The ATLAS Search for Supersymmetry and its Connection

The ATLAS Search for Supersymmetry and its Connection to Dark Matter PPC 2008 2 ND INTERNATIONAL WORKSHOP ON THE INTERCONNECTION BETWEEN PARTICLE PHYSICS AND COSMOLOGY http: //www. icepp. s. u-tokyo. ac. jp/~kenta/analysis. html http: //home. slac. stanford. edu/pressreleases/20060821. htm Sven Vahsen (Lawrence Berkeley Lab) for the ATLAS Collaboration PPC 2008, Albuquerque, New Mexico Sven Vahsen

Is the Dark Matter Supersymmetric? http: //home. slac. stanford. edu/pressreleases/20060821. htm • What can we learn from the LHC? PPC 2008, Albuquerque, New Mexico Sven Vahsen ~0 ? 1 2

• If new physics discovered at LHC — WIMP DM candidate? — What fraction of DM? • Try to measure enough to compute — Mass m. WIMP — Cross-section σWIMP-nucleon — Relic density W WIMP • Compare with — Astro Particle Physics (m, σ) — Experimental Cosmology W WIMP = W DM ? Cross-section (cm 2) How the LHC can help ZEPLIN-2 WIMP Mass [Ge. V/c 2] Potential for major impact on our understanding of Dark Matter! Will discuss ATLAS prospects if new physics is SUSY PPC 2008, Albuquerque, New Mexico Sven Vahsen 3

The Standard Model of Particle Physics • Successful theory of fundamental interactions since early 1970 s • Survived numerous experimental tests • Only Higgs missing • LHC built to look for Higgs and Physics beyond the Standard Model… H PPC 2008, Albuquerque, New Mexico Sven Vahsen 4

Supersymmetry (SUSY) g ~ G G • Well motivated extension of the Standard Model • Standard Model particles have supersymmetric partners — Differ by 1/2 unit in spin ~ ~ • EW-gaugino+higgsino mixing 2 charginos +/-1, 2 4 neutralinos 01, 2, 3, 4 PPC 2008, Albuquerque, New Mexico Sven Vahsen 5

Why We Like SUSY • Keeps corrections to Higgs mass small. Requires wino and stop masses ~ few hundred Ge. V • Unifies gauge couplings at large Q 2. Requires sparticle masses ~ few hundred Ge. V • Can provide plausible WIMP Dark matter candidates. Cosmological arguments prefer WIMP mass ~ hundred Ge. V • Highest mass limits from Tevatron ~ 300 & 400 Ge. V (gluinos, squarks) without SMSUSY With SUSY with SUSY Energy [Ge. V] PPC 2008, Albuquerque, New Mexico Sven Vahsen W. de Boer, C. Sander Phys. Lett. B (2004) 6

R-Parity, Stable LSP • Consider only R-parity conserving SUSY models • Even number of SUSY particles for every vertex — SUSY particles always produced in pairs — Get decay chains as below — Lightest SUSY particle (LSP) cannot decay, hence potential WIMP Dark Matter candidate ~0 1 q q~R g~ p_ q q PPC 2008, Albuquerque, New Mexico ~ g ~ q. L p q Sven Vahsen ~ 02 ~ 01 ~ l l l 7

m. SUGRA and dark matter • m. SUGRA — SUSY masses unify at GUT-scale m 0, m 1/2 — tanβ, A 0, sign(μ) — Neutralino LSP • Four regions with W NEUTRALINO W DM due to enhanced annihilation in early universe Pseudo-projection – no units! PPC 2008, Albuquerque, New Mexico Sven Vahsen 8

• to start operation this year The CERN Beschleuniger Komplex • design: 7 -Te. V proton-on-proton p √s=14 Te. V p • 5 -Te. V this year (see talk by Ulrich Parzefall) LHC CMS LHCB SPS ALICE PPC 2008, Albuquerque, New Mexico ATLAS CERN (Meyrin) Rüdiger Schmidt Bullay Oktober 2007 Sven Vahsen 9 9 9

Will the LHC be a SUSY Factory? • If SUSY exists at the Te. V scale, expect copious production of squarks and gluinos • Just QCD, nearly independent of SUSY model PPC 2008, Albuquerque, New Mexico • σ (pp SUSY) calculated at NLO • √s=14 Te. V, m. SUSY ~ 0. 5 -1. 0 Te. V σpp SUSY ~ 1 -100 pb Sven Vahsen 10

