Accessing nucleon structure from Euclidean spacetime
Chris Monahan
Institute for Nuclear Theory
HOW FAST DO PARTONS TRAVEL?
How is the momentum of a fast-moving nucleon distributed amongst its constituents?
WHERE DOES THE SPIN OF A PROTON COME FROM?
How do position and longitudinal momentum of
a parton correlate in a fast-moving nucleon?
PDF UNCERTAINTIES
From J. Butterworth et al., J.P.G 43 (2016) 023001
Nuclear landscape
Early universe
Neutron stars
LUX
LHCb 12 GeV
EIC @ JLab
EXPERIMENTAL EXTRACTION
From PDG 2016
EXPERIMENTAL EXTRACTION
PDF4LHC15 (NNLO), J. Phys. G 43 (2016) 023001
PDFs FROM EUCLIDEAN SPACETIME
An unsolved almost-solved challenge
Decompose cross-section
Hadronic contribution
in turn, expressed in terms of structure functions DIS
Decompose cross-section
Hadronic contribution
in turn, expressed in terms of structure functions DIS
parton distribution functions (PDFs)
Defined as
where
Renormalised PDFs
Satisfy DGLAP evolution
PDFs (GPDs)
Mellin moments of PDFs
related to matrix elements
of local twist-two operators
Moving beyond three moments is very challenging
Cannot reconstruct PDFs from only three moments MOMENTS OF PDFs
Detmold et al., Eur. Phys. J. C 3 (2001) 1 Detmold et al., Phys. Rev. D 68 (2001) 034025 Detmold et al., Mod. Phys. Lett. A 18 (2003) 2681
Nucleon axial charge
Controls:
● nucleon-nucleon force
● free neutron β-decay
● early Universe composition Experimental value
● cold neutron decay
MOMENTS OF PDFs:
AXIAL CHARGE
M.Constantinou, PoS(CD15) 009 2015
Nucleon axial charge
MOMENTS OF PDFs:
AXIAL CHARGE
C.C.Chang et al (CalLat), 1710.06523 E.Berkowitz et al (CalLat), 1704.01114
SPACELIKE DISTRIBUTIONS
Matrix elements of spacelike nonlocal operators
● Quasi distributions
● Pseudo distributions
● Lattice “cross-sections”
SPACELIKE DISTRIBUTIONS
X. Ji, PRL 110 (2013) 262002 X. Ji, Sci.Ch. PMA 57 (2014) 1407
A.Radyushkin, PLB 767 (2017) 314 A.Radyushkin, PRD 96 (2017) 034025
Y.-Q. Ma & J.-W. Qiu, 1404.6860
Y.-Q. Ma & J.-W. Qiu, IJMP 37 (2015) 1560041 Y.-Q. Ma & J.-W. Qiu, 1709.03018
SPACELIKE DISTRIBUTIONS
H.-W. Lin et al, PRD 91 (2015) 054510 C. Alexandrou et al., PRD 92 (2015) 014502 J.-H. Zhang et al., arXiv:1702.00008
J.-W. Chen et al., NPB 911 (2016) 246
See also:
H.-W. Lin et al (LP3), 1708.05301
C. Alexandrou et al (ETMC), NPB 923 (2017) 394
K. Orginos et al, 1706.05373
Defined as
Recall
Related to light-front PDFs via
QUASI DISTRIBUTIONS
X. Ji, PRL 110 (2013) 262002 X. Ji, Sci.Ch. PMA 57 (2014) 1407
GENERAL PROCEDURE
GENERAL PROCEDURE
J.-W. Chen et al, NPB 12 (2016) 004 T. Ishikawa et al, arXiv:1609.02018 J.-W. Chen et al, 1706.01295 C. Alexandrou et al, NPB 923 (2017) 394
T. Ishikawa et al, 1707.03107 X. Ji et al, 1706.08962 J.-W. Chen et al, NPB 12 (2016) 004 T. Ishikawa et al, arXiv:1609.02018 X. Ji et al, PRD 92 (2015) 034006
C.E. Carlson, M. Freid, PRD 095 (2017) 094504 X. Xiong et al, 1705.00246 X. Ji et al, NPB 924 (2017) 366
X. Ji, PRL 110 (2013) 262002 X. Xiong et al, PRD 90 (2014) 014051 X. Ji et al, arXiv:1506.00248 H.-W. Lin et al, 1708.05301
GENERAL PROCEDURE:
GENERAL CHALLENGES
Power-divergence must be controlled
GENERAL PROCEDURE:
GENERAL CHALLENGES
Power-divergence must be controlled Large momentum required:
- discretised Fourier transform
- control power-suppressed corrections
GENERAL PROCEDURE:
GENERAL CHALLENGES
Power-divergence must be controlled Large momentum required:
- discretised Fourier transform
- control power-suppressed corrections Renormalisation and continuum limit:
- perturbative truncation uncertainties - discretisation effects
GENERAL PROCEDURE:
GENERAL CHALLENGES
Power-divergence must be controlled Large momentum required:
- discretised Fourier transform
- control power-suppressed corrections Renormalisation and continuum limit:
- perturbative truncation uncertainties - discretisation effects
Matrix elements extracted from Euclidean correlator - identical to that extracted from LSZ reduction
GENERAL PROCEDURE:
GENERAL CHALLENGES
R. Briceno, M. Hansen & CJM, PRD 96 (2017) 014502
Matrix elements extracted from Euclidean correlator - identical to that extracted from LSZ reduction
EUCLIDEAN CORRELATORS
Agnostic matrix elements
Spacelike distributions assumed identical in Euclidean and Minkowski space First calculation to work strictly in Euclidean space found no IR divergence!
