Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Applications of the WW-type approximation to SIDIS
Kemal Tezgin
University of Connecticut
Based on: [Arxiv:1807.10606]
Co-authors: S. Bastami, H. Avakian, A. V. Efremov, A. Kotzinian, B. U. Musch,B. Parsamyan, A. Prokudin, M. Schlegel, G. Schnell, P. Schweitzer
DIS 2019, Torino, 8-12 April 2019
April 10, 2019
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 1 / 15
SIDIS Cross Section
Consider the SIDIS process l + N → l ′ + h + X
Θ
����������������������
����������������������
����������
����������
z−axis
hφS
φh
Ph
l’
l
q
HADRON PRODUCTION PLANE
LEPTON SCATTERING PLANE
N
S
S
In single photon exchange approximation, SIDIS cross section can beexpressed by 18 structure functions (SFs) [Kotzinian ’95, Mulders andTangerman ’96, Review: Bacchetta et al. JHEP 02 (2007)]
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 2 / 15
SIDIS Cross Section
TMD factorization for Ph⊥ � Q: Structure functions can be described byconvolutions of TMDs and FFs. Generically
C[ω f D
]= x
∑a
e2a
∫d2k⊥d2P⊥δ(2)(zk⊥+P⊥−Ph⊥) ωf a(x , k2
⊥) Da(z ,P2⊥)
1 8 structure functions → twist-2 (factorization X)2 8 structure functions → twist-3 (factorization assumed)
SIDIS CS can be expressed by1 8 twist-2 TMDs: f1, g1, h1, f
⊥1T , g
⊥1T , h
⊥1T , h
⊥1L, h
⊥1
2 16 twist-3 TMDs: e, eL, eT , f⊥, gT , ...
3 2 twist-2 FFs: D1,H⊥1
4 4 twist-3 FFs: D⊥,G⊥,H,E
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 3 / 15
What is WW Approximation?
First application of EoM [Wandzura and Wilczek ’77]:
gqT (x)︸ ︷︷ ︸
twist−3
= < qq >︸ ︷︷ ︸related to gq
1 (x)
+< qgq >︸ ︷︷ ︸”tilde”term
The approximation is ∣∣∣< qgq >
< qq >
∣∣∣� 1.
Neglecting ”tilde” terms is known to be WW approximation
gqT (x) =
∫ 1
x
dy
ygq
1 (y) + gqT (x)
WW≈∫ 1
x
dy
ygq
1 (y)
Analogously [Jaffe and Ji ’92]
hqL(x) = 2x
∫ 1
x
dy
y2hq1 (y) + hqL(x)
WW≈ 2x
∫ 1
x
dy
y2hq1 (y)
WW approximation is supported by:Instanton model of QCD vacuum [Balla et al. NPB 510 (1998)]Lattice data [Gockeler et al. PRD 63 (2001)]Quark models [Bag, LFCM, Spectator, ...]
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 4 / 15
WW Approximation
Illustration
g2(x) =1
2
∑a
e2a
(g aT (x)− g a
1 (x))
WW≈ d
dx
[x
∫ 1
x
dy
yg1(y)
]
Figure: The structure function xg2(x) in WWapproximation at Q2 = 7.1GeV 2 vs. dataE143[PRD 58 (1998)], E155[PLB 553 (2003)]
Figure: The structure function xg2(x) in WWapproximation at Q2 = 2.4GeV 2 vs. HERMESdata[EPJC 72 (2012)].
holds with 40% or better [Accardi et al. JHEP 11 (2009)]
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 5 / 15
WW-type Approximation
EoM allow us to decompose twist-2 h⊥1L, g⊥1T and all twist-3 TMDs, FFs intoqq and qgq-matrix elements.
