Recent Results from Tommaso Lari INFN Milano SUSY searches in - - PowerPoint PPT Presentation
Interpreting the LHC data, Trieste 2019/05/28 Recent Results from Tommaso Lari INFN Milano SUSY searches in ATLAS On behalf of the ATLAS Collaboration WHAT WE ARE LOOKING FOR Mixing of winos, binos, zinos, higgsinos Thats a lot of new
Interpreting the LHC data, Trieste 2019/05/28
Tommaso Lari INFN Milano On behalf of the ATLAS Collaboration
❖
That’s a lot of new particles and decay modes we can potentially discover
❖
Couplings are fixed by the symmetry with SM
❖
Masses are unknown, until we figure out how the symmetry is broken Mixing of winos, binos, zinos, higgsinos
predicted and natural
mass required to be light by naturalness.
particle stable, pair production
3
Bullet Cluster (astro-ph/0608407)
χ
~
˜ H
˜ tL
˜ bL ˜ tR
˜ g
natural SUSY decoupled SUSY
˜ W
˜ B
˜ Li, ˜ ei ˜ bR
˜ Q1,2, ˜ u1,2, ˜ d1,2
arXiv:1110.6926v1
❖
Strong production - light flavour squarks and gluinos
❖
Strong production - third generation squarks
❖
Electroweak production - sleptons, charginos, neutralinos Focus on recent results
4
Susy particle mass [GeV] 500 1000 1500 2000 2500 Cross-section [pb]
5 −
10
4 −
10
3 −
10
2 −
10
1 −
10 1 10
2
10
3
10
4
10 = 13 TeV s NLO + NLL, pp,
(higgsino-like)
2
χ ∼
± 1
χ ∼ (wino-like)
2
χ ∼
± 1
χ ∼ g ~ g ~ q ~ q ~ t ~ t ~
p p
¯ t t
1
1
¯ t t
100 events in run2
5
❖ High cross section => probing high masses ❖ R-parity conservation => lightest particle (LSP) stable, missing momentum ❖ High pT jets
❖ Each curve is a different
model, combining limits from relevant channels
❖ All curve but yellow have
neutralino LSP (mass on y- axis)
❖ For light LSP, limits in
1800-2250 GeV
❖ For small g-LSP mass
difference, mass down to 1000-1200 GeV allowed
6
~
ATLAS SUSY summary plots
7
Two same sign (SS) or three leptons : clean channel, small SM backgrounds Several target models :
˜ b ˜ b ˜ χ±
1
˜ χ⌥
1
p p t ˜ χ0
1
W t ˜ χ0
1
W
˜ t1 ˜ t1 ˜ χ0
2
˜ χ±
1
˜ χ0
2
˜ χ±
1
p p t W W ⇤ ˜ χ0
1
¯ t W ⌥ W ⇤ ˜ χ0
1
g ˜ g ˜ χ±
1
˜ χ0
2
˜ χ±
1
˜ χ0
2
p p q q W Z ˜ χ0
1
q q W Z ˜ χ0
1
˜ g ˜ g ˜ t ˜ t p p t λ00
313
b d t b d
Several (“cut and count”) selections target different model / mass regions :
2 leptons (e,µ) with pT > 20 GeV Either SS or with third lepton with pT > 10 GeV
139 fb-1 ATLAS-CONF-2019-015
SR n` nb nj Emiss
T
[GeV] meff [GeV] Emiss
T
/meff SUSY Rpv2L 2 (`±`±) 6 (pT > 40 GeV)
g ! t˜ t⇤
1, ˜
t⇤
1 ! qq0 (00 , 0)
˜ g ! t¯ t ˜
1, ˜ 1 ! 3q (00 , 0)
˜ g ! q ¯ q ˜
1, ˜ 1 ! qq0` (0 , 0)
Rpc2L0b 2 (`±`±) = 0 6 (pT > 40 GeV) > 200 > 1000 > 0.2 ˜ g ! q ¯ q0WZ ˜
1
Rpc2L1b 2 (`±`±) 1 6 (pT > 40 GeV)
˜ b1 ! tW ˜
1
Rpc2L2b 2 (`±`±) 2 6 (pT > 25 GeV) > 300 > 1400 > 0.14 ˜ b1 ! tW ˜
1
˜ g ! t¯ t ˜
1
Rpc3LSS1b 3 (`±`±`±) 1 no cut but veto 81 GeV < me±e± < 101 GeV > 0.14 ˜ t1 ! tW±(W⇤) ˜
1
❖ Irreducible from MC, the main ones (ttW, ttZ, WZ) compared to data in
dedicated validation regions
❖ Charge flip electrons : 2 OS ee in data reweighted by measured (in Z
events) charge flip probability
❖ Fake and non prompt leptons : matrix method (loose selection in data,
plus loose-to-tight probabilities measured in control samples)
8
ATLAS-CONF-2019-015
9
No significant excess
Signal selections are not orthogonal ATLAS-CONF-2019-015
1 10
2
10
3
10
Events
Data Total uncertainty Fake/non-prompt Charge-flip VV, 3t, 4t t H, t t t(W)Z, t Z t t W t t WZ WW, ZZ, VH, VVV
ATLAS Preliminary
= 13 TeV, 139 fb s
VRttV VRWZ4j VRWZ5j Rpc2L0b Rpc2L1b Rpc2L2b Rpc3LSS1b Rpv2L
0.5 1 1.5 Data/SM
10
ATLAS-CONF-2019-015
1000 1200 1400 1600 1800 2000 2200 ) [GeV] g ~ m( 500 1000 1500 2000 2500 ) [GeV]
