Stop NLSPs. David Shih Rutgers University. Based on: Kats & DS ( ) Kats, Meade, Reece & DS ( )

Transcription

1 Stop NLSPs David Shih Rutgers University Based on: Kats & DS ( ) Kats, Meade, Reece & DS (0.6444)

2 This year, the LHC has performed extremely well. ATLAS and CMS have collected over 5/fb of data each!!

3 Limits, limits everywhere... ATLAS and CMS have searched for SUSY far and wide...

4 Limits, limits everywhere bino wino higgsino Zrich higgsino hrich chargino slepton flavor dem slepton tau rich sneutrino gluino sbottom stop Summary of limits on GGM NLSP types with gluino production Kats, Meade, Reece & DS (0.6444) Gluino mass GeV

5 Limits, limits everywhere bino wino higgsino Zrich higgsino hrich chargino slepton flavor dem slepton tau rich sneutrino gluino sbottom stop Summary of limits on GGM NLSP types with gluino production Kats, Meade, Reece & DS (0.6444) Gluino mass GeV Where is SUSY hiding??

6 Today s Talk

7 Today s Talk Where could current LHC searches be missing something?

8 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production)

9 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation

10 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra

11 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra Multiple final states

12 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra Multiple final states }Stop NLSPs realize all of these

13 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra Multiple final states }Stop NLSPs realize all of these In today s talk, we will

14 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra Multiple final states }Stop NLSPs realize all of these In today s talk, we will motivate and describe the stop NLSP scenario

15 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra Multiple final states }Stop NLSPs realize all of these In today s talk, we will motivate and describe the stop NLSP scenario discuss the current constraints on stop NLSPs from Tevatron and LHC

16 Today s Talk Where could current LHC searches be missing something? Low mass, low cross section (e.g. EW SUSY production) 3rd generation Squeezed spectra Multiple final states }Stop NLSPs realize all of these In today s talk, we will motivate and describe the stop NLSP scenario discuss the current constraints on stop NLSPs from Tevatron and LHC suggest ways to improve searches for stop NLSPs

17 : Combined exclusion limits for simplified SUSY models with m( χ 0 ) = 0(left)andMSUGRA/CMSSM models with tan β = 0, A 0 = 0andµ>0(right). Motivations Degenerate squarks are highly constrained by LHC searches: squark mass [GeV] Squark-gluino-neutralino model, m( # 0 ) = 0 GeV Tevatron, Run I CDF, Run II D0, Run II ATLAS 0 lepton 0 combined CL s observed 95% C.L. limit CL s median expected limit Expected limit ±! 00 data PCL 95% C.L. limit - $ L dt =.04 fb,! SUSY! SUSY! SUSY s=7 TeV = 0.0 pb = 0. pb! SUSY = 0 pb LEP = pb ~ q gluino mass [GeV] [GeV] m / Σ fb MSUGRA/CMSSM: tan% = 0, A 00 ATLAS $ 0 q ~ (600) ~ q (000) -, ± L dt =.04 fb s=7 TeV LEP # D0 ~ g, ~ q, tan %=3, µ<0,. fb CDF ~ g, ~ q, tan %=5, µ<0, fb Theoretically excluded ~ q (400) Tevatron Limit - - = 0, µ>0 0 0 lepton 0 combined CL s observed 95% C.L. limit CL s median expected limit Expected limit ±! ATLAS j CMS high HT Reference point ATLAS 4jHM CMS high MHT 00 data PCL 95% C.L. limit NLO Rate ~ g (600) ~ g (00) ~ g (000) ~ g (800) M squark GeV m 0 [GeV]

18 Motivations However, a single stop can be much lighter. Cross section reduced by a factor of ~0. Σ t t pb tt Tevatron Σ t t pb tt 7 TeV LHC m t m t m t GeV m t GeV As we will see, stop decays usually lead to top-like final states. Light stops could be hiding in the top sample!

19 Motivations Other reasons for considering light stops include Loosely motivated by SUSY naturalness. See Papucci et al (0.696) and Sundrum et al (0.6770) for a recent discussion Light stops can occur even in flavor-blind mediation schemes, through RGEs and L-R splitting Light stops are also possible in models of effective SUSY, for instance those with composite st/nd generations Summary: light stops are theoretically possible, detecting them presents an interesting experimental challenge, and currently they represent a big blind spot for LHC searches.

