Determination of the Jet Energy Scale at the Collider Detector at Fermilab

TL;DR

The paper presents a comprehensive methodology to determine the jet energy scale at CDF by applying a sequence of data-driven and MC-driven corrections to map calorimeter jet energies to the parent parton energies. It integrates η-dependent flattening, an absolute energy scale derived from MC, and corrections for multiple interactions, out-of-cone energy, and the underlying event, with extensive validation using γ-jet, Z-jet, dijet, and tt̄ hadronic-W samples. The study quantifies systematic uncertainties from single-particle response, fragmentation, calibration stability, and MC modeling, achieving a JES uncertainty of a few percent at high pT and up to several percent at low pT. The results demonstrate good agreement between data and MC across multiple processes and jet cone sizes, confirming the reliability of the JES for precision measurements such as top-quark mass and jet cross sections at the Tevatron. The methodologies and cross-checks provide a framework for jet calibration in hadron collider experiments and highlight the importance of robust MC validation.

Abstract

A precise determination of the energy scale of jets at the Collider Detector at Fermilab at the Tevatron $p\bar{p}$ collider is described. Jets are used in many analyses to estimate the energies of partons resulting from the underlying physics process. Several correction factors are developed to estimate the original parton energy from the observed jet energy in the calorimeter. The jet energy response is compared between data and Monte Carlo simulation for various physics processes, and systematic uncertainties on the jet energy scale are determined. For jets with transverse momenta above 50 GeV the jet energy scale is determined with a 3% systematic uncertainty.

Figures (45)

• Figure 1: Elevation view of one half of the CDF detector displaying the components of the CDF calorimeter: CEM, CHA, WHA, PEM and PHA.
• Figure 2: Mean invariant mass of $Z\to e^+e^-$ candidates, $\langle M(ee) \rangle$, versus run number for events with $86<M(ee)<98$ GeV/$c^2$. Shown are the values for events with both electrons in the central calorimeter (full circles) and for events with one electron in the central and one in the plug calorimeter (open circles). The dashed lines indicate a $\pm 0.3\%$ uncertainty around $91.1$ GeV/$c^2$.
• Figure 3: Mean energy observed in the CHA/WHA for CMUP and CMX muons with $p_T>$20 GeV/$c$ from $W\to\mu\nu_\mu$ candidate events versus run number. The CMUP muons are confined to $|\eta|<0.6$ and thus only sensitive to the central part of the CHA. The CMX muons cover the region $0.6<|\eta|<1.0$ and probe the outer part of the CHA plus the innermost part of the WHA. The dashed lines indicate a 1.5% uncertainty.
• Figure 4: Illustration of the target tower for the electromagnetic (EM) and the hadronic (HAD) sections used in the tuning of charged hadrons. The yellow (lightest shading) region is the target tower. The "x" marks the impact point of the track. The yellow and cyan (light and medium shading) region is the signal region and the dark blue (darkest shading) illustrates the background region. The horizontal axis represents the $\eta$ direction and the vertical represents the $\phi$ direction.
• Figure 5: Fractional energy observed in the central calorimeter as a function of incident particle momenta. The top row shows $\langle E_{CEM}/p \rangle$, $\langle E_{CHA}/p \rangle$ and $\langle (E_{CEM}+E_{CHA})/p \rangle$ for data signal (triangles) and background (histogram) and for single track MC simulation (open circles). The bottom row shows the same distributions for data after background subtraction (full circles) and MC simulation (open circles).