a given pTqp_T^q bin scaled by the jacobian should be the same as the number of vacuum jets in a given pTvp_T^v bin when we assume jets to be ideally ordered as described above. Note that a relation similar to Eq. (9) was used in [37] as an experimental proxy for jet energy loss. The two coincide in the limit where dpTq/dpTv→1dp_T^q / dp_T^v \to1, i.e., slowly varying energy loss with pTp_T, in which case the two equations determine similar quantities (modulus the assignment pTq(pTv)=pTv−ΔpT(pTv)=pTv(1−Sloss(pTv))p_T^q(p_T^v) = p_T^v - \Delta p_T(p_T^v) = p_T^v(1 - S_{\mathrm{loss}}(p_T^v))). We explore this further in § 6.2. The equation used in [37] is a differential equation and thus, like Eq. (9), also needs an initial condition to be specified.4

Generation and analysis details

The samples used in this work were generated with Jewel 2.3 [9, 45], and with the Hybrid Model5 [41, 47, 48, 49, 50] for model comparison of part of the results.

Jewel is a Monte Carlo event generator that implements jet evolution both in vacuum and in a medium, describing the QCD evolution of hard partons and their scatterings with the medium in a perturbative framework. Jewel interleaves the final state parton shower with scatterings with a QCD medium with leading-log correct relative contributions for elastic and inelastic scattering, such that vacuum-like and medium-induced emissions are treated in the same perturbative language. The interplay of both types of radiation is governed by formation times – all emissions are due to happen, but the one with the longest formation time is discarded. With this construction and using virtuality as an ordering variable, radiative energy loss is included while accounting for the usual vacuum parton shower. Additionally, Jewel implements destructive interference between emissions induced by successive medium scatterings via a probabilistic formulation of the QCD Landau-Pomeranchuk-Migdal (LPM) effect.

Jewel simulates the medium as the boost-invariant longitudinal expansion of an ideal quark-gluon gas. Jewel has the option of storing the four-momenta of the partonic constituents of the medium that were scattered (recoils) and also their initial stage (thermal) counterparts in the event record. This information is used to perform a background subtraction bettering the quantitative understanding of medium response and allowing for the reconstruction of a heavy-ion jet that takes into account medium response while discarding uncorrelated background. In this work, we use the event-wise algorithmic procedure introduced in [51] to subtract the background prior to jet reconstruction. The cutoff distance for subtraction is set to ΔRij=0.5\Delta R_{ij}=0.5. All remaining Jewel parameters concerning recoiling partons are set to their default values.

In the Hybrid Model [41, 47, 48, 49, 50], jet evolution inside a droplet of QGP is described by separating the strongly coupled dynamics of QGP and of the jet-QGP interactions, both governed by the medium temperature scale TT, from the weakly coupled dynamics governing the production and showering of high energy partons. Since the initial virtuality of the parton that showers and forms a jet is much larger than any scale associated with the medium (e.g. its temperature), the model assumes that its evolution proceeds as in vacuum, i.e., the parton shower branching is unmodified by the presence of the plasma. Specifically, the evolution is given by the parton shower implemented in PYTHIA 8.244 [52] and includes initial state radiation. As a consequence, the modification of jets is only due to the interaction of each parton in the jet with the strongly coupled medium. The rate of parton energy loss is modeled by the energy loss of light quark jets in the strongly coupled plasma of N=4N = 4 supersymmetric Yang-Mills theory (SYM) which has been computed holographically [53, 54]. This rate is applied to each parton of the already developed shower, using a formation time estimate for the path length travelled by each of them. Moreover, it is assumed that the all the differences between the strongly coupled limit of N=4N = 4 SYM theory and QCD can be enconded into a single parameter which is fitted to jet and hadron RAAR_{AA} data [50].

Each event is embedded within a droplet of QGP which expands and cools as described by relativistic viscous hydrodynamics, providing an evolving space-time background which is averaged over collisions within a given centrality class. The holographic energy loss rate is computed in the local rest frame of this fluid and at its local temperature. The jet propagates through the expanding cooling medium until it reaches a region where the temperature has dropped below a critical temperature TcT_c at which point no further energy loss occurs. The value for this temperature is varied within a given range 145<Tc<170145 < T_c < 170 MeV, gauging theory uncertainties of the model.

The energy and momentum deposited by the energetic partons of the shower in the medium excites a wake, which then evolves hydrodynamically. When the QGP and the jet-induced wake reach the freeze-out temperature, they are hadronized via the the Cooper-Frye prescription [55]. This prescription ensures that by boosting the fluid in its wake in its direction, the jet enhances the production of soft particles in its direction and depletes it in the opposite direction, relative to what would be obtained for the case without any jet wake. In the Hybrid Model, where the resulting Cooper-Frye distribution is negative it is implemented by adding hadrons with this distribution but with a negative energy. These "negative particles" are subtracted by using the same procedure employed for the thermal momenta in JEWEL.

JEWEL samples were generated for s=5.02\sqrt{s} = 5.02 TeV, for both pp and PbPb ([0–10]% centrality) collisions. These samples consist of 10610^6 dijet events and 10610^6 γ+\gamma+jet events. For both dijet and γ+\gamma+jet events, the EPPS16NLO nuclear PDFs [56] (nPDFs) are used for PbPb collisions. For dijet events, for the most part of the work, the CT14NLO proton PDFs [57] are used for pp collisions. All sets are provided by LHAPDF [58]. The particles produced in the hard scattering have a minimum transverse momentum cutoff of 20 GeV and the medium is generated with τi=0.6\tau_i = 0.6 fm, Ti=0.55T_i = 0.55 GeV and Tc=0.17T_c = 0.17 GeV. Initial state radiation and hadronization are handled by the PYTHIA 6 [59] framework that underlies JEWEL.

Importantly, pp samples for γ+\gamma+jet events are generated with the EPPS16NLO nPDFs, which are isospin averaged, as initial conditions. Hence, instead of the usual pp collisions, vacuum samples are obtained from the unphysical scenario of (average)nucleon-(average)nucleon collisions with no QGP. This is an attempt at isolating the role of quenching effects in the differences between vacuum and medium samples by minimizing the role of nuclear effects (isospin [60] and nPDFs), which is significant for γ\gamma+jet events [37].

Note that, for PbPb events, both JEWEL and the Hybrid Model only generate a single (average)nucleon-(average)nucleon collision per event. The cross-section estimate per event already takes this into account, since it is weighted by the isospin averaged nuclear PDFs.

To illustrate the role of nuclear effects on the inclusive jet and γ\gamma+jet spectra we show, in the top panels of Figure 1, the ratio between vacuum jet spectra obtained from events generated with isospin averaged nPDFs and events generated with proton PDFs. It is clear that the nuclear effects are non-negligible both in absolute magnitude of the cross-section and in the shape of the spectrum. This is especially pronounced for γ\gamma+jets, where the isospin effect in particular is expected to be greater since, at leading-order, the hard matrix elements always include an initial state quark or anti-quark. On the bottom panels of Figure 1 we show the impact of nuclear effects on an observable like the jet RAAR_{AA}. At large pTp_T and for γ\gamma+jets, this translates into a factor ∼1.5\sim1.5 difference. Since nuclear effects are not as significant for dijet events, we will, for the most part, use dijet vacuum samples obtained from simple pp collisions. This is true except for when drawing comparisons with γ\gamma+jet results in § 5.3, for consistency.

Figure 1. Top: Sensitivity of the vacuum γ\gamma-jets (left) and inclusive jets (right) spectra to nuclear effects. Bottom: Sensitivity of RAAR_{AA} to nuclear effects. The spectra labeled with pp+nPDFs result from (average)nucleon-(average)nucleon collisions generated with isospin averaged nuclear PDFs.

The Hybrid Model samples were generated for s=5.02\sqrt{s}=5.02 TeV, for both pp and PbPb ([0−-5]% centrality) collisions. These samples consist of ∼106\sim10^6 dijet events. The EPS09LO nuclear PDFs [61] are used for PbPb collisions and the NNPDF2.3 PDFs for pp collisions. Importantly, the particles produced in the hard scattering have a minimum transverse momentum cutoff of 100 GeV. For more details on the generated samples see [46].

