CLASeg1 pol'proton analysis

1
CLAS-eg1 pol.-proton analysis
H.Avakian (JLab)
semi-SANE Collaboration Meeting April 21,
2005
2
SIDIS studies require

• Factorization in SIDIS
• Applicability of partonic description at JLab
  energies (HT lt10-15)
• low W and Q2 (direct production of resonances)
• low MX in ep-gtehX (resonances in target
  fragment, FSI)
• Contamination from non-DIS processes
• target fragmentation
• exclusive production (vector mesons)

3
SIDIS Target fragmentation
xFgt0 (current fragmentation)
xFlt0 (target fragmentation, TFR)
Rapidity (hgt1) and xF cuts are on average
related to z cut
What is the fraction of target fragmentation in
forward hemisphere?
What is the contribution to various observables
(multiplicities, asymmetries)?
4
SIDIS Missing mass of pions in ep?epX
p0
p-
p
n
D
D0
Applying MXgt1.4 (Q2gt1.5,W2gt4) we reduce the
fraction of where resonances in the target
fragment.
5
ALU from CLAS 4.3 and 5.7GeV data
DIS-cuts Q2gt1 GeV2 Wgt2 GeV y lt 0.85 0.5 lt z lt 0.8
5.7GeV
No significant dependence on beam energy
4.3 GeV
No significant ALU dependence on Mx beyond the
neutron.
6
SIDIS contamination from exclusive 2p
Major part of exclusive 2 pions are from r
Understanding of exclusive 2 pion production at
large W will help to understand single pion
measurements
7
SIDIS analysis strategy
Current Use MC(LUND) and data comparison to
define kinematic regions (in z,Mx,) where
contributions from non-DIS processes and HT
effect are not significant within statistical
accuracy of specific measurement.

• Future (also included in 12 GeV program)
• Perform studies of all contributing mechanisms,
  including
• Higher Twist effects
• Target fragmentation
• Exclusive channels

8
SIDIS MC and data analysis
histo-data files (P.Bosted) contain counts per
bin, occupy small disk space and are used for
data analysis.
N(i,j,k,l,m,n,h) array in a data file per run
used
i1,7 Q2-bins 0.453, 0.645, 0.919, 1.31, 1.87,
2.66, 3.79, 5.40
j1,6 x-bins 0.12-0.48
n 1-p, 2-p-, 3-p0
k1,8 z-bins 0.1-0.9
h1 helicity h2 helicity-
l1,6 PT-bins 0.0-1.2
m1,12 f-bins 0.0-360.0
The same structure created using the PEPSI-MC
with CLAS acceptance and smearing included
9
PEPSI MC asymmetries
GRVS set used
ALL for 3 pions from MC histo-data files
Testing extraction procedure Extract Du and Dd
using p0 Asymmetries on proton and deuteron
extracted
10
SIDIS factorization studies
A1pp- A1paA1p-
___________- 1a
LUND-MC used to get a

• A1 inclusive, from pp- sum and p0 are
  consistent (in range 0.4ltzlt0.7 )
• A1p dependence can serve an important check of
  HT effects and applicability of simple partonic
  description.
• There is an indication that A1p of pp- is
  lower than inclusive at large z.

11
SIDIS factorization studies
LUND-MC used to get a
ep?epX

• A1 inclusive and from pp- (HERMES published)
  show similar trend.
• Low A1p for p p- will lead to positive Ds
  (require more studies)

12
A1p -for p p - , p 0
HERMES
Multiplicity of p0 (pp-) is spin
independent.(if no strangeness) and provides a
unique test of partonic description

• A1 for p0 (pp-) can be a source of
  information on Ds/s
• (Frankfurt, Strikman et al. 1989)

13
g1/f1 PT-dependence
Constant in perturbative limit
Asymmetries from kT-odd (f1-, h1-, gT..) and
kT-even (g1) distribution functions are expected
to have a very different behavior
14
Precision measurements of parton distribution and
fragmentation functions in hard scattering
kinematics require

• MC generator based on LUND-MC (PYTHIA,JETSET)
  tuned for JLab, including
• Semi-Inclusive DIS
• Exclusive Processes
• T-odd Distribution and Fragmentation
• Radiative Processes

After testing with CLAS this MC may provide basis
for precision measurements of PDFs and FFs at
JLab (SANE?)
15
Summary
Precision measurement of parton distribution
functions will require separation of
contributions from different mechanisms and
global analysis of spin and azimuthal asymmetries
for exclusive and semi-inclusive final states
measured in a wide range of PT,f,x,Q2
16
support slides
17
Longitudinally polarized target future
measurements at CLAS
Hunf-5Hfav
Hunf-1.2Hfav
Hunf0

• Sensitive to unfavored fragmentation for p - and
  p 0
• Study the Collins fragmentation mechanism with
  long.pol.target
• p 0 SSA not sensitive to HT and diffractive r0
  production

18
Extracting h1L from p AUL
For Collins fragmentation use chirally invariant
Manohar-Georgi model (Bacchetta et
al) Distribution functions from cQSM from Efremov
et al
eg1
Systematic error from unknown ratio of favored
and un favored Collins functions (R
H1d?p/H1u?p), band correspond to -2.5ltRlt0
p- and p0 SSA will give access to h1Ld ( If R
-1 neutron data could be used)
19
Extracting h1 and h1L from p AUT and AUL
For Collins fragmentation use chirally invariant
Manohar-Georgi model (Bacchetta et
al) Distribution functions from cQSM from Efremov
et al
Systematic error from unknown ratio of favored
and un favored Collins functions (R
H1d?p/H1u?p), band correspond to -2.5ltRlt0
p- and p0 SSA will give access to h1d/h1Lu ( If R
-1 neutron data could be used)
20
Measuring the Q2 dependence of SSA
ssinfLU(UL) FLU(UL) 1/Q (Twist-3)
For fixed x, 1/Q behavior expected
Wide kinematic coverage allows to check the
higher twist nature of beam and longitudinal
target SSAs
21
Polarized target HERMES vs CLAS at 5.7GeV
GRSV95
x3 difference inltQ2gt may account for 15-20
in A1p at low x
22
SIDIS factorization studies
No significant variation observed in z
distributions of p for different x ranges
Double spin asymmetry for different samples
consistent, indicating no significant
contribution from non-DIS processes
CLAS data at 6GeV are consistent with
factorization and partonic description for
variety of observables with unpolarized and
polarized targets.