1
χ ∼ m(
)
exp
σ 1 ± Expected Limit ( )
SUSY theory
σ 1 ± Observed Limit ( [arXiv:1706.03731]
SS/3L obs. 36 fb
) < m(Z)
1χ ∼ ,
2χ ∼ m( Δ ) < m(W ),
2χ ∼ ,
± 1χ ∼ m( Δ )
1χ ∼ ) < m( g ~ m( ))/2
1χ ∼ ) + m(
1 ±χ ∼ ) = (m(
2χ ∼ ))/2, m(
1χ ∼ ) + m( g ~ ) = (m(
1 ±χ ∼ ; m(
1χ ∼ qqWZ → g ~ production, g ~ g ~
All limits at 95% CL
=13 TeV, 139 fb s Preliminary ATLAS
600 800 1000 1200 1400 1600 1800 2000 2200 ) [GeV] g ~ m( 400 600 800 1000 1200 1400 1600 1800 2000 ) [GeV] t ~ m(
)
exp
σ 1 ± Expected Limit ( )
SUSY theory
σ 1 ± Observed Limit ( [arXiv:1706.03731]
SS/3L obs. 36 fb
) + m(t) t ~ ) < m( g ~ m( d b → t ~ , t t ~ → g ~ production, g ~ g ~
All limits at 95% CL
=13 TeV, 139 fb s Preliminary ATLAS
˜ g ˜ g ˜ χ±
1
˜ χ0
2
˜ χ±
1
˜ χ0
2
p p q q W Z ˜ χ0
1
q q W Z ˜ χ0
1
˜ g ˜ g ˜ t ˜ t p p t λ00
313
b d t b d
Susy particle mass [GeV] 500 1000 1500 2000 2500 Cross-section [pb]
5 −
10
4 −
10
3 −
10
2 −
10
1 −
10 1 10
2
10
3
10
4
10 = 13 TeV s NLO + NLL, pp,
(higgsino-like)
2
χ ∼
± 1
χ ∼ (wino-like)
2
χ ∼
± 1
χ ∼ g ~ g ~ q ~ q ~ t ~ t ~
100 events in run2
11
❖ The simplest model has a stop
decaying directly to top LSP (blue curve)
❖ If charginos and heavy
neutralinos are in the decay chain, more complex final states
❖ Typical limits range from 400
GeV (small mass difference) to 1000 GeV (light LSP)
❖ Here focus on some recent
results
12
ATLAS SUSY summary plots
❖ Direct decay to LSP, but mass difference not enough
for on-shell top quark
❖ Limits 36 fb-1 : 420-580 GeV on stop mass ❖ Soft decay products, difficult S/B discrimination
˜ t ˜ t W W p p ˜
1
b ` ⌫ ˜
1
b q q
˜ t1 → b f f ˜ χ0
1
˜ t1 → b W ˜ χ0
1
˜ t
1
→ t ˜ χ
1
∆m > ∆m > m˜
t
1
∆m > mW + mb ∆m > 0 ∆m > m˜
t
m > mW +
∆m = m˜
t1 − m˜ χ0
1
0 100 200 300 100 200 0 100
˜ t1 → c˜ χ0
1
∆ m > mt
m˜
t1 < m˜ χ0
1
] m˜
t1 [GeV]
] m˜
χ0
1 [GeV]
13
ATLAS-CONF-2019-017 139 fb-1
Pre-selection Discriminating variables
14
❖ Neural network to discriminate signal
and background
❖ Top MC normalized to data at low NN
classifier values (+ tighter MT cut to increase purity) ATLAS-CONF-2019-017
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
bWN
NN 1 10
2
10
3
10
4
10
5
10 Events
Total SM 2L t t 1L t t W+jets Single top Multi-boson +V t t )=(450,300) GeV
1
χ ∼ , t ~ m( )=(500,380) GeV
1
χ ∼ , t ~ m(
ATLAS Simulation Preliminary
= 13 TeV, 139.0 fb s
Preselection
❖ Cut and count for discovery ❖ Shape of NN output for
exclusion
❖ Considerable extension of previous limits ❖ Sensitive to part of the 4-body decay region as well
15
ATLAS-CONF-2019-017
0.4 0.5 0.6 0.7 0.8 0.9 1
bWN
NN 2 − 1 − 1 2 Significance
10
2
10
3
10
4
10
5
10 Events
Data Total SM t t +V t t W+jets Single top Multi-boson
)=(500,380) GeV
1
χ ∼ ,
1
t ~ m(
ATLAS Preliminary
= 13 TeV, 139.0 fb s
200 300 400 500 600 700 800 ) [GeV]
1
t ~ m( 100 200 300 400 500 600 700 800 ) [GeV]
1
χ ∼ m(
) > m(t)
1
χ ∼ ) - m(
1
t ~ m( ) > m(W) + m(b)
1
χ ∼ ) - m(
1
t ~ m( ) < 0
1
χ ∼ )- m(
1
t ~ m(
1
χ ∼ t →
1
t ~ ,
1
χ ∼ bW →
1
t ~ ,
1
χ ∼ bff' →
1
t ~ production,
1
t ~
1
t ~
)
th
σ 1 ± Observed limit ( )
exp
σ 1 ± Expected limit ( JHEP 06 (2018) 108
ATLAS Preliminary
= 13 TeV, 139.0 fb s Limit at 95% CL
˜ b ˜ b ˜ χ±
1
˜ χ⌥
1
p p t ˜ χ0
1
W t ˜ χ0
1
W
t1 ˜ t1 ˜ χ0
2
˜ χ±
1
˜ χ0
2
˜ χ±
1
p p t W W ⇤ ˜ χ0
1
¯ t W ⌥ W ⇤ ˜ χ0
1
χ0
1
χ±
1
χ0
2
t1
ATLAS-CONF-2019-015
600 650 700 750 800 850 900 950 1000 1050 1100 ) [GeV] b ~ m( 200 400 600 800 1000 ) [GeV]
1χ ∼ m(
)
exp
σ 1 ± Expected Limit ( )
SUSY theory
σ 1 ± Observed Limit ( [arXiv:1706.03731]
SS/3L obs. 36 fb
) + 100 GeV
1χ ∼ ) < m(t) + m( b ~ m( ) + 100 GeV
