RECONSTRUCTING PARTON
DISTRIBUTION FUNCTIONS FROM
THEIR COORDINATE SPACE
BEHAVIOR
Light Cone Meeting,
Jefferson Lab, May 14-19, 2018
Simonetta Liuti
The aim of our work is to pose (and possibly answer) the
following questions:
1. What information from PDF fits is relevant to constrain/test/validate
lattice calculations?
2. What PDF-related quantities we would like lattice QCD to compute?
5/15/18 3
1. Can we quantify the accuracy that lattice calculations should
have in order to have a direct impact on PDF fits
2. To which extent available lattice results agree with the results of
global PDF fits?
….extend to Generalized Parton Distributions!
The Review poses (and possibly answers some of) the
following questions:
F. Yndurain, PLB 1978 Rep. Prog. In Physics 1983
Early attempts: Reconstructing DIS structure functions from their Mellin moments, using the orthogonal polynomials technique
N=3 N=8 Mn(Q2) = Z 1 0 dx xn 2F2(x, Q2) Mellin Moment
Z 1 0 dx Bkm(x)F2(x, Q2) = (m+ 1)! k! m kX l=0 ( 1)l l!(m k l)! Mk+l+2(Q 2)
In a nutshell….
… replace Mellin Moment with Orhtogonal Polynomial Moment
Mn(Q2) =
Z 1
0
0 2 4 6 8 10 12 14 16 18 0 0.2 0.4 0.6 0.8 1 Polynomials x 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.2 0.4 0.6 0.8 1 11th Bernstein Polynomial x 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 x F2P 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 x F2P Structure Function
Integrand of Bernstein Moment Bernstein Polynomial -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 x Q2 = 1.8 GeV2 F2P
GPD/PDF CORRELATION FUNCTION
W⇤0⇤(x,⇠, ) = 1 2 Z dz (2⇡)e ixP+z hP ,⇤0 | q¯(0) W 0,z q z | P,⇤i zT=0,z+=0 .q
,
Λ
γk’= k –
Δ
.
λ
’
q'=q+
Δ
,
Λ
γ’
P’= P-
Δ
,
Λ
’
P
,
Λ
k
,
λ
t Q2 x, ξ hGPDs involve two types of distance
5/15/18 9H
q(x
,
0
,
) =
Z
dz
2
⇡
e
ixP+zh
P
,
⇤
0|
q
¯
(0)
+q
(
z
)
|
P,
⇤
i
zT=0¯
q
+†(0
,
b
)
q
+(
z
,
b
)
!
⇢
(0
,
b
;
z
,
b
)
x distribution Fourier transform of non-diagonal density distribution in z
Hu(x,0,Δ) Fq(t ⌘ 2) = Z dxHq(x,0, ) ! ⇢q(b) O. Gonzalez-Hernandez et al., PRC88 (2013) x dependence t large t small
5/15/18 11
h
P
,
⇤
0|
q
¯
(0)
+W
(0
, z
)
q
(
z
)
|
P,
⇤
i
zT =0 P’ 0 z -b P P-P’=Δ bReconstructing GPDs from a finite number of moments
0.0 1.0 2.0 3.0 4.0 Hu-dat =0.0 t=-0.0 N=2 N=7 0.0 1.0 2.0 3.0 4.0 t=-0.1 0.0 0.5 1.0 1.5 2.0 2.5 t=-0.2 0.0 0.5 1.0 1.5 2.0 2.5 t=-0.3 0.0 0.2 0.4 0.6 0.8 x 0.0 0.4 0.8 1.2 1.6 t=-0.5 0.0 0.2 0.4 0.6 0.8 1.0 x 0.0 0.4 0.8 1.2 1.6 t=-1.0Bernstein polynomials based analysis
H. Honkanen et al., EPJC(2007)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 Hu-dat Q 2 =4 GeV2 =0.18 t=-0.035 0.0 0.5 1.0 1.5 2.0 2.5 3.0 =0.18 t=-1.00 0.0 0.5 1.0 1.5 2.0 2.5 =0.25 t=-0.073 0.0 0.5 1.0 1.5 2.0 2.5 =0.25 t=-1.00 0.0 0.5 1.0 1.5 2.0 2.5 =0.36 t=-0.18 0.0 0.5 1.0 1.5 2.0 2.5 =0.36 t=-1.00 0.0 0.2 0.4 0.6 0.8 x 0.0 0.5 1.0 1.5 2.0 2.5 =0.53 t=-0.53 0.0 0.2 0.4 0.6 0.8 1.0 x 0.0 0.5 1.0 1.5 2.0 2.5 =0.53 t=-1.00
(Using lattice moments from LHPC, Ph. Hägler et al., 2005, 2006, 2008)
3 moments from LHPC 8 moments from param.
