<%BANNER%>

Standard Model Higgs Boson Discovery Potential in the Decay Channel H - > ZZ(*) - > 4 mu with the CMS Detector

Permanent Link: http://ufdc.ufl.edu/UFE0021068/00001

Material Information

Title: Standard Model Higgs Boson Discovery Potential in the Decay Channel H - > ZZ(*) - > 4 mu with the CMS Detector
Physical Description: 1 online resource (140 p.)
Language: english
Creator: Drozdetski, Alexei Alexandrovic
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: cms, csc, detector, higgs, lhc, model, muon, standard, strip
Physics -- Dissertations, Academic -- UF
Genre: Physics thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Compact Muon Solenoid (CMS) is a general purpose detector at the Large Hadron Collider (LHC) currently under construction at CERN with start-up date in 2008. The putative Higgs boson is the most plausible solution for the mystery of mass in the Standard Model of elementary particles and its discovery is one of the prime goals for the LHC. Higgs boson search. Our research presents an optimized strategy for a Higgs boson search in its four-muon decay channel, H- > ZZ*- > 4mu, also known as a golden decay channel. The method automatically ensures an optimal signal-to-background ratio for any mass, at which the Higgs boson might appear. The most important theoretical and instrumental systematic errors are taken into account and our search was conducted in a broad range of possible Higgs boson masses. Muon reconstruction. We developed an algorithm for a fast and efficient muon track segment reconstruction in Cathode Strip Chambers. Designed to be CPU-efficient, the algorithm is specifically targeted for High Level Trigger purposes. The segment-finding efficiency and the spatial resolution attainable with the proposed algorithm and the required CPU time were validated using the cosmic ray data taken by the CMS in 2006. The results of validation showed the efficiency and spatial resolution attainable with the algorithm are well within the High Level Trigger requirements and the algorithm's timing performance is by far superior to all algorithms previously used in CMS.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Alexei Alexandrovic Drozdetski.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Mitselmakher, Gena.
Local: Co-adviser: Korytov, Andrey.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021068:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021068/00001

Material Information

Title: Standard Model Higgs Boson Discovery Potential in the Decay Channel H - > ZZ(*) - > 4 mu with the CMS Detector
Physical Description: 1 online resource (140 p.)
Language: english
Creator: Drozdetski, Alexei Alexandrovic
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: cms, csc, detector, higgs, lhc, model, muon, standard, strip
Physics -- Dissertations, Academic -- UF
Genre: Physics thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Compact Muon Solenoid (CMS) is a general purpose detector at the Large Hadron Collider (LHC) currently under construction at CERN with start-up date in 2008. The putative Higgs boson is the most plausible solution for the mystery of mass in the Standard Model of elementary particles and its discovery is one of the prime goals for the LHC. Higgs boson search. Our research presents an optimized strategy for a Higgs boson search in its four-muon decay channel, H- > ZZ*- > 4mu, also known as a golden decay channel. The method automatically ensures an optimal signal-to-background ratio for any mass, at which the Higgs boson might appear. The most important theoretical and instrumental systematic errors are taken into account and our search was conducted in a broad range of possible Higgs boson masses. Muon reconstruction. We developed an algorithm for a fast and efficient muon track segment reconstruction in Cathode Strip Chambers. Designed to be CPU-efficient, the algorithm is specifically targeted for High Level Trigger purposes. The segment-finding efficiency and the spatial resolution attainable with the proposed algorithm and the required CPU time were validated using the cosmic ray data taken by the CMS in 2006. The results of validation showed the efficiency and spatial resolution attainable with the algorithm are well within the High Level Trigger requirements and the algorithm's timing performance is by far superior to all algorithms previously used in CMS.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Alexei Alexandrovic Drozdetski.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Mitselmakher, Gena.
Local: Co-adviser: Korytov, Andrey.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021068:00001


This item has the following downloads:


Full Text





THE STANDARD MODEL HIGGS BOSON DISCOVERY POTENTIAL IN THE
DECAY CHANNEL H -+ ZZ* -+ 4/1 WITH THE C'\!iS DETECTOR



















By

ALEXEY ALEXANDROVICH DROZDETSKIY


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007

































2007Alexey Alexandrovich Drozdetskiy





























Happiness in my life is to do whatever I do in the best way I can for others,

and especially for one -for Maria.