Let us specify the details of this work’s analysis. In γ\gamma+jet events, the hard scattered photon is identified solely based on its pTp_T, i.e., the highest pTp_T photon from each event which passes pT≥50p_T \ge50 GeV and ∣y∣≤2.37|y| \le2.37 is chosen. Note that JEWEL does not implement QED radiation off either initial- or final-state quarks and, therefore, most photons not coming from the hard scattering come from the π0\pi^0 decay. The photon tagging is done prior to subtracting the background in PbPb events, keeping the photon out of the process. After the subtraction procedure, for both types of events, the final state particles are filtered according to pT≥0.5p_T \ge0.5 GeV (unless otherwise stated as in § 5.2) and ∣y∣≤3.3|y| \le3.3 and then given as input to the anti-ktk_t algorithm [62] for jet reconstruction with varying jet radius, as implemented in FASTJET [63]. All jets have a minimum cut of pTjet≥30p_T^{jet} \ge30 GeV (≥130\ge130 GeV) in JEWEL (Hybrid Model) and are required to have ∣yjet∣≤2.8|y^{jet}| \le2.8. The γ\gamma-tagged jets are required to be azimuthally separated by Δϕ≥7π/8\Delta\phi\ge7\pi/8 from the hard photon. For results using dijet events, the jet sample is inclusive.

Validation of the quantile procedure as a proxy for average energy loss

First, we validate the quantile procedure as an adequate proxy for jet energy loss away from the zero dispersion (and ideally ordered) case described in § 2. We do this by checking that the quantile procedure gives the same information as the pTp_T of the photon does in γ\gamma+jet events. Events with such a topology are particularly useful when validating any type of energy loss claim given that the transverse momentum of the photon is a proxy for the transverse momentum of the parton that initiated the jet found back-to-back with it. Given that nuclear effects (isospin and nuclear PDF) significantly alter the shape of the PbPb spectrum for γ\gamma+jet events, vacuum samples are generated with nuclear PDFs, which are isospin averaged, as initial conditions (see § 3).

In [36], the authors confirmed the validity of the quantile procedure for R=0.4R=0.4 jets produced back-to-back with a ZZ boson in the absence of medium response contributions to the jet. Here, we extend the validation to the more realistic case where medium response is accounted for (as described in § 3) and for multiple jet radii. The reasoning for this validation goes as follows. For a given photon pTγp_T^\gamma produced in the hard scattering, a distribution of jet pTp_T, pTvp_T^v for vacuum samples and pTmp_T^m for medium samples, is produced. The ratio between the mean value of these distributions ⟨pTm⟩/⟨pTv⟩\langle p_T^m\rangle/ \langle p_T^v\rangle (solid curves on the left plot of Figure 2) gives us a proxy for average jet energy loss, given that we are comparing jets originated by hard partons with the same pTp_T but that developed differently (vacuum and medium). Note that this is not the same as the average of the ratio of these two quantities, which is inaccessible in both JEWEL and in experimental data. The mean pTp_T values are given by

Figure 2. Left: Dependence of average jet energy loss proxies on jet radius RR – ratio of mean values of reconstructed pTp_T (dots) and ratio of average quantile pTp_T to mean vacuum pTp_T (solid). Right: Ratio between mean pTqp_T^q and mean pTmp_T^m. Vacuum samples include nuclear effects (isospin averaged nuclear PDFs).

⟨pTv/m⟩∣pTγ=∫dpT dNv/mdpT∣pTγpT.(10)\left.\langle p_T^{v/m}\rangle\right|_{p_T^\gamma} = \left.\int dp_T\,\frac{dN^{v/m}}{dp_T}\right|_{p_T^\gamma} p_T . \tag*{(10)}

where dNXdpT∣pTγ\left.\frac{dN^X}{dp_T}\right|_{p_T^\gamma} is the normalized transverse momentum distribution of jets for a given bin of photon transverse momentum pTγp_T^\gamma. On the other hand, the quantile procedure determines pTq=pTq(pTv)p_T^q = p_T^q(p_T^v) for a given pTvp_T^v, which should relate pp and AA jets that, on average, were originated by hard partons with the same pTp_T. One can then take the distribution of vacuum jets for a given pTγp_T^\gamma and calculate an hypothetical medium pTp_T distribution from it by applying pTq(pTv)p_T^q(p_T^v). The average of this distribution is then given by

⟨pTq⟩∣pTγ=∫dpT dNvdpT∣pTγpTq(pT).(11)\left.\langle p_T^q\rangle\right|_{p_T^\gamma} = \left.\int dp_T\,\frac{dN^v}{dp_T}\right|_{p_T^\gamma} p_T^q(p_T). \tag*{(11)}

The ratio ⟨pTq⟩/⟨pTv⟩\langle p_T^q\rangle/ \langle p_T^v\rangle (solid curves in Figure 2) is a proxy for the average jet energy loss for a given photon pTγp_T^\gamma, which should be similar to ⟨pTm⟩/⟨pTv⟩\langle p_T^m\rangle/ \langle p_T^v\rangle (dots in Figure 2). Note, again, that this is not the same as the mean ratio of pTp_T values. In this case, that would correspond to an average value of QAAQ_{AA}, which, although very similar to the quantile proxy we show here, is not the right quantity for this validation. With this being said, on the left-hand plot of Figure 2, we see that the agreement between the two average jet energy loss proxies is reasonable, although the quantile procedure systematically results in a smaller average energy loss than that estimated using the photon pTp_T.

On the right-hand plot of Figure 2, we directly show the comparison between ⟨pTq⟩\langle p_T^q\rangle and ⟨pTm⟩\langle p_T^m\rangle. A quantile procedure which perfectly quantified the correspondence between mean medium and vacuum jet pTp_T would result in a ratio of exactly 1. This ratio approaches 1 for larger the jet radii and for larger jet pTp_T, saturating at about a 1% difference. The largest deviation, which does not exceed ∼5%\sim5\%, is for R=0.2R = 0.2 jets. We that note that medium response is a large source of fluctuations pushing us away from the ideally ordered scenario. Looking at the left-hand plot of Figure 2, one can also see how jet energy loss varies with jet radius, for radii from 0.2 to 1. By using information from the photon’s pTp_T, we could already see (solid curves) that the ratio ⟨pTm⟩/⟨pTv⟩\langle p_T^m\rangle/ \langle p_T^v\rangle increases as one increases RR. Although this is a model dependent statement (see § 5.1 for model comparison), this means that larger jets lose a smaller fraction of their energy, on average. One can interpret this as due to medium response and to the recapture of medium-induced radiation compensating for the fact that increasing the jet radius increases the number of jet components that can lose energy. Both proxies have the same evolution with jet radius.

Jet energy loss dependence on jet radius, minimum particle pTp_T and colour charge

In the following subsections, we use QAAQ_{AA} as a proxy for average fractional jet energy loss and investigate its evolution with jet radius, its sensitivity to medium response and minimum particle pTp_T and, lastly, use it to evaluate the colour charge dependence of energy loss. Only on this last study are γ+\gamma+jet samples used. For the remaining two, the results are based on inclusive jet samples.

Jet radius

In Figure 3 we show the evolution of QAAQ_{AA} with jet radius for inclusive jet samples with (bottom panels) and without (top panels) medium response. For the purpose of evaluating model dependence, we compare two radically different implementations of jet physics in the presence of QGP - Jewel 2.3 [9, 45] (left) and the Hybrid Model [41, 47, 48, 49, 50] (right). Note that the curves for the Hybrid Model (right) start at 200 GeV. Despite the difference in parton-QGP interaction in these models, the trend of QAAQ_{AA} with jet radius and pTp_T is the same when considering events without medium response (top panels in Figure 3). Even in absolute value the two models predict similar results. The ordering of QAAQ_{AA} with jet radius is such that larger jets lose a larger fraction of their energy, on average. This implies, in particular for Jewel, that the increase in jet components that can lose energy dominates over the recapture of medium-induced radiation. When including medium response (bottom panels in Figure 3), however, the models predict different orderings with jet radius and only agree approximately in absolute value for large enough pTp_T and moderate radii R∼0.4,0.6R \sim0.4, 0.6. For Jewel the behaviour with jet radius is exactly inverted with respect to the case where medium response is absent, which is naturally attributed to the reconstruction of part of the energy lost to the medium inside the jet. This had already been observed for jet suppression (RAAR_{AA}) [64]. Nevertheless, the results we show for QAAQ_{AA} allow for this statement to be made quantitatively with regards to average jet energy loss. As for the Hybrid model, the rate of energy recovery by reconstructing part of the wake inside the jet is not very pronounced. Thus, for small to moderate jet radii, the energy recovery from the wake does not outgain the increase in lost energy due to the increase in number of jet components. It is only for large enough radii R≳0.6R \gtrsim0.6 that this happens, causing the ordering with jet radius to invert like in Jewel. Furthermore, with regards to the sensitivity of QAAQ_{AA} to the jet radius, in the absence of medium response, Jewel predicts a similar radius dependence to the Hybrid Model’s for jet radii R≲0.6R \lesssim0.6. For larger jet radii, the dependence is milder for Jewel. When medium response is included, the radius dependence of Jewel is pronouncedly larger than that of the Hybrid Model. This difference in behaviours can be explained by different rates of recapture of energy. Without medium response, jets in Hybrid Model have no way of recapturing lost energy, such that increasing the jet radius only increases the number of

Figure 3. Dependence of QAAQ_{AA} with jet radius for Jewel 2.3 (left) and for the Hybrid Model (right), with (bottom) and without (top) medium response. Vacuum samples come from pp collisions.

components of the vacuum parton shower that experience energy loss. On the other hand, for JEWEL, radius dependence starts saturating at intermediate jet radii (R∼0.6R \sim0.6) because of the recapture of medium-induced emissions. When including medium response, JEWEL jets with increasing radius, recapture energy at a significantly faster rate than in the Hybrid Model6. In fact, in [51] it was shown that JEWEL’s medium response is overpopulated in semi-hard portion of the particle spectrum, possibly a consequence of the recoiling medium particles not suffering any type of shower evolution or re-scattering. In contrast, in the Hybrid Model, medium response is composed of mostly very soft particles and spreading to large distances [41]. This means that, not only is the portion of lost energy that the jet can recover from the wake limited, but also that one needs to increase the jet radius significantly to be able to recapture a substantial amount of that energy.

