Liang Shang, Henan Normal University IHEP 2016.06.14 Explanation of the ATLAS Z-peaked excess in the NMSSM Liang Shang, Henan Normal University IHEP 2016.06.14

I Monte Carlo simulation in collider physics 1. Why do we need to do MC simulation? 2. How to do MC simulation in collider physics? 3. Which kind of work can MC simulation do? II Work: Explanation of the ATLAS Z-peaked excess in the NMSSM 1. The ATLAS experimental report for Z-peaked excess 2. The explanation of the Z-peaked excess III Introduction: Implement New Analysis into CheckMATE 1. Why do we use CheckMATE for analysis? 2. Information putting into CheckMATE 3. Easiest to do one task oneself IV Summary

I Monte Carlo simulation in collider physics

1.Why do we need to do MC simulation? Theory Experiment 𝜎𝐿𝜖=𝑁

1.Why do we need to do MC simulation? Monte Carlo methods (or Monte Carlo experiments) are a broad class of computational algorithms that rely on repeated random sampling to obtain numerical results. Monte Carlo methods are mainly used in three distinct problem classes: optimization, numerical integration, and generating draws from a probability distribution.

2.How to do MC simulation in collider physics? a. PDF CTEQ6L1, … b. Hard scattering MadGraph/MadEvent, SHERPA, … c. Parton Shower Hadronization PYTHIA d. Detector Delphes, PGS, GEANT4, … FastJet MG/FR School, Beijing, May 22-26, 2013, Johan Alwall

a. PDF b. Hard scattering c. Parton Shower Hadronization d. Detector CTEQ6L1, … b. Hard scattering MadGraph/MadEvent, SHERPA, … c. Parton Shower Hadronization PYTHIA d. Detector Delphes, PGS, GEANT4, … FastJet

a. PDF b. Hard scattering c. Parton Shower Hadronization d. Detector CTEQ6L1, … b. Hard scattering MadGraph/MadEvent, SHERPA, … c. Parton Shower Hadronization PYTHIA d. Detector Delphes, PGS, GEANT4, … FastJet

a. PDF b. Hard scattering c. Parton Shower Hadronization d. Detector CTEQ6L1, … b. Hard scattering MadGraph/MadEvent, SHERPA, … c. Parton Shower Hadronization PYTHIA d. Detector Delphes, PGS, GEANT4, … FastJet

a. PDF b. Hard scattering c. Parton Shower Hadronization d. Detector CTEQ6L1 b. Hard scattering MadGraph/MadEvent, SHERPA, … c. Parton Shower Hadronization PYTHIA d. Detector Delphes, PGS, GEANT4, … FastJet

3. Which kind of work can MC simulation do? arXiv:1507. 08471

3. Which kind of work can MC simulation do? arXiv:1511.05386

3. Which kind of work can MC simulation do? arXiv:1206.3865

3. Which kind of work can MC simulation do?

II Example-- Work Explanation of the ATLAS Z-peaked excess in the NMSSM arXiv: 1504.07869

arXiv:1503.03290

1. The ATLAS report general gauge mediation models

Event selection

Result 3.0σ

The ATLAS Z-peaked excess 29.6 12 29 - 10.6 = 18.4

End of the Story? ATLAS: Excess! 2~6 Jets +MET : Null! q 2~6 Jets +MET : Null! Multi-Jets +MET : Null!

Not GGM! Multi-Jets +MET : Null! 2~6 Jets +MET 0 GeV for GGM Z-peaked Excess in GGM

3. Explanation ( 𝑊 , 𝐻 𝑢 , 𝐻 𝑑 ) ( 𝑊 , 𝐻 𝑢 , 𝐻 𝑑 )

Work flow Fastjet MadGraph/MadEvent Pythia Delphes Parton Level Event Hardron Level Event Detector Level Event CheckMATE arXiv:1312.2591 .root

CheckMATE

The explanation in NMSSM By gluino pair production

III CheckMATE --Add New Analysis into CheckMATE

CheckMATE Authors: Manuel Drees, Herbi Dreiner, Jong Soo Kim, Daniel Schmeier, Jamie Tattersall, Krzysztof Rolbiecki