ACKNOWLEDGMENTS

Friendship is unnecessary, like philosophy, like
art... It has no survival value; rather it is one
of those things that give value to survival.
-C. S. Lewis

This is probably one of the hardest parts of the dissertation. And not because I have

hard time in remembering all the people with whom I have worked, who helped me and

supported me, but because no -, i-ill!-." limits for dissertation size would ever be able to

accommodate the list of all of them. It's so for a very simple reason: for example many

people from the University were working with me, we used something produced before by

others, our families and close to us people were loving and patient enough to allow us to

quench our thirst in research, etc... and we can continue to expand the list this way on

and on. Some -w it takes just a few people to connect any one of us with any one else

through common work, friends, etc. So it's clear what an impossible task it is to list on a

few pages all the people.

And I would like to thank all, who helped, supported and worked with me, and

influenced me knowingly or unknowingly: I remember all in my mind and my heart. And

if paper is not large enough, they are. I'm grateful to all, including University of Florida

(UFL, Gainesville, USA) professors, post docs, students, and not less all the wonderful

office managers and personal; members of the C'\ S collaboration (for countless work hours

and out-of-work meetings and discussions, help and support); people from Novosibirsk

State University (Russia) and Budker Institute (Novosibirsk, Russia); all teachers and

friends in my life.

Still, there are a few people I would like to thank a little more than others: Guenakh

Mitselmakher and Andrey Korytov my supervisors, for all their experience and

knowledge they passed onto me, both on physics and not less importantly on life in

general. I thank them for ahv--i-i giving me the freedom to get my own grasp on a topic.

And maybe even more importantly I would like to thank them for their constant kind and









friendly attitude I'm enjo, 1 for many years. These are wonderful people and I'm very

happy to learned from and to have been with them.

Others from UFL HEP group with whom I was happy to work with and spend

free time together: Darin Acosta (who also was my co-supervisor for some time), Rick

Cavanaugh, Victor Barashko, Bobb11 Scurlock, K,.- i Kotov, Yura Pakhotin, Sergo

Jindariani, Yura Oksuzian and Sasha Sherstnev (and their families).

Some C'\ S collaboration colleagues inspired me: Paris Sphicas, Daniel Denegri,

Salavat Abdullin, Sasha Nikitenko, and Bob Cousins... Thank you.

My "out-of-work-group" friends, some of whom I have known since I was 7 years old:

Igor, Eldar, Olga, Katerina, and Dasha (and their families).

I can expand this list and continue forever...

I would like to return to the thought I wrote in the beginning: all people in my life

are very important and I thank each of them from the bottom of my heart and will keep

them there for alv--,.









TABLE OF CONTENTS


page


ACKNOW LEDGMENTS ................................. 4

LIST OF TABLES ....................... ............. 9

LIST OF FIGURES ....................... ........... 10

ABSTRACT ................................. ...... 16

CHAPTER

1 INTRODUCTION ...................... .......... 18

2 THE STANDARD MODEL HIGGS BOSON ......... ........ ... 20

2.1 The Standard Model: Electroweak Symmetry Breaking. ............ 20
2.2 The Standard Model Higgs Boson Mass: Current Experimental and Theoretical
Constraints . . . . . . . .. 22


2.2.1 Theoretical Constraints .......
2.2.2 Searches at Large Electron-Positron
2.2.3 Searches at Tevatron .. ......
2.2.4 Electroweak Precision Measurements
2.3 The Standard Model Higgs Boson at LHC
2.3.1 Production .. ............
2.3.2 D ecays . . . .
2.4 The Standard Model Higgs Boson Searches
R each . . . . .
2.5 Issues in the Standard Model .. .....
2.6 Supersymmetry Higgs .. .........


Collider






at C'\!S


(LEP)


Summary:


Discovery


3 LARGE HADRON COLLIDER ............

4 COMPACT MUON SOLENOID (C'\!S) DETECTOR

4.1 Introduction . . . . . .
4.2 General Description .................
4.3 The Superconducting Magnet ...........
4.4 Inner Tracking System ...............
4.5 Electromagnetic Calorimeter (ECAL) .......
4.6 Hadron Calorimeter (HCAL) ............
4.7 M uon System .. ................
4.7.1 Barrel Muon System .. ..........
4.7.2 Endcap Muon System ............
4.7.3 Resistive Plate C(! iinlers ..........
4.8 M uon Ti1 r- . . .... . .









