Journal Club Meeting Minutes (2/19/2018)


Paper: A search for the rare decay of the Standard Model Higgs boson to dimuons in pp collisions at sqrt(s)=13 TeV with the ATLAS Detector


* How does the Higgs mechanism work? 
   * Spontaneous symmetry breaking from the Higgs Potential
   * Linear sigma model? *conversation if it’s necessarily a linear sigma model*


* Why do we need a Higgs Mechanism to get fermion masses?
   * SU(2) x U(1)
   * SM isn’t chiral. Regular mass term combines LH field with RH field. If they don’t transform in counter-setting ways, need something to get the mass.
  

   * Gauge invariance. Lorentz symmetry. 
        
* Is it necessary to have a Higgs mechanism to explain fermions? In original Lagrangian, already have mass term for fermions. Why not put the mass of the bosons? Lagrangian is still gauge invariant.
   * Why should it exist for the bosons? Multiple reasons. (e.g. For WW scattering, it goes to infinity without Higgs.)
   * Definitional term for fermion mass must include ψ_L and ψ_R terms.
   * Ground state: Can add 0 for fermions.
   * LH fermions and RH fermions aren’t interchangeable.


Higgs was to solve unitary problem for bosons. Happy accident that it worked for fermions.


* Why do we care about the coupling to the second generation?
   * Reference shown by Patrick: https://arxiv.org/pdf/1302.3229.pdf
      * Higgs doublet proposals. 
      * Higgs-dependent Yukawa couplings, Section 7: Acquiring mass higher dimensional operators. Explains why fermion masses are so different.
         * Tree coupling of Higgs is to the top. Others are present, but suppressed.


Can acquire mass with something other than the Higgs, but can’t just throw in fermion mass term.




Branching Ratio of Higgs vs Mass of Higgs. Bbar and ccbar look similar, but ccbar lower.
  

Image from http://inspirehep.net/record/921445/plots 


















Drew Feynman diagram with three-point (h^0 to ffbar)



Drew Feynman diagram for ggF and VBF (dominant processes)
  



Image from https://www.researchgate.net/figure/Feynman-diagrams-for-the-leading-production-modes-ggF-VBF-and-VH-where-the-V-VH-and_fig1_269339306 




Table 2
* VBF signal region. >= 2 jets. Neither jets should be b-tagged. Expect light quark coupling to vector boson should become themselves. (Initial quarks should be light.)
        








Drew Feynman diagram for VH and ttH (subdominant processes, should be mostly rejected)




  

Image from https://www.nikhef.nl/~vcroft/HiggsFeynmanDiagrams.html 


* ttH should be explicitly rejected by the no b-jet. Can technically study H → μμ with ttH, but ATLAS cuts it out because it has 2 b-jets.


  



Image from https://www.researchgate.net/figure/Feynman-diagrams-for-the-leading-production-modes-ggF-VBF-and-VH-where-the-V-VH-and_fig1_269339306


* ZH is almost certainly rejected by VBF cuts. Z goes to jets would be more central. VBF looks for two forward jets (from topology). Z to neutrinos would fail MET cut. Z to leptons [did not catch rest of sentence]




Table 1
* There’s a lot of room for analysis to improve (e.g. reducing ttbar background). Z to 2 jets will survive, but Z to neutrinos would be removed.
* Drew backgrounds
   * Z+jets
      * What fraction of Z decays go to muons? 3%
   * Z/γ (Drell-Yan) -- Virtual neutral vector 


  



* MET < 80GeV. Why is the upper limit so high? We can do a tighter cut. Probably affects efficiency. (ggF cut is very small)




Figure 1.  Right plot: Dimuon pT. ttbar has real missing energy, need to suppress those backgrounds.


Table 2
* What are the HL-LHC projections for H → μμ?
   * Reference: http://cdsweb.cern.ch/record/2319741/files/ATL-PHYS-PUB-2018-006.pdf 
* Most sensitive category is VBF categories.
* ttH too small. 
* VH not too bad, very distinguishable signature that you can explore. Diboson will be a problem, but can do weird analysis; variance mass of the Higgs would help (is what affects the discriminants of BDT).
        
        
General question for when reading papers: What should you pay attention to when reading the MC section?
* Section 3: Data and MC Samples
   * Data: How much data are we using? What triggers are used? Are there any unusual triggers?
      * For this paper, we are using standard triggers.
   * Simulation: Interesting to see the parameters used, renormalization scales, refactorization scales.
   * Example: Pg. 3, second to last paragraph. “Signal samples were generated…”
      * Important: Which mass are we using? How do they normalize? Which parameters do we use?
   * Example: Pg. 3, last paragraph. “Higgs boson production in the ggF process…”
      * Which generator? → Separated into two parts. Which generator matrix element and which PDF? How are they normalized?
         * Common for ATLAS to combine multiple generators to get better results. → One generator for events and one to do the calculations.
      * Which tuning is used? Tuning is important for parameters. What PDF is used for tuning?
         * ANZLO tuning stands for ATLAS Z boson measurement for Next Leading Order
   * Useful for phenomenologists
* Which QCD scale is used? 
   * Various choices: Z pT, dilepton invariant mass. Usually set scale to certain energy; dynamic scale. 




Drew ttbar background
  

Image from http://inspirehep.net/record/851277/plots 


* Helpful to think of which initial partons generate ttbar. 
   * Typical: qq, qg (higher order)
   * PDF: Gluon high fraction, quark low fraction
      * Gluons: large penalty if there’s loops, so low production






Allowed for qq to Higgs


*Argument for muon muon collider*


Top background -- Mostly killed by b-rejection. Z not as much (2 orders of magnitude more); Z also has larger cross section in general.


Look at ttbar cross section and probability that both go to muons.