Minimum particle pTp_T

Heavy-ion jets have a larger soft component then their pp counterparts, since both medium response and parton energy loss contribute to an increase of soft particles within the reconstruction cone. As a result (Figure 4), when including medium response, increasing the minimum particle pTp_T cut increases the energy lost by jets. This effect is very small for R=0.4R = 0.4 jets in the Hybrid model, owing to the large angular spread of the wake. Naturally, the increase is greater the larger the jet radius. This is because a larger jet will in principle be more populated in the softest part of the particle spectrum, thus suffering a larger relative energy cut. This radius dependence is much more pronounced in JEWEL than in the Hybrid Model, which is consistent with what we saw in § 5.1. As was seen in [41], the implementation of the medium response in the Hybrid Model underestimates the yield of semi-hard particles, leading to an excess of soft particles. This explains why, especially for R=0.8R = 0.8, we have a jump from the curve corresponding to a minimum pTp_T of 0.5 GeV to the one corresponding to 1 GeV, which seems to be the range where the pTp_T of most medium response particles lies.

Image not extracted

Figure 4. QAAQ_{\mathrm{AA}} sensitivity to different minimum required pTp_T of final state particles for JEWEL 2.3 (left) and for the Hybrid Model (right). Vacuum samples are pp collisions.

In the absence of medium response, the sensitivity to minimum pTp_T of the particles is almost non existent unless one goes to very large values of this cut, as can be seen for the curve corresponding to pT=2.5p_T = 2.5 GeV, which corresponds to a smaller energy loss than the remaining values for minimum pTp_T. Naturally, this arises from the particle constituents of a jet having a harder spectrum in vacuum than in medium. Interestingly enough, at this large a value of minimum pTp_T and when including medium response, the jet radius dependence of QAAQ_{AA} (Figure 5) is significantly reduced for both models, being almost washed out in JEWEL (left) except for the smaller jet radii R≲0.4R \lesssim0.4. In this case, the Hybrid Model (right) predicts a QAAQ_{AA} with a radius dependence which resembles what we saw for jets in the absence of medium response in the top right panel of Figure 3. This is expected, since at such a large cut most of the contribution from medium response is discarded.

Figure 5. Evolution of QAAQ_{AA} jet radius for JEWEL 2.3 (left) and for the Hybrid Model (right), for an extreme value of minimum required pTp_T of final state particles pT,min⁡=2.5p_{T,\min}=2.5 GeV). Medium response is included in both models. Vacuum samples are pp collisions.

Colour charge

We now address the dependence of jet energy loss on the colour charge of the parton, quark or gluon, that initiates a jet. Both perturbative radiative [16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26] and collisional [15] partonic energy losses are proportional to the Casimir colour factor of the propagating parton. This implies that the energy loss of a gluon is larger, by a factor CA/CF=9/4C_A/C_F=9/4, than that of a quark. In the absence of QGP, the amount of energy beyond a given jet reconstruction radius has been shown to not follow Casimir scaling [65, 66] since the dependence on the Casimir colour factor of the initiating parton is diluted as the parton cascade, dominated by gluon radiation, develops. When QGP is present, gluon initiated jets lose more energy than quark initiated jets by a factor substantially smaller than the CA/CFC_A/C_F Casimir scaling [67, 68].

To establish whether the large observed difference between the γ+\gamma+jet RAAR_{AA}, where quark jets are dominant, and the more gluon rich inclusive jet RAAR_{AA} can be explained by the colour charge of the parton initiating a jet, one cannot ignore the existence of pTp_T bin migration exacerbated by sensitivity of RAAR_{AA} to the steeply falling nature of the jet pTp_T spectrum. The γ+\gamma+jet and inclusive jets RAAR_{AA} could be different not because of different energy loss of quark and gluon jets, but simply because the steepnesses of the jet pTp_T spectra are different.

QAAQ_{AA} deconvolutes the effects of pTp_T migration and spectral shape influence from jet energy loss. In Figure 6 we show both RAAR_{AA} and QAAQ_{AA} for γ+\gamma+jets, inclusive jets and pure samples of quark jets (γ+\gamma+q) and gluon jets (γ+\gamma+q). These two last samples were obtained by selecting, at generation level, only the relevant (quark or gluon) channel. The difference between γ+\gamma+jets and inclusive jets in QAAQ_{AA} is significantly smaller than for RAAR_{AA} indicating that a significant part of the observed effect in RAAR_{AA} is due to pTp_T bin migration. Interestingly, the difference in energy loss of pure samples of quark and gluon initiated jets (γ+\gamma+q and γ+\gamma+g) is partly washed away when comparing γ+\gamma+jets and inclusive jets. In fact, inclusive jet QAAQ_{AA} is much closer to quark initiated jet QAAQ_{AA} than to gluon initiated jet QAAQ_{AA} for all pTp_T. While inclusive jets reconstructed in pp collisions are, for pT≲200p_T \lesssim200 GeV, a mixture of quark and gluon initiated jets with fractions of roughly 40/60 %, the larger energy loss of gluon jets in AA leads to a (survivor) bias that leads to a sample of reconstructed jets increasingly dominated by quark jets in each bin of pTp_T. The large steepness of the pTp_T spectrum further exaggerates this effect. The success of simple parametric models [69, 70] that encode such a variation of the q/g fraction supports this observation. Nevertheless, it remains to be precisely determined the degree to which q/g fractions in AA collisions are modified with respect to pp on both the theory [32, 68, 71, 72] and the experimental sides [73, 74]. Hence, our results in Figure 6 show one or both of the following statements to be true: in-medium q/g fractions are significantly modified due to energy loss; if inside a sample of jets one can distinguish two sub-populations of jets with known fractions and with different energy losses, then the resulting QAAQ_{AA} will differ from the weighted sum of the individual QAAQ_{AA} of each sub-population.

Figure 6. Dependence of RAAR_{AA} (left) and QAAQ_{AA} (right) on the hard process originating the jets. Vacuum samples include nuclear effects (isospin averaged nuclear PDFs).

The large difference between inclusive jets and γ+g\gamma+ g QAAQ_{AA} is in apparent contradiction with what is seen for RAAR_{AA} on the left-hand side plot of Figure 6, where for pT≲200p_T \lesssim200 GeV the two results are fairly similar. Nevertheless, note that the difference in spectrum steepness and the difference in energy loss affect the hierarchy of RAAR_{AA} in opposite directions. On the one hand, one would expect purely from the spectrum shape, while fixing the same energy loss, that inclusive RAAR_{AA} would be smaller than that of γ+g\gamma+g, based on the larger steepness of the former. Note that this difference in steepness is dependent on the initial partonic states that go into the leading order matrix element and not only on the final state, jet initiating partons. This means the shape of the inclusive spectrum does not necessarily follow a weighted sum of the γ+q\gamma+q and γ+g\gamma+g shapes according to the estimated q/g fractions. The energy loss of γ+g\gamma+g, on the other hand, is greater than that of inclusive jets as per the rough estimation of 40/60% for q/g fraction. This would result in a larger RAAR_{AA} for inclusive jets than for purely gluon initiated jets. Hence, what we see for pT≲200p_T \lesssim200 GeV is a cancellation of these two effects. When calculating QAAQ_{AA} the energy loss is the dominant effect, so the two curves are no longer similar at pT≲200p_T \lesssim200 GeV. At large enough pTp_T the steepness values of both spectra are similar, so the difference in RAAR_{AA} and QAAQ_{AA} are in agreement.

