ATLAS Heavy Flavour production Looking towards Run 2 Heavy - - PowerPoint PPT Presentation

▶
atlas heavy flavour production looking towards run 2
SMART_READER_LITE
LIVE PREVIEW

ATLAS Heavy Flavour production Looking towards Run 2 Heavy - - PowerPoint PPT Presentation

Mar 06, 2024 265 likes •775 views

ATLAS Heavy Flavour production Looking towards Run 2 Heavy Flavour at the LHC IPPP Durham 21/4/2016 Josh McFayden ATLAS Heavy Flavour production Looking towards Run 2 into Heavy Flavour at the LHC IPPP

slide-1
SLIDE 1

ATLAS Heavy Flavour production 
 Looking towards Run 2

  • Heavy Flavour at the LHC


IPPP Durham 21/4/2016

Josh McFayden

slide-2
SLIDE 2

ATLAS Heavy Flavour production 
 Looking towards Run 2 into

  • Heavy Flavour at the LHC


IPPP Durham 21/4/2016

Josh McFayden

slide-3
SLIDE 3

Josh McFayden | HF @ LHC | 21/4/2016

Chiara already showed some of the Run 1 results and 


HF-related issues that were discovered.

  • I will show a few more recent results and prospects as we

get further into Run 2:

  • A new Run 1 measurement on inclusive di-bjet production
  • MC generator setups for Run 2
  • Some early Run 2 results that are affected by HF production
  • Prospects for better HF production measurements in the future

3

Overview

slide-4
SLIDE 4

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Hot off the press

4

slide-5
SLIDE 5

Josh McFayden | HF @ LHC | 21/4/2016 5

7 TeV bb di-jet cross section

New 7 TeV ATLAS result on di-b-jet production.

One jet with pT > 270 GeV required due to trigger. b-jets with pT > 20 GeV and ∆R = 0.4. Template fit used to extract true b-b contribution.

mbb step at ~500 GeV due to “turn-on” of flavour creation. STDM-2013-03

slide-6
SLIDE 6

Josh McFayden | HF @ LHC | 21/4/2016 6

7 TeV bb di-jet cross section

New 7 TeV ATLAS result on di-b-jet production.

One jet with pT > 270 GeV required due to trigger. b-jets with pT > 20 GeV and ∆R = 0.4. Template fit used to extract true b-b contribution.

Large ΔΦ region is dominated by flavour creation and

underestimated by NLO predictions.

STDM-2013-03

slide-7
SLIDE 7

Josh McFayden | HF @ LHC | 21/4/2016

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

[pb] R(b,b) ∆ d (Zbb) σ d 0.05 0.1 0.15 0.2 0.25 0.3

a Data

  • 1

= 7 TeV, 4.6 fb s MCFM aMC@NLO 5FNS aMC@NLO 4FNS ALPGEN+HJ SHERPA

ATLAS 2 b-jets ≥ Z+

Data NLO

0.5 1 R(b,b) ∆

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Data LO multileg

0.4 0.6 0.8 1 1.2

7

What does it tell us?

These results use the same dataset…

How much do we learn about V+HF from inclusive di-b-jets?

Seems like trends might be different?

Is the large leading jet requirement good/bad?

slide-8
SLIDE 8

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Generator setups for Run 2

8

slide-9
SLIDE 9

Josh McFayden | HF @ LHC | 21/4/2016 9

Generators for Run 2

Generator Matrix Element Flavour number scheme Slicing/ filtering Comments Sherpa2.1 NLO@2j LO@4j 5fl pT(V) & HF

Known mismodelling

  • f low pT large η jets

Sherpa2.2 NLO@2j LO@4j 5fl pT(V) & HF

Improvement in low pT large η jets

MG+Py8 A LO@4j 5fl N-parton

Mismodelling of jet pT (too hard)

MG+Py8 B LO@4j 5fl HT & HF

NLO PDF, difgerent shower settings. (still too hard)

MG5_aMC+Py8 FxFx NLO@2j 5fl TBD

Very promising - some N-jets mismodelling

Alpgen LO@5j 4fl N-parton & 
 b/c/light

The new old! HFOR can be problematic.

slide-10
SLIDE 10

Josh McFayden | HF @ LHC | 21/4/2016

What appears to be ~small slope at 7 TeV seems to

becomes much more significant at 13 TeV.