20 Light Stop Scenarios

21 Production: Light Stop Scenarios

22 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled

23 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks

24 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks will consider both in today s talk

25 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks Decay: will consider both in today s talk

26 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks Decay: neutralino LSP (addressed by CDF stop search) will consider both in today s talk

27 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks Decay: neutralino LSP (addressed by CDF stop search) will consider both in today s talk sneutrino LSP (addressed by D0 stop search)

28 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks Decay: neutralino LSP (addressed by CDF stop search) will consider both in today s talk sneutrino LSP (addressed by D0 stop search) gravitino LSP (no dedicated search exists yet!!)

29 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks Decay: neutralino LSP (addressed by CDF stop search) will consider both in today s talk sneutrino LSP (addressed by D0 stop search) gravitino LSP (no dedicated search exists yet!!) will focus on this scenario in today s talk

30 Light Stop Scenarios Production: direct production of light stop, all other colored sparticles decoupled production through heavier gluinos or squarks Decay: will consider both in today s talk neutralino LSP (addressed by CDF stop search) sneutrino LSP (addressed by D0 stop search) gravitino LSP (no dedicated search exists yet!!) t t + G (m t >m t) will focus on this scenario in today s talk t W + + b + G (m t m t) Stop NLSP

31 Minimal Stop NLSP Scenario!%((! %&&'*)9 (:++%)#(!#$%&'()* 4'563'78 3

32 Relevant Analyses

33 Relevant Analyses Measurements of ttbar cross section Tevatron (~5/fb) and LHC (35/pb and /fb) Dilepton and lepton+jets channels With and without b-tagging Cut and count only

34 Relevant Analyses Measurements of ttbar cross section Tevatron (~5/fb) and LHC (35/pb and /fb) Dilepton and lepton+jets channels With and without b-tagging Cut and count only CDF stop search t b χ +, χ+ + ν χ 0

35 Relevant Analyses Measurements of ttbar cross section Tevatron (~5/fb) and LHC (35/pb and /fb) Dilepton and lepton+jets channels With and without b-tagging Cut and count only CDF stop search t b χ +, χ+ + ν χ 0 D0 stop search (initial cut-and-count selection only) t b + ν

36 Relevant Analyses Measurements of ttbar cross section Tevatron (~5/fb) and LHC (35/pb and /fb) Dilepton and lepton+jets channels With and without b-tagging Cut and count only CDF stop search t b χ +, χ+ + ν χ 0 D0 stop search (initial cut-and-count selection only) t b + ν ATLAS search for ttbar+met (35/pb and /fb) T t + X

37 Relevant Analyses Measurements of ttbar cross section Tevatron (~5/fb) and LHC (35/pb and /fb) Dilepton and lepton+jets channels With and without b-tagging Cut and count only CDF stop search t b χ +, χ+ + ν χ 0 D0 stop search (initial cut-and-count selection only) t b + ν ATLAS search for ttbar+met (35/pb and /fb) T t + X ATLAS and CMS searches for jets+met, lepton+jets+met, lepton+bjets+met, OS dileptons+met

38 Relevant Analyses Measurements of ttbar cross section Tevatron (~5/fb) and LHC (35/pb and /fb) Dilepton and lepton+jets channels With and without b-tagging Cut and count only CDF stop search t b χ +, χ+ + ν χ 0 D0 stop search (initial cut-and-count selection only) t b + ν ATLAS search for ttbar+met (35/pb and /fb) T t + X ATLAS and CMS searches for jets+met, lepton+jets+met, lepton+bjets+met, OS dileptons+met Apart from the CDF stop search, we have not reinterpreted analyses with sophisticated kinematic discriminants. These include xsec measurements, top mass measurements, and the D0 stop search.

39 Our Modus Operandi Generate events with combination of Pythia 8 for initial hard process, showering and hadronization Homemade code to decay stops to W+b+gravitino. ( Available at: ) Reconstruct events using basic homemade detector sim (jet algorithms via FastJet, lepton isolation, b-tagging requirements) Code up relevant analyses, validate them on publicly available results (ttbar; stop signal where applicable). Correct by scale factors where necessary (~0-30%). Infer limits on stop NLSPs using published backgrounds and systematic errors.