For a more precise statement of the dependence of jet energy loss on the colour charge of the initiating parton, we show on the left-hand side panel of Figure 7 the average absolute energy loss calculated using the quantile procedure for gluon jets (solid lines) and for quark jets (dashed lines), for a given value of photon pTp_T. Making the energy loss a function of photon pTp_T ensures that we avoid the bias coming from comparing quark and gluon jets with the different initial pTp_T in vacuum. The first interesting observation is that the jet radius dependence of energy loss is much more pronounced in gluon jets than in quarks jets. This is to be expected as per previous arguments, since a gluon radiating more means gluon initiated jets have a larger angular spread. Then, it is clear that gluon jets lose more energy than quarks jets for a wide range of jet radii. The ratio between gluon and quark jet energy loss for different radii is shown on right-hand side panel of Figure 7. It is stable around ∼1.5\sim1.5 independently of both jet radius and jet pTp_T. The Casimir scaling of single partons (the black line at 9/49/4) is substantially diluted in jets, with gluon jets not losing much more energy than quarks jets. Interestingly, the holographic calculations of single parton energy loss [47, 53, 54] that underlie the Hybrid Model predict a single parton scaling which is much milder than Casimir scaling (∼1.31\sim1.31) and which is much closer to our result obtained for in-medium jets. Note that this same result could be obtained by considering the energy loss proxy in Eq. (10) obtained simply by using the photon pTp_T to match mean jet pTp_T in vacuum and medium rather than the one obtained through quantile matching. Indeed, although we don’t show it here, this provides another testament to the quantile procedure’s robustness, since the result obtained using Eq. (10) shows the same features as the right-hand side plot of Figure 7.

Image not extracted

Figure 7. Evolution with jet radius of ⟨ΔpT⟩=⟨pTγ⟩−⟨pTA⟩\langle\Delta p_T\rangle= \langle p_T^\gamma\rangle-\langle p_T^A\rangle (see Eqs. (10) and (11)) (left) and of the ratio between average energy losses of gluon and quark initiated jets (right), for each bin of pTγp_T^\gamma. The black horizontal lines corresponds to what is expected from single parton Casimir scaling (solid) and from holography based calculations (dashed) [47, 53, 54]. Vacuum samples include nuclear effects (isospin averaged nuclear PDFs).

Towards a QAAQ_{AA} measurement

An obvious experimental obstacle to the calculation of the quantile function as defined in Eq. (2) is the requirement to integrate jet spectra up to arbitrarily large pTp_T, which is naturally not possible in experiment. However, doing the integration with a cutoff at +∞+\infty is not actually a requirement, since one can cast Eq. (2) into a form that takes into account pTp_T cutoffs in both spectra. This is done by considering Eq. (2) at two different values of pTvp_T^\mathrm{v} and subtracting the two resulting equations, giving Σpp(pT,1v,pT,2v)=ΣAA(pTq(pT,1v),pTq(pT,2v))\Sigma^{\mathrm{pp}}(p_{T,1}^\mathrm{v},p_{T,2}^\mathrm{v}) = \Sigma^{AA}(p_T^\mathrm{q}(p_{T,1}^\mathrm{v}),p_T^\mathrm{q}(p_{T,2}^\mathrm{v})). Hence, if the pp spectrum has a cutoff at a value pT,2v=pT,cvp_{T,2}^\mathrm{v}=p_{T,\mathrm{c}}^\mathrm{v} then the upper integration limit of ΣAA\Sigma^{AA} should be pT,cq=pTq(pT,cv)p_{T,\mathrm{c}}^\mathrm{q}=p_T^\mathrm{q}(p_{T,\mathrm{c}}^\mathrm{v}):

∫pTvpT,cvdpTdσppdpT=∫pTq(pTv)pT,cqdpTdσAAdpT.(12)\int_{p_T^\mathrm{v}}^{p_{T,\mathrm{c}}^\mathrm{v}} \mathrm{d}p_T \frac{\mathrm{d}\sigma^{\mathrm{pp}}}{\mathrm{d}p_T} = \int_{p_T^\mathrm{q}(p_T^\mathrm{v})}^{p_{T,\mathrm{c}}^\mathrm{q}} \mathrm{d}p_T \frac{\mathrm{d}\sigma^{AA}}{\mathrm{d}p_T}. \tag*{(12)}

However, because the right-hand side depends on the unknown quantile function evaluated at two different points (pT,cvp_{T,\mathrm{c}}^\mathrm{v} and pTvp_T^\mathrm{v}), the equation is underdetermined. This is expected, since Eq. (12) is an integral equation and therefore needs an initial condition that sets the value for pT,cqp_{T,\mathrm{c}}^\mathrm{q}. For values of pTvp_T^\mathrm{v} much smaller than pT,cvp_{T,\mathrm{c}}^\mathrm{v}, the sensitivity to the upper integration limit for the medium spectrum is insignificant and one can go around this problem by setting pT,cq=pT,cvp_{T,\mathrm{c}}^\mathrm{q}=p_{T,\mathrm{c}}^\mathrm{v}. For values comparable to pTcp_T^\mathrm{c}, however, the value of this limit becomes relevant. This is clearly seen in the left plot of Figure 8, where we see that setting the wrong value for pT,cqp_{T,\mathrm{c}}^\mathrm{q} (blue curve – overestimate it by imposing same cutoff; orange curve – underestimate it by setting it to 90% of the correct cutoff pT,cqp_{T,\mathrm{c}}^\mathrm{q}) results in a significant reduction in QAAQ_{AA} accuracy for a range of about 150 GeV below the vacuum cutoff.

Figure 8. Left: QAAQ_{AA} sensitivity to different values of pT,cqp_{T,c}^{q} in Eq. (12). Right: Comparison of QAAQ_{AA} and Q~AA\widetilde{Q}_{AA} for multiple jet radii. Vacuum samples include nuclear effects (isospin averaged nuclear PDFs).

Cutoff dependence

Possible workarounds for this problem include extracting pTq(pT,cv)p_T^\mathrm{q}(p_{T,\mathrm{c}}^\mathrm{v}) from an interpolated quantile function calculated via Monte Carlo at larger pTp_T than what is experimentally accessible. Another option, which can be fully calculated from data, is to use a variant of the quantile function also introduced in the original paper [36], the pseudo-quantile p~Tq\widetilde{p}_T^\mathrm{q}, to get an estimate of the initial condition pT,cq=pTq(pT,cv)p_{T,\mathrm{c}}^\mathrm{q}=p_T^\mathrm{q}(p_{T,\mathrm{c}}^\mathrm{v}). The way to calculate the pseudo-quantile is to equate the spectra themselves instead of their cumulative versions:

dσAAdp~Tq(p~Tq(pTv))=dσppdpTv(pTv).(13)\frac{\mathrm{d}\sigma^{AA}}{\mathrm{d}\widetilde{p}_T^\mathrm{q}}\left(\widetilde{p}_T^\mathrm{q}(p_T^\mathrm{v})\right) = \frac{\mathrm{d}\sigma^{\mathrm{pp}}}{\mathrm{d}p_T^\mathrm{v}}\left(p_T^\mathrm{v}\right). \tag*{(13)}

Interestingly enough, under the same assumptions as for the quantile function pTqp_T^\mathrm{q} in Eq. (8), one gets:

1−p~Tq(pTv)pTv=1−Q~AA(pTv)≈⟨ϵ⟩(pTv)pTv.(14)1-\frac{\widetilde{p}_T^\mathrm{q}(p_T^\mathrm{v})}{p_T^\mathrm{v}} = 1-\widetilde{Q}_{AA}(p_T^\mathrm{v}) \approx\frac{\langle\epsilon\rangle(p_T^\mathrm{v})}{p_T^\mathrm{v}}. \tag*{(14)}

Hence, despite having a clear conceptual difference, the quantile and the pseudo-quantile functions are in good agreement under the assumption of small and only mildly pTp_T-dependent energy loss. Indeed, by looking at Eq. (9), we see that it is equivalent to Eq. (13) in the limit where we have some constant energy loss shift ϵ0\epsilon_0

pTq(pTv)⟶pTv+ϵ0.(15)p_{T}^{q}(p_{T}^{v}) \longrightarrow p_{T}^{v}+\epsilon_0 . \tag*{(15)}

It is then expected that when calculating Q~AA\widetilde{Q}_{AA} and comparing it to QAAQ_{AA} (right-hand side plot in Figure 8), the two are not strikingly different for most of the pTp_T range and that they are in great agreement at large enough pTp_T and for all jet radii. By looking at the red curve of the left-hand side plot of Figure 8, we then confirm that using p~T,cq=p~Tq(pTc)\widetilde{p}_{T,c}^{q}=\widetilde{p}_{T}^{q}(p_{T}^{c}) as an estimate for the initial condition pT,cq=pTq(pT,c)p_{T,c}^{q}=p_{T}^{q}(p_{T,c}) works well.