1χ ∼ ) = m(
1 ±χ ∼ , m(
1 ±χ ∼ t →
1b ~ production,
1b ~
1b ~
All limits at 95% CL
=13 TeV, 139 fb s Preliminary ATLAS
550 600 650 700 750 800 850 900 950 ) [GeV] t ~ m(
3 −10
2 −10
1 −10 1 [pb] * BR σ
)
1χ ∼ m( ≈ )
1 ±χ ∼ )+100 GeV ; m(
1χ ∼ )=m(
2χ ∼ )-275 GeV ; m(
1t ~ ) = m(
1χ ∼ ; m(
1χ ∼ W* →
1 ±χ ∼ ,
1 ±χ ∼
±tW →
1t ~ production,
1t ~
1t ~
All limits at 95% CL
=13 TeV, 139 fb s Preliminary ATLAS
t ~ t ~ → pp Theoretical uncertainty Expected limit Observed limit σ 1 ± Expected σ 2 ± Expected [arXiv:1706.03731]
SS/3L obs. 36 fb
❖ Signature is one Z lepton pair, a third lepton, jets, large missing
transverse momentum
17
˜ t1 ˜ t1 ˜ χ0
2
˜ χ0
2
p p t ˜ χ0
1
h t ˜ χ0
1
Z
˜ t2 ˜ t2 ˜ t1 ˜ t1 p p Z ˜ χ0
1
b f f 0 Z ˜ χ0
1
b f 0 f
❖ Same as the model in previous slide, but the χ02 decays
to χ01 Z and χ01h (same BR) instead of χ 1 W
❖ lightest stop in 4-body decay region (ΔM = 40 GeV,
favoured in bino-stop coannihilation Dark Matter scenarios)
❖ Look for second stop, heavier but cleaner signature
stat ± 3
139 fb-1 ATLAS-CONF-2019-016
❖ Four signal regions, binned in one
key variable (ETMiss or pTll) for exclusion fits
❖ For each model, one SR targeting
large mass and one small mass (boosted or low pT Z)
❖ Backgrounds : ❖ ttZ and WZ normalized to data in
dedicated control regions
❖ other irreducible background
from MC
❖ fake lepton from data (matrix
method)
18
ATLAS-CONF-2019-016
❖ As usual, background
modeling checked in independent validation regions
❖ Background consistent
with data in each SR
19
ATLAS-CONF-2019-016
20
˜ t1 ˜ t1 ˜ χ0
2
˜ χ0
2
p p t ˜ χ0
1
h t ˜ χ0
1
Z
˜ t2 ˜ t2 ˜ t1 ˜ t1 p p Z ˜ χ0
1
b f f 0 Z ˜ χ0
1
b f 0 f
ATLAS-CONF-2019-016
Stop1 4-body decay limits
❖ Lower cross sections => we probe lower masses => potentially large backgrounds ❖ Electroweak states might be much lighter than strongly interacting ones.
Higgsinos most directly linked to naturalness
21
2 leptons + MET 2 bosons + MET
˜ ±
1
˜
2
W ∗ Z∗ p p ˜
1
q q ˜
1
` ` j
ISR + MET + soft leptons
wino NLSP, bino LSP
) [GeV]
2
χ ∼ ,
1 ±
χ ∼ m(
100 200 300 400 500 600 700
) [GeV]
1
χ ∼ m(
50 100 150 200 250 300
Expected limits Observed limits WW 2l
arXiv:1403.5294 ATLAS-CONF-2019-008
via
− 1
χ ∼
+
1
χ ∼
WZ 2l+3l
arXiv:1403.5294 arXiv:1712.08119 arXiv:1803.02762 arXiv:1806.02293
±
l
±
+l γ γ Wh lbb+2jbb+l
arxiv:1812.09432
via
2
χ ∼
± 1
χ ∼
All limits at 95% CL Preliminary ATLAS
=8,13 TeV, 20.3-139 fb s March 2019
)
1
χ ∼ ) = m(
2
χ ∼ m( ) + m( Z )
1
χ ∼ ) = m(
2
χ ∼ m( ) + m( h )
1
χ ∼ ) = m(
2
χ ∼ m(
22
ATLAS SUSY summary plots
❖ Two searches looked at the χ 1χ02 =>
WZ+MET case with 36 fb-1.
❖ The one based on “recursive Jigsaw”
variables found some excess in the selections targeting low mass signals
❖ We followed up on that using the full
run2 dataset
23
Signal region SR2`_High SR2`_Int SR2`_Low SR2`_ISR Total observed events 1 19 11 Total background events 1.9 ± 0.8 2.4 ± 0.9 8.4 ± 5.8 2.7+2.8
−2.7
Signal region SR3`_High SR3`_Int SR3`_Low SR3`_ISR Total observed events 2 1 20 12 Total background events 1.1 ± 0.5 2.3 ± 0.5 10 ± 2 3.9 ± 1.0
1.39σ 1.99σ 2.13σ 3.02σ 2 leptons 3 leptons
stat ± 3
arXiv:1806.02293
❖ 3-lepton channel ❖ Recursive Jigsaw [PRD 95 (2017)
035031] :
❖ Assume a specific decay chain ❖ Perform a series of Lorentz boosts between
frames
❖ Determine unknowns (like pzmiss) with
Jigsaw rules
❖ Provides 4-vector of each particle in the
assumed decay chain
❖ A new search uses “conventional”
variables that emulates the RJ variables (i.e. are very correlated)
❖ pT(I) replaced by ETMISS, ISR replaced
by sum of all jets in lab frame, etc.