Important to make new measurements at 13 TeV.

10

7TeV vs 13TeV | Z+jets

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG+Py8 A MG+Py8 B aMC@NLO FxFx 10−4 10−3 10−2 10−1 1 Transverse momentum of 1st jet dσ/dp⊥ [pb/GeV] 100 200 300 400 500 600 700 0.6 0.8 1 1.2 1.4 p⊥(1st jet) [GeV] MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 10−3 10−2 10−1 1 10 1 Transverse momentum of 1st jet dσ/dp⊥ [pb/GeV] 100 200 300 400 500 600 700 0.6 0.8 1 1.2 1.4 p⊥(1st jet) [GeV] Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-11
SLIDE 11

Josh McFayden | HF @ LHC | 21/4/2016 11

Looking back…

)

2

,jet

1

(jet φ ∆ 0.5 1 1.5 2 2.5 3 Data / Pred 0.9 1 1.1

0.5 1 1.5 2 2.5 3

rad

  • 2

x10 π Events / 5000 10000 15000 20000 25000

Data 2012 =1.0) µ VH(bb) ( Diboson t t Single top Multijet W+hf W+cl W+l Z+hf Z+cl Z+l

ATLAS = 8 TeV s

  • 1

L dt = 20.3 fb

∫

[GeV]

W T

p 50 100 150 200 250 Data / Pred 0.9 1 1.1

50 100 150 200 250

Events / 5 GeV 10000 20000 30000 40000 50000 60000 70000 80000

Data 2012 =1.0) µ VH(bb) ( Diboson t t Single top Multijet W+hf W+cl W+l Z+hf Z+cl Z+l

ATLAS = 8 TeV s

  • 1

L dt = 20.3 fb

∫

7 TeV VH(→bb) analysis saw large mismodelling of Δφ(j,j)

and discrepancies between generators in m(j,j).

Difference between Sherpa and Alpgen

Are in a better position with our Run 2 generators setups?

JHEP01(2015)069

slide-12
SLIDE 12

Josh McFayden | HF @ LHC | 21/4/2016

Some systematic difference between Sherpa and

MG5_aMC+Py8 CKKW-L

The data sits between the two. The modelling seems to be improved but much less data.

12

7TeV vs 13TeV | Z+jets

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG5 aMC+Py8 CKKW-L A MG5 aMC+Py8 CKKW-L B MG5 aMC+Py8 FxFx 1 10 1 dσ/d|∆φ| [pb] 0.5 1 1.5 2 2.5 3 0.6 0.8 1 1.2 1.4 |∆φ|(1st jet, 2nd jet) MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 10 1 Azimuthal distance of leading jets dσ/d|∆φ| [pb] 0.5 1 1.5 2 2.5 3 0.6 0.8 1 1.2 1.4 |∆φ|(1st jet, 2nd jet) Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-13
SLIDE 13

Josh McFayden | HF @ LHC | 21/4/2016

Some systematic difference between Sherpa and

MG5_aMC+Py8 CKKW-L

The data seems to prefer the MG5_aMC+Py8 CKKW-L shape.

13

7TeV vs 13TeV | Z+jets

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG5 aMC+Py8 CKKW-L A MG5 aMC+Py8 CKKW-L B MG5 aMC+Py8 FxFx 10−3 10−2 10−1 dσ/dm [pb/GeV] 200 400 600 800 1000 0.6 0.8 1 1.2 1.4 m(1st jet, 2nd jet) [GeV] MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 10−2 10−1 Invariant mass of leading jets dσ/dm [pb/GeV] 200 400 600 800 1000 0.6 0.8 1 1.2 1.4 m(1st jet, 2nd jet) [GeV] Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-14
SLIDE 14

Josh McFayden | HF @ LHC | 21/4/2016

Main difference between Sherpa2.1 and MG5_aMC CKKW-L is in

the rate

From early analyses we see that the data seems to prefer the higher rate. Shape deviations are more important.