Equivalence between 1−QAA1-Q_{AA} and SlossS_{\mathrm{loss}}

As discussed at the end of § 2, a relation similar to our Eq. (9) was used in [37] to estimate jet energy loss. Renaming the variables, it reads

dσAAdpTs(pTs)=dσppdpTv(pTv)(1+dΔpTsdpTv),(16)\frac{d\sigma^{AA}}{dp_{T}^{s}}(p_{T}^{s})=\frac{d\sigma^{pp}}{dp_{T}^{v}}(p_{T}^{v})\left(1+\frac{d\Delta p_{T}^{s}}{dp_{T}^{v}}\right), \tag*{(16)}

where pTs(pTv)=pTv−ΔpTs(pTv)=(1−Sloss(pTv))pTvp_{T}^{s}(p_{T}^{v})=p_{T}^{v}-\Delta p_{T}^{s}(p_{T}^{v})=(1-S_{\mathrm{loss}}(p_{T}^{v}))p_{T}^{v}. Setting ΔpTq(pTv)≡pTv−pTq(pTv)=(1−QAA(pTv))pTv\Delta p_{T}^{q}(p_{T}^{v})\equiv p_{T}^{v}-p_{T}^{q}(p_{T}^{v})=(1-Q_{AA}(p_{T}^{v}))p_{T}^{v}, the differential equation defining the quantile matching in Eq. (9) can be re-written as

dσAAdpTq(pTq)=dσppdpTv(pTv)(1−dΔpTqdpTv)−1.(17)\frac{d\sigma^{AA}}{dp_{T}^{q}}(p_{T}^{q})=\frac{d\sigma^{pp}}{dp_{T}^{v}}(p_{T}^{v})\left(1-\frac{d\Delta p_{T}^{q}}{dp_{T}^{v}}\right)^{-1}. \tag*{(17)}

Looking at Eqs. (16) and (17), we see that both agree to first order in dΔpT/dpTvd\Delta p_T/dp_T^v, thus computing similar proxies (pTs≈pTq)(p_T^s\approx p_T^q) for a mildly pTp_T-dependent energy loss. Evidently, both quantities agree with the pseudo-quantile as defined in Eq. (13) to zeroth order in dΔpT/dpTvd\Delta p_{T}/dp_{T}^{v} (see right plot of Figure 8). Nevertheless, if energy loss has a stronger pTp_{T} dependence than in JEWEL 2.3, then the two equations may give significantly different results. Additionally, this dependence can be artificially introduced if one incorrectly sets the initial condition. This is clearly seen in Figure 9 (left plot), where we show the difference between the two energy loss proxies (ΔpTq−ΔpTs)/pTv=(pTs−pTq)/pTv(\Delta p_{T}^{q}-\Delta p_{T}^{s})/p_{T}^{v}=(p_{T}^{s}-p_{T}^{q})/p_{T}^{v}. For the blue curve we estimate the initial condition using the pseudo-quantile as defined in Eq. (13), which we have seen provides an accurate approximation to the quantile at large enough pTp_{T} (see Figure 8). In this case, Eqs. (16) and (17) calculate almost exactly the same quantity, with the largest difference (∼0.1%\sim0.1\% of the lost energy) being at lowest end of the pTp_{T} range (∼100 GeV\sim100\ \mathrm{GeV}). This is expected since, as can be seen on the right plot of Figure 9, the derivative appearing on the right hand side of both equations is quite small – it is always below 0.1 and decreases with increasing pTvp_{T}^{v}. If we provide an inaccurate estimate for the initial condition, e.g., 0.9 of the previous value, then the two quantities ΔpTs\Delta p_{T}^{s} and ΔpTq\Delta p_{T}^{q} separate from each other, with the largest difference reaching ∼2.5%\sim2.5\% of the lost energy. Again, we can explain this by looking at the right plot of Figure 9, where we see the derivative increasing past a value for which Eqs. (17) and (16) agree to first order. Despite these differences, it is clear the quantities 1−QAA1-Q_{AA} and SlossS_{\mathrm{loss}} should not be too different from each other when applied to experimental data and when reasonably estimating the initial condition. Nevertheless, the quantile procedure has a much clearer physical motivation and, although not explored here, it provides more solid groundwork for the study of jet energy loss dispersion.

Image not extracted

Figure 9. Left: Comparing energy loss proxies ΔpTq=pTv−pTq\Delta p_{T}^{q}=p_{T}^{v}-p_{T}^{q} in Eq. (17) and ΔpTs=pTv−pTs\Delta p_{T}^{s}=p_{T}^{v}-p_{T}^{s} in Eq. (16) for two differential values of initial condition given by p~Tq(pTv=700 GeV)\tilde{p}_{T}^{q}(p_{T}^{v}=700\ \mathrm{GeV}) (reasonable estimate) and by 0.9p~Tq(pTv=700 GeV)0.9\tilde{p}_{T}^{q}(p_{T}^{v}=700\ \mathrm{GeV}) (underestimate). Right: Derivative of the energy loss function calculated through Eq. (17) for the same two values of initial condition. Vacuum samples include nuclear effects (isospin averaged nuclear PDFs).

Jet substructure observables after quantile correction

The main purpose of the quantile matching procedure is to suppress the pTp_T migration bias associated with observable comparison within the same pTjetp_T^{jet} bin. By estimating average jet energy loss, one compares vacuum and medium jets in matched pTp_T bins thus comparing jets that were, on average, initiated by hard partons with the same pTp_T. Naturally, studying how sensitive to pTp_T migration bias an observable is depends on the correlation of such observable with jet pTp_T. This result was showed in the original paper [36], where the authors saw the impact of pTp_T migration on the ratio between jet mass and jet transverse momentum mjet/pTjetm^{jet}/p_T^{jet}. Let us extend this analysis to other jet observables and also include medium response as implemented in JEWEL. In Figure 10, we show the distribution of four jet substructure observables for jets in a fixed pTp_T bin (dashed lines) and for quantile matched pTp_T bins (solid lines). The list of interesting jet substructure observables is quite extensive [75], but we decided to chose a few which illustrate both observables whose distributions are significantly corrected and those which are practically insensitive to the pTp_T migration.

Image not extracted

Figure 10. Distribution of four jet substructure observables for vacuum, medium without recoils and medium with recoils. The solid lines correspond to jets with pTjet∈(100,200)p_T^{\rm jet} \in(100, 200) GeV and the dashed lines to pTq∈(79,170)p_T^q \in(79, 170) GeV (medium w/o recoils) and pTq∈(84,177)p_T^q \in(84, 177) GeV (medium w/ recoils). Vacuum samples are pp collisions.

Let us first focus on the groomed jet radius rgr_g (top left corner in Figure 10). To calculate rgr_g, one first reclusters a jet using the Cambridge-Aachen (C/A) algorithm and then calculates the angular distance ΔRij\Delta R_{ij} of the first branching/subjet that satisfies the SoftDrop condition [76]

min⁡[pT,i,pT,j]pT,i+pT,j>zcut(ΔRijR)β(18)\frac{\min[p_{T,i},p_{T,j}]}{p_{T,i}+p_{T,j}} > z_{\mathrm{cut}}\left(\frac{\Delta R_{ij}}{R}\right)^\beta \tag*{(18)}

where zcut=0.1z_{\mathrm{cut}} = 0.1 and β=0\beta= 0. We see that for rgr_g the contribution of pTp_T migration is quite significant both with and without medium response. In fact, the apparent narrowing of jets from vacuum to medium is reduced when correcting the pTp_T bin with quantile matching. Continuing with an observable which conveys information about the width of a jet, we now look at the jet girth gSDg_{SD} (top right corner in Figure 10) after applying SoftDrop:

gSD=∑i∈jetSDziΔRi,jet,(19)g_{SD} = \sum_{i \in\mathrm{jet}_{SD}} z_i \Delta R_{i,\mathrm{jet}}, \tag*{(19)}

where ziz_i is the fraction of transverse momentum of the constituent ii with respect to jet pTp_T and ΔRi,jet\Delta R_{i,\mathrm{jet}} is the angular distance of the constituent with respect to the jet axis. Applying SoftDrop to groom the jet amounts to discarding the softest branch at each step of the C/A sequence until the condition in Eq. (18) is met. Looking at the result, the solid curves (quantile matched bins) are significantly deviated from the dashed ones (unmatched bins), such that the apparent narrowing of jets is, like for rgr_g, reduced. This is true whether or not we include medium recoils.

