Документ взят из кэша поисковой машины. Адрес оригинального документа : http://zebu.uoregon.edu/~uochep/docs/docs01/Snow01_cal.pdf Дата изменения: Wed Oct 17 20:53:11 2001 Дата индексирования: Tue Oct 2 10:52:47 2012 Кодировка: Поисковые слова: вторая космическая скорость

Calorimeter Resolution Studies
www.slac.stanford.edu xorg lcd calorimeter snow01 frey.pdf

Ray Frey, U. of Oregon Snowmass 2001

I. Single-particle resolution and shower studies Gary Bower and Ron Cassell, SLAC
See www.slac.stanford.edu xorg lcd calorimeter snow01 bower.pdf

II. W Si EM calorimeter dynamic range III. Energy ow ultimate" jet resolutions

1


First Look at W Si Dynamic Range Requirement Want to measure MIPs 400 keV mm in Si and dense EM showers due to 250 500 GeV Bhabha e EM showers in W are dense: Rm = X021:2MeV=Ec= 9:1 mm but is modi ed signi cantly by sampling layers 1. EGS setup 2. Results 3. Conclusions

2


EGS Setup Use the G. Lindstrom recommendations for Ecut in thin sampling layers. Good accuracy with nite CPU time. Reduced Ecut, Pcut in thin regions near the Si Step size small 0:3 everywhere
~6 X 0.5 mm 0.4 mm 0.5 mm >3X

W bulk

W thin 100 keV 100 keV

Si 100 keV 20 keV

W thin 100 keV 500 keV

W bulk 500 keV 1000 keV

e

Ecut = 500 keV Pcut = 500 keV

3


Fraction of total EDEP in 1cm 1cm : Ee = 100 GeV

1. Broad shower max in depth 6:5 1 X0 2. Fraction of energy in central 1cm 1cm is independent of Ee

Results not sensitive to these
4


EDEP fraction in center pixel as function of pixel size mm:

need big pixel size reduction to change dynamic range requirement signi cantly

5


So for 1cm1 cm : 250 GeV Bhabha MIP = 340MeV=0:16MeV = 2100 11 bits +3 bits for MIP over threshold +2 bits formargin =16 bits decrease in pixel area by 100 gives only 2-3 bits reduction Results depend on sampling gap thickness: For SD, 2.5 mm W + 2.5 mm gap Rm =0:9 cm ! 1:9 cm

6


EFlow: Jet particles have modest energy Tracker has best resolution Role of calorimeter: separate neutrals and measure their energy
=K

vs 's and

=K

vs K 0's L

=0 separation x e.g. 3-d DOCA depends on B , R, and physics pro cess
Figures of merit: BR2=Rm and segmentation and similarly for the HAD Requirements:
x

Rm similarly for HAD x Rm sim. HAD tracking of MIPs in EM and HAD shower pattern recognition 0 E for 's, KL's, n's
x
7


Side bene ts: aid to tracker trajectory non-IP pointing electron and muon id. Bhabha acolinearity Lum. measurement

8


Separation of Cluster and nearest charged track Large Det: BR2 =12 T-m2, Rm =2:1 cm, e+e, ! ZZ !jets Cluster is due to a :
photon clus-trk: dz vs drphi 14
d2D Nent = 846 Mean x = 0.7084 Mean y = 0.2393

dr (cm)

dr Nent = 846 Mean = 0.7729

350

RMS = 1.874

12

RMS x = 1.753 RMS y = 0.6093

300

10 250 8

200

6

150

4

100

2

50

0 0

2

4

6

8

10

12

14

0 0

2

4

6

8

10

12

14

Cluster is due to a :
photon clus-trk: dz vs drphi 14
d2D Nent = 1213 Mean x = 4.745 Mean y = 3.82

dr (cm)

dr Nent = 1213 Mean = 7.721 RMS = 3.619

12

RMS x = 3.557 RMS y = 3.599

12

10

10

8

8

6

6

4

4

2

2

0 0

2

4

6

8

10

12

14

0 0

2

4

6

8

10

12

14

9


Can the proposed calorimeters meet the requirements ? TESLA and SD W Si with x 1 cm: Likely Yes L : Is scint. tile, x 5 cm , 1 mm thick, possible ? If so, does this segmentation su ce ? Can it be reduced ? This is work in progress: full sims., pattern recog., algorithms, etc. Can also imagine a parameterized approach for fast sims. In the meantime: What is optimal EFlow performance for a given detector ? Assume ideal charged-neutral identi cation

Use single-particle resolutions from tracker charged or cal. neutrals.
10


Key quantities to evaluate jet performance: Jet energy resolution and jet-jet mass. Explicitly omit ISR and beamstrah. Selections: Neutral clusters:
E

0:1 GeV, j cos j 0:9

Tracks: j cos j 0:9 Thrust: j cos j 0:8 For jet energy resolution, use e+e, ! qq: Demand 2 jets; put both in plot

11


LCD SD Detector Ultimate" Jet Energy Resolution, e+e, ! qq
E_jet (GeV)
250
Ej Nent = 1438 Mean = 93.45 RMS = 11.4 Chi2 / ndf = 28.52 / 7 Constant = 203.4 ± 8.942 Mean = 99.01 ± 0.07008 Sigma = 1.892 ± 0.0632