14

7TeV vs 13TeV | Z+b(b)

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG5 aMC+Py8 CKKW-L A MG5 aMC+Py8 CKKW-L B MG5 aMC+Py8 FxFx 10−4 10−3 10−2 10−1 dσ(Zb)/dpT /Nb-jets [pb/GeV] 10 2 0.6 0.8 1 1.2 1.4 b-jet pT MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 10−3 10−2 10−1 1 Z+ ≥ 1 b-jet dσ(Zb)/dpT/Nb-jets [pb/GeV] 10 2 0.6 0.8 1 1.2 1.4 b-jet pT Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-15
SLIDE 15

Josh McFayden | HF @ LHC | 21/4/2016

Perhaps larger differences observed at 7 TeV than 13 TeV.

We have enough data to constrain the MC prediction here. Improvement by going to NLO.

15

7TeV vs 13TeV | Z+b(b)

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG5 aMC+Py8 CKKW-L A MG5 aMC+Py8 CKKW-L B MG5 aMC+Py8 FxFx 1 10 1 dσ(Zb)/d∆φ(Z, b)/Nb-jets [pb] 0.5 1 1.5 2 2.5 3 0.6 0.8 1 1.2 1.4 ∆φ(Z, b) MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 1 10 1 Z+ ≥ 1 b-jet, pT(Z) > 20 GeV dσ(Zb)/d∆φ(Z, b)/Nb-jets [pb] 0.5 1 1.5 2 2.5 3 0.6 0.8 1 1.2 1.4 ∆φ(Z, b) Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-16
SLIDE 16

Josh McFayden | HF @ LHC | 21/4/2016

The familiar plot… Systematic shape differences observed

Sherpa seems to do a better job of modelling the shape of the

low ΔR(b,b) region.

16

7TeV vs 13TeV | Z+b(b)

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG5 aMC+Py8 CKKW-L A MG5 aMC+Py8 CKKW-L B MG5 aMC+Py8 FxFx 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 dσ(Zbb)/d∆R(b, b) [pb] 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0.6 0.8 1 1.2 1.4 ∆R(b, b) MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 0.2 0.4 0.6 0.8 1 Z+ ≥ 2 b-jets dσ(Zbb)/d∆R(b, b) [pb] 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0.6 0.8 1 1.2 1.4 ∆R(b, b) Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-17
SLIDE 17

Josh McFayden | HF @ LHC | 21/4/2016

Sherpa ~flat in m(b,b), some shape deviation from other

17

7TeV vs 13TeV | Z+b(b)

ATLAS data, √s=7 TeV Sherpa 2.1 Sherpa 2.2 MG5 aMC+Py8 CKKW-L A MG5 aMC+Py8 CKKW-L B MG5 aMC+Py8 FxFx 10−3 10−2 dσ(Zbb)/dm(b, b) [pb/GeV] 50 100 150 200 250 300 350 0.6 0.8 1 1.2 1.4 m(b, b) [GeV] MC/Data Sherpa 2.1 MG+Py8 A MG+Py8 B ATLAS simulation preliminary, √s = 13 TeV 10−3 10−2 Z+ ≥ 2 b-jets dσ(Zbb)/dm(b, b) [pb/GeV] 50 100 150 200 250 300 350 0.6 0.8 1 1.2 1.4 m(b, b) [GeV] Ratio to Sherpa 2.1

7 TeV - ratio wrt data 13 TeV - ratio wrt Sherpa2.1 ATL-PHYS-PUB-2016-003

slide-18
SLIDE 18

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Practicalities…

18

slide-19
SLIDE 19

Josh McFayden | HF @ LHC | 21/4/2016

Average CPU time/event Sherpa has the highest CPU time/evt.

