1. Report on p0 decay width: analysis updates

2. Analysis updates: Clock frequency correction . Flux is ~2.4% up
Tagger energy corrections were introduced
More statistics added: 130nA and 110nA runs were added to previously used 100nA runs. Statistics increased by factor of 2 and stat. err dropped from 0.28eV to 0.18eV

3. Analysis updates: Error budget was updated
Efficiency was recalculated (non-resonant gg background was added)
Background from w and r± decays was subtracted from p0 yield
To make sure that skim data are not biased, raw data files were analyzed. p0 yield fit parameters including lifetime are stable

4. Response function of Hycal:
Observed elasticity (i.e. EHycal/ETAGGER ratio) during snake scan when beam was centered on W2016
What do we measure during snake scan ? Shape of experimentally observed response function:
*) resp. function itself
*) accidentals (with false Tagger hits)
*) electronics noise
*) “leakage”

5. How to distinguish (tail and zero) of response function from the background. (Can we do it in the future run to decrease our systematics ?)
Put TAC behind Hycal during snake scan

6. Response function (MC) with and w/o TAC during snake
Hycal only
Hycal with TAC
TAC itself has energy resolution (MC) ~ 7%/Ö E

7. To study response function it was split into elasticity regions:
X = min – 0.2
X = 0.2 – 0.5
X = 0.5 – 0.8
X = 0.8 – 0.9

8. Fraction of events in elasticity tails was estimated for runs with different beam current: 20, 30, 50, 60, 95 pA
Hycal resp. function itself
is assumed to be independent from beam current.
Asymptotic values from
the following simple fit
were used as a most close
to the real values:
Y = c1/x + c0

9. What does MC reproduce for response function:
Snake scan
Centered beam

10. Data VS MC (percentage of events in the tail of response function):
Difference between data and MC ~ 0.5 – 0.9%
1.3GeV
4.4GeV
X range MC
Data (asympt.)
Min – 0.2
0.02 ± 0.00
0.15 ± 0.03
0.2 – 0.5
0.23 ± 0.04
0.5 – 0.8
0.07 ± 0.01
0.40 ± 0.05
0.8 – 0.9
0.38 ± 0.02
0.61 ± 0.05
X range MC
Data (asympt.)
Min – 0.2
0.03 ± 0.01
0.06 ± 0.02
0.2 – 0.5
0.02 ± 0.00
0.14 ± 0.04
0.5 – 0.8
0.08 ± 0.01
0.26 ± 0.03
0.8 – 0.9
0.40 ± 0.02
0.57 ± 0.04

11. Tail of response function comes mostly from wrapping:
Snake scan data might be used as a unique measure of wrapping thickness between modules. Any other measurements of this thickness could not have enough accuracy.

12. “tagging ratio” during snake, MC:
“tagging ratio” means number of events with enough energy and position close to beam impact point divided by number of total events with normal beam
Elasticity cut 0.7
Elasticity cut 0.4
edges / wrapping
edges / wrapping

13. “tagging ratio” during snake, data:
Dips near edges may slightly
vary from one module to another
Elasticity cut 0.4
Elasticity cut 0.7
edges / wrapping

14. Estimated contribution from response function to error budget
Systematic error comes from inconsistency between data and MC response functions (not from response function shape itself).
Since MC may underestimate tails of response function up to 0.9%, we have to introduce correction factor of for MC efficiency and 0.5% sys. error comes from this item.
More precise measurement of Hycal response function would decrease this uncertainty

15. Lifetime sensitivity to single gamma energy cut value
Method 1: we have to vary cut value and check how result (with no additional correction for efficiency) depends on it.
10% variation from cut value (0.5GeV) was used as a probe.
This cut variation is changing result by 0.2%
Used cut
Cut variation

16. Lifetime sensitivity to single gamma energy cut value
Method 2: vary cut value and check how result (with corrected efficiency) depends on it.
Final result shows 0.2% fluctuations from the mean value (it is probably statistical only fluctuations)

17. Lifetime sensitivity to single gamma energy cut value
Conclusion:
error budget contribution from this item is £ 0.2%

18. Contribution from misalignment:
Beam VS Hycal position uncertainty: 0.2mm uncertainty value gives negligible syst. Error
Beam slope stability: 0.12mrad uncertainty value gives syst. error £ 0.1%
Beam width stability: varying beam parameters according to superharp scan data gives 0.3% syst. error
Hycal Z uncertainty: 1.5cm uncertainty value gives 0.4% syst. error

19. w and r± sources of p0 background were subtracted

20. The following sources of quasielastic p0s were studied:
Coherent w: g A -> w A
Incoherent w: g A -> w N A’
g A -> w D A’
Incoherent r±: g A -> r± N A’
g A -> r± D A’
With consequent decays:
w/r0 -> p0 g; w -> p0 p- p+; r± -> p0 p±

21. Simulations: t’ distributions for w and r production used
t’ = |t – tmin|

22. Simulations: production angle distributions for w and r used

23. Simulations: cos(qGJ) distributions for w and r used
1+x2
1-x2

24. Distributions obtained from MC were scaled and subtracted from the data
Scale factor for MC:
Kscale = F data / FMC
FMC = Nevent NC / (L Br s)
L – number of atoms per sq. unit
Br – decay branching ratio
s – production cross section

25. p0 elasticity distribution observed (crystal part of Hycal) with simulated background from w and r+ -
w and r+ -
contribution
r+ - only
contribution

26. p0 elasticity distribution with subtracted background from w and r+ -
Elastic peak
parameters
improved

27. dN/dq for original data and with background subtraction (on the level of gg invariant mass spectrum, bin by bin basis)
original
rad. width changes
by ~ -0.8%;
more calculations
will be performed
Subtracted part

28. Non-resonant 2g background was added to the MC
Old method: number of reconstructed p0 events was calculated in mass window
New method: background with linear shape and line parameters obtained from the data (individually for each T-counter and each q-bin) was added to MC distributions and the same fitting procedure was performed as for the data
data MC

29. Non-resonant 2g background affects on efficiency
New efficiency curve
for 5.2GeV
Difference between new
and old efficiencies
Rad. Width goes +1%

30. New formfactors (with more advanced calculations) were introduced into analysis

31. Imaginary part of formfactors were introduced:
Interference term was modified: sin(f) appeared
ds/dq interference =
Ö [ cos(f)· ( Re(Prim)· Re(Coh) + Im(Prim)· Im(Coh) ) +sin(f) · ( Re(Prim) · Im(Coh) + Im(Prim)· Re(Coh) ) ]

32. Reanalyzing lifetime with all corrections applied:
c2 was estimated for Primakoff region (0o…0.5o) separately to make sure quality of the fit in most sensitive region.
Old formfactors:
(Primakoff region (0o-0.5o) c2 = 1.12
New formfactors:
(Primakoff region (0o-0.5o) c2 = 1.05
Better agreement

33. Error budget table Target (thickness and density) ……... ………. 0.1%
Target material (impurity) …………... ………. -0.4%
Photon beam flux ……………………. ……… 1.1%
Trigger efficiency …………………… ………. +0.1%
ADC channels status during run time ………. negligible
Photon beam energy uncertainty …… %
Photon beam flux distribution within E-counters ……………………... ….…… 0.1%
p0 branching ratio …………………….. ………. Negligible
Energy cut for single g ……………….. ………. 0.2%
Energy cut for p0 ……………………... ………. negligible
p0 production angle resolution uncertainty ………. 0.25%

34. Error budget table MC eff. simulations (statistical) accuracy …… 0.3%
Selection of “best in time” beam candidates %
Hycal z uncertainty …………… %
Beam position uncertainty ……… negligible
Beam direction uncertainty ……… %
Beam divergence uncertainty ……… %
Timing cut (tdiff) ……… %
Hycal response function ……… %
Background separation: (currently used value) %
Uncertainty in theoretical parameters ………. ?

35. Summary Sources of background were taken into account (most of them)
Error budget is close to be completed
Rad. width (to be updated):
8.18eV ± 2.2% stat. ± 2.1% syst.

36. Plans: Update incoherent amplitude: ds/dq shape update
Include Fermi motion
(production angle and energy spread) ?
Introduce different efficiency for primakoff /coherent and incoherent production
Check lifetime stability to variations of formfactor parameters. More formfactor updates will be performed by Sergey
Use of LG part of Hycal for better incoherent estimation

37. Spare slides

38. Fermi motion angular smearing (p0 production ):
Simulated smearing
for q = 0.25o, 1o and 2o
dw/dp (Carbon)