40 D0 stop search (hep-ex/ ) Integrated luminosity: 5.4/fb Search for stop pair production with t b + ν Initial selection Exactly one OS (e,mu) with pt > (5,0) GeV MET>7 MET>0 or DeltaPhi(e,mu)<.8 (to reject Z->tautau) This is all we were able to simulate from this search. The final selection involves using multiple composite discriminants, optimized separately for each point in parameter space.

41 D0 publication did not provide enough information (e.g. the definition of these discriminants) to allow us to reinterpret this search. It would be very interesting for D0 to apply this analysis to stop NLSP, it could a strong limit!!

42 CDF Stop Search (hep-ex/09.308) Integrated luminosity:.7/fb Search for stop pair production with t b χ +, χ+ + ν χ 0 Basic selection: two leptons with pt > 0 GeV at least two jets with ET > (5,) GeV There is also a 0 btag selection, but the limits for this are not shown. MET > 0 GeV >= btags Final limits are based on comparing signal and background distributions for the reconstructed stop mass variable

43 b-tagged Channel Non-b-Tagged Channel t b χ +, χ+ + ν χ 0 The reconstructed stop mass is computed using the leptons, the two highest ET jets, and the MET as follows: Fix a chargino mass, perform reconstruction for different values (in the paper, 05.8 GeV and 5.8 GeV). Events per.7 fb in the signal region. Top Z/γ +jets Diboson W +jets Total Data b-tag 49.0± ± ±0..8± ±7. 57 no tag 5.± ± ±.3 9.8± ± Treat invisible ν χ 0 as coming from a pseudoparticle with mass 75 GeV and width 0 GeV. Scan over pseudoparticle phi directions. TABLE I: The expected event yields from SM processes with the total uncertainties and the observed numbers of events in the signal region. Pair jets and leptons to minimize summed invariant masses b-tagged Channel Events/0 GeV/c 30 M~ Minimize chi-squared, varying =3.5 GeV/c t observables 5 M within their experimental resolutions, =05.8 GeV/c ±! to obtain M 0=47.6 GeV/c! 0 best fit reconstructed top mass 5 ~ ~ Signal # 6.0 excluded t t Non-b-Tagged Channel - CDF Run II data (.7 fb ) Top (M Z/$ * + Jets Other SM =7.5 GeV/c t ) Events/0 GeV/c ~ ~ Signal # 6.0 excluded t t M~ t M! =3.5 GeV/c ± M! 0 =05.8 GeV/c =47.6 GeV/c - CDF Run II data (.7 fb ) Top (M Z/$ * + Jets Other SM =7.5 GeV/c t ~ Reconstructed t Mass (GeV/c ) 6 a) ) GeV/c 0! b) m( ) GeV/c CDF Run II (.7 fb ) ± 0 BR (! #! $l)=.0 ± 0 BR (! #! $l)=0.50 ± 0 BR (! #! $l)=0.5 ± 0 BR (! #! $l)=0. Observed 95% CL ± m(! )=05.8 GeV/c ± 0 BR (! #! $l)=.0 ± 0 BR (! #! $l)=0.50 ~ ± BR( t #! b)= ± m(! )=5.8 GeV/c ) ~ Reconstructed t Mass (GeV/c ) 0 m(! 60 ± 0 BR (! #! $l)= Excluded by LEP ~ m( t ) 70 GeV/c