Assuming| < qgq > | � | < qq > |
We obtaintwist-2:
g⊥(1)1T (x)
WW−type≈ x
∫ 1
x
dy
yg1(y), h
⊥(1)1L (x)
WW−type≈ −x2
∫ 1
x
dy
y2h1(y).
twist-3:
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 6 / 15
WW-type Approximation
As a result of WW-type approximation, all SIDIS SFs (twist-2 & twist-3)can be expressed in terms of 8 basis functions:
1 6 leading-twist TMDs: f1, g1, h1, f⊥
1T , h⊥1 , h
⊥1T
2 2 leading-twist FFs: D1,H⊥1
Our goal is to check where the WW-type approximation works and where itdoes notWe use state-of-the-art parametrizations for the basis functions [MSTW, DSS,Anselmino et al., Barone et al., Lefky and Prokudin; Gauss Ansatz used, parameters fixedusing lattice data Hagler et al.]
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 7 / 15
WW-type Approximation at Leading Twist
At leading twist, WW-type approximation is useful for 2 structure functions:
Fcos(φh−φS )LT and F
sin(2φh)UL .
Fcos(φh−φS )LT = C
[h·~k⊥MN
g⊥1TD1
].
First application of WW-type approximation [Kotzinian, Parsamyan, Prokudin,
PRD 73 (2006)]Results are compatible with the recent preliminary COMPASS data[Parsamyan, PoS DIS2013 (2013) 231].
-210 -110 1
-0.2
-0.1
0
0.1
0.2preliminaryCOMPASS +h
Proton 2010 data
)Sϕ-
hϕco
s(
LTA
x
PRD73:114017(2006)
arXiv:0806.3804 [hep-ph]PRD79:094012(2009)
PRD73:114017(2006) updated
-210 -110 1
-0.2
-0.1
0
0.1
0.2COMPASS preliminary-h
)S
-ϕh
cos(
ϕLT
Ax
PRD73:114017(2006)PRD73:114017(2006) updatedarXiv:0806.3804 [hep-ph]PRD79:094012(2009)
Proton 2010 data
quark-diquark model [Kotzinian, arXiv:0806.3804] and light-cone quarkconstituent model [Boffi et al., Phys. Rev. D 79, 094012], are also displayed forcomparison.
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 8 / 15
WW-type Approximation at Leading Twist
Similarly, in WW-type approximation: Fsin(2φh)UL = C
[ω h⊥1LH
⊥1
]Results [Avakian et al., PRD 77 (2008)] are compatible with preliminary COMPASS
[Parsamyan, PoS DIS2017 (2018) 259] and JLab π0 [Jawalkar et al., PLB 782 (2018)]data
2−10 1−10
0.05−
0
0.05
+hpreliminaryCOMPASS
HERMES PRL 84(2000) D(y)-rescaledPRD 77, 014023(2008)
hφsi
n 2
UL
A
x2−10 1−10
0.05−
0
0.05
−h > 0.2z
x
π0
0.0 0.2 0.4 0.6 0.8 1.0
-0.04
-0.02
0.00
0.02
0.04
x
AUL,⟨y⟩
sin(2Φh)
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 9 / 15
WW-type Approximation at Subleading Twist
Subleading twist → Complexity. Example (expression symbolic):
Fcos(φS )LT =
MN
Q
[gT D1 + h1 E + eT H⊥1 + g⊥1T D⊥ + e⊥T H⊥1 + f ⊥1T G⊥
]WW-type approximation → Crucial simplification
Fcos(φS )LT
WW−type≈ MN
Q
[gT D1
]∣∣∣∣gT→g1
Results are compatible with COMPASS data [Parsamyan, PoS DIS2013 (2013)231].
COMPASS preliminaryProton 2010 data
-210 -110 1
-0.2
-0.1
0
0.1
0.2h+
Sco
sϕ
LTA
x
arXiv:0806.3804 [hep-ph]
π+
π-
0.00 0.05 0.10 0.15 0.20 0.25 0.30
-0.06
-0.04
-0.02
0.00
0.02
0.04
x
ALTcos(ΦS)
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 10 / 15
WW-type Approximation at Subleading Twist
Fcos(2φh−φS )LT =
MN
Q
[eT H⊥1 +g⊥1T D⊥+e⊥T H⊥1 +f ⊥1T G⊥+g⊥T D1 +h⊥1T E
]WW-type approximation → Crucial simplification
Fcos(2φh−φS )LT
WW−type≈ MN
Q
[g⊥T D1
]∣∣∣∣g⊥T →g1
Results are not in contradiction with COMPASS data [Parsamyan, PoS DIS2013(2013) 231].