Tevatron LHC • Win twice when moving to LHC • σSUSY increases 20000(!) for mgluino=400 Ge. V • S/B improves • SUSY-discovery challenge — reject SM by factor of ~1011 — understand SM events that survive SUSY selection PPC 2008, Albuquerque, New Mexico Sven Vahsen 11

Experimental Signature ~0 1 q q~R g~ p_ q q ~ g ~ q. L p q • Two sparticles initially • Cascade decays down to LSP: jets, leptons ~ 02 ~ 01 ~ l l l m. SUGRA bulk region • LSP escapes undetected: large ETmiss Canonical SUSY signature: ETmiss, high-p. T jets, often leptons PPC 2008, Albuquerque, New Mexico Sven Vahsen 12

A Toroidal LHC Apparatus (ATLAS) Solenoid Hadronic Tile Calorimeter Muon spectrometer Inner Detector PPC 2008, Albuquerque, New Mexico Electromagnetic Calorimeter Sven Vahsen Toroid 13

Generic SUSY signature: Missing ET + jets • Event selection — — — m. SUGRA bulk region Jets 1, 2 with p. T>100 Ge. V Jets 3, 4 with p. T> 50 Ge. V ETMISS > 100 Ge. V ETMISS > 0. 2 Meff Transverse Sphericity > 0. 2 veto isolated leptons 1 fb-1 • Plot Effective Mass variable — Meff = Σ|p. Ti| + ETmiss Events with several hard, (often miss-measured) QCD jets: Monte Carlo predictions have large systematic uncertainties PPC 2008, Albuquerque, New Mexico Sven Vahsen Excess at large Meffective potential discovery of SUSY 14

Measuring SUSY Backgrounds • Estimate Standard Model background passing SUSY selection using data-driven techniques • Example: Select Z→ll and replace charged leptons by neutrinos • Obtain shape of ETMISS, Meffective 1 fb-1 PPC 2008, Albuquerque, New Mexico Sven Vahsen 15

L ~ 1 fb-1: SUSY Discovery Potential • Full simulation of backgrounds • Includes expected systematic (JES, background estimation) • Good chance of finding Te. V scale SUSY with 1 fb-1 of data — Dream scenario! • SUSY at higher mass scales could still show up later, but would make detailed studies difficult Tevatron Exclusion! PPC 2008, Albuquerque, New Mexico Sven Vahsen 16

L > 1 fb-1: What exactly did we discover? • Inclusive studies provide first hints of where in SUSY parameters — Meffective Mass scale — Relative Significance in 0, 1, 2 lepton channels m 0, m 1/2 — 3 rd generation tan • Is it really SUSY? want spin, hard at LHC. See talk by Martin White this afternoon. PPC 2008, Albuquerque, New Mexico Sven Vahsen 17

Mass reconstruction • Most promising: Opposite sign, same flavor di-leptons from single neutralino decay m. SUGRA bulk region 1 fb-1 l+ l- • Subtract background (from Standard Model and SUSY itself) using flavor information — e+e- + μ+μ– - e+μ- - e-μ+ (after efficiency correction) • Low background, relatively high statistics PPC 2008, Albuquerque, New Mexico • Sven Vahsen Position of mass-edge sensitive to combination of sparticle masses 18

Fitting for edges after flavor subtraction • Method sensitive to any sleptons lighter than 2 nd neutralino Bulk region Low mass point 0. 5 fb-1 1 fb-1 SUSY Two light sleptons, Coannihilation region Standard Model 18 fb-1 PPC 2008, Albuquerque, New Mexico Sven Vahsen l+ l- 19

Just add quarks • Use same dilepton events, similar mass-edge extraction w/ mlq and mllq • Use position of all edges to fit for sparticle masses Measured ~ q. L ~0 q 2 l l~ l Monte Carlo ~ 01 m. SUGRA bulk region, 1 fb-1 • • Fit assumes we know mass hierarchy, e. g. from di-lepton edge shape Otherwise model-independent Need more data for precise masses Quite sensitive to mass differences PPC 2008, Albuquerque, New Mexico Sven Vahsen 20

