Jet Reconstruction in Heavy Ion Collisions

TL;DR

This work assesses jet reconstruction in heavy-ion collisions using jet-area/median background subtraction within FastJet, comparing kt, C/A, anti-kt, and filtered-C/A under RHIC and LHC conditions. It demonstrates that local background estimation and area-based subtraction can recover jets with small offsets and manageable dispersion, with anti-k_t and C/A(filt) performing best overall. Quenching effects are generally modest, though C/A(filt) shows some vulnerability at the LHC; centrality and background fluctuations are accounted for, and unfolding considerations are emphasized. The study also discusses fake-jet issues, contrasting inclusive and exclusive analyses and highlighting the role of origin distributions and dijet topology in estimating fake rates.

Abstract

We examine the problem of jet reconstruction at heavy-ion colliders using jet-area-based background subtraction tools as provided by FastJet. We use Monte Carlo simulations with and without quenching to study the performance of several jet algorithms, including the option of filtering, under conditions corresponding to RHIC and LHC collisions. We find that most standard algorithms perform well, though the anti-kt and filtered Cambridge/Aachen algorithms have clear advantages in terms of the reconstructed transverse-momentum offset and dispersion.

Figures (16)

• Figure 1: Graphical representation of the 4 different background-estimation ranges we shall consider: the Global range, the Strip range ${\cal S}_\Delta(j)$, the Circular range ${\cal C}_\Delta(j)$ and the Doughnut range ${\cal D}_{\delta,\Delta}(j)$. The last three are local ranges with a position depending on the jet being subtracted. See the text for detailed definitions.
• Figure 2: Distribution of the background density $\rho$ per unit area (left) and its intra-event fluctuations $\sigma$ (right). It has been obtained from 5000 HYDJET events with RHIC (AuAu, $\sqrt{s_{NN}} = 200\,\mathrm{GeV}$) and LHC (PbPb, $\sqrt{s_{NN}} = 5.5\,\mathrm{TeV}$) kinematics. The background properties have been estimated using the techniques presented in section \ref{['sec:subtraction']}, using the $k_t$ algorithm with $R_\rho=0.5$, and keeping only the jets with $|y|<1$ (excluding the two hardest).
• Figure 3: Matching efficiency for reconstructed jets as a function of the jet $p_t$. Left: RHIC, right: LHC. These results are independent of the choice of background-subtraction range in the heavy-ion events, since background subtraction does not enter into the matching criterion. Here and in later figures, the label "unquenched" refers to the embedded $pp$ event; the background is always simulated including quenching.
• Figure 4: Effect of the choice of range on the average $p_t$ shift, $\Delta p_t$, as defined in eq. (\ref{['eq:deltapt']}). Left: RHIC, right: LHC. In this figure and those that follow, the yellow band corresponds to 1% of the $p_t$ of the hard jet.
• Figure 5: Distribution of $\Delta p_t$ (red histograms) for each of our 4 jet algorithms, together with a Gaussian (black curve) whose mean (solid vertical line) and dispersion are equal to $\left\langle \Delta p_t \right\rangle$ and $\sigma_{\Delta p_t}$ respectively.