❖
24
ATLAS-CONF-2019-020 139 fb-1
LAB
CM ISR
S
V
Lab State Decay States Visible States Invisible States
compressed tree
,jets)
miss T
(E φ Δ
0.5 1 1.5 2 2.5 3
(ISR,MET) φ Δ
0.5 1 1.5 2 2.5 3
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Simulation Preliminary ATLAS
) = (200,100) GeV
1
χ ∼ ,
± 1
χ ∼ /
2
χ ∼ m(
❖ Same selection as RJ 2015-16 paper in the emulated variables ❖ In SR-low, find the same 2015-16 events; a bit more data and background events for
SR-ISR, with similar significance
❖ Main irreducible background is WZ, estimated from low mT CR, validated in VRs ❖ Fake lepton background estimated from anti-ID lepton based selection times the ID/
antiID ratios measured in data control samples (fake factor method)
25 3 leptons, one SFOS pair from the Z, b-veto
ATLAS-CONF-2019-020
Selection Criteria Low-mass Region p`1
T [GeV]
p`2
T [GeV]
p`3
T [GeV]
mT [GeV] Emiss
T
[GeV] Hboost [GeV]
m3`
eff
Hboost psoft
T
psoft
T +m3` eff
CR-low > 60 > 40 > 30 ∈ (0, 70) > 40 > 250 > 0.75 < 0.2 VR-low > 60 > 40 > 30 ∈ (70, 100)
> 0.75 < 0.2 SR-low > 60 > 40 > 30 > 100
> 0.9 < 0.05 ISR Region p`1
T [GeV]
p`2
T [GeV]
p`3
T [GeV]
mT [GeV] Emiss
T
[GeV] |∆φ ⇣ Emiss
T
, jets ⌘ | R ⇣ Emiss
T
, jets ⌘ p jets
T
[GeV] psoft
T
[GeV] CR-ISR > 25 > 25 > 20 < 100 > 60 > 2.0 ∈ (0.55, 1.0) > 80 < 25 VR-ISR > 25 > 25 > 20 > 60 > 60 > 2.0 ∈ (0.55, 1.0) > 80 > 25 VR-ISR-small psoft
T
> 25 > 25 > 20 > 60 > 60 > 2.0 ∈ (0.55, 1.0) < 80 < 25 VR-ISR-small R ⇣ Emiss
T
, jets ⌘ > 25 > 25 > 20 > 60 > 60 > 2.0 ∈ (0.30, 0.55) > 80 < 25 SR-ISR > 25 > 25 > 20 > 100 > 80 > 2.0 ∈ (0.55, 1.0) > 100 < 25
❖ No significant excess in full
run2 dataset
26
ATLAS-CONF-2019-020
❖ Rare but clean decay - best channel at low mass ❖ Both W hadronic and leptonic decays targeted ❖ Events divided in categories
˜ χ±
1
˜ χ0
2
h p p ˜ χ0
1
W ˜ χ0
1
γ γ
(a)
27
2 photons, 105 < m(γγ) < 160, ET(γ)/m(γγ) cuts
ATLAS-CONF-2019-019
5 10 15 20
] GeV [
T miss
E
S
1 2
Data / Bkg
1 −
10 1 10
2
10
3
10
4
10
5
10
6
10
7
10
GeV Events /
Data γ γ +jets γ SM Higgs boson γ V γ γ V
⊕ Syst.
) = (200,0.5) GeV
1χ ∼ ,
2χ ∼ /
± 1χ ∼ ( m ,
1χ ∼ h
1χ ∼
±W →
2χ ∼
± 1χ ∼ ) = 1 MeV G ~ ( m ) = 150 GeV,
1χ ∼ ( m , G ~ h G ~ Z →
1χ ∼
1χ ∼ ) = 1 MeV G ~ ( m ) = 150 GeV,
1χ ∼ ( m , G ~ h G ~ h →
1χ ∼
1χ ∼
Preliminary ATLAS
= 13 TeV, 139 fb s
❖ Non resonant background from sideband fits ❖ SM Higgs from MC ❖ No significant excess
28
ATLAS-CONF-2019-019
110 120 130 140 150 160
[GeV]
γ γ
m
5 10 15 20 25 30
Events / 5 GeV
Category 3
Data Non-resonant Bkg Fitted Signal SM Higgs Total
Preliminary ATLAS
= 13 TeV, 139 fb s
One of the twelve categories
1 −
10 1
[fb] ε × A × σ =
vis BSM
σ
C a t e g
y 1 C a t e g
y 2 C a t e g
y 3 C a t e g
y 4 C a t e g
y 5 C a t e g
y 6 C a t e g
y 7 C a t e g
y 8 C a t e g
y 9 C a t e g
y 1 C a t e g
y 1 1 C a t e g
y 1 2
Observed limit Expected limit limit σ 1 ± Expected limit σ 2 ± Expected
Preliminary ATLAS
T miss
+ E γ γ →
T miss
+ E
125 GeV
h → pp
Limits at 95% CL
= 13 TeV, 139 fb s
29
ATLAS-CONF-2019-019 200 300 400 500 600 700 800
) [GeV]
2
χ ∼ /
± 1
χ ∼ ( m
50 100 150 200 250 300 350
) [GeV]
1
χ ∼ ( m
Preliminary ATLAS
= 13 TeV, 139 fb s
h,
1
χ ∼
±
W
1
χ ∼ →
2
χ ∼
± 1
χ ∼
) = 100%
1
χ ∼ h →
2
χ ∼ ) = BR(
1
χ ∼
±
W →
± 1
χ ∼ BR( Limits at 95% CL
) + 125 GeV
1
χ ∼ ) < m(
2
χ ∼ /
1 ±
χ ∼ m(
γ γ W , [arXiv:1812.09432] b 1lb , [arXiv:1812.09432] b 0lb , [arXiv:1812.09432]
±
l
±
l , [arXiv:1812.09432] γ γ 1l Observed Expected
