1. Higgs studies at the upgraded ATLAS
Fabiola Gianotti (CERN)
Preliminary remarks:
Higgs studies at the LHC have moved from discovery phase to precise measurements
ATLAS and CMS were able to run in much harsher conditions than the design
(~ 40 pile-up events at the beginning of fill, design was ~23) maintaining excellent
performance (trigger, efficiency/resolution for physics objects, ..)
 one of the assets for the fast Higgs discovery
 foundation for future upgrades
We should not underestimate the capability of hadron colliders to perform precise
measurements (“hammer vs scalpel” model is too a rough description, contradicted
by the beautiful Tevatron legacy of precise measurements)
HL-LHC: full exploitation (with moderate additional cost) of the huge LHC investment

2. Robustness against pile-up
Note: number of reconstructed primary
vertices is ~ 60% number of number of
interactions per crossings
Re-optimized e± identification using
pile-up robust variables
> 95% identification efficiency
~ flat vs pile-up
excellent agreement data/MC
 crucial for H 4l
Z ee data and MC
ETmiss resolution vs pile-up in Z μμ
before/after pile-up suppression with tracks
Stability of EM calorimeter response vs time
(and pile-up) during 2012 run better than 0.1%
Number of reconstructed primary vertices

3. Characterizing the new particle: first couplings measurements (examples …)
Sept. 2012
(July data)
Explore tension SM-data from H γγ
different production modes (VBF, ggF)
μγγ=1.8 ± 0.5
New particles in the gg  H and H γγ loops ?
BR (H invisible or undetected) < 0.84 at 95% CL
Couplings to fermions kF weakly constrained by direct H  ττ , bb;
indirect constraints from ggF (tt loop) indicate it’s non-vanishing

4. different production mechanisms (hence couplings)
Characterizing the new particle: first couplings measurements (examples …)
Sept. 2012
(July data)
Explore tension SM-data from H γγ
different production modes (VBF, ggF)
μγγ=1.8 ± 0.5
New particles in the gg  H and H γγ loops ?
Precision now increasing thanks to larger dataset (> 2.5x)  targeting exclusive
final states (VBF, VH, etc.) with dedicated analyses to increase sensitivity to
different production mechanisms (hence couplings)
BR (H invisible or undetected) < 0.84 at 95% CL
Couplings to fermions kF weakly constrained by direct H  ττ , bb;
indirect constraints from ggF (tt loop) indicate it’s non-vanishing

5. Spin and CP: from angular distributions of decay products
December 2012
H  4l : 0+ vs 0-
H  γγ : 0+ vs 2+
0.5σ
1.4σ
2+ disfavoured
at 91% CL
(gg G)
0- disfavoured
at 99.5% CL
(pure 0+ vs
pure 0-)
0.5σ
2.8σ

6. With full 2011-2012 dataset: Further ahead (present LHC plans):
~5 σ from each of H γγ (already achieved/ATLAS), H lνlν, H 4l per experiment
~3 σ from H ττ and ~3 σ from W/ZH  W/Zbb per experiment (already achieved at Tevatron)
Separation 0+/2+ and pure O+/O- at 4σ level combining ATLAS and CMS ?
Some couplings to %
 first ATLAS results with full datased will be released in coming 1-2 weeks
Further ahead (present LHC plans):
: shut-down (LS1)
: √s ~ 13 TeV, L ~ 1034, ~ 100 fb-1
2018: shut-down (LS2)
: √s ~ 14 TeV, L ~ 2x1034, ~300 fb-1  pile-up evts/x-ing
: shut-down (LS3)
: √s ~14 TeV, L ~ 5x1034, ~3000 fb-1 (HL-LHC)  pile-up evts/x-ing
Note: 2012 pile-up conditions: evts/x-ing
Remarks:
Most of ATLAS detector upgrades (new tracker, new calorimeter electronics) needed
to run beyond ~ 2020 (integrated radiation, aging ..) even if LHC is not upgraded in LS3
Results shown here very preliminary (work of a few months, input to European Strategy)
Assumptions: the upgraded tracker at HL-LHC performs as well as present tracker at LHC;
for calorimeters, deterioration of resolution due to higher pile-up included

7. With fb-1 :
Mass can be measured to 0.1% (~ 100 MeV) dominated by e/μ/γ E-scale systematics
Spin/CP can be determined to > 5σ for a pure 0+ state.
If mixed state  HL-LHC needed to probe non-maximal CP-violation in the Higgs sector
Assuming SM ΓH and one scale factor for fermion/vector sector
kF, kV to 3-9% (2-4%) with 300 (3000) fb-1 per experiment
Note: theoretical errors (not included) dominate !
Without assumptions on ΓH, only ratios of couplings can be
measured at LHC. Without constraints, typical precisions are
10-50% with ~ 300 fb-1
5-25% with 3000 fb-1
Statistical error, or systematics scaling with statistics,
dominate in most cases  big gain with 3000 fb-1
W/Z H bb not included; H ττ only VBF  conservative
Γt/Γg and ΓZ/Γγ give access to gg  H and H γγ loops
Γτ/Γμ gives insight into 2nd/3rd generation couplings
per experiment
Experimental precision will challenge theoretical uncertainties in several cases
 improved theoretical calculations needed

