Last week CERN hosted a four day workshop addressing the interface of "Jet quenching studies in theory and experiment".
Rosi Reed presented an overview of the jet measurements in ALICE. During the second day Marta Verweij discussed the methods used in jet measurements with ALICE.
In particle physics, a jet is a collimated beam of particles that have all fragmented from the same parton, where a parton can be either a quark or a gluon. By gathering all, or a significant fraction of the parton’s energy, we can determine what the kinematics of the original parton were even though we cannot directly measure it. Experimentally the measured particles are grouped together by a jet algorithm. The kinematic properties of the measured jets are correlated to the kinematic properties of the original hard parton.
In Pb-Pb collisions we use jets as a probe of the quark gluon plasma (QGP). The partons that eventually fragment into jets are created in the very early stages of the collision, prior to formation of the QGP. They then traverse the medium and interact with the constituents, quarks and gluons, of the medium. Due to this interaction with the QGP, the partons lose energy by radiating gluons, which is defined as “jet quenching”. It is expected that the structure of the spray of hadrons is modified with respect to the vacuum case due to jet quenching. This modification will be reflected in the resulting jet energy spectrum and jet shape. By looking at a number of different observables, we can test the available models of energy loss within the QGP and better understand the exact properties of the QGP.
An event display of a jet within the ALICE EMCal acceptance, with EMCal towers containing energy shown in blue.
In ALICE we measure charged jets, formed from charged tracks measured in the Time Projection Chamber (TPC) and the Inner Tracking System (ITS), and full jets, formed both from tracks and clusters within the Electromagnetic Calorimeter (EMCal). The EMCal clusters allow us to measure the neutral energy component of the jet. These clusters are corrected for hadronic contamination to avoid energy double counting as charged particles will also deposit some energy in the EMCal. The ALICE tracking and calorimeter system allow us to make a very low constituent cut on the tracks and clusters, which allows us to gather a larger fraction of the energy of the initial parton.
Here we will discuss a small fraction of the jet measurements that are being done in ALICE. The first of these measurements is the full jet cross-section measurement from the 2.76 TeV p-p data taken early last year (see Fig.1) . In order to understand how jets are modified in Pb-Pb collisions, it is important to understand the baseline measurements. For all measurements, the technique of unfolding is used to correct the raw measured spectrum for detector (and in the case of Pb-Pb collisions background) effects. In unfolding, simulation is used to quantify all the detector effects on a jet-by-jet basis which is then put into a single response matrix. This matrix is then inverted in order to determine the “true” spectrum from the measured spectrum. For the baseline pp measurement, a bin-by-bin unfolding technique was used where each bin in jet pT was corrected assuming that they are independent of one another. The paper for this measurement can be found at arXiv:1301.3475.
Fig.1: Full jet cross-section measurement at 2.76 TeV.
Measuring jets in Pb-Pb collisions has the added difficulty of dealing with the underlying event. There is an overall increase in the measured jet energy since the jet finder will cluster the background particles in addition to those that have come from the initial parton. The challenge is to separate the particles originating from the hard parton from the soft underlying event.
The background energy density of the underlying event is determined on an event-by-event basis by clustering the event with the kT jet finder, which is dominated by the soft physics rather than the hard physics, and using the median density of this collection as the energy density so that outliers do not play a large role.
In addition, the background itself has fluctuations, which means that we cannot know precisely how much energy should be removed from a given jet. The size of the fluctuations are determined by randomly throwing well defined probes in the event and looking how much the background fluctuates at the position of the probe. By repeating this many times, the response of jets to the fluctuating underlying event is obtained. The background subtracted jet spectrum is then corrected for both the detector effects and these fluctuations in unfolding. The ALICE paper detailing our background determination can be found at DOI: 10.1007/JHEP03(2012)053.
Fig. 2: Spectra from four different centrality ranges. This type of measurements allow us to look at the evolution of the spectra versus centrality
We have applied this technique to charged jets in ALICE from Pb-Pb collisions from 2010 and 2011.
The spectra shown in Figure 2 are from four different centrality ranges, which allow us to look at the evolution of the spectra versus centrality. If the centrality percentile is lower, the medium density is higher and therefore jet quenching effects are expected to be stronger. It can be observed that for the most central collisions, 0-10%, the jet yield is lower than in peripheral collisions, 50-80%: Jet Suppression. Even though these spectra show significant suppression, at least at the jet sizes used in this analysis, the fragmentation of the jets seems to be very similar to vacuum fragmentation as we can see from the ratio of the spectra of R = 0.2 to R = 0.3 jets as shown in Figure 3.
Fig.3: Ratio of the spectra of R = 0.2 to R = 0.3 jets
The techniques outlined in the paragraphs above were applied to a full jet analysis in Pb-Pb collisions. Since the fluctuations were larger, in this case we applied a bias to the jet spectrum of requiring that the leading track have at least 5 GeV/c of momentum. This removes a lot of the combinatorial jets created from the background, which helps with unfolding stability. It can introduce a bias in the fragmentation, but in the range that we show in Figure 4, the difference is minimal. In this figure we show the nuclear modification factor (RAA) for the most central events from the data measured in 2011. Both the pp reference and the Pb-Pb measurement have the leading track bias. Here we can see that jets are suppressed, climbing to a maximum RAA of about 0.5, which is very similar to what has been seen in charged hadrons, charged jets and by the other experiments of the LHC.
Fig.4:"Fig. 4: Nuclear modification factor for R = 0.2 full jets in 0-10% central events."