❖ Relatively small cross section ❖ Until recently, we could not probe staus - run2 cross section and luminosity
larger, but triggers tighter, so it was not trivial to improve on run1 result
[pb] σ
2 −
10
1 −
10 1 10
)= 0 GeV
1χ ∼ m( Observed 95% CL Expected 95% CL ) σ 1 ± NLO Theory (
[GeV]
τ ∼
m 100 150 200 250 300 [pb] σ
2 −
10
1 −
10 1 10
)=60 GeV
1χ ∼ m( )=20 GeV
1χ ∼ m( ATLAS
=8 TeV, 20.1 fb s
[GeV]
τ ∼m 100 150 200 250 300
)=80 GeV
1χ ∼ m( )=40 GeV
1χ ∼ m(
[GeV]
τ ∼m 100 150 200 250 300
)=100 GeV
1χ ∼ m( 1
χ ∼ τ × 2 →
R,L∼
R,L +τ ∼ Simplified Model:
4-fold degeneracy : eReR+eLeL+µRµR+µLµL 2-fold degeneracy : τRτR+τLτL
30
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ arXiv:1509.07152
ATLAS SUSY summary plots
❖ had-had channel only ❖ two selections combined for
limits
❖ Backgrounds : ❖ QCD from ABCD method ❖ W+jets normalized in CR ❖ Others (VV, top, Z+jets) from
MC, checked in VRs
31
trigger
pT(tau1)
pT(tau2)
MET asymmetric di-tau 95 GeV 60-75 GeV
40 GeV 150 GeV
ATLAS-CONF-2019-018 139 fb-1
❖ QCD from ABCD method ❖ W+jets : normalized in CR, VR checks mT
shape
❖ Z+jets from MC, VR inverts mT2 ❖ VV from MC, VR inverts m(ττ) ❖ top from MC, VR inverts b-veto
32
ATLAS-CONF-2019-018
Tau-id and charge (mT2 , ET
miss)
[GeV]
Multi-jet VR−E Multi-jet VR−F
Used for validation and systematics
≥ 2 loose ts 2 tight ts (OS) < 2 medium ts (OS)
Multi-jet CR−A
Used for nominal ABCD method
Multi-jet CR−B Multi-jet CR−C SR−lowMass
T T
(>7075-150) (30-70<150) (10-30<150)
lowMass
70 75 80 85 90 95 100 [GeV]
T2
m 50 100 150 200 250 300 350 Events / 5 GeV
data SM Total W+jets Multi-jet Top quark Multi-boson Z+jets Higgs ) = (120, 1) GeV
1
χ ∼ , τ ∼ m( ) = (280, 1) GeV
1
χ ∼ , τ ∼ m(
= 13 TeV, 139 fb s
ATLAS Preliminary
WVR post-fit
Events
1 10
210
310
410
510
Data SM Total Multi-jet Top quark Z+jets Multi-boson Higgs W+jets ) = (120, 0) GeV 1 χ ∼ , τ ∼ m( ) = (280, 0) GeV 1 χ ∼ , τ ∼ m(post-fit
= 13 TeV, 139 fb s
Preliminary ATLAS
TVR-lowMass ZVR-lowMass VVVR-lowMass TVR-highMass ZVR-highMass VVVR-highMassData/SM
1 2
33
ATLAS-CONF-2019-018
80 100 120 140 160 180 200 220 [GeV]
T2
m 2 4 6 8 10 12 Events / 30 GeV
data SM Total Multi-jet Multi-boson W+jets Top quark Z+jets Higgs ) = (120, 1) GeV
1
χ ∼ , τ ∼ m( ) = (280, 1) GeV
1
χ ∼ , τ ∼ m(
= 13 TeV, 139 fb s
ATLAS Preliminary
SR-highMass post-fit
70 75 80 85 90 95 100 105 110 115 120 [GeV]
T2
m 2 4 6 8 10 Events / 10 GeV
data SM Total Multi-jet Multi-boson W+jets Top quark Z+jets Higgs ) = (120, 1) GeV
1
χ ∼ , τ ∼ m( ) = (280, 1) GeV
1
χ ∼ , τ ∼ m(
= 13 TeV, 139 fb s
ATLAS Preliminary
SR-lowMass post-fit
❖ trigger and backgrounds limit the low mass sensitivity ❖ significant area excluded, also for a single left eigenstate ❖ no exclusion for stau right (3 times less cross section) yet
34
2-fold degeneracy : τRτR+τLτL
τLτL only
ATLAS-CONF-2019-018 ~~ ~ ~ ~~
100 150 200 250 300 350 400 450 500
) [GeV] τ ∼ m(
50 100 150 200 250 300
) [GeV]
1
χ ∼ m(
)
1χ ∼ ) < m( τ ∼ m(
)
exp
σ 1 ± Expected Limit ( )
SUSY theory
σ 1 ± Observed Limit (
1
χ ∼ τ × 2 →
τ ∼
+ R,L
τ ∼
Preliminary ATLAS
=13 TeV, 139 fb s All limits at 95% CL SR-combined
100 150 200 250 300 350 400 450 500
) [GeV] τ ∼ m(
50 100 150 200 250 300
) [GeV]
1
χ ∼ m(
)
1χ ∼ ) < m ( τ ∼ m (
)
exp
σ 1 ± Expected Limit ( )
SUSY theory
σ 1 ± Observed Limit (
1
χ ∼ τ × 2 →
τ ∼
+ L
τ ∼
Preliminary ATLAS
=13 TeV, 139 fb s All limits at 95% CL SR-combined
❖ Trigger on neutralinos (MET) recoiling
against an ISR jet
❖ Target small mass differences - soft
lepton pairs !