8. Rare channels only measurable at HL-LHC with 3000 fb-1
1-lepton final state
ttH  ttγγ
~ 200 S events expected with 3000 fb-1
S/B ~ 0.2
Rate can be measured to better than 20%
Gives access to ttH coupling
H  μμ
BR ~ 2x10-4
σxBR ~ 10 fb at 14 TeV
S/B ~ 0.2%
Significance ~ 6σ with 3000 fb-1
Gives acces to Higgs couplings to
second-generation fermions

9. Higgs self-coupling ~ 30% measurement of λ/λSM may be achieved with
mH2 = 2  v2
Higgs self-coupling
Triple-Higgs couplings measured from
double Higgs production
Negative interference with independent
production of two Higgs
σ (H HH) ~ 35 fb at 14 TeV
Most promising channel: HH  bbγγ
Expect S~ 15 events, B~ 24 events for 3000 fb-1
Background dominated by ttH ttγγ
Significance:~ 3σ per experiment
HH bbττ also promising  needs further study
~ 30% measurement of λ/λSM may be achieved with
two channels and two experiments

10. Vector-boson scatteringq W H
Without Higgs, VV scattering violates
unitarity at m(WW) ~ TeV
Does the new particle fix the SM problems at high energy ?
Fully or only partially (e.g. if H is non-standard) ?
 need √s ~ 14 TeV and 3000 fb-1
Search for ZZ  4l resonances in
pp ZZjj (with high-mass forward jet tags)
in a model including a Higgs.
With 3000 fb-1:
limits in the TeV range
SM ZZjj can be measured to ~ 10%
Energy upgrade of LHC very useful in
this case

11. Physics potential of HL-LHC is much more than just Higgs
Is a scenario which only modifies the Higgs couplings by few % with no other
observable effects (new particles/phenomena) realistic ?
If new physics exists at the TeV scale: HL-LHC extends mass reach of the LHC
by typically 30% (15%) for single (pair-produced) particles.

12. SPARES

13. SPIN and CP
5σ separation 2+/0+ and 0+/0- for CP-pure state with 300 fb-1
a1,a2 : CP-even (a2=0 here)
a3 : CP-odd
Max
CP-violation

14. Is the Higgs mass stabilized by New Physics ?
ATLAS results on stop-pair searches
as presented at September 2012 SPC

15. ATLAS results on stop-pair searches TODAY
September 2012 SPC

16. Present status of stop searches
With full dataset: expect to cover stop masses up to ~ 700 GeV

17. BIG THANKS to the LHC team !
The first LHC proton-proton run ( ) is over
Luminosity delivered to ATLAS since the beginning
2012:
23 fb-1
at 8 TeV
2011
5.6 fb-1
at 7 TeV
2010
0.05 fb-1
4th July seminar
and ICHEP
Peak luminosity:
~ 7.7 x1033 cm-2 s-1
Total:
28.6 fb-1 delivered
26.9 fb-1 recorded
BIG THANKS to the LHC team !

18. The BIG challenge in 2012: PILE-UP
Experiment’s
design value
(expected to be
reached at L=1034 !)
Z μμ event from 2012 data with 25 reconstructed vertices
Z μμ

19. Present LHC upgrade plans

20. SM Higgs production cross-section and decay modes
Most sensitive channels (decreasing order)
for 120 < m < 130 GeV:
H ZZ* 4l, H γγ, H WW* lνlν
H ττ
W/ZH W/Z bb
Challenges: tiny rates, small S/B
√s=7  8 TeV: signal cross-section increases by
~1.3 for m ~ 125 GeV; similar increase for main
backgrounds  expected increase in sensitivity: 10-15%
Huge efforts and progress from theory community to compute NLO/NNLO cross-sections
for Higgs production (today: ~ 15% uncertainty on dominant ggF process, 4% for B(H γγ)) and for (often complex !) backgrounds

22. ~ 7 σ All channels together Maximum significance at mH ~ 125 GeV
December 2012
p-value: it measures consistency of data with background-only expectation
H bb
H ττ
H WW  lνlν
H ZZ  4l
H γγ
Combination
H 4l: σ
H γγ: 6.1 σ  discovery in a single channel
Total: ~7 σ
Maximum significance at mH ~ 125 GeV
~ 7 σ

23. Spin studies : H γγ
From distribution of polar angle θ* of the di-photon system in the Higgs rest frame
Compare θ* distribution in the region of the peak for:
spin-0 hypothesis: flat before cuts
spin-2 hypothesis: ~ 1+6cos2θ* +cos4θ* for Graviton-like (gg  G production)
Fit to data gives: compatibility within 0.5σ with spin-0 and 1.4σ with spin-2 ggG
(i.e. spin-2 disfavoured at 91% CL)

24. Spin-parity studies : H 4l
From distributions of 5 production and decay angles combined in BDT or Matrix Element (MELA) discriminants
Spin-parity studies : H 4l
0+ vs 0-
G-like spin-2
gg production
0+ vs 2-
0+ vs 2-
Expected separation: ~ 1.7σ
Compatibility with data:
0+ : 0.15 σ
2- : 1.9 σ
0+ vs 0-
Expected separation: ~ 1.7σ
Compatibility with data:
0+ : 0.5 σ
0- : 2.8 σ

25. Mass and signal strengths
December 2012
mH = GeV ± 0.7 GeV
= ± 0.3 (stat) ± 0.6 (syst)
μ = 1.35 ± 0.24
= 1.35 ± 0.19 (stat) ± 0.15 (syst)