As for observables that are not as sensitive to pTp_T migration (i.e. that are only weakly correlated with pTp_T in the pTp_T range we consider), let us look at yet another observable quantifying the angular spread of a jet – the distance ΔR\Delta R between the two subjets in the C/A reclustering sequence that satisfy the following dynamical grooming [77] condition:

κ(a)=1pTjetmax⁡i∈C/A seq.[zi(1−zi)pT,i(θiR)a].(20)\kappa^{(a)} = \frac{1}{p_T^{jet}} \max_{i \in\mathrm{C/A\ seq.}} \left[z_i(1-z_i)p_{T,i}\left(\frac{\theta_i}{R}\right)^a\right]. \tag*{(20)}

By setting a=1a = 1 (kTk_T Drop) we are tagging the branching with the largest relative transverse momentum kT∼κ(1)pTjetk_T \sim\kappa^{(1)} p_T^{\rm jet}. Looking at bottom left corner of Figure 10, we see that in the absence of medium response this observable is only slightly sensitive to pTp_T migration (orange curves). However, when including medium response (green curves) the pTp_T correlation seems to be washed out since the observable is very little modified after quantile matching the pTp_T bin. Finally, in lower right corner of Figure 10, we show the momentum dispersion inside a jet, DpTD_{p_T}, which quantifies how hard or soft the jet fragmentation is:

DpT=∑i∈jetpT,i2pTjet.(21)D_{p_T} = \frac{\sqrt{\sum_{i\in\mathrm{jet}} p_{T,i}^{2}}}{p_T^{\mathrm{jet}}}. \tag*{(21)}

Whether or not we include medium recoils, this observable is almost insensitive to pTp_T migration.

Conclusions

Performing an unbiased comparison of proton-proton and heavy-ion jet populations is essential to be able to infer quantitative knowledge about the transport properties of the underlying QGP. In this work, and following the groundwork laid by [36], we took a step in this direction by further exploring how the quantile procedure can be used as a tool to mitigate the pTp_T migration bias inherent to usual jet comparisons by providing a proxy for average fractional jet energy loss (QAAQ_{AA}).

Comparing QAAQ_{AA} with the jet energy loss proxy obtained by using the photon’s pTp_T as a reference in γ\gamma+jet events, we concluded that the agreement between the two is reasonable, even in the presence of a large source of fluctuations, the medium response, especially pronounced for larger jet radii. One should note, however, that the former systematically underestimates energy loss compared to the latter. The evolution of QAAQ_{AA} with jet radius was also studied for two different models of jet evolution in heavy-ion collisions – Jewel and the Hybrid Model. In the absence of medium response, the two models exhibit the same behaviour of QAAQ_{AA} with jet radius – larger jets lose more energy –, indicating, for Jewel, that the increase in jet components contributing to energy loss dominates over the recapture of medium-induced radiation inside the jet cone. When including medium response, however, the difference in implementations appears quite significant since the two models predict quite different behaviours with jet radius. Overall, the conclusion is that medium jets in Jewel recapture energy at a significantly larger, and quite possibly exaggerated, rate than in the Hybrid Model, since in the latter the medium response is not only softer but also more widely distributed in angle. For similar reasons, jets in Jewel exhibit a QAAQ_{AA} which is much more sensitive to minimum particle pTp_T than in the Hybrid Model. By studying this sensitivity at a larger jet radius, we clearly see that the pTp_T of particles in the medium response of the Hybrid Model is dominated by the range ∼(0.5,1)\sim(0.5, 1) GeV. With an extreme cut at 2.5 GeV, the medium response seems to be removed in the Hybrid model, while this is not the case in Jewel.

The strength of colour charge dependence was also evaluated by comparing the RAAR_{AA} and QAAQ_{AA} of γ\gamma+jets and inclusive jets. This lead to the conclusion that spectrum shape severely impacts this comparison when looking at RAAR_{AA}, since disentangling this effect from the actual energy loss using QAAQ_{AA} reveals that colour charge dependence is not as strong as predicted by the large observed difference in jet suppression. By further comparing these results with the QAAQ_{AA} of pure samples of quark jets (γ\gamma+q) and of gluon jets (γ\gamma+q), we conclude that one or both of the following statements must be true: the in-medium q/gq/g fractions are modified with respect to the vacuum due to the survivor bias induced by gluon jets losing more energy than quarks jets; the QAAQ_{AA} of a population of jets containing two sub-populations with substantially different energy losses is not trivially given by the weighted sum of their individual energy losses. Additionally, we compared the ratio between the absolute energy loss of gluon jets and that of quarks jets with the expected Casimir scaling of single parton energy loss. The result is substantially smaller than the Casimir ratio CF/CAC_F/C_A and it closer to the value predicted by holographic calculations of parton energy loss (e.g. [47, 54]).

The experimental obstacle posed by the requirement to integrate the jet pTp_T spectrum up to +∞+\infty to calculate QAAQ_{AA} was addressed. In fact, this requirement can be translated into the need for an initial condition of the equation defining the quantile function, which can be quite accurately estimated, for large enough pTp_T, using the pseudo-quantile function [36], which uses the differential cross-sections directly in the computation. A comparison between QAAQ_{AA} and SlossS_{\mathrm{loss}}, which was used in [37], as jet energy loss proxies was drawn. Put simply, we find no advantage in using SlossS_{\mathrm{loss}} rather than QAAQ_{AA} to estimate jet energy loss, given that the two observables involve the same technical difficulties (namely providing an accurate initial condition) and QAAQ_{AA} has a clear physical motivation from which one can build more complex, rigorous energy loss studies. At the end, a simple exploration of the size of pTp_T migration in a few substructure observables was carried out, illustrating the potential of the quantile procedure to quantify jet modifications due to interaction with a QGP in a more unbiased way.

As a last note, we point out a clear limitation of QAAQ_{AA} as currently defined. It is certainly tempting to calculate QAAQ_{AA} in bins of a given jet substructure observable, that we will refer to as XX. In doing so, we would be going beyond just an average value of energy loss, partially resolving energy loss dispersion by correlating it with a single jet substructure observable. However, doing so would effectively get us into the same type of bias that exists when comparing jets in bins of equal reconstructed pTp_T. In particular, this would mean that we would be confounding energy loss of jets with the modification of the distribution of XX in the medium with respect to the vacuum. A possible solution would be to calculate a quantile matching in XX, assuming its bin migration to be monotonic. Nevertheless, bin migration in XX and in pTp_T are not, in principle, independent. Hence, to calculate them separately would be to distort this dependence, obtaining results which do not reflect actual energy loss dependence on substructure. Having said this, the only rigorously correct procedure to take is to calculate a quantile matching in pTp_T and XX simultaneously. However, this problem is undetermined since we would have a single (quantile) equation for two unknown (quantile) functions. In order for this to be possible, further assumptions about the nature of bin migration in XX and its dependence on pTp_T would need to be made.

Acknowledgments

We thank Daniel Pablos and Arjun Srinivasan Kudinoor for kindly assisting us with the Hybrid model samples and for discussions. This work is supported by European Research Council project ERC-2018-ADG-835105 Yoct oLHC. It as also been supported by OE - Portugal, Fundação para a Ciência e a Tecnologia (FCT), I.P., under projects CERN/FIS-PAR/0032/2021 (http://doi.org/10.54499/CERN/FIS-PAR/0032/2021) and EXPL/FIS-PAR/0905/2021 (https://doi.org/10.54499/EXPL/FIS-PAR/0905/2021). L.A. was supported by FCT under contract 2021.03209.CEE CIND. L.L. was supported by grant id 46225 from EXPL/FIS-PAR/0905/2021. J.M.S. was supported by OE - Portugal, Fundação para a Ciência e Tecnologia (FCT) under contract PRT/BD/152262/2021.