19

Resource use by generator

slide-20
SLIDE 20

Josh McFayden | HF @ LHC | 21/4/2016

Average CPU time/event Sherpa has the highest CPU time/evt. Longer than full

simulation!

20

Resource use by generator

slide-21
SLIDE 21

Josh McFayden | HF @ LHC | 21/4/2016

Average CPU time/event Sherpa has the highest CPU time/evt. Longer than full

simulation!

Dominated by high

pT(V) slices and filter efficiencies.

NLO significantly

more CPU consuming than LO.

B- and C-hadron

filters have large effect.

21

Resource use by generator

slide-22
SLIDE 22

Josh McFayden | HF @ LHC | 21/4/2016

Average CPU time/event Sherpa has the highest CPU time/evt. Longer than full

simulation!

Dominated by high

pT(V) slices and filter efficiencies.

NLO significantly

more CPU consuming than LO.

B- and C-hadron

filters have large effect.

22

Resource use by generator

Not sustainable to scale MC statistics with integrated Lumi

slide-23
SLIDE 23

Josh McFayden | HF @ LHC | 21/4/2016 23

HFOR…

Generator Matrix Element Flavour number scheme Slicing/ filtering Comments Sherpa2.1 NLO@2j LO@4j 5fl pT(V) & HF

Known mismodelling

  • f low pT large η jets

Sherpa2.2 NLO@2j LO@4j 5fl pT(V) & HF

Improvement in low pT large η jets

MG+Py8 A LO@4j 5fl N-parton

Mismodelling of jet pT (too hard)

MG+Py8 B LO@4j 5fl HT & HF

NLO PDF, difgerent shower settings. (still too hard)

MG5_aMC+Py8 FxFx NLO@2j 5fl TBD

Very promising - some N-jets mismodelling

Alpgen LO@5j 4fl N-parton & 
 b/c/light

The new old! HFOR can be problematic.

slide-24
SLIDE 24

Josh McFayden | HF @ LHC | 21/4/2016

Generator Matrix Element Flavour number scheme Slicing/ filtering Comments Sherpa2.1 NLO@2j LO@4j 5fl pT(V) & HF

Known mismodelling

  • f low pT large η jets

Sherpa2.2 NLO@2j LO@4j 5fl pT(V) & HF

Improvement in low pT large η jets

MG+Py8 A LO@4j 5fl N-parton

Mismodelling of jet pT (too hard)

MG+Py8 B LO@4j 5fl HT & HF

NLO PDF, difgerent shower settings. (still too hard)

MG5_aMC+Py8 FxFx NLO@2j 5fl TBD

Very promising - some N-jets mismodelling

Alpgen LO@5j 4fl N-parton & 
 b/c/light

The new old! HFOR can be problematic.

24

HFOR…

All 5-flavour scheme except Alpgen → HFOR needed.

slide-25
SLIDE 25

Josh McFayden | MC Perf | 12/4/2016

Modelling

Significant V+jets modelling improvements since Run 1. Quite large HF variations but also large uncertainty on the data.

  • HFOR

Conspicuous absence of 4-flavour samples due to difficulty in HFOR.

Can we do better here?

  • Systematics

Is the variation between several 5-fl scheme MCs sufficient for

systematic uncertainty evaluations?

Should we always add comparison to 4-fl scheme predictions? Usually we take normalisation from a control region so only shape/

acceptance variations are significant.

Are there any other uncertainties that are not taken into account with

this prescription?

25

MC summary

slide-26
SLIDE 26

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Recent analysis

26

slide-27
SLIDE 27

Josh McFayden | HF @ LHC | 21/4/2016

Search for VH production

via a heavy resonance.

Interpreted HVT model.

