[Paper Review] On the Feasibility and Utility of ISR Tagging

David Krohn, Lisa Randall+1 more|arXiv (Cornell University)|Jan 4, 2011

Particle physics theoretical and experimental studies13 citations

TL;DR

This paper proposes ISR tagging—identifying initial-state radiation jets on an event-by-event basis in hadron collider data—to extract kinematic information about new physics beyond the Standard Model. By exploiting the distinct kinematic properties of ISR jets in supersymmetric di-squark and di-gluino production, the authors achieve ~40% tagging efficiency with ~10% mistag rate, enabling mass reconstruction of squarks and gluinos within 20% of their true values using a novel kinematic method.

ABSTRACT

The production of new particles at a hadron collider like the LHC is always accompanied by QCD radiation attributable to the initial state (i.e. ISR). This tends to complicate analyses, so ISR is normally regarded as a nuisance. Nevertheless, we show that ISR can also be valuable, yielding information that can help in the discovery and interpretation of physics beyond the Standard Model. To access this information we will introduce new techniques designed to identify ISR jets on an event-by-event basis, a process we term ISR tagging. As a demonstration of their utility, we will apply these techniques to SUSY di-squark (di-gluino) production to show that they can be used to identify ISR jets in roughly 40% (15%) of the events, with a mistag rate of around 10% (15%). We then show that, through the application of a new method which we will introduce, knowledge of an ISR jet allows us to infer the squark (gluino) mass to within roughly 20% of its true value.

Motivation & Objective

• To demonstrate that initial-state radiation (ISR) jets, typically treated as a nuisance, can be systematically tagged and used to extract new physics information.
• To develop event-by-event ISR jet tagging techniques applicable to pair-produced BSM particles decaying into jets and invisible particles.
• To enable precise mass measurements of new physics particles (e.g., squarks, gluinos) using kinematic correlations with ISR jets, independent of the full SUSY spectrum.
• To show that ISR tagging can be implemented with low mistag rates and high efficiency even in challenging scenarios with near-degenerate LSP masses.

Proposed method

• Identify ISR jets by selecting the hardest jet among the Nf+1 leading jets in an event, where Nf is the number of final-state radiation (FSR) jets.
• Apply three tagging criteria: (1) smallest transverse momentum among the Nf+1 hardest jets; (2) largest angular separation from the FSR jet system; (3) smallest invariant mass when combined with the FSR jet system.
• Use a ranking system for multiple jets passing criteria, prioritizing those satisfying Eq. (1) over (2) or (3).
• Apply a new kinematic method based on the average sign of the FSR momentum projection along the transverse ISR direction to infer the BSM particle mass.
• Simulate events at 14 TeV LHC using Pythia 6.422 with MLM matching, cluster jets using anti-kT algorithm with R=0.7 (di-squark) and R=0.4 (di-gluino), and analyze tagging performance in 0.1×0.1 η-φ cells.
• Use Monte Carlo samples to compute tagging efficiencies and mistag rates across different SUSY spectra, including degenerate LSP cases.

Experimental results

Research questions

• RQ1 Can ISR jets be reliably identified on an event-by-event basis in BSM processes involving pair-produced particles decaying into jets and invisible particles?
• RQ2 What kinematic properties distinguish ISR jets from FSR jets in high-multiplicity final states at the LHC?
• RQ3 To what extent can knowledge of the ISR jet enable accurate reconstruction of the mass of the produced BSM particle, independent of other model parameters?
• RQ4 How robust are ISR tagging techniques in scenarios with near-degenerate LSP masses, which complicate standard mass measurements?

Key findings

• ISR tagging achieves a tagging efficiency of approximately 40% for di-squark production and 15% for di-gluino production, even in the presence of near-degenerate LSP masses.
• The mistag rate remains low at around 10% for di-squark and 15% for di-gluino production, indicating high reliability of the tagging method.
• The proposed kinematic method allows reconstruction of the squark or gluino mass to within roughly 20% of its true value, without assumptions on the rest of the SUSY spectrum.
• The method successfully identifies the BSM mass scale as the point where the average sign of FSR momentum projection along the transverse ISR direction crosses zero, corresponding to a balanced boost.
• Tagging performance is stable across different SUSY spectra, including cases with LSP masses close to the produced particle mass, demonstrating robustness.
• The results show that ISR, traditionally a source of background, can be repurposed as a powerful tool for new physics discovery and precision measurement.