Jet Reconstruction Research Articles

Studies on a hadronic calorimeter with MPGD technology for a future Muon Collider experiment

In the context of the European strategy for particle physics, the Multi-TeV Muon Collider has emerged as a compelling alternative for advancing our understanding of the Standard Model, after the full exploitation of the High-Luminosity LHC. The physics programme at the Muon Collider includes precise measurements in the Higgs boson sector and the search for new physics at the TeV scale. Achieving these goals relies on accurate full event reconstruction, including the identification and precise four-momentum estimation of various particles. The Particle Flow (PF) algorithm is one of the most suitable approach for this task, exploiting information from tracking, calorimeter, and muon detectors for particle identification and measurements of their momenta/energies. Tracking detectors measure charged particle momenta, while calorimeters provide energy measurements for photons and neutral hadrons. Therefore, a combination of an exceptional tracking system and high-granularity calorimeters is necessary. However, one of the biggest challenges for a future experiment at the Muon Collider is to discriminate the product of the μμ collisions from the intense beam-induced-background (BIB), due to the unstable nature of muons, whose decay products interact with the detector material. To address this, an innovative hadronic calorimeter (HCAL) based on Micro Pattern Gas Detectors (MPGDs) is proposed. MPGDs offer robust technology for high radiation environments and a high granularity for precise spatial measurements. Dedicated studies are needed to assess and optimize the performance of an MPGD-based HCAL, including the development of medium-scale prototypes for performance measurements. The response of HCAL to incoming particles is examined through Monte Carlo simulations using Geant4, comparing the performance of digital and semi-digital readouts, with energy resolution as the figure of merit. The simulated geometry will be integrated into the Muon Collider software to study its impact on jet reconstruction within the full apparatus and in the presence of BIB. The simulation will be also validated through the test of a small-size calorimeter cell equipped with advanced resistive MPGD technologies, namely resistive MicroMegas, resistive μRWELL and RP-WELL.

Reconstructing jets in the Phase-2 upgrade of the CMS Level-1 Trigger with a seeded cone algorithm

The Phase-2 Upgrade of the CMS Level-1 Trigger (L1T) will reconstruct particles using the Particle Flow algorithm, connecting information from the tracker, muon, and calorimeter detectors, and enabling fine-grained reconstruction of high level physics objects like jets. We have developed a jet reconstruction algorithm using a cone centred on an energetic seed from these Particle Flow candidates. The implementation is designed to find up to 16 jets in each Xilinx Ultrascale+ FPGA, with a latency of less than 1 µs, and event throughput of 6.7 MHz to fit within the L1T system constraints. Pipelined processing enables reconstruction of jet collections with different cone sizes for little additional resource cost. The design of the algorithm also provides a platform for additional computation using the jet constituents, such as jet tagging using neural networks. We will describe the implementation, its jet reconstruction performance, computational metrics, and the developments towards jet tagging.

Fast b-tagging at the high-level trigger of the ATLAS experiment in LHC Run 3

The ATLAS experiment relies on real-time hadronic jet reconstruction and b-tagging to record fully hadronic events containing b-jets. These algorithms require track reconstruction, which is computationally expensive and could overwhelm the high-level-trigger farm, even at the reduced event rate that passes the ATLAS first stage hardware-based trigger. In LHC Run 3, ATLAS has mitigated these computational demands by introducing a fast neural-network-based b-tagger, which acts as a low-precision filter using input from hadronic jets and tracks. It runs after a hardware trigger and before the remaining high-level-trigger reconstruction. This design relies on the negligible cost of neural-network inference as compared to track reconstruction, and the cost reduction from limiting tracking to specific regions of the detector. In the case of Standard Model HH → bb̅bb̅, a key signature relying on b-jet triggers, the filter lowers the input rate to the remaining high-level trigger by a factor of five at the small cost of reducing the overall signal efficiency by roughly 2%.

Machine Learning for Particle Flow Reconstruction at CMS

We provide details on the implementation of a machine-learning based particle flow algorithm for CMS. The standard particle flow algorithm reconstructs stable particles based on calorimeter clusters and tracks to provide a global event reconstruction that exploits the combined information of multiple detector subsystems, leading to strong improvements for quantities such as jets and missing transverse energy. We have studied a possible evolution of particle flow towards heterogeneous computing platforms such as GPUs using a graph neural network. The machine-learned PF model reconstructs particle candidates based on the full list of tracks and calorimeter clusters in the event. For validation, we determine the physics performance directly in the CMS software framework when the proposed algorithm is interfaced with the offline reconstruction of jets and missing transverse energy. We also report the computational performance of the algorithm, which scales approximately linearly in runtime and memory usage with the input size.

