Long-range azimuthal correlations in 2.76 & 13 TeV pp with ATLAS - - PowerPoint PPT Presentation

Long-range azimuthal correlations in 2.76 & 13 TeV pp with ATLAS

Andy Buckley

University of Glasgow

MPI@LHC 2015, Trieste, 2015-11-26

1/9

Introduction to the ridge

The “near-side ridge” phenomenon has been one of the most prominent and enduring physics puzzles at the LHC. Expected/hoped-for effect in pA and AA collisions, from collective flow theory paradigm. Discovery in high-multiplicity pp was a surprise! Ridge in p + Pb due to global sinusoidal modulation of particle

• production. . . same in pp?

Today show ATLAS’ latest ridge measurements, cf. arXiv:1509.04776

2/9

• Phys. Rev. Lett. 110, 182302 (2013),

arXiv:1212.5198

Datasets

pp data taken with ATLAS: 4 pb−1 at 2.76 TeV and 14 nb−1 at 13 TeV. Pile-up low in both cases: µ ∼ 0.5 at 2.76 TeV and ∼ 0.04 at 13 TeV. Charged tracks with pT > 300 MeV and |η| < 2.5 used as input to correlation measures. Only tracks with pT > 400 MeV counted in Nrec

ch .

rec ch

N 50 100 150 Events / 3 1 10

10

10

10

10

10

10 ATLAS =2.76 TeV s rec ch N 50 100 150 200 1 10

ATLAS =13 TeV s

3/9

Correlation observables 1: S, B, and C in ∆η,∆φ

Raw observable is 2p correlation C(∆η, ∆φ) = S(∆η, ∆φ)/B(∆η, ∆φ). S and B are distributions of particle a, b separations for a & b in same event and mixed events respectively. Division cancels acceptance effects and systematics. Not explicitly unfolded, but tracking efficiencies used as weights 1/ǫ(pa

T, ηa) ǫ(pb T, ηb).

φ ∆

2 4

η ∆

• 4
• 2

2 4

) φ ∆ , η ∆ C(

0.95 1 1.05 1.1 ATLAS =13 TeV s <5.0 GeV

a,b T

0.5<p <30

rec ch

N ≤ 10 φ ∆ 2 4 η ∆

• 4
• 2

2 4

) φ ∆ , η ∆ C(

0.98 1 1.02 ATLAS =13 TeV s <5.0 GeV

a,b T

0.5<p 120 ≥

rec ch

N

Dominant structure is the dijet system with “this” jet around (0, 0) and the “other” jet’s far ridge in ∆η at ∆φ ∼ π. Near-side ridge at ∆φ ∼ 0.

4/9

Correlation observables 2: per-particle yields

To focus on long-range ridge effects, integrate over large |∆η| to define S, B, and C functions in ∆φ only, e.g. S(∆φ) = 5

2

d|∆η| S(∆η, ∆φ) Useful to convert to a per-particle correlation yield: Y(∆φ) = C(∆φ) ·

• B(∆φ)d∆φ

πNa = S(∆φ) ˆ B(∆φ)

• πNa

Y measures the average number

• f long-range correlation

partners per “trigger” particle a at a given ∆φ. Results:

2 4

) φ ∆ Y( 0.5 0.6 0.7 =13 TeV s =2.76 TeV s ATLAS <20

rec ch

N ≤ |<5.0 η ∆ 2.0<| <5.0 GeV

a,b T

0.5<p (a) 2 4 ) φ ∆ Y( 2.3 2.35 2.4 2.45 ) φ ∆ Y( )+G φ ∆ (

periph

FY (0)

periph

+ FY

ridge

Y ) φ ∆ (

templ

Y =13 TeV s <50

rec ch

N ≤ 40 (b) 2 4 ) φ ∆ Y( 2.9 2.95 3 3.05 3.1 =2.76 TeV s <60

rec ch

N ≤ 50 (c) 2 4 ) φ ∆ Y( 3.4 3.45 3.5 3.55 3.6 =13 TeV s <70

rec ch

N ≤ 60 (d)

φ ∆

2 4

) φ ∆ Y(

4 4.1 4.2 =2.76 TeV s <80

rec ch

N ≤ 70 (e)

φ ∆

2 4

) φ ∆ Y(

5.5 5.6 5.7 =13 TeV s 90 ≥

rec ch

N (f)

Increasing modulation with Nch fills in the near-side minimum ⇒ produces the ridge, and narrows + heightens the far-side peak.

