1. Small Extra Dimension Graviton Detection at ATLAS
Looking at the Graviton as it decays through G*->Z0 Z0 ->e-e+e-e+
Audrey Todhunter
Case Western Reserve University
Dominik Dannheim
Columbia University

2. The Graviton
Massive graviton resonances are predicted by models using extra dimensions
-resonances could be well separated in mass
The graviton couples universally (ee,,,,Z0Z0, etc…)
-Products could be detected at ATLAS
Analysis used the Randall-Sundrum model, which predicts the mass of the first resonance by:
mn =xn(k/MPl)
We chose k/MPl=0.01 to give resonances on the TeV scale

3. This analysis:
ppG*Z0Z0e+e-e+e-
against a Standard Model background
qqZ0Z0e+e-e+e-
The branching ratio Br for Z0Z0e+e-e+e- is only 3.4%*3.4% = 11.56% for this decay process

4. Signal and Background Simulation
-Pythia event generator
-mass points at 300GeV, 500GeV, 700GeV, 1TeV, and 1.5TeV, as well background processes around each point
-want to know how well these signals can be detected at ATLAS Geant4 used in a full simulation of the ATLAS reconstruction

5. Properties of simulated signal and background
Process
Inv. mass of G* resonance (GeV)
(·Br) (fb)
Nev events generated
G*Z0Z0e+e-e+e-
300
4.807
1480
500
0.7347
1918
700
0.0972
1677
1000
0.0167
1999
1500
1899
qqZ0Z0e+e-e+e-
200
9.068
400
1.527
1986
800
0.167
1678
1200
1876
Properties of simulated signal and background

6. Particle Selection
We can identify electrons by their characteristics detected in the EM calorimeter and the central tracker
Transverse energy eT > 20 GeV
Likelihood > 0.9 being electron rather than pion
2e+e- pairs in each entry
- best combination of particles decided by comparing reconstructed Z0 mass to known Z0 mass (91.1GeV)

7. Invariant Z0 mass from a 300GeV signal: from electron pairs closest to Z0 mass within each interaction

8. Graviton Reconstruction
-electrons chosen for Z0 mass used to reconstruct invariant G* mass
-signal scaled for 100 fb-1 luminosity (1 year running at design luminosity)
NEv = (·Br) · L
imposing a weight factor on the number of simulated events

9. Signal at 500GeV with a Gaussian fit sigma parameter used to look at events within a mass window of ±3

10. Signal and background added and fit with [p0]e(-(x-[p1])2/2[p2] 2 + [p3]e([p4]x+[p5])

14. Potential discovery of resonances from G*Z0Z0e+e-e+e- at ATLAS
count number of signal and background events within the ±3 window
need to have Nsig>5√ (NBG) or Nsig > 10 for a statistically significant signal
number of events can be increased by raising luminosity

15. Luminosity needed for Nsig>5√ (NBG) or Nsig > 10
L=100fb-1: one year running at design luminosity

16. Conclusions
-for the process G*Z0Z0e+e-e+e- the upper mass limit is slightly below 700GeV
-higher discovery potential at increased luminosity
-other decay channels are possible for the graviton which may have higher upper limits for detection