44 b-tagged Channel Non-b-Tagged Channel t b χ +, χ+ + ν χ 0 The reconstructed stop mass is computed using the leptons, the two highest ET jets, and the MET as follows: Fix a chargino mass, perform reconstruction for different values (in the paper, 05.8 GeV and 5.8 GeV). Events per.7 fb in the signal region. Top Z/γ +jets Diboson W +jets Total Data b-tag 49.0± ± ±0..8± ±7. 57 no tag 5.± ± ±.3 9.8± ± Treat invisible ν χ 0 as coming from a pseudoparticle with mass 75 GeV and width 0 GeV. Scan over pseudoparticle phi directions. TABLE I: The expected event yields from SM processes with the total uncertainties and the observed numbers of events in the signal region. Pair jets and leptons to minimize summed invariant masses b-tagged Channel Events/0 GeV/c 30 M~ Minimize chi-squared, varying =3.5 GeV/c t observables 5 M within their experimental resolutions, =05.8 GeV/c ±! to obtain M 0=47.6 GeV/c! 0 best fit reconstructed top mass 5 ~ ~ Signal # 6.0 excluded t t Non-b-Tagged Channel - CDF Run II data (.7 fb ) Top (M Z/$ * + Jets Other SM =7.5 GeV/c t ) Events/0 GeV/c ~ ~ Signal # 6.0 excluded t t M~ t M! =3.5 GeV/c ± M! 0 =05.8 GeV/c =47.6 GeV/c - CDF Run II data (.7 fb ) Top (M Z/$ * + Jets Other SM =7.5 GeV/c t ~ Reconstructed t Mass (GeV/c ) 6 a) ) GeV/c 0! b) m( ) GeV/c CDF Run II (.7 fb ) ± 0 BR (! #! $l)=.0 ± 0 BR (! #! $l)=0.50 ± 0 BR (! #! $l)=0.5 ± 0 BR (! #! $l)=0. Observed 95% CL ± m(! )=05.8 GeV/c ± 0 BR (! #! $l)=.0 ± 0 BR (! #! $l)=0.50 ~ ± BR( t #! b)= ± m(! )=5.8 GeV/c ) ~ Reconstructed t Mass (GeV/c ) 0 m(! 60 ± 0 BR (! #! $l)= Excluded by LEP ~ m( t ) 70 GeV/c

45 b-tagged Channel Non-b-Tagged Channel t b χ +, χ+ + ν χ 0 The reconstructed stop mass is computed using the leptons, the two highest ET jets, and the MET as follows: Fix a chargino mass, perform reconstruction for different values (in the paper, 05.8 GeV and 5.8 GeV). Events per.7 fb in the signal region. Top Z/γ +jets Diboson W +jets Total Data b-tag 49.0± ± ±0..8± ±7. 57 no tag 5.± ± ±.3 9.8± ± Treat invisible ν χ 0 as coming from a pseudoparticle with mass 75 GeV and width 0 GeV. Scan over pseudoparticle phi directions. TABLE I: The expected event yields from SM processes with the total uncertainties and the observed numbers of events in the signal region. Pair jets and leptons to minimize summed invariant masses b-tagged Channel Events/0 GeV/c 30 M~ Minimize chi-squared, varying =3.5 GeV/c t observables 5 M within their experimental resolutions, =05.8 GeV/c ±! to obtain M 0=47.6 GeV/c! 0 best fit reconstructed top mass 5 0 ~ ~ Signal # 6.0 excluded t t Not a rigorous process, but it 5 gives a useful discriminant... Non-b-Tagged Channel - CDF Run II data (.7 fb ) Top (M Z/$ * + Jets Other SM =7.5 GeV/c t ~ Reconstructed t Mass (GeV/c ) ) Events/0 GeV/c ~ ~ Signal # 6.0 excluded t t M~ t M! =3.5 GeV/c ± M! 0 =05.8 GeV/c =47.6 GeV/c - CDF Run II data (.7 fb ) Top (M Z/$ * + Jets Other SM =7.5 GeV/c t ~ Reconstructed t Mass (GeV/c ) 6 a) ) GeV/c 0! b) m( ) GeV/c 0 m(! CDF Run II (.7 fb ) ± 0 BR (! #! $l)=.0 ± 0 BR (! #! $l)=0.50 ± 0 BR (! #! $l)=0.5 ± 0 BR (! #! $l)=0. Observed 95% CL ± m(! )=05.8 GeV/c ± 0 BR (! #! $l)=.0 ± 0 BR (! #! $l)=0.50 ± 0 BR (! #! $l)=0.5 Excluded by LEP ~ ± BR( t #! b)= ~ m( t ) 70 GeV/c ± m(! )=5.8 GeV/c )

46 CDF Stop Search To reinterpret this search as a limit on stop NLSP, we computed the reconstructed stop mass distribution for stop NLSPs and for the benchmark stop scenarios used by CDF, using their method. By matching distributions and comparing to the CDF limits on the benchmark scenarios, we obtained the limit on the effective cross section for stop NLSPs. This procedure obviates the need to understand the limit-setting procedure and systematic uncertainties. It should also cancel out many of the systematic biases from our simulations. A M GeV M GeV B case NLSP m t m t m χ ± gravity-mediated m χ 0 BR A B C D