COMPASS preliminaryProton 2010 data
-210 -110 1
-0.2
-0.1
0
0.1
0.2h+
)S
-ϕh
cos(
2ϕLT
A
x
arXiv:0806.3804 [hep-ph]
π+
π-
0.00 0.05 0.10 0.15 0.20 0.25 0.30-0.010
-0.005
0.000
0.005
0.010
x
ALTcos(2Φh-ΦS)
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 11 / 15
WW-type Approximation at Subleading Twist
Fcos(φh)LL =
MN
Q
[eL H⊥1 + g1 D⊥ + g⊥L D1 + h⊥1L E
]In WW-type approximation
F cosφh
LL
WW−type≈ MN
Q
[g⊥L D1
]∣∣∣∣g⊥L →g1
WW-type results [Anselmino et al., Phys. Rev. D74 (2006)] are compatible withCOMPASS preliminary data [Parsamyan, PoS DIS2017 (2018) 259].
2−10 1−10
0.1−
0.05−
0
0.05
0.1 +hpreliminaryCOMPASS
PRD 74, 074015(2006)NPA 945(2016) 153
hφco
s
LL
A
x2−10 1−10
0.1−
0.05−
0
0.05
0.1 −h > 0.1z
x
A model study [Mao et al., Nucl. Phys. A945 (2016)] is also displayed forcomparison.
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 12 / 15
WW-type Approximation at Subleading Twist
F sinφhUL =
MN
Q
[hL H⊥1 + g1 G⊥ + f ⊥L D1 + h⊥1L H
]In WW-type approximation
F sinφh
UL
WW−type≈ MN
Q
[hL H⊥1
]∣∣∣∣hL→h1
We observe discrepancies with HERMES π± [Airapetian et al., Phys. Lett. B622
(2005)] and JLAB π0 [Jawalkar et al., Phys. Lett. B782 (2018)] data(underestimate π+, can not describe π0).
π+
π-
0.00 0.05 0.10 0.15 0.20 0.25 0.30
-0.05
0.00
0.05
x
AUL,<y>
sin(Φh)
π0
0.0 0.2 0.4 0.6 0.8 1.0-0.10
-0.05
0.00
0.05
0.10
x
AUL,⟨y⟩
sin(Φh)
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 13 / 15
WW-type Approximation at Subleading Twist
F sinφh
LU = MQ
[e H⊥1 + f1 G⊥ + g⊥ D1 + h⊥1 E
]WW−type≈ 0
0
0.05
0.1
0.15
0 0.1 0.2 0.3 0.4 0.5 0.6
CLAS, π+
x
Avakian et al, PRD69 (2004) 112004ALUsin φh(x)
Figure: CLAS data shows a non-zero asymmetry
F cosφh
UU = MQ
[hH⊥1 + f1D
⊥+ f ⊥D1 +h⊥1 H]
WW≈ M
Q
[hH⊥1 + f ⊥D1
]∣∣∣∣h→h⊥1 ,f
⊥→f1
Figure: HERMES data [Airapetian et al., Phys. Rev. D87 (2013)]
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 14 / 15
Conclusions
twist-3 PDFs: WW approximation works, supported by experiment for gT ,instanton calculus, lattice QCD
twist-2 TMDs: applicable for gear worms g⊥1T , h⊥1L, in agreement with data
Due to the complexity of SIDIS SFs at subleading-twist, a guideline is muchneeded. Predictions were made for all SIDIS asymmetries at subleadingtwist.
Observations: compatible with data in several cases (e.g. AcosφS
LT , ...)
Not in a agreement with data in some cases (e.g. Asinφh
LU , ....)
More work is needed to reliably test where WW-type approximation works
If we find it works: it will motivate theoretical studies to explain thesuppression mechanism for qgq. If we find it does not work: it will help toidentify large qgq terms, and stimulate studies to find out why they arelarge. We learn in any case!
Mathematica package: https://github.com/prokudin/WW-SIDIS
Kemal Tezgin (University of Connecticut) WW-type approximation to SIDIS April 10, 2019 15 / 15