future gpd measurements - uvacollab : gateway...like’ version of the same process, that depends on...

Future GPD MeasurementsShort, Medium and Long TermRalf Kaiser, University of Glasgow

Topics

• Current & Future GPD Experiments

• HERMES, CLAS, CLAS12 COMPASS, PANDA

• Future Perspectives for GPD Measurements

Major GPD Experiments - Timeline

3

HERMES

2008 2010 2015 2020 2025

COMPASS

PANDA

JLAB JLAB 12 GeV

(c) R

.Kai

ser 2

008

Major GPD Experiments - Timeline

3

HERMES

2008 2010 2015 2020 2025

COMPASS

PANDA

JLAB JLAB 12 GeV

EIC

(c) R

.Kai

ser 2

008

HERMES @ DESY

4

!"#$%&'($)$*

!!

!"

!

#

"

!+,-

"

!

$

./0+123+45

,-6078!3

9!+:4

:,;<=>

9?,

0,;<!@021+42!

7,;<=>

7,;@=A

'!2'B,-6078!3

:+8.9;,.60'3

'!83-2C8!;D->E

,6.2!+0848!

9!+:4;,-6078!3

4!+FF8!;-2923,2'8;-<

# ! " % & ' ( ) * +

06F184

3488.;'.648

-2923,2'8;-G

>HBI;F"?

"J

$

4!9

9848,42!!8,2+.

DVCS measurements with Recoil Detector in 2006/7 yielded about as much data as 1995-2005; data are being analysed.

HERMES DVCS BCA

5JHEP 0806:066,2008

HERMES DVCS TTSA

6JHEP 0806:066,2008

Limits on Ju/Jd

7JHEP 0806:066,2008

HERMES Recoil Detector

8

!"#$#%&'($()$#*

+,$(*&

-##./%0

1#*&2/&'($3

2)/4/56/0"$07/8(-#%%()$#*&9/%0

-:&-#../;,$#*

<=7%0>$(%?

-#%%()$#*>

<:&.,@(*>&$7%0>$(%5>)/%$/..,$#*?

2)/4/&A

2/&'($3&B@C*/8

="/%!D,..(82),$$(*/%0-",;C(*

2)/4/&E

2/&'($3

=FG9H&2(%>#*>

2/&'($3&4*,;(<I.%?

-(..

=,*0($

Talk by A.Mussgiller tomorrow

Kinematic Plane MC & Data

9

DVCS MC Events TDR 2002 DVCS Candidate Event Data 2007DVCS, BH, ADVCS, background

Recoil Detector - Combined PID

10

JLAB Experiments

11

Hall A

CLAS12 Detector

12

CentralDetector

ForwardDetector

CLAS12 DVCS BSA

13

Phys.Rev.Lett.87:182002,2001. arXiv:0711.0755 (submitted to PRL)

CLAS12 DVCS BSA Projection and GPD H

14

2000 hrs at L=1035 cm-2s-1

CLAS12 DVCS BSA Projection and GPD H

14

2000 hrs at L=1035 cm-2s-1 Projection for GPD H

CLAS 6 Experiment E08-021

15

Courtesy A.Sandorfi

CLAS 6 Experiment E08-021

16

Courtesy A.Sandorfi

DVCS TTSA and GPD E

17

!UT ~ sin("!"S) cos(") Im{ F2H - F1E + … }+..

V.Guzey

E=0

• Transverse Asymmetry is large and has strong sensitivity to GPD E

• CLAS 6 experiment scheduled for 25 days in 2011

µ’

p’ µ

DVCS µp ! µ’p’"

" ECal1 + ECal2

#" $ 10°

Nµ=2.108/SPS cycle (duration 5.2s, each 16.8s)

Possibility for an increase of intensity?

2.5m cryogenic target to be designed and built

4m long Recoil proton detector to insure exclusivity to be designed and built

+ additional calorimeter ECal0 at larger angle

- 2011: long H2 target -! later: transversely polarized

COMPASS Setup for DVCS Measurements

18

Present 1m long Recoil Proton Detector in COMPASS for the hadron program (spectroscopy)

COMPASS - Present Recoil Detector

19

Analysis through dependence in ! and Q2 Absolute cross section measurements

At COMPASS with µ+ and µ- ! access both ImH and ReH

with DVCS + BH with polarized and charged leptons and unpolarized target µ

p µ

p BH calculable DVCS

+

"

! µ’ µ

#* #

p

DVCS BC&SA at COMPASS

20

Beam Charge and Spin Asymmetry at Eµ = 100 GeV COMPASS prediction

BC&SA !