47 ATLAS ttbar+met ATLAS-CONF (35/pb) hep-ex/ (.04/fb) Search for fermionic T->t+MET in lepton+jets final state. Exactly one lepton with pt > 0 5 (e) At least four jets with pt > 0 5 MET > mt > 0 50 No btag requirement Large xsecs => limits in GeV range. Mass [GeV] A m(t)-m(a 0 ) < m(t) 3pb pb Expected Limit (±) Obs. Limit (Theory Unc.) CDF Exclusion ATLAS! - L dt=.04 fb s=7 TeV Excl. $ BR(TT#ttA 0 A 0 ) pb pb T Mass [GeV]

48 Acceptances (relative to ttbar) $&/3, &'0-(4!#$&/3,$&'0-(4!./0'5!#!$%&'() *+#,%-.(/0!#!$% (.%'(.)5!#$%&'( )*+$,-'./0 Generally, we find that the efficiency of standard cut-and-count analyses is around the same for stops as for tops. ATLAS ttbar+met does better because of its mt cut!

49 Results: Tevatron (Kats & DS )!#$&/3,$&'0-(4!./0'5 $&/3, &'0-(4!#$%&'( )*+$,-'./0 Standard cut and count techniques only. Basically no limit.

50 Results: Tevatron (Kats & DS )!#$&/3,$&'0-(4!./0'5 $&/3, &'0-(4!#$%&'( )*+$,-'./0!#$&/3,$&'0-(4 67/4$80&& -'(39&/-:(/739 Best limit from Tevatron: mstop > 50 GeV Use of sophisticated discriminant was essential!!

51 Results: LHC (Kats, Meade, Reece, DS ) 5 4 ATLAS ttbar+met will have good coverage over widest range of stop mass. Σexcluded Σstop 3 0 m t ATLAS tt xsec 0.7fb ATLAS ttmet fb CMS jetsmet fb CMS jetsmet.fb CMS OS dileptons fb Tevatron limit A surprise: a number of /fb SUSY searches could be sensitive to light stops! Dashed lines indicate ultra-low acceptances (~ ) where we don t trust our simulation of the signal tails. M stop GeV There are still no firm LHC limits on direct stop pair production. Stop could still be lighter than the top??!!

52 Stop NLSPs w/ Squark Production squarkbinostop case squarkwinostop case Msquarks GeV Msquarks GeV CMS SS dileptons fb CMS jetsmet.fb ATLAS jetsmet fb CMS jetsmet.fb ATLAS jetsmet fb M stop GeV CMS SS dileptons fb CMS jetsmet.fb ATLAS jetsmet fb CMS jetsmet.fb ATLAS jetsmet fb M stop GeV 4t (*) +jets+met, 4t (*) +jets+met Now events have multiple tops, SS tops. 3t (*) +b+jets+met, SS dileptons becomes a major constraint! t (*) +b+jets+met Figure 5: Current best limits on the simplified parameter spaces for stop NLSPs produced via squarks, as described in Table 4 (top row). Msquark > In both cases, GeV the EW-ino mass has been fixed to Somewhat weaker non-top-rich scenarios (M squark + M stop )/. The shaded region either has M stop >M squarks or has no on-shell 3-body decay

53 Stop NLSPs w/ Gluino Production Mgluino GeV CMS SS dileptons fb CMS OS dileptons fb CMS jetsmet.fb ATLAS jetsmet fb ATLAS 68jetsMET.3fb ATLAS jetsmet fb CMS jetsmet fb M stop GeV 4t (*) +MET Mgluino > GeV Somewhat weaker non-top-rich scenarios gluinos, described in Table 4 (bottom row). The diagonal is positioned to allow on-shell g t t. Figure 6: Current best limits on the simplified parameter space for stop NLSPs produced from

54 4:6)7$8 566)7$8!# Suggestions for future analyses 456)7$8 ;<$)',&#+$,+$)%&++)/?)'<$) B)+)&)A//*)*+(,%#&'/,0 496)7$8 #&,&#')%&++ -.')#$#%&'(#)3 C</.)&#*)D$+E# DFG)H4)I5666J)69966KL)<$=<MNN6N9OH Q/.)*/#R')$$#)<&$)'/)A$' '<$)(/%-#&'/,(+),A<'0