2. DETERMINATION OF PDF
FROM COORDINATE SPACE
BEHAVIOR
Example: Unpolarized PDFs
5/15/18 14
Parton x-momentum distribution is the Fourier transform of a non-diagonal one body density distribution in coordinate z-space
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0.0001 0.001 0.01 0.1 1 PDFs
Uv + Dv Q2 = 150 GeV2 Collaboration and SOM PDFs
x ABM CT10 CJMid 6x6
¯
q
+†(0)
q
+(
z
)
!
⇢
(0
,
z
)
E. Askanazi, K. Holcomb, S. L., J. Phys. G 42, no. 3, 034030 (2015) [arXiv:1411.2487 [hep-ph]].
F
2(
x
) =
X
qxf
1q(
x
)
f
1q(
x
) =
Z
dz
2
⇡
e
ixP+zh
P,
⇤
|
q
¯
(0)
+q
(
z
)
|
P,
⇤
i
zT =0Symmetry: C-parity
¯
q
(
x
) =
q
(
x
)
q
(
x
) + ¯
q
(
x
)
q
(
x
)
q
¯
(
x
)
Collins and Soper, ‘80s
C-odd C-even
Symmetric with respect to x=0 Anti-symmetric with respect to x=0
hP,⇤0 | q¯(0) +q z | P,⇤i + hP,⇤0 | q¯(0) +q z | P,⇤i = i 2P+ Z 1 0 dxqs(x) sin(xz) hP,⇤0 | q¯(0) +q z | P,⇤i hP,⇤0 | q¯(0) +q z | P,⇤i = 1 2P+ Z 1 0 dxqv(x) cos(xz) Tc Ts
T
s(
z
)
=
i
2
M
Z
1 0dxq
(
x
) sin(
xz
) =
i
2
M
M
2z
+
1
3!
M
4z
3+
...
M
2=
Z
1 0dxx q
(
x
)
M
4=
Z
1 0dxx
3q
(
x
)
T
c(
z
)
=
1
2
M
Z
1 0dxq
s(
x
) cos(
xz
) =
i
2
M
M
1+
1
2
M
3z
2+
...
M
1=
Z
1 0dx q
(
x
)
M
3=
Z
1 0dxx
2q
(
x
)
T
sT
cRegge tail z
-α+1proton size
0 0.1 0.2 0.3 0.4 0.5 0.6 0 5 10 15 20 25 30 Q(z) z CT10 u CT10 xmin=0.1
Space-time picture of the correlation function
! At small distances - z- < R - the virtual photon scatters
incoherently from the individual partons. This region is dominated by bulk properties such as the average
momentum.
! At larger distances - z- >> R - the interaction is given by
coherent scattering over several partons. The virtual photon converts into a q-qbar pair covering a distance z
-z- zo
z3
z+
2), 3) 4)
3. HOW WELL CAN WE RECONSTRUCT
THE X DEPENDENT FUNCTION (PDF/
Information from lattice: Mellin Moments
W. Detmold et al., Moment u-d µ2= 4 GeV2 Linear extrapolation Chiral extrapolation Phenomenology CT10 M1 1 1 1 M2 0.262 0.18(3) 0.169 M3 0.0843 0.05(2) 0.0536 M4 0.0340 0.02(1) 0.0221 Mod.Phys.Lett. A18 (2003) Eur.Phys.J.3 (2001),u-d
Mn(Q2) = Z 1 0 dx xn 2F2(x, Q2)LHPC (Ph. Hägler et al.) Moment u-d µ2= 4 GeV2 Linear extrapolation Chiral extrapolation Phenomenology CT10 M1 1 1 1 M2 / 0.200(-7/+9) 0.169 M3 / / 0.0536 M4 / / / Phys.Rev. D77 (2008) Moment u-d µ2= 4 GeV2 mπ=352 MeV Chiral extrapolation Phenomenology CT10 M1 1 1 1 M2 0.206(14) 0.157(10) 0.169 M3 0.078(16) / 0.0536 M4 / / 0.0221
G. Bali et al. PoS LATTICE2015 (2016)
Moment u-d µ2= 4 GeV2
mπ=physical Hägler: Chiral extrapolation Phenomenology CT10 M1 1 1 1 M2 0.140(21) 0.157(10) 0.169 M3 / / 0.0536 M4 / / 0.0221
ETMC (C. Alexandrou et al., Phys Rev D93, 2016)
u-d
LHPC (J. Green et al., Phys. Lett. B, 2014 )
Moment u-d µ2= 4 GeV2 mπ=physical (largest source sink) Hägler: Chiral extrapolation Phenomenology CT10 M1 1 1 1 M2 0.208(24) 0.157(10) 0.169 M3 / / 0.0536 M4 / / 0.0221
Reconstructed pdf: using its own moments up
to x
5
no error shown
Using Detmold et al.
Weigl and Mankiewicz, PLB 1996
Error analysis in progress
ii) What PDF-related quantities would we like lattice QCD to
compute?
Sources of error: • lattice moments
• Regge tail (CT10)
• unknown region
A careful analysis needs
to be done and it’s on its
way!
First stab at GPDs…
0.4 0.2 0 0.2 0.4 0.6 0.8 1 0.1 0 0.1 0 0.5 x ⇠ Hu ( x, ⇠ ,t) t= 0.1GeV2reconstructed 0.2 0 0.2 0.4 0.6 1 0.4 0.2 0 0.2 0.4 0.6 0.8 1 0.1 0 0.1 0 1 x ⇠ Hu ( x, ⇠ ,t) t= 0.1GeV2 0 0.2 0.4 0.6 0.8 1 1.2 1 Reconstructed Model x x ξ ξ H(x,ξ,t) H(x,ξ,t)0.4 0.2 0 0.2 0.4 0.6 0.8 1 0.1 0 0.1 0 0.5 x ⇠ Hu ( x, ⇠ ,t) t= 0.1GeV2reconstructed 0.2 0 0.2 0.4 0.6 1 0.4 0.2 0 0.2 0.4 0.6 0.8 1 0.1 0 0.1 0 1 x ⇠ Hu ( x, ⇠ ,t) t= 0.1GeV2 0 0.2 0.4 0.6 0.8 1 1.2 1 Perturbative evolution x ξ H(x,ξ,t)
" Pseudo-PDFs interpolate between PDFs evaluated on the LF and lattice
QCD space-like correlations
" Pseudo-PDFs are the Fourier transforms of zT(z3)-dependent generalized
Ioffe time distributions,
" For the ratio reduces to
" is the Ioffe time distribution conjugate to
M(
ν
,
z
32)
M(0,
z
32)
z
3→
0
M
(
ν
,
z
32)
M
(
ν
, 0)
f
(
x
)
M
(
ν
, 0)
-0.2 0 0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 Im M(z - ,z 3 ) z- (fm) Experimental data (CT10) Orginos et al. (2017) Haegler et al., LHPC 2007
0 2 4 6 8 10 12 0.1 1 10 Im M( ,z 3 ) z32 (fm2) 0 2 4 6 8 10 12 1 2 3 4 5 6 7 8 9 Im M( ,z3 ) z32 (fm2)
Evolution in z
T Non-perturbative region ν=5 ν=5 transitionConclusions and Outlook
" We presented results on the reconstruction of PDFs from their Mellin
moments which are available in lattice QCD
" Lattice QCD moments evaluation complements pseudo-PDFs
analysis
" Both methods need assumptions on small x behavior
" Lattice QCD moments evaluation can be extended to GPDs
" Error analysis is crucial