µ’ µ

"* "

p

Q2

xBj

7

6

5

4

3

2

0.05 0.1 0.2

6 month data taking in 2010 250cm H2 target 25 % global efficiency

COMPASS BC&SA Projection

21Courtesy N.d’Hose

COMPASS Plans for DVCS Measurements

22

• 2008 test with muon beam (last week)

• DVCS feasibility study in the COMPASS environment, study existing recoil detector, calorimeter performance, extrapolate to larger, dedicated setup

• End of 2008 submit proposal for 2009-2015 to SPSC

• 2009 request for 30 days GPD pilot run

• From 2011 run with long recoil detector and liquid hydrogen target, later also with transverse target

PANDA @ FAIR

23

PANDA @ FAIR

232.7 m£ STFC Grant for the Dipole Magnet for Glasgow

24

• PANDA can measure the ‘cross channel’ or ‘time-like’ version of the same process, that depends on the same GPDs

• More precisely on Generalised Distribution Amplitutes, introduced by M.Diehl et.al. to describe the inverse process [PRL.81:1782 (1998)].

γ

γ

Transition Distribution Amplitudes

25

• The same factorisation proof as for DVCS does not hold for the crossed channel.

• Alternative approach: Transition Distribution Amplitudes• TDAs extend the GPD concept to transitions

[B.Pire, L.Szymanowski, PLB 622 (2005) 83, J.P.Lansberg et al. Nucl.Phys. A782 (2007) 16-23]

• Impact parameter space interpretation as for GPDs• Fourier transform gives a transverse picture of the pion cloud

in the proton

Transition Distribution Amplitudes

26

3

at large !. In the following, we shall therefore limit our-selves to the computation of the cross section for reaction(1) in this region. At large !, the ERBL regime (xi > 0)covers most of the integration domain. Therefore it is le-gitimate to approximate the cross section only from theERBL contribution, i.e. when the integration range ofthe momentum fractions is restricted to [0, 2!].

The di!erential cross section for unpolarised protonsand antiprotons is calculated as usual using Eq. (11).from the averaged-squared amplitudes,

|M!!! |2 =14

!

spsp̄

M!spsp̄

(M!!

spsp̄)!. (16)

|M00|2 vanishes at the leading-twist accuracy, as in thenucleon-form-factor case. The same is true for |M+"|2and |M0+|2, etc., since the x and y directions are notdistinguishable when "2

T is vanishing. We then define|MT |2 ! |M++|2 + |M""|2.

To compute I, we need to choose models for the DAsand the deduced TDAs. For the sake of coherence withexperimental data, we shall use reasonable parametrisa-tions of CZ [15] and KS [16], which are both based on ananalysis of QCD sum rules. For CZ, they are

V p(xi) = "as[11.35(x21 + x2

2) + 8.82x23 " 1.68x3 " 2.94],

Ap(xi) = "as[6.72(x22 " x2

1)], (17)T p(xi) = "as[13.44(x2

1 + x22) + 4.62x2

3 + 0.84x3 " 3.78],

and for KS (which we used in Fig. 2)

V p(xi) = "as[17.64(x21 + x2

2) + 22.68x23 " 6.72x3 " 5.04],

Ap(xi) = "as[2.52(x22 " x2

1) + 1.68(x2 " x1)], (18)T p(xi) = "as[21.42(x2

1 + x22) + 15.12x2

3 + 0.84x3 " 7.56],

and we evaluate our model TDAs from Eq. (13). Thisgives I # 1.28 · 105 for CZ and I # 2.15 · 105 for KS; thisyields an induced uncertainty of order 3 for our estimatesof the cross section. In the following, we use #s = 0.3 assuggested in [15].

Altogether, we have the following analytic results forthe dominant ERBL contribution:

|MT |2 =(4$#s)4(4$#em)f4

N

542f2"

2(1 + !)|I|2

!Q6. (19)

From this, we straightforwardly obtain d#dt , whose W 2

evolution is displayed on Fig. 2 (a) at "T = 0 for the twoextreme values of meson longitudinal momentum whereone may trust the soft pion limit, corresponding to pz

" = 0or |pz

"| = M/3 in the laboratory frame (! = 1 or 1/2).For the process (1), the averaged-squared amplitude is:

|Mp̄p#$+$""0 |2 =14

!

sp,sp̄,!,!!

M!spsp̄

1Q2

L!!! 1Q2

(M!!

spsp̄)!,

with L!!! = e2Tr(p/$"%/(&)p/$+%/%(&$)). Integrating on thelepton azimuthal angle "$, we have

"d"$|Mp̄p#$+$""0 |2 = |MT |2

2$e2(1 + cos2 '$)Q2

, (20)

0.01

0.1

1

10

100

6 12 18 24 30

d!

/d

t|"

T=

0 (

nb

/Ge

V2)

W2 (GeV

2)

|pz#|=0

|pz#|=310 MeV

0.01

0.1

1

10

100

5 10 15 20

d!