5 CATHODE STRIP CHAMBERS (CSC) ....

5.1 Introduction .................
5.2 C! I ,inl er Mechanical Design ........
5.3 Electronic Design ..............

6 FAST ALGORITHM FOR TRACK SEGMENT
IN CATHODE STRIP CHAMBERS ......


AND


HIT


RECONSTRUCTION


6.1 Introduction . . . . . . . . .
6.2 Algorithm Principles .. ... .. .. .. ... .. .. .. .. ... .. .. .
6.2.1 ProtoSegment Building .........................
6.2.2 RecHits and Their Coordinates ....................
6.2.3 RefinedSegm ent .. ........................
6.3 Algorithm Validation at MTCC .. ....................
6.3.1 ProtoSegm ents . . . . . . . .
6.3.2 Spatial Resolution . . . . . . .
6.3.2.1 Resolution for high-pr muons .. .............
6.3.2.2 Resolution for high-pr muons with tabulated pedestals .
6.3.2.3 Resolution for low-pr, or highly-inclined, muons ......
6.3.2.4 Resolution for high-pr muons with charge-dependent sigmas
6.3.3 RefinedSegm ents .. .......................
6.4 C conclusions . . . . . . . . .
6.5 R atio M ethod . . . . . . . . .

7 SEARCH STRATEGY FOR THE STANDARD MODEL HIGGS BOSON IN
THE H -- ZZ(*) -+ 4/ DECAY CHANNEL USING M(4p)-DEPENDENT CUTS.

7.1 Introduction . . . . . . . . .
7.2 Physics Processes and Their Simulation .. .................
7.2.1 Signal: H ZZ(*) 4i .............
7.2.2 Background: tt .... . . ...............
7.2.3 Background: (Z/7*)bb 2pbb .. .................
7.2.4 Backgrounds: qq i Z/7*Z/7* z 4p and qq Z/7*Z/7* 2p27 .
7.3 Ti,--,- i and Offline Muon Reconstruction .. ................
7.4 Higgs Search Strategy .. . .......................
7.4.1 Introductory Remarks on Significance .. ..............
7.4.2 Optimization of the M(4p)-Dependent Cuts .. ..........
7.4.3 System atic Errors . . . . . . .
7.4.3.1 Uncertainties in the background .. ............
7.4.3.2 Significance with the background uncertainties included .
7.4.4 Local Significance and Overall Statistical Fluctuation Probability .
7.5 Sum m ary . . . . . . . . .
7.6 On the true significance of a local excess of events .. ............









8 CONCLUSIONS ................... ............... 122

APPENDIX

A RELATIVE CONTRIBUTIONS OF T- AND S-CHANNELS TO THE ZZ(*)
4/1 PROCESS ................... ......... ...... 124

B ZZ DISCOVERY WITH < 1FB-1 WITH CM\S IN ZZ(*) 4/ MODE ..... 126

C MEASURING MUON RECONSTRUCTION EFFICIENCY FROM DATA ... 127

D SENSITIVITY OF THE MUON ISOLATION CUT EFFICIENCY TO THE
UNDERLYING EVENT UNCERTAINTY ......... ........ .... 129

E GENETIC ALGORITHM FOR RECTANGULAR CUTS OPTIMIZATION 130

REFERENCES ....................................... 131

BIOGRAPHICAL SKETCH ................... .......... 139









LIST OF TABLES
Table page

2-1 The Standard Model elementary particles. .................. .... 20

7-1 The LO/NLO cross sections for various Higgs boson masses and backgrounds,
corresponding number of events with four muons surviving the generator level
preselection cuts (see section 7.2) calculated for L = 30fb-1, and the number
of simulated events ............... ........... .. .. 91

7-2 Summary of the results: number of signal and background events in a window
used for a counting experiment with the M(4p)-dependent cuts. Systematic
error on the background is normalized to the Z -- 2p process (6KNLo/KNLO
is not included); three different significance without systematic errors included:
the SL estimator for the Log Likelihood Ratio (LLR) built for the full M(4p)
spectrum, ScL LLR estimator built for a counting experiment approach, and
the S, true significance for the counting experiment approach; the final result
for S,, now including all systematic errors. .................. .. 108









LIST OF FIGURES
Figure page

2-1 Divergent (top three) WW cross-section graphs and their cancellation (bottom
two graphs) .................... ................... .. 22

2-2 Theoretical bounds on the SM Higgs mass as a function of the cut-off scale. .22