1.Why do we use CheckMATE for analysis? Analysis frame: in analysis frame Focus on how to add cuts 1. Final objects: “Loose”, “Tight”, “Medium” electron in ATLAS “Combined” and “Segment-tagged”, “Combined+Segment-tagged” muon in ATLAS 2. kinematical variables “ 𝑀 𝑇 ” “ 𝑀 𝑇2 ” “ 𝑀 𝑇2 W ” 3. Root integrated TLorentzVector class: Δ𝜙= 𝜙 1 − 𝜙 2 , Δ𝑅= Δ 𝜙 2 +Δ 𝜂 2 , vector1.M(vector2) Histogramming Statistical functions 4. Helper functions CountSignalEvent: trigger CheckMATE evaluation filterPhaseSpace: simplify basic cuts - avoid multiple loops! Ref: arXiv:1312.2591, arXiv: 1503.01123

2.Information putting into CheckMATE 1. Author and analysis card information 1. Analysis Name, c.m.s, luminosity 2. Author name, email 2.Properties of Final States Isolation criteria for lepton Jet clustering algorithm B-jet tagging efficiency 3. Information of Signal regions Signal name Observed number of events, SM expectation and associated error Questions and Answered by ./AnalysisManager 4. Information of cuts Basic cut Cut for phase space: Pt and eta Isolation condition and Overlap removal Event selection cutting M, HT, … C++ Initialize() Finalize() Analyze()

A atlas_zpeak 3. Example：ATLAS Z-peak Excess 1. ./AnalysisManager 2. arXiv: 1504.07869 and 1312.2591 2. A 3. The description for the new analysis atlas_zpeak

arXiv: 1503.03290 General Information 4. Name of Signal Region and events’ number arXiv: 1503.03290 General Information

1. Information for isolation criteria Track 400MeV DR=0.3 PT < 16%

2. Information for jets’ construction PT>20GeV

3. Information for b-jets yes 60

initialize(): initialize variables and open files, etc. Success! 4. Information of cuts Basic cut Cut for phase space: Pt and eta Isolation condition and Overlap removal condition Event selection cutting M, HT, … C++ initialize(): initialize variables and open files, etc. finalize(): free pointers and close file, etc. analyze(): core atlas_zpeak.cc

1. Type of lepton and cuts for phase space

2. Overlap Removal

3. Isolation Condition

Event selection – example: M()

Add new Analysis into CheckMATE Ref: http://checkmate.hepforge.org/AnalysesList/ValidatedList.html

SUMMARY I Monte Carlo simulation in collider physics 1. Why do we need to do MC simulation? Monte Carlo simulation studies of physics processes at colliders 2. How to do MC simulation in collider physics? MadGraph/MadEvent, Pythia, Delphes, Root, CheckMATE, … 3. Which kind of work can MC simulation do? Constraint, Prediction, Improve, Explanation II Work: Explanation of the ATLAS Z-peaked excess by gluino pair in the NMSSM 1. The ATLAS experimental report for Z-peaked excess Process, identification, selection, results 2. The explanation of the Z-peaked excess 𝑚 𝑔 =750𝐺𝑒𝑉, 𝑚 𝜒 2 0 =650𝐺𝑒𝑉, 𝑚 𝜒 1 0 =565𝐺𝑒𝑉 can explain at 1.3𝜎 III Introduction: Implement new analysis into CheckMATE 1. Why do we use CheckMATE for analysis? Final objects, kinematical variables, Root integrated, Helper functions 2. Information putting into CheckMATE Questions and Answered, analysis code in C++ 3. Easiest to do one task oneself Atlas_zpeak

We ARE JUST ON THE WAY Thank you.

2.How to do MC simulation in collider physics? MET

arXiv:1307.6346

2.How to do MC simulation in collider physics? arXiv:1312.2591 arXiv:0802.1189

Why do we need to do MC simulation?

1. Brief introduction of SUSY ( 𝐵 , 𝑊 3 , 𝐻 𝑢 0 , 𝐻 𝑑 0 ) ( 𝜒 1 0 , 𝜒 2 0 , 𝜒 3 0 , 𝜒 4 0 ) MET SM MSSM 𝑊 + 𝐻 + 𝜒 1 + 𝜒 2 +

Result arXiv:1504.07869

By squark pair production

Conclusion