References

  1. [1]PHENIX Collaboration, K. Adcox et al., Suppression of hadrons with large transverse momentum in central Au+Au collisions at √sNN = 130-GeV, Phys. Rev. Lett. 88 (2002) 022301, [nucl-ex/0109003].pubmed.ncbi.nlm.nih.gov/11801005
  2. [2]STAR Collaboration, J. Adams et al., Experimental and theoretical challenges in the search for the quark gluon plasma: The STAR Collaboration’s critical assessment of the evidence from RHIC collisions, Nucl. Phys. A 757 (2005) 102–183, [nucl-ex/0501009].
  3. [3]STAR Collaboration, J. Adam et al., Measurement of inclusive charged-particle jet production in Au + Au collisions at √sNN =200 GeV, Phys. Rev. C 102 (2020), no. 5 054913, [arXiv:2006.00582].
  4. [4]B. Muller, J. Schukraft, and B. Wyslouch, First Results from Pb+Pb collisions at the LHC, Ann. Rev. Nucl. Part. Sci. 62 (2012) 361–386, [arXiv:1202.3233].DOI
  5. [5]ATLAS Collaboration, M. Aaboud et al., Measurement of the nuclear modification factor for inclusive jets in Pb+Pb collisions at √sNN = 5.02 TeV with the ATLAS detector, Phys. Lett. B 790 (2019) 108–128, [arXiv:1805.05635].DOI
  6. [6]CMS Collaboration, A. M. Sirunyan et al., First measurement of large area jet transverse momentum spectra in heavy-ion collisions, JHEP 05 (2021) 284, [arXiv:2102.13080].DOI
  7. [7]W. Busza, K. Rajagopal, and W. van der Schee, Heavy Ion Collisions: The Big Picture, and the Big Questions, Ann. Rev. Nucl. Part. Sci. 68 (2018) 339–376, [arXiv:1802.04801].DOI
  8. [8]L. Apolinário, Y.-J. Lee, and M. Winn, Heavy quarks and jets as probes of the QGP, Prog. Part. Nucl. Phys. 127 (2022) 103990, [arXiv:2203.16352].DOI
  9. [9]K. C. Zapp, F. Krauss, and U. A. Wiedemann, A perturbative framework for jet quenching, JHEP 03 (2013) 080, [arXiv:1212.1599].DOI
  10. [10]P. Caucal, E. Iancu, A. H. Mueller, and G. Soyez, Vacuum-like jet fragmentation in a dense QCD medium, Phys. Rev. Lett. 120 (2018) 232001, [arXiv:1801.09703].DOI
  11. [11]Y. Mehtar-Tani, J. G. Milhano, and K. Tywoniuk, Jet physics in heavy-ion collisions, Int. J. Mod. Phys. A 28 (2013) 1340013, [arXiv:1302.2579].
  12. [12]G.-Y. Qin and X.-N. Wang, Jet quenching in high-energy heavy-ion collisions, Int. J. Mod. Phys. E 24 (2015), no. 11 1530014, [arXiv:1511.00790].DOI
  13. [13]L. Cunqueiro and A. M. Sickles, Studying the QGP with Jets at the LHC and RHIC, Prog. Part. Nucl. Phys. 124 (2022) 103940, [arXiv:2110.14490].DOI
  14. [14]L. Apolinário, Y.-T. Chien, and L. Cunqueiro Mendez, Jet substructure, Int. J. Mod. Phys. E 33 (2024), no. 07 2430003.
  15. [15]S. Wicks, W. Horowitz, M. Djordjevic, and M. Gyulassy, Elastic, inelastic, and path length fluctuations in jet tomography, Nucl. Phys. A 784 (2007) 426–442, [nucl-th/0512076].DOI
  16. [16]M. Gyulassy, P. Levai, and I. Vitev, NonAbelian energy loss at finite opacity, Phys. Rev. Lett. 85 (2000) 5535–5538, [nucl-th/0005032].DOI
  17. [17]M. Gyulassy, P. Levai, and I. Vitev, Reaction operator approach to nonAbelian energy loss, Nucl. Phys. B 594 (2001) 371–419, [nucl-th/0006010].DOI
  18. [18]R. Baier, Y. L. Dokshitzer, A. H. Mueller, S. Peigne, and D. Schiff, Radiative energy loss of high-energy quarks and gluons in a finite volume quark - gluon plasma, Nucl. Phys. B 483 (1997) 291–320, [hep-ph/9607355].
  19. [19]R. Baier, Y. L. Dokshitzer, A. H. Mueller, and D. Schiff, Radiative energy loss of high-energy partons traversing an expanding QCD plasma, Phys. Rev. C 58 (1998) 1706–1713, [hep-ph/9803473].DOI
  20. [20]R. Baier, Y. L. Dokshitzer, A. H. Mueller, and D. Schiff, Medium induced radiative energy loss: Equivalence between the BDMPS and Zakharov formalisms, Nucl. Phys. B 531 (1998) 403–425, [hep-ph/9804212].DOI
  21. [21]B. G. Zakharov, Fully quantum treatment of the Landau-Pomeranchuk-Migdal effect in QED and QCD, JETP Lett. 63 (1996) 952–957, [hep-ph/9607440].DOI
  22. [22]B. G. Zakharov, Radiative energy loss of high-energy quarks in finite size nuclear matter and quark - gluon plasma, JETP Lett. 65 (1997) 615–620, [hep-ph/9704255].DOI
  23. [23]X.-N. Wang and X.-f. Guo, Multiple parton scattering in nuclei: Parton energy loss, Nucl. Phys. A 696 (2001) 788–832, [hep-ph/0102230].
  24. [24]A. Majumder, The In-medium scale evolution in jet modification, arXiv:0901.4516.
  25. [25]P. B. Arnold, G. D. Moore, and L. G. Yaffe, Photon emission from quark gluon plasma: Complete leading order results, JHEP 12 (2001) 009, [hep-ph/0111107].DOI
  26. [26]P. B. Arnold, G. D. Moore, and L. G. Yaffe, Photon and gluon emission in relativistic plasmas, JHEP 06 (2002) 030, [hep-ph/0204343].DOI
  27. [27]Y. Mehtar-Tani, C. A. Salgado, and K. Tywoniuk, Anti-angular ordering of gluon radiation in QCD media, Phys. Rev. Lett. 106 (2011) 122002, [arXiv:1009.2965].
  28. [28]Y. Mehtar-Tani, D. Pablos, and K. Tywoniuk, Cone-Size Dependence of Jet Suppression in Heavy-Ion Collisions, Phys. Rev. Lett. 127 (2021), no. 25 252301, [arXiv:2101.01742].
  29. [29]CDF Collaboration, D. Acosta et al., Study of jet shapes in inclusive jet production in p̄p collisions at √s = 1.96 TeV, Phys. Rev. D 71 (2005) 112002, [hep-ex/0505013].
  30. [30]Y. He, L.-G. Pang, and X.-N. Wang, Bayesian extraction of jet energy loss distributions in heavy-ion collisions, Phys. Rev. Lett. 122 (2019), no. 25 252302, [arXiv:1808.05310].DOI
  31. [31]J. Wu, W. Ke, and X.-N. Wang, Bayesian inference of the path-length dependence of jet energy loss, Phys. Rev. C 108 (2023), no. 3 034911, [arXiv:2304.06339].arxiv.org/abs/2304.06339
  32. [32]S.-L. Zhang, E. Wang, H. Xing, and B.-W. Zhang, Flavor dependence of jet quenching in heavy-ion collisions from a Bayesian analysis, Phys. Lett. B 850 (2024) 138549, [arXiv:2303.14881].DOI
  33. [33]M. Connors, C. Nattrass, R. Reed, and S. Salur, Jet measurements in heavy ion physics, Rev. Mod. Phys. 90 (2018) 025005, [arXiv:1705.01974].
  34. [34]J. Brewer, Q. Brodsky, and K. Rajagopal, Disentangling jet modification in jet simulations and in Z+jet data, JHEP 02 (2022) 175, [arXiv:2110.13159].DOI
  35. [35]C. Andres, J. Holguin, R. Kunnawalkam Elayavalli, and J. Viinikainen, Minimizing Selection Bias in Inclusive Jets in Heavy-Ion Collisions with Energy Correlators, arXiv:2409.07514.arxiv.org/abs/2409.07514
  36. [36]J. Brewer, J. G. Milhano, and J. Thaler, Sorting out quenched jets, Phys. Rev. Lett. 122 (2019), no. 22 222301, [arXiv:1812.05111].DOI
  37. [37]ATLAS Collaboration, G. Aad et al., Comparison of inclusive and photon-tagged jet suppression in 5.02 TeV Pb+Pb collisions with ATLAS, Phys. Lett. B 846 (2023) 138154, [arXiv:2303.10090].DOI
  38. [38]J. G. Milhano and K. C. Zapp, Origins of the di-jet asymmetry in heavy ion collisions, Eur. Phys. J. C 76 (2016), no. 5 288, [arXiv:1512.08107].arxiv.org/abs/1512.08107
  39. [39]K. Rajagopal, A. V. Sadofyev, and W. van der Schee, Evolution of the jet opening angle distribution in holographic plasma, Phys. Rev. Lett. 116 (2016), no. 21 211603, [arXiv:1602.04187].DOI