2-3 Expected and observed 95'. CL cross section ratios for the combined CDF and
DO an lv-- and the expected 95'. CL ratios for the CDF and DO experiments
alone. ...... .... .... ...................... 24

2-4 The Ax2 curve derived from high-Q2 precision electroweak measurements. 24

2-5 Typical diagrams for all relevant Higgs boson production mechanisms at leading
order at the LHC. ................... .............. 26

2-6 Higgs production cross sections at the LHC for the various production mechanisms
as a function of the Higgs mass. .................. ..... 26

2-7 Total decay width and branching ratios of the dominant decay modes of the
SM Higgs boson particle. .................. ........... 27

2-8 The integrated luminosity needed for the 5 discovery of the inclusive Higgs
boson production. .................. ............. ..29

2-9 The signal significance as a function of the Higgs boson mass for 30 fb-1 of the
integrated luminosity for the different Higgs boson production and decay channels. 29

3-1 Accelerator complex at CERN. .................. ...... 33

4-1 An exploded view of the C'\ S detector. ................... .... .37

5-1 Quater-view of the C'\ I Detectors. Cathode Strip C!('i ,i ers of the Endcap
Muon System are highlighted ................... . .. 46

5-2 The ME2 cathode strip chambers. The outer ring consists of 36 ME2/2 chambers,
each spanning 100 in Q; while 18 20-degree ME2/1 chambers form the inner ring.
The chambers overlap to provide contiguous coverage in . . .... 47

5-3 A schematic view of a C'\ S cathode strip chamber made of seven trapezoidal
panels ...................... ........ ... .... ... 48

5-4 A schematic view of a single gap illustrating the principle of a CSC operation. 48

5-5 Mechanical design of the C'\!S Cathode Strip Chambers (exploded view). ... 49

5-6 CSC gas gain vs high voltage. .................. ..... 51

5-7 ME2/1 chamber singles rate vs high voltage (the overall sensitive area of all six
planes in this chamber is ~9.5 m2). .................. .... .51









5-8 Signal in ME1/1 chambers for not tilted and tilted wires. ........... ..53

5-9 An overview of all custom-made CSC trigger and readout electronics. Functionality
of various boards is described in the text. ................ ..... 53

5-10 Muon signals as seen at the AFEB amplifier output. ............. ..55

5-11 A schematic event di-pli showing anode signals in a six-plane chamber. . 55

5-12 Basic functional diagram of a Cathode Front-End Board. ........... ..56

5-13 Muon hit signals from six nearby strips. Four curves are actual oscillograms. .. 56

5-14 A simplified schematic illustrating the idea behind the comparator network. .57

5-15 CLCT (Cathode Local C!i urged Track) is a pattern of half-strip hits consistent
with a muon track. .................. ... ......... 57

6-1 The noise distribution for a three-sample-sum in the absence of a signal for the
largest ME23/2 chambers. .................. ......... 69

6-2 The three-strip cluster charge distribution for the largest ME23/2 chambers. 69

6-3 The noise distribution for a single-sample in the absence of a signal for the largest
M E23/2 chambers. .................. ... ......... 70

6-4 The noise distribution for a two-sample sum in the absence of a signal for the
largest ME23/2 chambers. .................. .. ....... 70

6-5 The 3rd-plane residuals for the five distinct Ix strip part ranges for the middle
part of the largest ME23/2 chambers. And a summary for a single-plane spatial
resolution as a function of a hit position across a strip. ............ ..73

6-6 The 3rd-plane residuals a's vs. five xl strip part ranges and overall six-plane
resolution vs. Ixl-part of a strip (in units of strip widths and microns). [Notations
are: black empty squares HV segment 1; red squares HV segment 2; triangles
up HV segment 3; triangles down HV segment 4; stars HV segment 5.] 74

6-7 The 3rd-plane residuals a's vs. |x/wl-part of a strip and overall six-plane resolution
vs. Ix/wl-part of a strip (in units of strip widths and microns). Same notations
as for Fig. 6-6 ............... ............. .. .. 75

6-8 The overall six-plane chamber resolution vs. five xl strip part ranges for dynamically
measured pedestals. Same notations as for Fig. 6-6. .............. ..77

6-9 The overall six-plane chamber resolution vs. five xl strip part ranges for calibrated
pedestals. Same notations as for Fig. 6-6. .................. ..... 77

6-10 Inclination angles for track segments reconstructed from di-strip CLCTs (\ Il. /1
chambers) . .................................... .. 77









6-11 The spatial resolution per six-plane chamber for tracks reconstructed from di-strip
CLCTs (M'I. 1/1 chambers) for five Ix strip part ranges. Same notations as for
Fig. 6-6. . . . . . .. . ... .. .. 77

6-12 Probabilities for different numbers of charge clusters found in association with
ME23/2 chamber ProtoSegments. ............... ...... 80

6-13 L -r charge clusters occupancy in association with ME23/2 chambers ProtoSegments. 80

6-14 The X2/dof distributions for a linear fit for ME23/2 chamber ProtoSegments
with six clusters, five clusters, and four clusters. ................. 81

6-15 Scatter plot of min(xj/dof) vs. x2/dof for six-cluster ME23/2 chamber ProtoSegments. 81

6-16 Occupancy for |x/w -coordinates of RefinedSegments. ............. ..82

6-17 Pools for all retained RecHits (i.e., RecHits associated with RefinedSegments.) .82

6-18 Induced charge distribution calculated according to Gatti for large and ME1/1
chamber geometries. .................. ... ......... 82

6-19 Ratio r versus a local coordinate x for large chambers calculated for variety of
strip widths in the assumption of the Gatti charge distribution for large chambers. 82

6-20 Correction that must be added to the measured ratio r to obtain the hit position
across a strip .................. ................. .. 84

6-21 An occupancy distribution for the 1st-order corrected coordinate xst. . 84

6-22 The second-order correction that must be added to xl to obtain the hit position
across a strip x2 .................. ................ .. 85

6-23 An occupancy distribution for the 2nd-order corrected coordinate 2. . 85

6-24 Sensitivity of the ratio method to electronic noise (in of strip width and mm). 86

6-25 Sensitivity of the ratio method to errors in electronic gain calibrations (in of
strip width and mm). ............... .......... .. .. 86

6-26 Sensitivity of the ratio method to uncertainties in cross talks between strips (in
of strip width and mm). ............... ...... .... 87

7-1 M(4p) distributions after generator-level cuts for tt, (Z/7*)bb, Z/7*Z/7*, and
MH 140 GeV/c2 (log scale). ............... ..... .... 110

7-2 Same as Figure 7-1, but on a linear scale. ................ ...... 110

7-3 Enhancement to the signal samples' cross sections due to interference effects
not accounted for at the generator level. ................ ...... 110









7-4 The MA(4p)-dependent NLO K-factor KNLO(M4,) for the ZZ -+ 4/1 process
evaluated with MCFM [98]. ............... ......... 110

7-5 Global muon reconstruction efficiency calculated from matching reconstructed
and true Monte Carlo muons in the barrel region vs. pr. ........... ..111

7-6 Global muon reconstruction efficiency calculated from matching reconstructed
and true Monte Carlo muons in the endcap region vs. momentum . .... 111

7-7 M(4p) distribution for m(H)=150 GeV/c2 and the fit described in the text. ... 111

7-8 M(4p) resolution vs. MH. ................. ........ 111

7-9 Comparison of different significance estimators for 1 background event: probability
of measuring significance S > So, background only case, NB = 1 event. ..... ..112

7-10 Same as Figure 7-9, but for NB = 10 events. .................. 112

7-11 Dependence of the tracker-based muon isolation cut on the least isolated muon
versus Higgs mass. .................. ............... .. 112

7-12 Dependence of the calorimeter-based cut on the least isolated muon versus Higgs
m ass. . . .. . . . . . . 112

7-13 Dependence of the pr cut on the second-lowest-pr muon versus Higgs mass. 113

7-14 Dependence of the M(4p) window cuts versus Higgs mass. . .... 113

7-15 First muon pair invariant mass distribution, M(Z1), after analysis cuts were
applied . ................... ............ ... . 113

7-16 Second muon pair invariant mass distribution, M(Z2), after analysis cuts were
applied. ... .. .. .. ... .. .... ...... .. .. .. ... .. .. .. 113

7-17 Lowest muon pr distribution, after analysis cuts were applied. . ... 114

7-18 Maximum distance in XY-plane between muon impact point coordinates distribution,
after analysis cuts were applied. .................. ...... 114

7-19 ScL vs. Higgs boson mass. .................. ........ 114

7-20 M(4p) invariant mass distribution for the three background subprocesses and a
Higgs boson signal at MH 150 GeV/c2, after applying cuts on muon isolation
and PT .................. .................. ... 114