Selection very similar to

SM VH(→bb) analysis

27

VH Resonances

(GeV)

Z’

m 1000 1500 2000 2500 3000 ) [pb] c , c b b

  • BR(H

× ZH)

  • BR(Z’

×

  • 2
  • 10

1

  • 10

1 10 Preliminary ATLAS

Analysis

b b

ll+J

  • 1

L dt = 3.2 fb

  • = 13 TeV

s

Observed Expected 1 s.d. ± Expected 2 s.d. ± Expected =1

V

HVT Model A, g =3

V

HVT Model B, g

ATLAS-CONF-2015-074

slide-28
SLIDE 28

Josh McFayden | HF @ LHC | 21/4/2016

Search for VH production

via a heavy resonance.

Interpreted HVT model.

Selection very similar to

SM VH(→bb) analysis

28

VH Resonances

(GeV)

Z’

m 1000 1500 2000 2500 3000 ) [pb] c , c b b

  • BR(H

× ZH)

  • BR(Z’

×

  • 2
  • 10

1

  • 10

1 10 Preliminary ATLAS

Analysis

b b

ll+J

  • 1

L dt = 3.2 fb

  • = 13 TeV

s

Observed Expected 1 s.d. ± Expected 2 s.d. ± Expected =1

V

HVT Model A, g =3

V

HVT Model B, g

High mass optimisation means a

focus on boosted regime.

ttbar and V+HF are dominant

backgrounds

ATLAS-CONF-2015-074

slide-29
SLIDE 29

Josh McFayden | HF @ LHC | 21/4/2016

Search for VH production

via a heavy resonance.

Interpreted HVT model.

Selection very similar to

SM VH(→bb) analysis

29

VH Resonances

(GeV)

Z’

m 1000 1500 2000 2500 3000 ) [pb] c , c b b

  • BR(H

× ZH)

  • BR(Z’

×

  • 2
  • 10

1

  • 10

1 10 Preliminary ATLAS

Analysis

b b

ll+J

  • 1

L dt = 3.2 fb

  • = 13 TeV

s

Observed Expected 1 s.d. ± Expected 2 s.d. ± Expected =1

V

HVT Model A, g =3

V

HVT Model B, g

High mass optimisation means a

focus on boosted regime.

ttbar and V+HF are dominant

backgrounds

Fit m(T)VH distribution to look for

resonances/set exclusion limits.

ATLAS-CONF-2015-074

slide-30
SLIDE 30

Josh McFayden | HF @ LHC | 21/4/2016

Search for VH production

via a heavy resonance.

Interpreted HVT model.

Selection very similar to

SM VH(→bb) analysis

30

VH Resonances

(GeV)

Z’

m 1000 1500 2000 2500 3000 ) [pb] c , c b b

  • BR(H

× ZH)

  • BR(Z’

×

  • 2
  • 10

1

  • 10

1 10 Preliminary ATLAS

Analysis

b b

ll+J

  • 1

L dt = 3.2 fb

  • = 13 TeV

s

Observed Expected 1 s.d. ± Expected 2 s.d. ± Expected =1

V

HVT Model A, g =3

V

HVT Model B, g

Very low stats in 2-btag high

stats region.

ATLAS-CONF-2015-074

slide-31
SLIDE 31

Josh McFayden | HF @ LHC | 21/4/2016

Signal region defined

by fat jet mass

Control regions defined

  • utside mjet window

Low mjet region used for

V+jets control region.

Significant

contribution from 
 V+HF .

Uncertainty on V+HF

production is the dominant systematic uncertainties!

Mismodelling observed

in high mass m(bb)…

31

VH Resonances | V+jet modelling

SR CR CR SR CR CR

slide-32
SLIDE 32

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Future measurements

32

slide-33
SLIDE 33

Josh McFayden | HF @ LHC | 21/4/2016 33

Future measurements

Measurements in the pipeline

8 TeV Jpsi+mu 13 TeV W/Z+HF

  • What else should we be

looking to measure?

Flavour-labelled kT-spittings?

  • F. Krauss
slide-34
SLIDE 34

Josh McFayden | HF @ LHC | 21/4/2016

  • F. Krauss

34

Future measurements

Measurements in the pipeline

8 TeV Jpsi+mu 13 TeV W/Z+HF

  • What else should we be

looking to measure?