/d

t d

Q2| "

T=

0 (

pb

/Ge

V4)

Q2 (GeV

2)

W2=5 GeV

2

W2=10 GeV

2

W2=20 GeV

2

FIG. 2: (a) Di!erential cross section d!/dt for p̄p! "!#0 asa function a W 2 for |pz

"| = 0 (lower curve) and |pz"| = M/3.

(b) Di!erential cross section d!/(dtdQ2) for p̄p! $+$!#0 asa function of Q2 for various beam energies.

from which we get, via Eq. (10) and integrating over '$

the di!erential cross section displayed in Fig. 2 (b).Although the cross sections are evaluated at "T = 0,

we do not anticipate any dramatic "T -dependence of theTDAs below a few hundred MeV, so that our estimatesare likely to be valid in a not-too-narrow "T region. Toevaluate a magnitude of the integrated cross section wetake as an example the kinematical region with W 2 = 10GeV2 in Fig.2b accompanied by the Q2 window 7 GeV2

< Q2 < 8 GeV2, which corresponds to ! $ 1/2 or to thepion momentum of the order 310MeV. Integrating overthis Q2-bin and in a t"bin corresponding to "T < 500MeV leads then to a cross section around 100 femtobarns.Such a cross section is sizable and seems to be accessibleto experimental setups such as PANDA with the designedvalue of the FAIR luminosity.

The calculations done till now and which involve theproton % $0 TDA are valid for the small t region. Letus however stress that - due to the charge symmetry -an identical result will be obtained in the small u re-gion but with the p̄% $0 TDA. In the laboratory frameat GSI-FAIR, this second region is quite di!erent fromthe previous one since the $0 meson is boosted in theforward direction. A precise detection of the particles

• Current models of TDA predict small cross section (~100 fb)

• Need excellent detector system to remove background

• Measurement feasible with PANDA J. P. Lansberg et al.Phys.Rev.D76:111502,2007

Present and Future ep-Facilities

27

1031

1032

1033

1034

1035

1036

1037

1038

1 10 100ECM [GeV]

Lum

inos

ity [

cm-2

s-1

]

(c) R

.Kai

ser 2

008

HALL A

CLAS

CLAS12

HERMES COMPASS H1

ZEUS

1000


27

1031

1032

1033

1034

1035

1036

1037

1038

1 10 100ECM [GeV]

Lum

inos

ity [

cm-2

s-1

]

(c) R

.Kai

ser 2

008

HALL A

CLAS

CLAS12

HERMES COMPASS H1

ZEUS

1000

EIC


27

1031

1032

1033

1034

1035

1036

1037

1038

1 10 100ECM [GeV]

Lum

inos

ity [

cm-2

s-1

]

(c) R

.Kai

ser 2

008

HALL A

CLAS

CLAS12

HERMES COMPASS H1

ZEUS

1000


27

1031

1032

1033

1034

1035

1036

1037

1038

1 10 100ECM [GeV]

Lum

inos

ity [

cm-2

s-1

]

(c) R

.Kai

ser 2

008

HALL A

CLAS

CLAS12

HERMES COMPASS H1

ZEUS

1000

ELIC

eRHICep@FAIR LHeC

Future ep-Facility

28

• Several concepts for future ep-facility: eRHIC at BNL, ELIC at JLab (together referred to as EIC), LHeC at CERN and ep@FAIR

• All designs use an existing machine and combine it with a second, new machine to a collider

• High luminosity, high energy, energy range

• GPDs are only part of the physics reach of such a facility

GPD Measurements - The Way Forward

29


29

DATA


29

DATA

LATTICE


29

DATA

LATTICEMODELS


29

DATA

LATTICE

MODELS


29

DATA

LATTICE

MODELS

HERMES


29

DATA

LATTICE

MODELS

HERMES

COMPASS


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

VGG


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

VGG

DUAL


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

VGG

DUAL

KMK-P


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

DUAL

KMK-P


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC KMK-P


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

?


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

Combined Effortsby Experimentalists

and Theorists required !!!


29

DATA

LATTICE

MODELS

HERMES

COMPASS

HALL A

CLAS EIC

PANDA

QCDSF

LHPC

Combined Effortsby Experimentalists

and Theorists required !!!

EU FP7 JRA ‘Hard Exclusive Reactions Spokesperson RK

Summary and Outlook

30

• Generalised Parton Distributions are promising to revolutionize our picture of the nucleon and to solve the spin puzzle.

• Present experiments at HERMES and JLab are playing a pioneering role.

• Future experiments especially at JLab after the upgrade, but also at COMPASS and FAIR will further complete the picture.

• Ultimately a future ep-facility with high luminosity and an energy range up to higher energies will be required to finalise the picture.

• All of this will only be successful in the combination of experiments, lattice calculations and GPD model fits to the data.

future gpd measurements - uvacollab : gateway...like’ version of the same process, that depends on...

Documents