Assuming model known - can we extract model parameters early on? • • • Input data: kinematic edges — Dilepton, (Di)lepton+jet, ~ ~ q 0 1 q R Scan of m. SUGRA parameter space Pseudo Experiments + Fit A 0 not well constrained, µ ambiguity Even in highly constrained model, ambiguous parameter determination with 1 fb-1 PPC 2008, Albuquerque, New Mexico Sven Vahsen m. SUGRA bulk region, 1 fb-1 21

Ultimate LHC precision (300 fb-1) • SPS 1 a Snowmass Point (m. SUGRA bulk) • older work using fast simulation • kinematic endpoints using leptons, Taus, jets, b-jets Polesello, Tovey JHEP 0405 (2004) 071 300 fb-1 PPC 2008, Albuquerque, New Mexico Sven Vahsen 22

Ultimate LHC precision (300 fb-1) 300 fb-1 100 fb-1 • m. SUGRA parameters well constrained at 300 fb-1 calculate m , σ W PPC 2008, Albuquerque, New Mexico Sven Vahsen 23

• Calculate neutralino mass m and cross section σ -nucleon • Compare with direct detection NOW 2009? 300 fb-1 Cross-section (cm 2) Comparing with Direct Detection BULK FOCUS ZEPLIN-2 COAN. FUNNEL 2012? WIMP Mass [Ge. V/c 2] 300 fb-1 log 10 (s psi / 1 pb) Same WIMP in lab and space? PPC 2008, Albuquerque, New Mexico Sven Vahsen 24

Comparing with Observational Cosmology No. MC Experiments • Calculate neutralino relic density • Precision comparable to that from cosmology: W = W DM ? 300 fb-1 W h 2 = 0. 192 ± 0. 005 (stat) ± 0. 006 (sys) • What fraction of the dark matter is neutralinos? • (simulated point is pre-WMAP) W h 2 Caveat: this precision depends on assuming very constrained SUSY breaking scenario. In more general scenario: much looser constraints. PPC 2008, Albuquerque, New Mexico Sven Vahsen 25

Model Independent Approach I • In calculations of m , σ W — m. SUGRA unification assumption constrains • MSSM analysis not assuming specific SUSY breaking scenario needs more measurements • e. g. to calculate W need — LSP mass — LSP mixing matrix — to establish which processes are relevant to LSP annihilation… — …and measure them PPC 2008, Albuquerque, New Mexico Nojiri, Polesello, Tovey JHEP 0603: 063 (2006) 300 fb-1 • SPA m. SUGRA bulk region point, analyzed as MSSM • Less restrictive on W : precision ~ 10 -20% at 300 fb-1 Sven Vahsen 26

Model Independent Approach II • • Several m. SUGRA points also analyzed as MSSM to evaluate LHC + ILC prospects by Baltz et al. Bulk region, 300 fb-1: - W precision in agreement with Polesello et al. - σ -nucleon not well constrained, as it depends on heavy Higgs mass (not observable at LHC in this model) Baltz, Battaglia, Peskin, Wizansky PRD 74: 103521 (2006) PPC 2008, Albuquerque, New Mexico Sven Vahsen 27

Less favorable scenarios • m. SUGRA bulk region points discussed are LHC friendly scenarios • Caveat 1: Situation may be less ideal, even with low SUSY mass scale • High tan Tau’s dominate • Small mass-gaps very soft leptons ~ • If slepton too heavy, depend on other 02 decays • Caveat 2: Part of SUSY particle spectrum could be heavy • Several scenarios considered in Baltz, Battaglia, Peskin, Wizansky PRD 74: 103521 (2006) • Progress will depend on SUSY scenario Nature has chosen PPC 2008, Albuquerque, New Mexico Sven Vahsen 28

Conclusion • If SUSY at Te. V-scale, LHC discovery possible with ~1 fb-1 — In this case, rates at LHC would be high, allowing detailed studies in large number of final states • To calculate m , σ W , need to understand much of SUSY phenomenology — Model assumptions can reduce needed measurements • Degree of progress at LHC will depend on Nature’s benevolence in breaking SUSY • Favorable scenarios suggest precise calculations of m , σ W possible with ~300 fb-1 of data • When combined with Astroparticle & Cosmology measurements, this would reveal the relation of the SUSY LSP to the dark matter PPC 2008, Albuquerque, New Mexico Sven Vahsen 29