Flavour-labelled kT-spittings?

slide-35
SLIDE 35

Josh McFayden | HF @ LHC | 21/4/2016 35

Future measurements

Measurements in the pipeline

8 TeV Jpsi+mu 13 TeV W/Z+HF

  • What else should we be

looking to measure?

Flavour-labelled kT-spittings?

slide-36
SLIDE 36

Josh McFayden | HF @ LHC | 21/4/2016 36

13 TeV W/Z+HF

Repeat of 7 TeV measurements

Add W+HF to Z+HF measurement More statistics

dR(b,b) with smaller statistical uncertainty

More differential measurements for W+HF Add dedicated measurement in boosted regime So far plan to have the following MC comparisons

Sherpa 2.2 5fl MG5_aMC+Py8 CKKW-L 5fl MG5_aMC+Py8 FxFx 5fl Alpgen+Py6 4fl

What should we add?

MG5_aMC NLO Z+bb 4-fl Sherpa 4-fl?

slide-37
SLIDE 37

Josh McFayden | HF @ LHC | 21/4/2016 37

J/ψ+μ ⇦ B+B ⇦ g→bb

A new measurement of B hadron pair production in

progress at √s = 8 TeV

Identifying the B → J/ψ(→ μ+μ−) + X and B → μ + X decay Fit to extract differential cross

sections in a number of kinematic variables.

Non-prompt J/ψ Non-prompt 3rd μ

New constraints particularly in the

region of close-by B-hadron pair production

Sensitive to g→bb production. No jet radius restriction.

slide-38
SLIDE 38

Josh McFayden | HF @ LHC | 21/4/2016 38

J/ψ+μ ⇦ B+B ⇦ g→bb

  • P. Davidson’s thesis

Signal

mass(μμ) [MeV] pseudo-proper lifetime [ps] d0 significance BDT output Extract non-prompt

J/ψ and 3rd μ

Repeat in each

kinematic bin

2-D fit to J/ψ mass

and lifetime.

Separate from

prompt J/ψ and fakes

2-D fit to 3rd μ 


d0-significance and BDT output.

Separate from

prompt, decay-in- flight and pileup μ.

slide-39
SLIDE 39

Josh McFayden | HF @ LHC | 21/4/2016 39

J/ψ+μ ⇦ B+B ⇦ g→bb

Signal cross section extracted as function of:

ΔR(J/ψ,μ) Δφ(J/ψ,μ) Δy(J/ψ,μ) pT(J/ψ,μ) m(J/ψ,μ) yboost(J/ψ,μ) m/pT pt/m

~20% xs

uncertaintiy in small ΔR region.

  • P. Davidson’s thesis
slide-40
SLIDE 40

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Summary

40

slide-41
SLIDE 41

Josh McFayden | HF @ LHC | 21/4/2016 41

Summary

Technical obstacles

HFOR HF-enhanced generation to increase efficiency?

  • New HF measurements in progress

Z+HF W+HF

More stats, more kinematic variables…

  • Close-by g→bb region of most interest:

BB hadron J/ψ+μ 0.2 kT track jets?

  • What else?!
slide-42
SLIDE 42

Josh McFayden | MC performance | 21/05/2015 Josh McFayden | HF @ LHC | 21/4/2016

Back-ups

42

slide-43
SLIDE 43

Josh McFayden | HF @ LHC | 21/4/2016

Signal region defined

by fat jet mass

Control regions defined

  • utside mjet window

Low mjet region used

for V+jets control region.

Significant

contribution from 
 V+HF .

Uncertainty on V+HF

production is still one

  • f the dominant

systematic uncertainties!

43

VH Resonances | V+jet modelling

00 200 300 400 00 Events / 25 GeV 100 200 300 400 500 600 700 800

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+c Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 1 tag, 0 add. tags

[GeV]

jet

m 100 200 300 400 500 Data/Pred 0.8 1 1.2

SR CR CR

slide-44
SLIDE 44

Josh McFayden | HF @ LHC | 21/4/2016

Signal region defined

by fat jet mass

Control regions defined

  • utside mjet window

Low mjet region used for

V+jets control region.