200

150

100

50

0 20

30

40

50

60

70

80

90 100 110 120

Ej =Ej

=0:15=

q

E

j
12


LCD Large Detector Ultimate" Jet Energy Resolution, e+e, ! qq
E_jet (GeV)
180 160 140 120 100 80 60 40 20 0 10 15 20 25 30 35 40 45 50 55 60
Ej Nent = 1196 Mean = 47.4 RMS = 4.679 Chi2 / ndf = 94.54 / 15 Constant = 144.3 ± 7.313 Mean = 49.19 ± 0.05005 Sigma = 1.266 ± 0.04893

Ej =Ej

=0:18=

q

E

j
13


Mass resolution for e+e, ! qq p equivalent to jet is energy resolution SD detector, s = 200 GeV Jet Energy
E_jet (GeV)
250
Ej Nent = 1384 Mean = 95.4 RMS = 10.3 Chi2 / ndf = 19.55 / 10 Constant = 233.3 ± 9.658 Mean = 99.23 ± 0.05947 Sigma = 1.857 ± 0.05596

200

150

100

50

0 20

30

40

50

60

70

80

90 100 110 120

Jet-Jet Mass
JJ Mass (GeV)

70 60 50 40 30 20 10

JJM Nent = 692 Mean = 191 RMS = 14.82 Chi2 / ndf = 12.55 / 10 Constant = 70.09 ± 4.324 Mean = 198.4 ± 0.1926 Sigma = 2.578 ± 0.151

0 120 130 140 150 160 170 180 190 200 210 220

Ej =Ej

=M

jj

=Mjj
14


Compare this idealized EFlow with perfect calorimeteronly jet reconstruction perfect "by hand" compensaps = 200 GeV tion, SD detector, EFlow:
JJ Mass (GeV)

70 60 50 40 30 20 10

JJM Nent = 692 Mean = 191 RMS = 14.82 Chi2 / ndf = 12.55 / 10 Constant = 70.09 ± 4.324 Mean = 198.4 ± 0.1926 Sigma = 2.578 ± 0.151

0 120 130 140 150 160 170 180 190 200 210 220

Calorimeter Only:
JJ Mass (GeV)

25

JJM Nent = 525 Mean = 187.3 RMS = 15.39 Chi2 / ndf = 40.96 / 41 Constant = 17.89 ± 1.196 Mean = 192.4 ± 0.5162 Sigma = 9.246 ± 0.4695

20

15

10

5

0 120

140

160

180

200

220

240

15


For Snowmass studies, jet-jet mass resolution for a many-jet process might be more relevant Expect resolution to be a ected by QCD": overlapping jets, imperfect assignments, radiation, etc. Chose e+e, !
ZZ

! hadrons:

Demand 4 jets Tag one 2-jet comb. MZ ; plot the other mass pair Have not tried anything to optimize the jet reconstruction or aleviate the multi-jet confusion

16


ps = 350 GeV
JJ Mass (GeV)

LCD SD Detector Ultimate" Jet-Jet Mass Resolution e+e, ! ZZ ! 4 jets
JJMt Nent = 814 Mean = 89.98 RMS = 32.75 Chi2 / ndf = 79.98 / 45 p0 = 49.09 ± 3.851 p1 = 86.68 ± 0.6064 p2 = 5.565 ± 0.499 p3 = 10.72 ± 1.708 p4 = 69.63 ± 2.254 p5 = 18.75 ± 1.776

60 50 40 30 20 10 0 0

20

40

60

80 100 120 140 160 180 200

Mjj =Mjj

=0:72=

q

E

Z
17


ps = 350 GeV
JJ Mass (GeV)

LCD Large Detector Ultimate" Jet-Jet Mass Resolution e+e, ! ZZ ! 4 jets
JJMt Nent = 813 Mean = 90.03 RMS = 31.92 Chi2 / ndf = 76.32 / 44 p0 = 50.42 ± 3.882 p1 = 86.66 ± 0.5646 p2 = 5.751 ± 0.489 p3 = 9.776 ± 1.931 p4 = 73.61 ± 2.058 p5 = 19.8 ± 2.321

60 50 40 30 20 10 0 0

20

40

60

80 100 120 140 160 180 200

Mjj =Mjj

=0:64=

q

E

Z
18


Last Comment EFlow, realized in a dense, highly-segmented calorimeter, appears likely to achieve substantial improvement in jet energy resolution Downside: Cost and complexity e.g. 100's of m2 of silicon What is the improved resolution good for ? A. known" cases: ZH vs ZZ vs WW H!WW vs H!ZZ HHZ: Higgs self-coupling top and WW reconstruction: angular dists.! anomalous couplings WW scatering at high energy SUSY decay chain reconstruction, etc.
19


B. The Unknown ! In general, a major advantage of e+e, is access to all nal states. Reconstruction of hadronic nal states can be di cult at hadron colliders. Complementarity suggests we pursue such modes with vigor !

20