Measurement of $$\mathcal{B}r(H \to Z\gamma )$$ at the 250 GeV ILC

The $e^+e^- \to HZ$ process with the subsequent decay of the Higgs boson $H \to Z\gamma$ is studied, where both $Z$ bosons are reconstructed in the final states with two jets. The analysis is performed using Monte Carlo data samples obtained with detailed ILD (The International Large Detector) detector simulation assuming an integrated luminosity of 2 ab$^{-1}$, beam polarizations of ${\cal{P}}_{e^-e^+} = (-0.8, +0.3)$, and center-of-mass energy of $\sqrt{s}$ = 250 GeV. The analysis is repeated for the case of two 0.9 ab$^{-1}$ data samples with polarizations ${\cal{P}}_{e^-e^+} = (\mp0.8, \pm0.3)$. Contributions of the potential background processes are studied using all available ILD MC event samples. The largest background comes from the $e^+e^- \to W^+W^-$ process supplemented by an energetic photon produced by initial state radiation. To suppress this background we require that at least one of the $Z$ bosons decays to $b$-jets. To reduce the jet reconstruction uncertainties the $M_{\Delta} = M(jj\gamma) - M(jj) + M(Z_{\rm nom})$ variable is used, where $M(Z_{\rm nom})$ = 91.2 GeV. The $M_{\Delta}$ distributions are obtained for the studied signal and backgrounds to estimate the expected accuracy of the ${\cal B}r(H \to Z\gamma)$ measurement. The accuracy is 22$\%$ for the option of the single polarization sample described above and deteriorate to 24$\%$ in case of the sample with two polarizations. The proposed method can be applied at any future $e^+e^-$ collider.

Design and simulation of a MPGD-based hadronic calorimeter for Muon Collider

The project of a Multi-TeV Muon Collider represents a unique opportunity to explore the high energy physics frontier and to measure with high precision the Higgs coupling with the other particles of the Standard Model as well as the trilinear and quadrilinear Higgs self-coupling, leading to a precise determination of the Higgs potential, in order to confirm the theoretical predictions of the SM and possibly to find evidences for new physics. One of the major challenges for the design and optimization of the technologies suitable for a Muon Collider experiment is represented by the high background induced by the decay of the muons coming from the beam. This contribution present the design of an innovative MPGD-based hadronic calorimeter (HCAL). The detector consists of a sampling calorimeter exploiting the Micro Pattern Gas Detectors (MPGDs) as active layers: MPGDs offer a fast and robust technology for high radiation environments and a high granularity for precise spatial measurements. Moreover, the detector is designed to optimize the jet reconstruction and for background suppression. The calorimeter is simulated using the Geant4 toolkit to support the detector R&D. The detector design and layout optimization supported by the simulation is described.

Performance studies of single-particle uncertainties and local cluster weighting calibration for Particle-Flow jets with the ATLAS detector at [formula omitted] = 13 TeV

In the ATLAS Experiment at the CERN Large Hadron Collider, jets play a central role in many physics analyses. In the last years, the standard jet reconstruction method shifted away from jets built using only calorimeter information towards jets that additionally leverage tracking information, so-called Particle-Flow jets. In this note, two performance studies will be discussed. They evaluate the application of the local cluster weighting calibration scheme to the inputs of Particle-Flow jets and the use of single-particle uncertainties with Particle-Flow jets.

R&D of a Novel High Granularity Crystal Electromagnetic Calorimeter

Future electron-positron collider experiments aim at the precise measurement of the Higgs boson, electroweak physics and the top quark. Based on the particle-flow paradigm, a novel highly granular crystal electromagnetic calorimeter (ECAL) is proposed to address major challenges from jet reconstruction and to achieve the optimal EM energy resolution of around 2–3%/E(GeV) with the homogeneous structure. Extensive R&D efforts have been carried out to evaluate the requirements and potentials of the crystal calorimeter concept from sensitive detection units to a full sub-detector system. The requirements on crystal candidates, photon sensors as well as readout electronics are parameterized and quantified in Geant4 full simulation. Experiments including characterizations of crystals and silicon photomultipliers (SiPMs) are performed to validate and improve the simulation results. The physics performance of the crystal ECAL is been studied with the particle flow algorithm “ArborPFA” which is also being optimized. Furthermore, a small-scale detector module with a crystal matrix and SiPM arrays is under development for future beam tests to study the performance for EM showers.

