1 Inelastic J/y differential cross sections: paper material A. Bertolin , R. Brugnera Outline: • short introduction • differential p t 2 cross section in z bins • differential z cross section in p t bins • y(2S) to J/y cross sections ratio • momentum flow along the J/y direction (new) • outlook ZEUS week 15-17 Feb. 2012 DESY, 15/2/2012
52
Embed
Inelastic J/ y differential cross sections: paper material
ZEUS week 15-17 Feb. 2012 DESY, 15/2/2012. Inelastic J/ y differential cross sections: paper material. A. Bertolin , R. Brugnera. Outline: short introduction differential p t 2 cross section in z bins differential z cross section in p t bins y (2S) to J/ y cross sections ratio - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Inelastic J/y differential cross sections: paper material
A. Bertolin, R. Brugnera
Outline:• short introduction
• differential pt2
cross section in z bins
• differential z cross section in pt bins
• y(2S) to J/y cross sections ratio• momentum flow along the J/y direction (new)• outlook
ZEUS week 15-17 Feb. 2012
DESY, 15/2/2012
2
previous ZEUS papers:
1. Measurement of inelastic J/y photoproduction at HERA DESY 97-147 (July 1997) Zeitschrift f. Physik C76 (1997) 4, 599-612 Alessandro B. PhD thesis + Riccardo B. 94 data
2. Measurements of inelastic J/y and y^prime photoproduction at HERA DESY-02-163 (September 2002)Europ. Phys. Journal C 27 (2003) 173-188 Alessandro B. + Riccardo B. 96-97 data
3. Measurement of Inelastic J/y Production in Deep Inelastic Scattering at HERADESY-05-071 (May 2005)European Physical Journal C44 (2005) 13-25Alessandro B. + Alexei A. + Igor K. (+ Leonid G. + Riccardo B.) 96-00 data
4. Measurement of J/y helicity distributions in inelastic photoproduction at HERADESY-09-077 (June 2009)JHEP12 (2009) 007Alessandro B. + Riccardo B. HERA I + HERA II
short introduction
3
paper material
as already stated several times we have to take into account several external constraints: we are in 2012, myself and Riccardo have other commitments ...
so we have to define some realistic goals for this paper:
• ds / dpt2 in z slices: shown as preliminary in DIS11, all mature PHP experiments have
measured it
• ds / dz in pt slices: inelasticity, z, is a key variable for J/y production
provide to the theorists an inelasticity distribution at “high p t”
• y(2S) to J/y cross section ratio: needed to evaluate the y(2S) ® J/y p p feed down
• study of the momentum flow (using vertex tracks) along / against the J/y direction of flight: theorists are telling us that this measurements is very significant, never done before in PHP
4
96/97 pos. data (20718 27889)
mbtake & GLOMU effic.: ok
num97v5.2 HERWIG MC
evtake + mbtake lumi: 38.0 pb-1
98/99 ele. data (30758 32906 )
mbtake & GLOMU effic.: ok
num98v5.0 HERWIG MC
evtake + mbtake lumi: 15.9 pb-1
99/00 pos. data (33125 37715 )
mbtake & GLOMU effic.: ok
num98v5.0 HERWIG MC
evtake + mbtake lumi: 60.2 pb-1
03/04 pos. data (45783 51245)
mbtake & GLOMU effic.: ok
num03t6.0 HERWIG MC
evtake + mbtake lumi: 36.9 pb-1
04/05 ele. data (52258 57123 )
mbtake & GLOMU effic.: ok
num05t3.0 HERWIG MC
evtake + mbtake lumi: 126.5 pb-1
06 ele. data (58207 59947)
mbtake & GLOMU effic.: ok
num06t4.0 HERWIG MC available
evtake + mbtake lumi: 53.3 pb-1
06/07 pos. data before L/MERs (60005 62639)mbtake & GLOMU effic: oknum07t4.1 HERWIG MC availableevtake + mbtake lumi: 137.5 pb-1S (HERA I) = 114.1 pb-1
diffractive events are generated at z » 1we measure the cross section for z < 0.9the overlap should be ZEROHOWEVER due to the finite z resolution some of the diffractive events are RECONSTRUCTED with z < 0.9
EPSOFT MC generated z
fit the reconstructed z distribution to estimate the amount of diffractive events left after the z < 0.9 cut
for the fit the only change with respect to the nominal analysis is done for the z range:• from 0.1 < z < 0.9• to 0.3 < z < 1
we:• remove 0.1 < z < 0.3 because there is no diffractive yield at low z, instead observe larger
non resonant background and expect contributions also from beauty and may be resolved• we add 0.9 < z < 1 to have more diffractive background and hence more “signal” for the fit
13
Proton diffractive dissociation subtraction
data
HERWIG MC
EPSOFT MC
even
ts /
bin
shape distorted by the E(FCAL) > 1 GeV and ³ 3 (vertex) tracks requirements
purpose of the fit: fractions of HERWIG MC and EPSOFT MC that best describe the data
14
Proton diffractive dissociation subtraction
even
ts fr
actio
n / b
in
HERWIG MC component
from fit
MC sumdata
outcome: HERWIG MC fraction for z < 0.9 is 93.9 %
data: stat. errors, MC: sys. errors, due for example to the hadronic energy resolution, with size comparable to the data stat. errors, NOT shown in the above plot
in the range 0.3 < z < 0.9, this fit, the result is: 0.9397566
in the nominal analysis z range, 0.1 < z < 0.9, the result is: 0.9397768
15
• 60 < W < 240 GeV• 0.3 < z < 0.9• pt > 1 GeV
• ³ 3 (vertex) tracks• E(FACL) > 1 GeV• HERA I + HERA II data
even
ts fr
actio
n / b
inControl plots for the MC mixture
EPSOFT MC
HERWIG and EPSOFT MC predictions are affected by (systematic) uncertainties due, for example, to the hadronic energy reconstruction NOT shown in these plots
even if not too large these uncertainties are of the size of the data stat. errors (HERA I + HERA II data)
in the signal part of the sys. error the W and pt HERWIG MC spectra and the hadronic energy resolution implemented in the MC, E-Pz(rec)-E-Pz(gen), will be varied accordingly
16
even
ts fr
actio
n / b
inControl plots for the MC mixture
the uncertainty on the muon chamber efficiency is not shown for the MC histograms, its size is
similar to the data statistical error, this uncertainty will be
included in the signal part of the sys. error
EPSOFT MC
17
0.75 < z < 0.9 0.6 < z < 0.75
0.45 < z < 0.6 0.3 < z < 0.45
Alessandro: black
Riccardo: blue
60 < W < 240 GeV
1 < pt2 < 100 GeV2
Cross section vs pt2 in z slices
18
Cross section vs pt2 in z slices
0.1 < z < 0.3 Alessandro: black
Riccardo: blue
two analyses are in very good agreement
19
Cross section vs z in pt slices
1 < pt < 2 GeV 2 < pt < 3 GeV
3 < pt < 4.5 GeV pt > 4.5 GeV
60 < W < 240 GeV
0.9
0.90.9
0.9
Alessandro: black
Riccardo: blue
two analyses are in very good agreement
20
2S to 1S cross section ratio
basic formulas:
with some algebra:
FULL details:http://www-zeus.desy.de/~bertolin/ZEUS_ONLY/zn-03004/node41.htmlZEUS Note of the HERA I paper
for the HERA I paper we used PDG2002:Data Br1SMu/5.88E-2/,Br2SMu/0.70E-2/,Br2S1S/55.7E-2/in PDG2010:Data Br1SMu/5.93E-2/,Br2SMu/0.77E-2/,Br2S1S/59.5E-2/
in today’s presentation PDG2010 values are being used
PDG2010 ≡Please use this CITATION: K. Nakamura et al. (Particle Data Group), Journal of Physics G37, 075021 (2010) and 2011 partial update for the 2012 edition.
21
2S to 1S cross section ratio vs z
two analyses are in very good agreement
22
2S to 1S cross section ratio vs pt
two analyses are in very good agreement
23
2S to 1S cross section ratio vs W
two analyses are in very good agreement
24
p flow against / along the J/y direction
• 60 < W < 240 GeV
• pt > 1 GeV
• 0.3 < z < 0.9 (0.1 < z < 0.3 removed because of the small amount of signal expected and large amount non resonant background observed)
• ³ 3 (central) vertex tracks• E(FCAL) > 1 GeV
J/y direction of flight in the lab.
• vertex tracks
• pt(min) > 150 MeV
• | h | < 1.75• do not consider the m+ and m- tracks• track and J/y same hemisphere: p projection along
the J/y gives a positive contribution to Palong• track and J/y opposite hemisphere: p projection
along the J/y gives a positive contribution to Pagainst
as discussed with F. Maltoni (UC Louvain, Be)
25
p flow against / along the J/y direction
goal: check if the Color Singlet Model, as implemented in the LO + PS HERWIG MC, can give a reasonable description of the energy flow along the J/y direction
the against direction is studies only “as a cross check”
reminder:• in the CSM you have only a J/y and a backward “hard” gluon (transverse momentum
conservation in PHP) … so along we expect almost nothing• in the Color Octet Model you have a J/y with some nearby hadronic activity (“soft” gluons)
and a backward “hard” gluon … some along activity should be visible
• clearly such an analysis would profit of large J/y pt (like in CMS) but theorist told us than a qualitative results in PHP would be very valuable anyway
26
p flow against / along the J/y direction
• have to measure two distributions: Pagainst and Palong
binning of each distribution: 0. 0.25 0.5 1. 1.5 2. 2.5 3. 4. 5. GeV
• have to measure vs pt(J/y)
pt(J/y) bins: 1. 1.4 1.9 2.4 3.4 4.2 10. GeV (bins used for the “inelastic J/y helicity paper”)
expect large statistical errors (like in the helicity paper)
Alessandro’s analysis:
(1) fit a number of J/y events for every pt(J/y) bin (6 invariant mass fits only)
(2) compute Pagainst / Palong distribution for events close to the J/y mass peak
(3) compute Pagainst / Palong distribution for events in the side bands
(4) knowing the background below the peak normalize properly the side bands contribution for Pagainst / Palong
(5) subtract (2) and (4) to get Pagainst / Palong for J/y events only
(6) build Pagainst / Palong for every pt(J/y) bin using MC: add HERWIG (94 %) and EPSOFT (6 %) MC predictions
(6) compare data and MC predictions both with normalization set to 1.
27
Riccardo’s analysis:
(1) build an invariant mass distribution for every pt(J/y) bin, 1. 1.4 1.9 2.4 3.4 4.2 10. , and for every p flow bin, 0. 0.25 0.5 1. 1.5 2. 2.5 3. 4. 5. , i.e. 6 x 9 distributions for Palong and 6 x 9 distributions for Pagainst
(2) fit them all to get Pagainst / Palong for J/y events only
(3) build Pagainst / Palong for every pt(J/y) bin using MC: add HERWIG (94 %) and EPSOFT (6 %) MC predictions
(4) compare data and MC predictions both with normalization set to 1.
every method has its own advantages and disadvantages, both have been used in the past … side band subtraction method has been for the inelastic J/y helicity paper
how does the two methods compare ?
p flow against / along the J/y direction
28
p flow against the J/y: data comparison
two analyses are in very good agreement for the Pagainst distribution
Alessandro: black
Riccardo: blue
29
p flow along the J/y: data comparison
two analyses are in very good agreement for the Palong distribution
Alessandro: black
Riccardo: blue
30
p flow along / against the J/y: MC comparison
along against
Alessandro: black
Riccardo: pink
two analyses are in very good agreement
31
p flow along / against the J/y: MC comparison
along against
two analyses are in very good agreement
32
p flow against the J/y
crosses: HERA I + HERA II data, stat. errors only
continuous line: HERWIG MC, CMS only
the HERWIG MC provides a reasonable description of the data
33
p flow along the J/y
crosses: HERA I + HERA II data, stat. errors only
continuous line: HERWIG MC, LO CMS + PS
the HERWIG MC provides a reasonable description of the data moreover the agreement improves as the J/y pt increases
momentum flow around the J/y does not show large deviations from the CMS picture
34
Cross section vs pt2 in z slices
35
Cross section vs pt2 in z slices
36
Cross section vs pt2 in z slices
37
Cross section vs z in pt slices
38
Cross section vs z in pt slices
39
Observation on cross sections
• proton diffractive dissociation background is subtracted. This is fundamental for the theorists;
• y(2S) feed down is not subtracted, an inclusive reconstruction of the y(2S) decay would be needed and we do not have it. But the 2S to 1S cross section ratios vs p t W and z are measured. Theorists (should) know how to correct for this. Moreover they never asked us to perform this subtraction;
• J/y from any B hadron decay are not subtracted:
• the ratio between B hadron to J/y to J/y from primary vertex is much smaller at HERA than at hadron colliders (CDF, D0, CMS, ATLAS)
• theorists never asked us to perform this subtraction
• no PHP experiment up to now has performed this subtraction
• a subtraction based on data is very hard
• the only subtraction we could do would be fully based on MC … likely the theorists could account for this better than us (by adding a beauty component to the QCD predictions)
• the cross section we quote, the number in nb, is corrected for this effect, as explained in the following slides
• we will quantify the size of this contribution
40
for cross section vs pt2 in z slices
for every Dpt2 bin:
• disregarding any difference in the correction factors for beauty:
s = N / CH L f.f. BR Dpt2 = s0
• taking explicitly the beauty correction factor into account:
s = N – Nb / CH L f.f. BR Dpt2 + Nb / Cb L f.f. BR Dpt
2
…
s = s0 [1+(Nb/N)(CH-Cb)/Cb]
J/y from b decay: effect of the quoted number of nb
outcome:
• if Nb/N << 1 s = s0
• if CH=Cb s = s0
41
J/y from b decay: effect of the quoted number of nb
continuous: 0.1 < z < 0.3
dashed: 0.3 < z < 0.45
dotted: 0.45 < z < 0.6
z > 0.6: at the event level negligible
usually < 10 events, 16 at most (low z low pt)
expected number of J/y from b decay (Pythia PHP inclusive beauty sample):
when Nb is largest Nb/N < 16/375 < 0.045
usually Nb/N < 10/400 < 0.025
total number of beauty MC events processed:
22.207.433
42
0.3 < z < 0.45:
CH=Cb s = s0
whatever amount of beauty we have in data the number in nb we quote for the cross section is correct
0.1 < z < 0.3:
s = s0 [1+(Nb/N)(CH-Cb)/Cb]
= s0 [1+ 0.25×(Nb/N)]
= s0 [1+ 0.25×0.045]
< s0 [1+ 1.2 %]
J/y from b decay: effect of the quoted number of nb
for z > 0.45 Nb/N is so small that the effect is negligible anyway
43
J/y from b decay: effect of the quoted number of nb
outcome:
beauty is included in the cross section and the quoted cross section, the number of nb, takes this contribution properly into account
any theorist can compute the J/y cross section, according to his preferred model, compute the J/y from beauty contribution, according to his preferred model, add the two numbers and compare with our data
44
2S to 1S cross section ratios
naive expectation of the LO CS model: flat ratios at 0.25(2) according to the latest PDG values
most of the central values are above 0.25, mild indications that the z ratio may not be flat
LO CMS prediction:s µ Gmm / m3
1S: 3096.6 MeV, 5.93 % x 92.9 keV2S: 3686.0 MeV, 0.77 % x 304 keV
45
Conclusions and outlook
the material for the paper has been presented … we have the feeling this is the best we can do keeping in mind the different constraints we have (we are not leaving in a word with an infinite amount of money and time)
last but not least:
• the computation of the systematic errors will have to be carried out
• a paper draft will be prepared
46
extra slides
47
luminosities tuning plots:• high z (z > 0.9): EPSOFT MC tuning• medium z (z < 0.8): HERWIG MC tuning diffractive fit:• E(FCAL) only• >= 3 tracks cross section vs pt2 cross section vs z 2S to 1S ratio
/!\ check latest .for p flow along and against
48
ff*BR*Lumi_tot=0.0975*0.0593*468300• ff for 60 < W < 240 GeV and Q2
max=1 GeV2
• BR = ( 5.93 ± 0.06 ) × 10−2
• Lumi_tot = 468.3 pb-1 = 468300 nb-1
PDG read on 27/1/2012Please use this CITATION: K. Nakamura et al. (Particle Data Group), Journal of Physics G37, 075021 (2010) and 2011 partial update for the 2012 edition.