  40. [40]J. Brewer, K. Rajagopal, A. Sadofyev, and W. Van Der Schee, Evolution of the Mean Jet Shape and Dijet Asymmetry Distribution of an Ensemble of Holographic Jets in Strongly Coupled Plasma, JHEP 02 (2018) 015, [arXiv:1710.03237].DOI
  41. [41]J. Casalderrey-Solana, D. Gulhan, G. Milhano, D. Pablos, and K. Rajagopal, Angular Structure of Jet Quenching Within a Hybrid Strong/Weak Coupling Model, JHEP 03 (2017) 135, [arXiv:1609.05842].
  42. [42]A. Takacs and K. Tywoniuk, Quenching effects in the cumulative jet spectrum, JHEP 10 (2021) 038, [arXiv:2103.14676].DOI
  43. [43]R. Baier, Y. L. Dokshitzer, A. H. Mueller, and D. Schiff, Quenching of hadron spectra in media, JHEP 09 (2001) 033, [hep-ph/0106347].DOI
  44. [44]PHENIX Collaboration, K. Adcox et al., Formation of dense partonic matter in relativistic nucleus-nucleus collisions at RHIC: Experimental evaluation by the PHENIX collaboration, Nucl. Phys. A 757 (2005) 184–283, [nucl-ex/0410003].
  45. [45]K. C. Zapp, JEWEL 2.0.0: directions for use, Eur. Phys. J. C 74 (2014), no. 2 2762, [arXiv:1311.0048].DOI
  46. [46]H. Bossi, A. S. Kudinoor, I. Moult, D. Pablos, A. Rai, and K. Rajagopal, Imaging the Wakes of Jets with Energy-Energy-Energy Correlators, arXiv:2407.13818.arxiv.org/abs/2407.13818
  47. [47]J. Casalderrey-Solana, D. C. Gulhan, J. G. Milhano, D. Pablos, and K. Rajagopal, A Hybrid Strong/Weak Coupling Approach to Jet Quenching, JHEP 10 (2014) 019, [arXiv:1405.3864]. [Erratum: JHEP 09, 175 (2015)].
  48. [48]J. Casalderrey-Solana, D. C. Gulhan, J. G. Milhano, D. Pablos, and K. Rajagopal, Predictions for Boson-Jet Observables and Fragmentation Function Ratios from a Hybrid Strong/Weak Coupling Model for Jet Quenching, JHEP 03 (2016) 053, [arXiv:1508.00815].
  49. [49]Z. Hulcher, D. Pablos, and K. Rajagopal, Resolution Effects in the Hybrid Strong/Weak Coupling Model, JHEP 03 (2018) 010, [arXiv:1707.05245].DOI
  50. [50]J. Casalderrey-Solana, Z. Hulcher, G. Milhano, D. Pablos, and K. Rajagopal, Simultaneous description of hadron and jet suppression in heavy-ion collisions, Phys. Rev. C 99 (2019), no. 5 051901, [arXiv:1808.07386].
  51. [51]J. G. Milhano and K. Zapp, Improved background subtraction and a fresh look at jet sub-structure in JEWEL, Eur. Phys. J. C 82 (2022), no. 11 1010, [arXiv:2207.14814].arxiv.org/abs/2207.14814
  52. [52]T. Sjöstrand, S. Ask, J. R. Christiansen, R. Corke, N. Desai, P. Ilten, S. Mrenna, S. Prestel, C. O. Rasmussen, and P. Z. Skands, An introduction to PYTHIA 8.2, Comput. Phys. Commun. 191 (2015) 159–177, [arXiv:1410.3012].DOI
  53. [53]P. M. Chesler and K. Rajagopal, Jet quenching in strongly coupled plasma, Phys. Rev. D 90 (2014), no. 2 025033, [arXiv:1402.6756].DOI
  54. [54]P. M. Chesler and K. Rajagopal, On the Evolution of Jet Energy and Opening Angle in Strongly Coupled Plasma, JHEP 05 (2016) 098, [arXiv:1511.07567].arxiv.org/abs/1511.07567
  55. [55]F. Cooper and G. Frye, Single-particle distribution in the hydrodynamic and statistical thermodynamic models of multiparticle production, Phys. Rev. D 10 (Jul, 1974) 186–189.DOI
  56. [56]K. J. Eskola, P. Paakkinen, H. Paukkunen, and C. A. Salgado, EPPS16: Nuclear parton distributions with LHC data, Eur. Phys. J. C 77 (2017), no. 3 163, [arXiv:1612.05741].DOI
  57. [57]S. Dulat, T.-J. Hou, J. Gao, M. Guzzi, J. Huston, P. Nadolsky, J. Pumplin, C. Schmidt, D. Stump, and C. P. Yuan, New parton distribution functions from a global analysis of quantum chromodynamics, Phys. Rev. D 93 (2016), no. 3 033006, [arXiv:1506.07443].DOI
  58. [58]A. Buckley, J. Ferrando, S. Lloyd, K. Nordström, B. Page, M. Rüfenacht, M. Schönherr, and G. Watt, LHAPDF6: parton density access in the LHC precision era, Eur. Phys. J. C 75 (2015) 132, [arXiv:1412.7420].DOI
  59. [59]T. Sjostrand, S. Mrenna, and P. Z. Skands, PYTHIA 6.4 Physics and Manual, JHEP 05 (2006) 026, [hep-ph/0603175].DOI
  60. [60]H. T. Li and I. Vitev, Jet charge modification in dense QCD matter, Phys. Rev. D 101 (2020) 076020, [arXiv:1908.06979].DOI
  61. [61]K. J. Eskola, H. Paukkunen, and C. A. Salgado, EPS09: A New Generation of NLO and LO Nuclear Parton Distribution Functions, JHEP 04 (2009) 065, [arXiv:0902.4154].DOI
  62. [62]M. Cacciari, G. P. Salam, and G. Soyez, The anti-kt jet clustering algorithm, JHEP 04 (2008) 063, [arXiv:0802.1189].DOI
  63. [63]M. Cacciari, G. P. Salam, and G. Soyez, FastJet User Manual, Eur. Phys. J. C 72 (2012) 1896, [arXiv:1111.6097].DOI
  64. [64]R. Kunnawalkam Elayavalli and K. C. Zapp, Medium response in JEWEL and its impact on jet shape observables in heavy ion collisions, JHEP 07 (2017) 141, [arXiv:1707.01539].DOI
  65. [65]M. Dasgupta, L. Magnea, and G. P. Salam, Non-perturbative QCD effects in jets at hadron colliders, JHEP 02 (2008) 055, [arXiv:0712.3014].DOI
  66. [66]M. Dasgupta, F. Dreyer, G. P. Salam, and G. Soyez, Small-radius jets to all orders in QCD, JHEP 04 (2015) 039, [arXiv:1411.5182].DOI
  67. [67]L. Apolinário, J. a. Barata, and G. Milhano, On the breaking of Casimir scaling in jet quenching, Eur. Phys. J. C 80 (2020), no. 6 586, [arXiv:2003.02893].DOI
  68. [68]J. Brewer, J. Thaler, and A. P. Turner, Data-driven quark and gluon jet modification in heavy-ion collisions, Phys. Rev. C 103 (2021), no. 2 L021901, [arXiv:2008.08596].DOI
  69. [69]M. Spousta and B. Cole, Interpreting single jet measurements in Pb + Pb collisions at the LHC, Eur. Phys. J. C 76 (2016), no. 2 50, [arXiv:1504.05169].DOI
  70. [70]A. Ogrodnik, M. Rybář, and M. Spousta, Flavor and path-length dependence of jet quenching from inclusive jet and γ-jet suppression, arXiv:2407.11234.DOI
  71. [71]F. Ringer, B.-W. Xiao, and F. Yuan, Can we observe jet PT-broadening in heavy-ion collisions at the LHC?, Phys. Lett. B 808 (2020) 135634, [arXiv:1907.12541].DOI
  72. [72]D. Pablos and A. Soto-Ontoso, Pushing forward jet substructure measurements in heavy-ion collisions, Phys. Rev. D 107 (2023), no. 9 094003, [arXiv:2210.07901].DOI
  73. [73]ALICE Collaboration, S. Acharya et al., Exploration of jet substructure using iterative declustering in pp and Pb–Pb collisions at LHC energies, Phys. Lett. B 802 (2020) 135227, [arXiv:1905.02512].DOI
  74. [74]CMS Collaboration, A. M. Sirunyan et al., Measurement of quark- and gluon-like jet fractions using jet charge in PbPb and pp collisions at 5.02 TeV, JHEP 07 (2020) 115, [arXiv:2004.00602].DOI
  75. [75]H. A. Andrews et al., Novel tools and observables for jet physics in heavy-ion collisions, J. Phys. G 47 (2020), no. 6 065102, [arXiv:1808.03689].arxiv.org/abs/1808.03689
  76. [76]A. J. Larkoski, S. Marzani, G. Soyez, and J. Thaler, Soft Drop, JHEP 05 (2014) 146, [arXiv:1402.2657].DOI
  77. [77]Y. Mehtar-Tani, A. Soto-Ontoso, and K. Tywoniuk, Dynamical grooming of QCD jets, Phys. Rev. D 101 (2020), no. 3 034004, [arXiv:1911.00375].

Paper details

vlnc:W_9tgkvqqn
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents