1/18

Sea Quark Longitudinal Spin: measurements at PHENIXDaniel Jumper

University of Illinois at Urbana-Champaign

For the PHENIX collaboration

CIPANP – Vail, CO – May 23, 2015 Proton Longitudinal Spin Structure Boson Probe for Polarized at RHIC PHENIX Measurements:

2/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Proton Spin Spin Composition of the Proton:

– Quark and anti quark spins

– Gluon spin

orbital angular momentum

Current Understanding DIS/SIDIS

some flavor separation

RHIC

and , with separation

5/2

3/2

01

5

3/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Accessing Sea Quark Spin

Advantages over DIS/SIDIS No fragmentation

functions

High

Constraints: Weak parity violation

’s couple to left handed and right handed

Known spin distributions

spin tends to be aligned with proton spin, anti-aligned

𝑊 ±

5/2

3/2

01

5

4/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Flavor Sensitivity at Forward Rapidity

A first order motivation: At large rapidity or

equations reduce:

Actual observable:

This and other higher order effects are accounted for in global fit

Proton Constituent Probability Distributions

5/2

3/2

01

5

5/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Spin at RHIC at up to 60% polarization

“Siberian Snake” Spin Flippers

5/2

3/2

01

5

6/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Spin at RHIC

Runs 2011-2013 2013 offers highest

luminosity

High polarization is important for figure of merit

5/2

3/2

01

5

7/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

PHENIX W Measurements

Central

Forward

5/2

3/2

01

5

8/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Central

Forward

PHENIX W Measurements5

/23

/20

15

9/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Central

Forward

PHENIX W Measurements5

/23

/20

15

10/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Central

Analysis Strategy: Reduce Background with

Relative Isolation Cut Energy ratio between candidate

and all hits in a cone

Factor of 10 reduction in Bkg

Fit Jacobian Peak Extrapolate background with Gaussian Process Regression technique

Fit signal + bkg model to data to extract signal yield

Calculate Asymmetry

5/2

3/2

01

5

11/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Central Run 13 final result just released!

Now available on arXiv: http://arxiv.org/pdf/1504.07451v1.pdf

5/2

3/2

01

5

12/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Forward 5

/23

/20

15

Challenge: Low signal with dominant background

13/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Forward

Challenge: Low signal with dominant background

Signal events dominated by other processes at low

Process cross sections vs from simulations

5/2

3/2

01

5

14/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Forward

Challenge: Low signal with dominant background

Signal events dominated by other processes at low

Limited resolution smears dominant low background

Process cross sections vs from simulations

5/2

3/2

01

5

15/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Forward

Challenge: Low signal with dominant background

Signal events dominated by other processes at low

Limited resolution smears dominant low background

Additional “fake muons” result from multiple scattering hadronsProcess cross sections vs from simulations

5/2

3/2

01

5

16/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Forward Analysis Strategy: Select High-Signal Data Subset

Multivariate Likelihood Ratio Calculation

Characterize Remaining Data Signal to Background Ratio Fit

Use simulations to estimate

distributions for components

Calculate & Correct Asymmetry Count yields for each helicity

Apply correction factors (Sig/Bkg ratio, efficiencies, …)

Likelihood Ratio

BackgroundSignal

SigBkg

5/2

3/2

01

5

17/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Forward

Preliminary Result

Currently moving toward a finalized result and publication

5/2

3/2

01

5

18/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Summary

Proton spin structure remains poorly constrained

’s as a weak probe, offer unique insight to ,

Forthcoming PHENIX results contribute to significantly improved constraints on these distributions Central results just released

Forward preliminary result available

5/2

3/2

01

5

19/18

Danie

l Jum

per - P

HE

NIX

W P

hysics

Backup

5/2

3/2

01

5