❖ An higgsino like LSP would have a
triplet of nearly mass degenerate states (χ01, χ02, χ 1)
35
˜ ±
1
˜
2
W ∗ Z∗ p p ˜
1
q q ˜
1
` ` j ˜ ` ˜ ` p p j ˜
1
` ˜
1
`
Low invariant mass Low mT2 139 fb-1 ATLAS-CONF-2019-014
stat ± 3
❖ Lepton threshold the lowest
event in ATLAS : 3(4.5) GeV for muons (electrons)
❖ Also using events where the
second “lepton” is an isolated Inner Detector track matched to an element of reconstructed electron or muon candidates (“lepton-track selection”)
36
flavour pT efficiency elec. 3-4 GeV 20% muons 2-3 GeV 35%
Efficiency of track selection for signal leptons
ATLAS-CONF-2019-014
10 20 30 40 50
[GeV]
T
p Lepton
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Efficiency
Muon Electron
Simulation Preliminary ATLAS =13 TeV s
37
ATLAS-CONF-2019-014
❖ 1L1T : background from same sign (SS) pairs ❖ 2L fakes from matrix method ❖ 2L Z+jets, top, dibosons : normalized to data in CR, validated in VR
38
VR for 1 lepton 1 track (same cuts as SR, except inverted Δφ (lepton,pTmiss)
ATLAS-CONF-2019-014
20 GeV / Events
1 10
210
310
410
510
Data Total SM )+jets τ τ → Z( , single top t t Diboson Fake/nonprompt Others
Preliminary ATLAS
= 13 TeV, 139 fb s m Δ
CR-tau-ewkino-low-E
[GeV]
τ τ
m
100 − 50 − 50 100 150 200
SM / Data
1 2
CR for Z(ττ) normalization
1 GeV / Events
1 −10 1 10
210
310
410
510
Data Total SM Data SS
Preliminary ATLAS
= 13 TeV, 139 fb s T 1 l VR-1
[GeV]
track l
m
1 2 3 4 5 6 7 8 9 10
SM / Data
1 2
❖ No significant
excess
39
ATLAS-CONF-2019-014
Events
1 −
10 1 10
2
10
3
10
Preliminary ATLAS
1 −
= 13 TeV, 139 fb s
Total SM Data
Fake/nonprompt , single top t t )+jets τ τ → Z( Diboson Others Data SS
[1, 2] [2, 3] [3.2, 5] [5, 10] [10, 20] [20, 30] [30, 40] [40, 60] [3.2, 5] [5, 10] [10, 20] [20, 30] [30, 40] [40, 60] [1, 2] [2, 3] [3.2, 5] [5, 10] [10, 20] [20, 30] [30, 40] [40, 60] [3.2, 5] [5, 10] [10, 20] [20, 30] [30, 40] [40, 60] [1, 2] [2, 3] [3.2, 5] [5, 10] [10, 20] [20, 30] [3.2, 5] [5, 10] [10, 20] [20, 30] [0.5, 1.0] [1.0, 1.5] [1.5, 2.0] [2.0, 3.0] [3.0, 4.0] [4.0, 5.0]
Significance
2 − 2
[GeV]
ll
m
µ SR [GeV]
ll
m
SR
miss T
high-E
[GeV]
ll
m
µ SR [GeV]
ll
m
SR
m ∆ , high-
miss T
low-E
[GeV]
ll
m
µ SR
ll
m
SR [GeV]
m ∆ , low-
miss T
low-E
[GeV]
track l
m SR-
T 1 l 1
40
ATLAS-CONF-2019-014
100 150 200 250 300 350
m(˜
2) [GeV]
1 5 10 50
∆m(˜
2, ˜ 1) [GeV]
Expected limit
±
1 excluded
ATLAS 13 TeV excluded
ATLAS Preliminary
√s = 13 TeV, 139 fb−1
ee/µµ, m`` shape fit All limits at 95% CL pp → ˜
2˜
±
1 , ˜ 2˜ 1, ˜
+
1 ˜
−
1 (Higgsino)
˜
2 → Z ∗˜ 1, ˜
±
1 → W ∗˜ 1
m(˜ ±
1 ) = [m(˜ 2) + m(˜ 1)]/2
100 150 200 250 300
m(˜ `L,R) [GeV]
0.5 1 5 10 50
∆m(˜ `, ˜
1) [GeV]
Expected limit (˜ µL) Observed limit (˜ µL) Expected limit (˜ µR) Observed limit (˜ µR) Expected limit (˜ eL) Observed limit (˜ eL) Expected limit (˜ eR) Observed limit (˜ eR)
ATLAS Preliminary
√s = 13 TeV, 139 fb−1
ee/µµ, m100
T2 shape fitAll limits at 95% CL pp → ˜ `+
L,R˜`−
L,R, ˜` → `˜
1, ` ∈ [e, µ] ❖ Searches for supersymmetry are pursued vigorously - several new
results have just been released with the full run2 dataset
❖ No hint of a signal, limits are extended both at high mass and
filling more difficult low mass scenarios
❖ First limits on degenerate and left stau direct production have been
set !