Crilin: A CRystal calorImeter with Longitudinal InformatioN for a future Muon Collider

The measurement of physics processes at new energy frontier experiments requires excellent spatial, time, and energy resolutions to resolve the structure of collimated high-energy jets. In a future Muon Collider, the beam-induced backgrounds (BIB) represent the main challenge in the design of the detectors and of the event reconstruction algorithms. The technology and the design of the calorimeters should be chosen to reduce the effect of the BIB, while keeping good physics performance. Several requirements can be inferred: i) high granularity to reduce the overlap of BIB particles in the same calorimeter cell; ii) excellent timing (of the order of 100 ps) to reduce the out-of-time component of the BIB; iii) longitudinal segmentation to distinguish the signal showers from the fake showers produced by the BIB; iv) good energy resolution (less than 10%/√E) to obtain good physics performance, as has been already demonstrated for conceptual particle flow calorimeters. Our proposal consists of a semi-homogeneous electromagnetic calorimeter based on lead fluoride crystals (PbF2) read out by surface-mount UV-extended Silicon Photomultipliers (SiPMs): the Crilin calorimeter. In this paper, the performance of the Crilin calorimeter in the Muon Collider framework for hadron jets reconstruction has been analyzed. We report the characterisation of individual components together with the development of a small-scale prototype, consisting of 2 layers of 3 × 3 crystals each.

End-to-end multi-particle reconstruction in high occupancy imaging calorimeters with graph neural networks

We present an end-to-end reconstruction algorithm to build particle candidates from detector hits in next-generation granular calorimeters similar to that foreseen for the high-luminosity upgrade of the CMS detector. The algorithm exploits a distance-weighted graph neural network, trained with object condensation, a graph segmentation technique. Through a single-shot approach, the reconstruction task is paired with energy regression. We describe the reconstruction performance in terms of efficiency as well as in terms of energy resolution. In addition, we show the jet reconstruction performance of our method and discuss its inference computational cost. To our knowledge, this work is the first-ever example of single-shot calorimetric reconstruction of {mathcal {O}}(1000) particles in high-luminosity conditions with 200 pileup.

Quantum clustering and jet reconstruction at the LHC

Clustering is one of the most frequent problems in many domains, in particular, in particle physics where jet reconstruction is central in experimental analyses. Jet clustering at the CERN's Large Hadron Collider (LHC) is computationally expensive and the difficulty of this task will increase with the upcoming High-Luminosity LHC (HL-LHC). In this paper, we study the case in which quantum computing algorithms might improve jet clustering by considering two novel quantum algorithms which may speed up the classical jet clustering algorithms. The first one is a quantum subroutine to compute a Minkowski-based distance between two data points, whereas the second one consists of a quantum circuit to track the maximum into a list of unsorted data. The latter algorithm could be of value beyond particle physics, for instance in statistics. When one or both of these algorithms are implemented into the classical versions of well-known clustering algorithms (K-means, Affinity Propagation and $k_T$-jet) we obtain efficiencies comparable to those of their classical counterparts. Even more, exponential speed-up could be achieved, in the first two algorithms, in data dimensionality and data length when the distance algorithm or the maximum searching algorithm are applied.

Studying hadronization at LHCb

LHCb is a fully instrumented forward spectrometer with particle identification and muon reconstruction covering the pseudorapidity (\etaη) range from 2 to 5. Its full jet reconstruction capability makes the LHCb experiment a suitable venue to explore jet substructure observables, particularly formation of hadrons and heavy quarkonia resonances inside jets. This contribution presents a brief overview of ongoing research program and discusses recent results on non-identified charged hadron distributions in Z-tagged jets and charmonium distributions within jets. Future works towards furthering the knowledge of hadronizion are also discussed.

Studying the mechanisms for strange particle production with ALICE at LHC

The main goal of the ALICE experiment is to study the physics of strongly interacting matter, focusing on the properties of the quark-gluon plasma (QGP). The relative production of strange hadrons with respect to non-strange hadrons in heavy-ion collisions was historically considered as one of the signatures of QGP formation. However, the latest results in proton-proton (pp) and proton-lead (p-Pb) collisions have revealed an increasing trend in the yield ratio of strange hadrons to pions with the charged-particle multiplicity in the event, showing a smooth evolution across different collision systems and energies. We present the new studies which are performed with the aim of better understanding the production mechanisms for strange particles and hence the strangeness enhancement phenomenon in small collision systems. In one of the recent studies, the very forward energy transported by beam remnants (spectators) and detected by the Zero Degree Calorimeters (ZDC) is used to classify events. The contribution of the effective energy and the particle multiplicity on strangeness production is studied using a multi-differential approach in order to disentangle initial and final state effects. In the second study, the origin of strangeness enhancement with multiplicity in pp has been further investigated by separating the contribution of soft and hard processes, such as jets, to strange hadron production. Techniques involving full jet reconstruction or two-particle correlations have been exploited. The results indicate that the increased relative strangeness production emerges from the growth of the underlying event, being disconnected from initial state properties, and suggest that soft (out-of-jets) processes are the dominant contribution to strange hadron production.

Searching for New Physics in Hadronic Final States with Run 2 Proton–Proton Collision Data at the LHC

The symmetries of the Standard Model give rise to the forces that act on particles, and the corresponding force mediators. While the Standard Model is an excellent description of particle interactions, it has known limitations; it is therefore important to search for new physics beyond the Standard Model, potentially indicating as-of-yet unknown symmetries of nature. The ATLAS and CMS collaborations have detailed physics programmes, involving a large number of searches for new physics in hadronic final states. As the start of Run 3 of the LHC is imminent, now is a good time to review the progress made and the status of hadronic searches during Run 2 at a centre-of-mass collision energy of s=13TeV. This review provides an overview of the motivations and challenges of hadronic final states at the LHC, followed by an introduction to jet reconstruction, calibration, and tagging. Three classes of searches for new physics in hadronic final states are discussed: di-jet searches, searches for missing transverse momentum in association with another object, and searches for hadronic di-boson resonances. The complementarity of these different analysis strategies is discussed, emphasising the importance of a varied hadronic physics programme in the search for new physics.

Particle flow with a hybrid segmented crystal and fiber dual-readout calorimeter

In the reconstruction of physics events at future e+e- colliders the calorimeter design has a crucial role in the overall detector performance. The reconstruction of events with many jets in their final state sets stringent requirements on the jet energy and angular resolutions. The energy resolution for jets with energy of about 45 GeV is required to be at the 4–5% level to enable an efficient separation of the W and Z boson invariant masses. We demonstrate in this paper how such a performance can be achieved by exploiting a particle flow algorithm tailored for a hybrid dual-readout calorimeter made of segmented crystals and fibers. The excellent energy resolution and linearity of such calorimeter for both photons and neutral hadrons (3%/√E and 26%/√E, respectively), inherent to the homogeneous crystals and dual-readout technological choices, provides a powerful handle for the development of a new approach for particle identification and jet reconstruction. While the dual-readout particle flow algorithm (DR-PFA) presented in this paper is at its early stage of development, it already demonstrates the potential of a hybrid dual-readout calorimeter for jet reconstruction by improving the jet energy resolution with respect to a calorimeter-only reconstruction from 6.0% to about 4.5% for 45 GeV jets.

Lightweight jet reconstruction and identification as an object detection task

We apply object detection techniques based on deep convolutional blocks to end-to-end jet identification and reconstruction tasks encountered at the CERN large hadron collider (LHC). Collision events produced at the LHC and represented as an image composed of calorimeter and tracker cells are given as an input to a Single Shot Detection network. The algorithm, named PFJet-SSD performs simultaneous localization, classification and regression tasks to cluster jets and reconstruct their features. This all-in-one single feed-forward pass gives advantages in terms of execution time and an improved accuracy w.r.t. traditional rule-based methods. A further gain is obtained from network slimming, homogeneous quantization, and optimized runtime for meeting memory and latency constraints of a typical real-time processing environment. We experiment with 8-bit and ternary quantization, benchmarking their accuracy and inference latency against a single-precision floating-point. We show that the ternary network closely matches the performance of its full-precision equivalent and outperforms the state-of-the-art rule-based algorithm. Finally, we report the inference latency on different hardware platforms and discuss future applications.