Significant

contribution from 
 V+HF .

Low statistics in tail

after 1-btag requirement

44

VH Resonances | V+jet modelling

00 200 300 400 00 Events / 25 GeV 100 200 300 400 500 600 700 800

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+c Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 1 tag, 0 add. tags

[GeV]

jet

m 100 200 300 400 500 Data/Pred 0.8 1 1.2

SR CR CR

600 200 800 2400 3000 Events / 100 GeV 50 100 150 200 250

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+c Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 1 tag, 0 add. tags > 145 GeV

jet

< 65 GeV || m

jet

m

[GeV]

VH

m 500 1000 1500 2000 2500 3000 Data/Pred 0.8 1 1.2 600 200 800 2400 3000 Events / 100 GeV

1

  • 10

1 10

2

10

3

10

4

10

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 1 tag, 0 add. tags < 145 GeV

jet

75 GeV < m

[GeV]

VH

m 500 1000 1500 2000 2500 3000 Data/Pred 0.5 1 1.5 2

slide-45
SLIDE 45

Josh McFayden | HF @ LHC | 21/4/2016

Signal region defined by fat jet mass

Control regions defined

  • utside mjet window

Low mjet region used for

V+jets control region.

Significant contribution

from 
 V+HF .

Very low statistics after

2-btag requirement

Will be worse with

10x luminosity!

45

VH Resonances | V+jet modelling

00 200 300 400 00 Events / 25 GeV 100 200 300 400 500 600 700 800

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+c Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 1 tag, 0 add. tags

[GeV]

jet

m 100 200 300 400 500 Data/Pred 0.8 1 1.2

SR CR CR

600 200 800 2400 3000 Events / 100 GeV 50 100 150 200 250

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+c Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 1 tag, 0 add. tags > 145 GeV

jet

< 65 GeV || m

jet

m

[GeV]

VH

m 500 1000 1500 2000 2500 3000 Data/Pred 0.8 1 1.2 600 200 800 2400 3000 Events / 100 GeV

1

  • 10

1 10

2

10

3

10

Data 2015 HVT(M=2.0 TeV) Diboson t t Single top VH W+b W+c W+l Z+b Z+c Z+l Uncertainty Pre-fit background

ATLAS Preliminary

  • 1

Ldt = 3.2 fb

  • = 13 TeV

s 1 lep., 2 tags, 0 add. tags < 145 GeV

jet

75 GeV < m

[GeV]

VH

m 500 1000 1500 2000 2500 3000 Data/Pred 0.5 1 1.5 2

slide-46
SLIDE 46

Josh McFayden | HF @ LHC | 21/4/2016 46

13TeV | Z+jets

ATLAS-CONF-2015-041

slide-47
SLIDE 47

Josh McFayden | HF @ LHC | 21/4/2016 47

Sherpa slicing

B-veto C-veto C-Filter B-veto B-Filter PtV0-70 PtV70-140 PtV140-280 PtV280-500 PtV500-700 PtV700-1000 PtV1000-2000 PtV2000+

24 samples!

slide-48
SLIDE 48

Josh McFayden | HF @ LHC | 21/4/2016 48

Di-b-jet template fit

  • ±

χ

ATLAS

∆

  • ∑
slide-49
SLIDE 49

Josh McFayden | HF @ LHC | 21/4/2016

LO->NLO

Sample according to the optimised integrators, but for each PS

point still have to calculate the ME, which is more expensive in the NLO case.

Also the unweighting efficiency is much worse for the

subtracted real in the NLO.

High pTV

The integrators do not adapt very well to the extreme phase

space regions in the high pT slices so the unweighting efficiency goes down

Also, in these regions more multi-parton ME calculations are

required instead of the cheap V+0p@NLO because they dominate the high pT region

49

Sherpa production time