THE WORRY
C.E. Carlson, M. Freid, PRD 095 (2017) 094504
Introduce a scalar, toy-model spacelike distribution
Momentum space correlation function:
perturbative QCD scalar toy model
Consider and compare:
1. LSZ reduction in Minkowski spacetime 2. Long time behaviour in Euclidean space
SCALAR TOY MODEL:
SPACELIKE DISTRIBUTION
R. Briceno, M. Hansen & CJM, PRD 96 (2017) 014502
Spacelike distributions assumed identical in Euclidean and Minkowski space First calculation to work strictly in Euclidean space found no IR divergence!
No fundamental challenge to, or problem with, this whole approach THE WORRY
C.E. Carlson, M. Freid, PRD 095 (2017) 094504
R. Briceno, M. Hansen & CJM, PRD 96 (2017) 014502
Power-divergence must be controlled GENERAL PROCEDURE:
GENERAL CHALLENGES
CJM & K. Orginos, JHEP 03 (2017) 116 CJM & K. Orginos, 1710.06466
CJM, 1710.04607
SMEARING
Deterministic evolution in new parameter - flow time - one-parameter mapping
- five-dimensional theory
Drives fields to minimise action - removes UV fluctuations Finite correlation functions remain finite
Correlation functions of “bulk” fields provide probe of underlying field theory GRADIENT FLOW
Narayanan & Neuberger, JHEP 0603 (2006) 064 Lüscher, Commun. Math. Phys. 293 (2010) 899
Lüscher & Weisz, JHEP 1102 (2011) 51 Luscher, JHEP 04 (2013) 123 Makino & Suzuki, arXiv:1410.7538
Deterministic evolution in new parameter - flow time - one-parameter mapping
- five-dimensional theory
Drives fields to minimise action - removes UV fluctuations Finite correlation functions remain finite
Correlation functions of “bulk” fields provide probe of underlying field theory GRADIENT FLOW
CJM, PoS(Lattice2015) 052 Narayanan & Neuberger, JHEP 0603 (2006) 064
Lüscher, Commun. Math. Phys. 293 (2010) 899
Lüscher & Weisz, JHEP 1102 (2011) 51 Luscher, JHEP 04 (2013) 123 Makino & Suzuki, arXiv:1410.7538
SMEARING
GRADIENT FLOW
Gradient flow is a smearing (smoothing) tool that:
● generates more continuum-like operators
● provides a method to fix smearing length scale
Flow time serves as a nonperturbative, rotationally-invariant cutoff Matrix elements of operators at fixed flow time are finite
Fixing the flow time (physical units) allows a continuum limit In essence: exchange lattice regulator for gradient flow regulator
CJM & K. Orginos, PRD 91 (2015) 074513
SMEARED QUASI DISTRIBUTIONS
Provides continuum limit
Defined as
Related to light-front PDFs via
Provided
Matching kernel satisfies
SMEARED QUASI DISTRIBUTIONS
CJM & K. Orginos, JHEP 03 (2017) 116 CJM, 1710.04607
Feynman diagrams at one loop in perturbation theory
Smeared quasi distribution Quasi distribution
MATRIX ELEMENTS IN PERTURBATION THEORY
CJM, 1710.04607
At one loop
where
Two regimes:
1. Local vector-current limit
2. Small flow-time limit
MATRIX ELEMENTS IN PERTURBATION THEORY
CJM, 1710.04607
Hieda & Suzuki, MPLA 31 (2016) 1650214
MATRIX ELEMENTS IN PERTURBATION THEORY
CJM, 1710.04607
SUMMARY
GENERAL PROCEDURE:
GENERAL CHALLENGES
Power-divergence must be controlled Large momentum required:
- discretised Fourier transform
- control power-suppressed corrections Renormalisation and continuum limit:
- perturbative truncation uncertainties - discretisation effects
Matrix elements extracted from Euclidean correlator - identical to that extracted from LSZ reduction
PDFs FROM EUCLIDEAN SPACETIME
Quasi distributions
Most theoretical issues generally under control
SMEARED QUASI DISTRIBUTIONS Finite continuum distributions
Looking forward: study systematics THE GRADIENT FLOW
Nonperturbative, gauge-invariant regulator
Matrix elements finite at fixed flow time
THANK YOU
cjm373@uw.edu
Systematic uncertainties
● finite lattice spacing
● finite volume
● unphysical pion masses
● excited state contamination
● Euclidean spacetime
● nontrivial renormalisation
LATTICE QCD Nonperturbative gauge-invariant regulator Rigorous definition of the path integral
quarks
gluons
SCALAR FIELD THEORY Scalar field theory
Exact solution possible with Dirichlet boundary conditions
Smearing radius
Interactions occur at zero flow time (i.e. in the original “boundary” theory):
guarantees that renormalised correlation functions remain finite.
CJM & K. Orginos, PRD 91 (2015) 074513
QCD QCD
Exact solution no longer possible (even with Dirichlet boundary conditions)
Smearing radius
Interactions occur at non-zero flow time: generalised BRST symmetry guarantees renormalised correlation functions remain finite.
EXPERIMENTAL EXTRACTION
From PDG 2016
Implemented nonperturbatively via discretised diffusion equation
and
ON THE LATTICE
Lüscher & Weisz, JHEP 1102 (2011) 51 Luscher, JHEP 04 (2013) 123
lattice gauge action
covariant lattice Laplacian