Theory Input for \(t\bar {t}j\) Experimental Analyses at the LHC

The precise measurement of the top quark mass, which is a fundamental SM parameter, constitutes one of the main goals of the LHC top physics program. One approach to measure this quantity uses the $\rho_\mathrm{s}$ distribution, an observable depending on the invariant mass of the $t\bar{t}j$ system. To fully exploit the experimental accuracy achievable in measuring top quark production cross sections at the LHC, the theory uncertainties associated to these measurements need to be well under control. To this end we present a study of the effect of varying the theoretical input parameters in the calculation of differential cross sections of the $t\bar{t}j$ process. Thereby we studied the influence of the jet reconstruction procedure, as well as the effect of various renormalization and factorization scale definitions and different PDF sets. The variation of the $R$ parameter in the jet reconstruction algorithm was found to have negligible influence on the scale variation uncertainty. A strong reduction of scale uncertainties and a better behaviour of the NLO/LO ratios using selected dynamical scales instead of a static one in the high energy tails of differential distributions was observed. This is particularly interesting in the context of the top quark mass measurements through the $\rho_\mathrm{s}$ distribution, in which the perturbative stability can be improved by applying the proposed dynamical scale definition.

Higgs and top physics reconstruction challenges and opportunities at FCC-ee

The Higgs bosons and the top quark decay into rich and diverse final states, containing both light and heavy quarks, gluons, photons as well as W and Z bosons. This article reviews the challenges involved in reconstructing Higgs and top events at the FCC-ee and identifies the areas where novel developments are needed. The precise identification and reconstruction of these final states at the FCC-ee rely on the capability of the detector to provide excellent flavour tagging, jet energy and angular resolution, and global kinematic event reconstruction. Excellent flavour tagging performance requires low-material vertex and tracking detectors, and advanced machine learning techniques as successfully employed in LHC experiments. In addition, the Z pole run will provide abundant samples of heavy flavour partons that can be used for calibration of the tagging algorithms. For the reconstruction of jets, leptons, and missing energy, particle-flow algorithms are crucial to explore the full potential of the highly granular tracking and calorimeter systems, and give access to excellent energy–momentum resolution and precise identification of heavy bosons in their hadronic decays. This enables, among many other key elements, the reconstruction of Higgsstrahlung processes with leptonically and hadronically decaying Z bosons, and an almost background-free identification of top quark pair events. Exploiting the full available kinematic constraints together with exclusive jet clustering algorithms will allow for the optimisation of global event reconstruction with kinematic fitting techniques.

Jet performance at the circular electron-positron collider

Jet reconstruction is critical for the precision measurement of Higgs boson properties and the electroweak observables at the CEPC. We analyze the jet energy and angular responses of benchmark 2- and 4-jet processes with fully simulated samples with the CEPC baseline detector geometry. We observe a relative resolution of 3.5% and 1% on the jet energy and angular measurement for jets in the detector barrel region (|cos θ| < 0.6) with energy greater than 60 GeV. Meanwhile, the jet energy/angular scale can be controlled within 0.5/0.01%. The differential dependences of the jet response on the jet direction and energy are extracted. We also analyze the impact on the jet responses induced by different jet clustering algorithms and matching criteria, which yields a relative difference of up to 8%.

Boosted Higgs boson jet reconstruction via a graph neural network

By representing each collider event as a point cloud, we adopt the graphic convolutional network (GCN) with focal loss to reconstruct the Higgs jet in it. This method provides higher Higgs tagging efficiency and better reconstruction accuracy than the traditional methods, which use jet substructure information. The GCN, which is trained on events of the $H+\text{jets}$ process, is capable of detecting a Higgs jet in events of several different processes, even though the performance degrades when there are boosted heavy particles other than the Higgs boson in the event. We also demonstrate the signal and background discrimination capacity of the GCN by applying it to the $t\overline{t}$ process. Taking the outputs of the network as new features to complement the traditional jet substructure variables, the $t\overline{t}$ events can be separated further from the $H+\text{jets}$ events.