[Paper Review] SUSY Without Prejudice at the 7 and 8 TeV LHC: Gravitino LSPs

M. Cahill-Rowley, JoAnne L. Hewett+2 more|arXiv (Cornell University)|Nov 29, 2012

Particle physics theoretical and experimental studies4 references11 citations

TL;DR

This study evaluates the LHC's sensitivity to the phenomenological MSSM (pMSSM) with gravitino LSPs at 7 and 8 TeV, using ATLAS SUSY search analyses. Despite reduced missing transverse energy (MET) in gravitino models, long-lived NLSPs significantly enhance model coverage, making the LHC more effective at excluding pMSSM parameter space than in neutralino LSP scenarios.

ABSTRACT

We have examined the capability of the LHC, running at both 7 and 8 TeV, to explore the 19(20)-dimensional parameter space of the pMSSM with neutralino(gravitino) LSPs and soft masses up to 4 TeV employing the ATLAS SUSY analysis suite. Here we present some preliminary results for the gravitino model set, following the ATLAS analyses whose data were publically available as of mid-September 2012. We find that the impact of the reduced MET, resulting from models with gravitino LSPs on sparticle searches is more than off-set by the detectability of the many possible long-lived NLSPs.

Motivation & Objective

• To assess the LHC's ability to probe the 20-dimensional pMSSM parameter space with gravitino LSPs using ATLAS SUSY search results.
• To compare model exclusion coverage between gravitino LSPs and neutralino LSPs in the pMSSM framework.
• To evaluate the impact of the 126 GeV Higgs boson mass constraint on model viability across both LSP scenarios.
• To investigate the role of long-lived NLSPs in enhancing exclusion reach despite reduced MET.
• To quantify the effectiveness of various ATLAS search channels—especially MET, multilepton, heavy flavor, and long-lived particle searches—on the gravitino model set.

Proposed method

• Employed a modified version of the PGS Monte Carlo tool to simulate ATLAS SUSY search signatures using public ATLAS background estimates and benchmark models.
• Reproduced key ATLAS analyses including 2-6 jets, multijet, 1-lepton, SS dilepton, heavy flavor, and multilepton searches at both 7 and 8 TeV.
• Generated a 20-parameter pMSSM model set with gravitino LSPs, incorporating cosmological constraints and fine-tuning considerations.
• Applied LHC search constraints sequentially, including MET-based, long-lived particle, HSCP, disappearing tracks, and rare decay searches.
• Used the Higgs mass constraint $m_h = 126 \pm 3$ GeV to define a subset of models for comparative analysis.
• Compared model exclusion fractions across search types and center-of-mass energies, quantifying coverage differences between gravitino and neutralino LSP cases.

Experimental results

Research questions

• RQ1 How does the exclusion power of the LHC at 7 and 8 TeV compare between pMSSM models with gravitino LSPs and those with neutralino LSPs?
• RQ2 To what extent do long-lived NLSPs in the gravitino LSP scenario enhance model coverage despite reduced MET?
• RQ3 How does the 126 GeV Higgs mass constraint affect the viability of pMSSM models with gravitino LSPs?
• RQ4 Which ATLAS search channels—MET-based, multilepton, heavy flavor, or long-lived particle searches—are most effective for gravitino LSP models?
• RQ5 What is the impact of center-of-mass energy (7 vs. 8 TeV) on model exclusion, particularly for compressed spectra?

Key findings

• At 8 TeV, 28.69% of the full gravitino LSP model set was excluded by 'vanilla' MET-based ATLAS searches, compared to 19.44% at 7 TeV.
• The 8 TeV multijet and SS dilepton searches were significantly more effective for gravitino models than for neutralino models, with 4.13% and 11.50% exclusions, respectively.
• Long-lived sparticle searches, particularly for HSCP and disappearing tracks, contributed substantially to model coverage, more than offsetting reduced MET in gravitino models.
• The multilepton and heavy flavor searches were more effective for gravitino LSPs due to higher frequency of light bino NLSPs and lighter third-generation squarks.
• The Higgs mass constraint $m_h = 126 \pm 3$ GeV reduced model coverage by only ~1.5 percentage points on average, with less degradation in the gravitino set than in the neutralino case.
• Approximately 1.3 thousand models were excluded by 7 TeV searches but remained viable at 8 TeV, indicating energy-dependent exclusion sensitivity.