41
❖
Naturalness (need light higgsinos) and relic density consideration make an 100% stop decay to the LSP not very
❖
For 300 GeV LSP mass, any stop mass is likely viable for some decay modes 43 Higgsino-like LSP Higgsino LSP Bino LSP Bino LSP Wino NLSP t N1 t N1 t N2 b C1 Z/h/W N1 t N2 t N1 b C1 Stealth stop hole
Hyperparameters of the neural network To increase statistics, the signal training sample has been generated by using truth with detector performance smearing (as in high luminosity LHC studies)
44
❖ Updated projections prepared for the European Strategy
process
400 600 800 1000 1200 1400 1600 1800 2000 [GeV] t ~ m 200 400 600 800 1000 1200 1400 [GeV]
1
χ ∼ m
Not explored by this analysis
1χ ∼ < m t ~ m
1χ ∼ + m t = m t ~ m
= 15%
syst
σ 95% CL exclusion, = 30%
syst
σ 95% CL exclusion, = 15%
syst
σ discovery, σ 5 95% CL exclusion
36.1 fb
1
χ ∼
1
χ ∼ t t → t ~ t ~
=14 TeV, 3 ab s ATLAS Simulation Preliminary Excluded by the 140 fb-1 3-body search presented earlier 45
) [GeV]
2
χ ∼ ,
1 ±
χ ∼ m( 200 400 600 800 1000 1200 1400 1600 ) [GeV]
1
χ ∼ m( 200 400 600 800 1000
Kinematically Forbidden )
exp
σ 1 ± 95% CL exclusion ( discovery σ 5
ATLAS
miss T
+ E b + b µ 1 e/ →
1
χ ∼ h
1
χ ∼
±
W →
2
χ ∼
1 ±
χ ∼ Simulation Preliminary
=14 TeV, 3000 fb s
All limits at 95% CL
200 300 400 500 600 700 800
) [GeV]
2χ ∼ /
± 1χ ∼ ( m
50 100 150 200 250 300 350
) [GeV]
1χ ∼ ( m
Preliminary ATLAS
˜ ±
1
˜
2
W h p p ˜
1
` ⌫ ˜
1
b b
46
1L channel ATLAS-PUB-2018-048
) [GeV]
2
χ ∼ ,
± 1
χ ∼ m( 500 600 700 800 900 1000 1100 1200 1300 1400 ) [GeV]
1
χ ∼ m( 200 400 600 800 1000 1200
Kinematically Forbidden
Z
) = m
1
χ ∼ ) - m(
2
χ ∼ ,
± 1
χ ∼ m(
ATLAS 13 TeV, 36 fb ), multi-bin
exp
σ 1 ± 95% CL exclusion ( discovery, inclusive σ 5
3L + MET final state →
1
χ ∼ Z
1
χ ∼
±
W →
2
χ ∼
± 1
χ ∼ Wino
=14 TeV, 3000 fb s
All limits at 95% CL
ATLAS Simulation Preliminary
2L+3L combination
47
ATLAS-PUB-2018-048
48
) [GeV]
± 1
χ ∼ m( 200 300 400 500 600 700 800 900 1000 ) [GeV]
1
χ ∼ m( 100 200 300 400 500 600 700 800
Kinematically Forbidden
W
) = m
1
χ ∼ )-m(
± 1
χ ∼ m(
ATLAS 13 TeV, 80 fb ), multi-bin
exp
σ 1 ± 95% CL exclusion ( discovery, inclusive σ 5
2L + MET final state →
1
χ ∼
1
χ ∼
+
W →
χ ∼
+ 1
χ ∼ Wino
=14 TeV, 3000 fb s
All limits at 95% CL Run-2 Uncertainties
ATLAS Simulation Preliminary
ATLAS-PUB-2018-048
˜ ±
1
˜ ⌥
1
W W p p ˜
1
` ⌫ ˜
1
` ⌫
49
) [GeV] τ ∼ m( 100 200 300 400 500 600 700 800 900 1000 ) [GeV]
1
χ ∼ m( 100 200 300 400 500 600 700 800
)
exp
σ 1 ± : 95% CL exclusion (
R,L
τ ∼ : 95% CL exclusion
L
τ ∼ : 95% CL exclusion
R
τ ∼ discovery σ : 5
R,L
τ ∼ discovery σ : 5
L
τ ∼ Baseline Uncertainties All limits at 95% CL )
exp
σ 1 ± : 95% CL exclusion (
R,L
τ ∼ : 95% CL exclusion
L
τ ∼ : 95% CL exclusion
R
τ ∼ discovery σ : 5
R,L
τ ∼ discovery σ : 5
L
τ ∼
=14 TeV, 3000 fb s
1
χ ∼ τ × 2 →
∼
+
τ ∼
1χ ∼
< m
τ ∼
m
ATLAS Simulation Preliminary
˜ τ ˜ τ p p ˜ χ0
1
τ ˜ χ0
1
τ
ATLAS-PUB-2018-048
100 150 200 250 300 350 400 450 500
) [GeV] τ ∼ m(
50 100 150 200 250 300
) [GeV]
1χ ∼ m(
)
1χ ∼ ) < m ( τ ∼ m (
)
expσ 1 ± Expected Limit ( )
SUSY theoryσ 1 ± Observed Limit (
1χ ∼ τ × 2 →
τ ∼
+ R,Lτ ∼
Preliminary ATLAS
=13 TeV, 139 fb s All limits at 95% CL SR-combined
50
100 200 300 400 500 ) [GeV]
± 1
χ ∼ m( 1 10 ) [GeV]
1
χ ∼ ,
± 1
χ ∼ m( Δ
All limits at 95% CL = 200 µ ,
