1. A Feasibility Study on Measuring a Strange Sea Asymmetry in the Proton at ATLAS Laura Gilbert Presentation for Oxford ATLAS RA Interview 18 th October 2007

2. Strange Sea Asymmetry analysis Attempting to place a limit on feasibility of detecting a strange sea momentum asymmetry at ATLAS. Mainly solo, self-directed analysis. pQCD predicts u, d distributions in the proton to be almost flavour symmetric. MNC and further experiments show that there are in fact large differences in their momentum fractions. Various theoretical models proposed. Meson Cloud model (MCM) seems physically intuitive as a way to explain observations. This implies an s, s asymmetry also. Phys.Lett. B590 (2004) 216-222: Ding & Ma Calculations from Meson Cloud Model – 2-body wavefunctions [Gaussian (thick) and power-law (thin)] x(s(x) - s(x)) s(x), s(x) distributions currently poorly constrained → this analysis will be useful for pdf studies regardless of asymmetry result. Ws at LHC produced from low-x sea quarks (x<~0.01), so an asymmetry would be hard to detect. However this kinematic region has not been probed by previous fixed-target experiments.

3. Signal Selection: W+D* Select W candidate Reconstruct D 0 →K - π + D 0 vertex displaced. Add prompt (soft) pion. Plot reconstructed D*-D 0 mass difference = 145.4 MeV(small intrinsic resolutions, manageable background) Consider 3 sign correlations: (K - with π +, K - with π B +, π B + with e - ) Consider backgrounds inc. Cabibbo suppressed wrong sign combinations Sample: 3 million of each W +, W -, generated with MC@NLO, ATLFAST. s c W g s g W c cg W s NLO Gluon production: 10% of total s c W NLO W production LO production: 77%

4. Selection Cuts Preliminary Cuts: –1 electron with p T >25GeV, |η|<2.4 –Missing E T >25GeV –Two oppositely signed tracks: assign one K, one π. –p T (K)>1.5GeV, p T (π)>1GeV –Third track: assign bachelor π B, p T (π B )>0.5GeV –π B charge opposite to e, opposite to K Optimised Cuts: –m(D 0 reco )- m(D 0 true )< 40MeV –Signed L xy > 0.35mm –π B impact parameter significance d 0 /σ(d 0 )<3 –d 0 (K)*d 0 (π)<0mm 2 –D 0 impact parameter <0.2mm –D* p T >6GeV, |η|<2.5 D0D0 D 0 : cτ=123μm K π L xy (L xy –ve if tracks point towards vertex) Reconstruct vertex: straight line approx d0(π)d0(π) d0(K)d0(K)

5. No. signal events =86±22 No “real” D*s in window = 76 No. W - events = 45 ±14 No “real” D*s = 40 No. W + events = 41 ±13 No “real” D*s = 36 Reconstructed Unsmeared Real D*s Signal Selection 3 million of each W+, W-, generated with MC@NLO, ATLFAST.3 million of each W+, W-, generated with MC@NLO, ATLFAST. two passing events are due to gluon splitting (s+c→W+g, g→cc)two passing events are due to gluon splitting (s+c→W+g, g→cc) ~1.5% inherent asymmetry from d+g→W+c diagrams (CompHep)~1.5% inherent asymmetry from d+g→W+c diagrams (CompHep) No. signal events =86±22 No “real” D*s in window = 76 No. W - events = 45 ±14 No “real” D*s = 40 No. W + events = 41 ±13 No “real” D*s = 36 Normalised to 1fb -1

6. Signal and Electroweak Backgrounds W→eν: Signal: 84±22 events/fb -1W→eν: Signal: 84±22 events/fb -1 –MC@NLO with ATLFAST: 84/6x10 6 events pass cuts W→τν: Signal: <8 events/fb -1 pass cuts at 95% CLW→τν: Signal: <8 events/fb -1 pass cuts at 95% CL MC@NLO with ATLFAST: 3/6x10 6 events pass cuts –Comphep, then MC@NLO with ATLFAST: 3/6x10 6 events pass cuts Z→ee: < 3 events/fb -1 pass cuts at 95% CLZ→ee: < 3 events/fb -1 pass cuts at 95% CL MC@NLO with ATLFAST: 0/2x10 6 events pass cuts –Comphep, then MC@NLO with ATLFAST: 0/2x10 6 events pass cuts Z→ττ: << 1Z→ττ: << 1 event /fb -1 likely –Inferred from above WW: HERWIG x- section: 3.5 events/fb -1 before cuts.WW: <1 event /fb -1 - HERWIG x- section: 3.5 events/fb -1 before cuts. WZ: HERWIG x-section: 0.45 events/fb -1 before cuts.WZ: <<1 event /fb -1 - HERWIG x-section: 0.45 events/fb -1 before cuts. ZZ: HERWIG x-section: 0.06: events/fb -1 before cuts.ZZ: <<1 event /fb -1 - HERWIG x-section: 0.06: events/fb -1 before cuts.

7. QCD backgrounds D* + fake W: Sample 5802 dijet + fake electron (W, Z, t, γ). σ=191μb <8 events/fb -1 (95% CL). Further study: increase sample size, cut on angle between D* and W in transverse plane.W + cc (bb), Z + cc (bb): in current samples (gluon splitting), mainly removed by ET cuts. <8 events/fb -1 (95% CL). Further study: increase sample size, cut on angle between D* and W in transverse plane. qqbar: –bb: MC@NLO ~3mb, in progress. –tt: MC@NLO ~0.8nb –cc: Not available at NLO. Pythia ~5mb. Current contribution <30 events/fb -1 (95% CL). Further reduction needed to drop this limit (0 events pass of 10 8 )

8. Steps Towards Full Simulation Generated and validated CSC MC@NLO W→e,μ ν to tight deadline. Insufficient numbers of events generated for full analysis.Generated and validated CSC MC@NLO W→e,μ ν to tight deadline. Insufficient numbers of events generated for full analysis. Used these samples to study data-like electron and muon ID and resolution, MET resolution, including separation of LO, NLO diagrams.Used these samples to study data-like electron and muon ID and resolution, MET resolution, including separation of LO, NLO diagrams. Further work needed on low p T track reconstruction.Further work needed on low p T track reconstruction. Probable LO contribution Probable NLO contribution Plot from DC3 sample 005250 (MC@NLO), v 11.0.42

9. Relevant Research Experience Work with full simulation in preparation for data taking, including electron and muon ID and resolution, MET resolution, Work in fast simulation on secondary vertex reconstruction, aspects of B physics, jets. Experience of working with QCD/QED channels similar to SUSY signals and backgrounds (esp. channels involving MET, leptons and jets) Computing: –Extensive experience with distribution releases –Worked with various grids (up/downloading, registering, software installaion) –Event generators and decay packages: Pythia, Herwig, MC@NLO, Sherpa, Comphep, EvtGen –Simulations: GEANT, ATLFAST (with HepVis viewer) –Coding in C++, Java, VB, python, XML –Designed Virtual Machine solutions (VMware), including novel and unconventional deployments.

10. Summary and Conclusions Signal selection looking promising compared to EW backgrounds QCD backgrounds likely to be more significant but we have further rejection possibilities to work with (MET, stronger electron isolation criteria, W/D* angular separation) Low p T track reconstruction efficiencies poorly understood. Back-of-envelope: to exclude null hypothesis to 95% CL at 1fb -1 we need around 60% asymmetry (80:20). 1fb -1 insufficient for convincing asymmetry calculations – probably need at least 100 fb -1. Experience gained whilst performing this analysis has given me a good background to reach rapid conclusions in work on SUSY signals.