5/9

Interpretation: yield fits

In p + Pb collisions, the ridge results from sinusoidal global modulation of single-particle azimuthal angle distributions. This new study uses template fitting of Y to investigate whether the pp ridge has the same origin: Ytempl(∆φ) = FYperiph(∆φ) + Yridge(∆φ) where Yridge(∆φ) = G[1 + 2 v2,2 cos(2∆φ)]. G is fixed by template normalisation = data; Yperiph taken from lowest-multiplicity data bin; F and v2,2 free parameters for χ2 fit. Results:

2 4

) φ ∆ Y( 0.5 0.6 0.7 =13 TeV s =2.76 TeV s ATLAS <20

rec ch

N ≤ |<5.0 η ∆ 2.0<| <5.0 GeV

a,b T

0.5<p (a) 2 4 ) φ ∆ Y( 2.3 2.35 2.4 2.45 ) φ ∆ Y( )+G φ ∆ (

periph

FY (0)

periph

+ FY

ridge

Y ) φ ∆ (

templ

Y =13 TeV s <50

rec ch

N ≤ 40 (b) 2 4 ) φ ∆ Y( 2.9 2.95 3 3.05 3.1 =2.76 TeV s <60

rec ch

N ≤ 50 (c) 2 4 ) φ ∆ Y( 3.4 3.45 3.5 3.55 3.6 =13 TeV s <70

rec ch

N ≤ 60 (d)

φ ∆

2 4

) φ ∆ Y(

4 4.1 4.2 =2.76 TeV s <80

rec ch

N ≤ 70 (e)

φ ∆

2 4

) φ ∆ Y(

5.5 5.6 5.7 =13 TeV s 90 ≥

rec ch

N (f)

Yperiph and Yridge as open points and blue line respectively; Ytempl in red fits several data features with 2 params on one sinusoid.

6/9

Interpretation: testing single particle modulation

If the ridge is formed by sinusoidal modulation of individual particle production, then v2,2 should factorise: v2,2(pa

T, pb T) = v2(pa T)v2(pb T).

Tested for 3 pb

T bins vs track

multiplicity Nrec

ch . Extract v2 from

combinations of pa,b

T bins in v2,2:

v2(pT1) = v2,2(pT1, pT2)/

• v2,2(pT2, pT2).

Results: top row shows fitted v2,2, middle shows v2. Latter shows clear agreement between pb

T and

substantial independence of multiplicity at both √s. Bottom row shows pT dependence of v2.

20 40 60 80 100 2,2

v

0.002 0.004 0.006 =2.76 TeV s ATLAS 20 40 60 80 100 120 0.002 0.004 0.006 =13 TeV s

rec ch

N

20 40 60 80 100 2

v

0.05 0.1 =2.76 TeV s

rec ch

N

20 40 60 80 100 120 0.05 0.1 <5.0 GeV

b T

0.5<p <1.0 GeV

b T

0.5<p <3.0 GeV

b T

2.0<p =13 TeV s |<5.0 η ∆ 2.0<| <5.0 GeV

a T

0.5<p [GeV]

a T

p

1 2 3 4 2

v

0.05 0.1 =2.76 TeV s =13 TeV s ATLAS |<5.0 η ∆ 2.0<| <5.0 GeV

b T

0.5<p <60

rec ch

N ≤ 50 [GeV]

a T

p

1 2 3 4 0.05 0.1 <50

rec ch

N ≤ 40 <80

rec ch

N ≤ 70 100 ≥

rec ch

N =13 TeV s

Extraction of v2(pa

T) showing stability

vs pb

T bin and Nrec ch .

7/9

Interpretation: modulation strength vs. Nrec

ch Relative size of ridge modulation ∼ Gv2,2/FYperiph(0) ⇒ study G and F vs Nrec

ch :

Different datasets study Nrec

ch dependence of Yperiph extraction by

subdividing the Nperiph

ch

∈ [0, 20] range. v2,2 fairly stable with Nrec

ch + linear growth of G + flattening of F

⇒ increase in ridge visibility with Nrec

ch

8/9

Interpretation: modulation strength vs. Nrec

ch Relative size of ridge modulation ∼ Gv2,2/FYperiph(0) ⇒ study G and F vs Nrec

ch :

Different datasets study Nrec

ch dependence of Yperiph extraction by

subdividing the Nperiph

ch

∈ [0, 20] range. v2,2 fairly stable with Nrec

ch + linear growth of G + flattening of F

⇒ increase in ridge visibility with Nrec

ch

8/9

Summary

◮ The near-side ridge is still a major puzzle in LHC physics. Still

there in high-multiplicity 13 & 2.76 TeV pp events!

◮ A new fit of the ∆φ modulation in these pp events shows excellent

consistency with a single 2∆φ Fourier mode, which not only produces the near side ridge but also beneficially modifies the away side peak with increasing Nrec

ch . ◮ Comparison of the v2,2 Fourier coefficient in ∆φ yields between pT

bins of trigger and partner tracks reveals single-particle modulation coefficients, v2, independent of Nrec

ch and pb T: consistent

with azimuthal modulation of individual particle production.

◮ It hence appears that the pp and p + Pb ridge phenomena have the

same source.. . but exactly what that is remains to be seen!

9/9