55 Suggestions: leptonic MT (for details, see ) mt: generalization of W transverse mass to events with double decay chains ending in invisible particles. (Barr, Lester, Stephens, Summers,...) mt has been used for measurements of top properties, but in all cases, the full event was used (leptons+bjets+met). Expect an endpoint at the top mass, but combinatorics is an issue. ttbar is the dominant background to stop NLSP, especially at the LHC. We propose computing mt using only the leptons and MET to reject ttbar background. Expect an endpoint at W mass and no combinatorial confusion. (figure from Cheng & Han )

56 Suggestions: leptonic MT Efficiency Comparison: M t 300 GeV Probability Top Background M t 00 GeV M t 300 GeV Background Rejection Leptonic M T Cut MET Cut M T Signal Efficiency Based on CMS OS dileptons+met search, after all cuts except for MET>75 GeV.

57 Long-lived stop NLSPs F TeV m cm mm m t GeV Light stop NLSPs actually tend to be long-lived. The promptlydecaying case which we have focused on here only happens for the lowest-possible SUSY-breaking scales. Long-lived stop NLSPs would lead to R-hadrons, kinked tracks, displaced jets and leptons. Well-motivated benchmark scenario for many interesting searches!

58 Remaining Theory Hurdles We find that efficiencies of current LHC searches to light stop NLSP are extremely low, largely due to hard jet and HT cuts. This is problematic, because: We don t trust Pythia for modeling of extra hard jets from ISR/FSR. Problems with Madgraph 4 implementation of gravitino LSP in simulating light stops. (4-body phase space?) Gravitino LSP not yet implemented in Madgraph 5. The current situation is unsatisfactory. Dedicated experimental searches will either have to wait for Madgraph 5 to catch up, or (our preference) design searches with higher signal efficiencies which are less sensitive to the kinematic tails.

59 Summary and Outlook We have reviewed the current constraints on stop NLSP, and suggested ways to improve analyses. The stop NLSP could still be lighter than the top! Current limits are estimated to be mstop > 50 GeV from the Tevatron. There are no firm LHC limits yet on direct stop production. LHC should be sensitive to mstop~300 GeV in the near future. No dedicated search exists yet for stop NLSP. Currently a blind spot at both the Tevatron and the LHC. This could be a good opportunity to discover supersymmetry hiding in our backyard!

60 The End

61 Gauge Mediation Hidden sector SUSY+... SU(3)xSU()xU() Visible sector: MSSM+... Gauge mediation is a very attractive scenario for the MSSM: Solves SUSY flavor problem Calculable framework Recently, a model-independent framework for GMSB was formulated, and the full parameter space was understood: General Gauge Mediation (Meade, Seiberg & DS; Buican, Meade, Seiberg & DS) LHC searches are now being designed with GGM in mind!

62 The NLSP Gravitino LSP is a universal prediction of gauge mediation models: m 3/ = F 3Mpl ( ev GeV) Lightest MSSM sparticle becomes the next-to-lightest superpartner (NLSP). MSSM{... NLSP gravitino LSP

63 NLSP Collider Signatures

64 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures.

65 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G

66 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G Decays can be prompt or delayed: τ NLSP F m 5 NLSP

67 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G Decays can be prompt or delayed: τ NLSP F m 5 NLSP Will focus on prompt case today

68 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G Decays can be prompt or delayed: τ NLSP F m 5 NLSP Will focus on prompt case today All SUSY cascade decays pass through the NLSP.

69 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G Decays can be prompt or delayed: τ NLSP F m 5 NLSP Will focus on prompt case today All SUSY cascade decays pass through the NLSP. So all events contain:

70 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G Decays can be prompt or delayed: τ NLSP F m 5 NLSP Will focus on prompt case today All SUSY cascade decays pass through the NLSP. So all events contain: high pt objects determined by the NLSP type

71 NLSP Collider Signatures In gauge mediation, the NLSP type largely determines the inclusive collider signatures. NLSP decays to the gravitino plus its SM partner. f f λ V µ h h G G G Decays can be prompt or delayed: τ NLSP F m 5 NLSP Will focus on prompt case today All SUSY cascade decays pass through the NLSP. So all events contain: high pt objects determined by the NLSP type missing energy