=14 TeV, 3000 fb s ATLAS Simulation Preliminary > 0 µ = 5, β production, tan
1
χ ∼
± 1
χ ∼ ,
± 1
χ ∼
± 1
χ ∼ ,
1
χ ∼
2
χ ∼ ,
± 1
χ ∼
2
χ ∼ Pure Higgsino discovery σ 5 Expected limit Disappearing tracks Soft leptons LEP2 exclusion Theory
100 150 200 250 300 350
m(˜
2) [GeV]
1 5 10 50
∆m(˜
2, ˜ 1) [GeV]
Expected limit
±
1 excludedATLAS 13 TeV excluded
ATLAS Preliminary
√s = 13 TeV, 139 fb−1
ee/µµ, m`` shape fit All limits at 95% CL pp → ˜
2˜±
1 , ˜ 2˜ 1, ˜+
1 ˜−
1 (Higgsino)˜
2 → Z ∗˜ 1, ˜±
1 → W ∗˜ 1m(˜ ±
1 ) = [m(˜ 2) + m(˜ 1)]/2˜ ±
1
˜
2
W ∗ Z∗ p p ˜
1
q q ˜
1
` ` j ATLAS-PUB-2018-048
200 400 600 800 1000 10
−4
10
−2
10 10
2
M2 (GeV) σ (pb) Case AI M1 < M2 < |µ| χ2
0χ1 ±
χ1
+χ1 −
χ1
0χ1 ±
(a)
200 400 600 800 1000 10
−4
10
−2
10 10
2
µ (GeV) σ (pb) Case AII M1 < |µ| < M2 χ2
0χ1 ±
χ3
0χ1 ±
χ1
+χ1 −
χ2
0χ3
χ1
0χ1 ±
χ1
0χ3
χ1
0χ2
(b)
Wino NLSP Higgsino NLSP
100 150 200 250 300 350 400
m(˜ `L,R) [GeV]
0.5 1 5 10 50
∆m(˜ `, ˜
1) [GeV]
Expected limit
`L,R excluded ATLAS 13 TeV ˜ `L,R excluded LEP ˜ eR excluded
ATLAS Preliminary
√s = 13 TeV, 139 fb−1
ee/µµ, m100
T2 shape fit
All limits at 95% CL pp → ˜ `+
L,R˜
`−
L,R, ˜
` → `˜
1, ` ∈ [e, µ]
53
100 150 200 250 300 350 400
m(˜
2) = m(˜
±
1 ) [GeV]
1 5 10 50
∆m(˜
2, ˜ 1) [GeV]
Expected limit
±
1 excludedATLAS Preliminary
√s = 13 TeV, 139 fb−1
ee/µµ, m`` shape fit All limits at 95% CL pp → ˜
2˜±
1 (Wino)˜
2 → Z ∗˜ 1, ˜±
1 → W ∗˜ 1m(˜
2) × m(˜ 1) < 0100 150 200 250 300 350 400
m(˜
2) = m(˜
±
1 ) [GeV]
1 5 10 50
∆m(˜
2, ˜ 1) [GeV]
Expected limit
±
1 excluded
ATLAS 13 TeV excluded ATLAS 8 TeV excluded
ATLAS Preliminary
√s = 13 TeV, 139 fb−1
ee/µµ, m`` shape fit All limits at 95% CL pp → ˜
2˜
±
1 (Wino)
˜
2 → Z ∗˜ 1, ˜
±
1 → W ∗˜ 1
m(˜
2) × m(˜ 1) > 0
54
3 leptons 2 leptons + 2jets Targeted by two papers :
❖ “conventional” discriminating variables [EPJ C78 (2018)
995, arXiv:1803.02762]
❖ Recursive Jigsaw Reconstruction [PRD 98 (2018) 092012,
arXiv:1806.02293] Complementary (similar sensitivity to target simplified model, but selecting different events / phase space) What is Recursive Jigsaw Reconstruction ?
❖ Assume a specific decay chain ❖ Perform a series of Lorentz boosts
between frames
❖ Determine unknowns (like pzmiss) with
Jigsaw rules
❖ Provides 4-vector of each particle in the
assumed decay chain References :
Jackson, Rogan, Santini, PRD 95 (2017) 035031 http://restframes.com
50 52 54 56 58 60 62 64 66 68 70 [GeV]
T2
m 20 40 60 80 100 120 140 Events / 2.5 GeV
data SM Total Multi-jet W+jets Z+jets Top quark Multi-boson Higgs ) = (120, 1) GeV
1
χ ∼ , τ ∼ m( ) = (280, 1) GeV
1
χ ∼ , τ ∼ m(
= 13 TeV, 139 fb s
Multi-jet VR-highMass post-fit
ATLAS Preliminary 50 55 60 65 70 75 80 85 90 95 100 [GeV]
miss T
E 50 100 150 200 250 Events / 10 GeV
data SM Total Multi-jet W+jets Z+jets Top quark Multi-boson Higgs ) = (120, 1) GeV
1
χ ∼ , τ ∼ m( ) = (280, 1) GeV
1
χ ∼ , τ ∼ m(
= 13 TeV, 139 fb s
Multi-jet VR-highMass post-fit
ATLAS Preliminary
❖ We provide a benchmark simplified
model limits
❖ For interpretations in other scenarios
we provide in publications
❖ Cross section limits as a function of
mass [for models with a different σ x BR but same final state]
❖ truth-level acceptances and
efficiencies [validate simplified simulation codes]
❖ model-independent limits on the
number of BSM events
60 arXiv:1812.09432