7-21 Expected excess significance ScL with L = 30 fb-1 for different Higgs boson
masses for MA(4p)-dependent and independent cuts. No systematic errors included. 115

7-22 Expected excess significance SL with L = 30 fb-1 for different Higgs boson
masses for Mf(4p)-dependent and independent cuts. No systematic errors included. 115









7-23 Luminosity required to reach a 5a event excess for different Higgs boson masses
for M(4p)-dependent and independent cuts. No systematic errors included. 115

7-24 The 95'. CL exclusion contours for the SM Higgs hypothesis. . ... 115

7-25 Combined systematic error on the number of background events due to PDF
and QCD scale uncertainties for the c(qq -+ ZZ -+ 4p) process at NLO. ..... ..116

7-26 Top: the factors KNLO(M4p) in MCFM and EffNLO calculations versus M(4p);
bottom: the difference between them. ............... .... 116

7-27 Muon isolation cut efficiency for random cone direction for Z-inclusive (dashed
lines) and for ZZ (solid lines) events. .................. ..... 116

7-28 An example of a possible bias in evaluating the significance of an event excess
due to a non-optimal choice of the signal window width. ........... .116

7-29 Uncertainties in the number of ZZ -+ 4/z background events in the signal region
window at different M(4p). The event count is referenced to the number of Z --
2p1 events . . . . . . . . . 117

7-30 Uncertainties in the number of ZZ -+ 4/z background events in the signal region
window at different M(44p). The event count in signal region, is calculated from
the number of ZZ -- 4p events in the range 100-700 GeV/c2 (excluding the
signal region window). ............... ............ .. 117

7-31 Significance vs. Higgs mass (with and without dF/K contribution). ...... .118

7-32 Integrated luminosity needed for 95'. CL exclusion, 3j, and 5a discovery versus
Higgs boson mass (with and without dF/K contribution). . . 118

7-33 Integrated luminosity needed for a 5a discovery of the Higgs boson versus its
mass for (with and without dF/K contribution). ................. 119

7-34 The background pdf and an example of one pseudo-experiment with a statistical
fluctuation appearing just like a signal. .................. .... 121

7-35 Profile of the ScL scan corresponding to the pseudo-experiment example shown
on the left. Green (inner) and yellow (outer) bands denote lc and 2 a intervals.
Spikes that can be seen are due to events coming in or dropping off the trial-window,
a feature of low-statistics searches. ................ .. .... 121

7-36 ScL cumulative probability density function. .................. 121

7-37 Local significance i. i~i!, I !- II i," from an observed value to the true significance
with a proper probabilistic interpretation. .................. .... 121

A-i ZZ background: t- and s- channel diagrams and s-channel contribution peak
around Zo mass after pre-selection cuts for fully simulated events. . ... 125









C-i Z -- 2/ invariant mass peak: built from the HLT muon and all other tracks,
the HLT muon and all other standalone muons and the HLT muon and all other
globally reconstructed muons. .................. ......... 128









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

THE STANDARD MODEL HIGGS BOSON DISCOVERY POTENTIAL IN THE
DECAY CHANNEL H -+ ZZ* -+ 4/p WITH THE C'\! S DETECTOR

By

Alexey Alexandrovich Drozdetskiy

August 2007

C('! i': Guenakh Mitselmakher
CoC'l! ,i: Andrey Korytov
Major: Physics

The Compact Muon Solenoid (C'\!S) is a general purpose detector at the Large

Hadron Collider (LHC) currently under construction at CERN with start-up date in 2008.

The putative Higgs boson is the most plausible solution for the n1'-- I. i- of mass in the

Standard Model of elementary particles and its discovery is one of the prime goals for the

LHC.

Higgs boson search. Our research presents an optimized strategy for a Higgs boson

search in its four-muon decay channel, H -+ ZZ(*) 4/, also known as a golden decay

channel. The method automatically ensures an optimal signal-to-background ratio for

any mass, at which the Higgs boson might appear. The most important theoretical and

instrumental systematic errors are taken into account and our search was conducted in a

broad range of possible Higgs boson masses.

Muon reconstruction. We developed an algorithm for a fast and efficient muon track

segment reconstruction in Cathode Strip C(i ihl'ers. Designed to be CPU-efficient, the

algorithm is specifically targeted for High Level Ti i- -. i purposes. The segment-finding

efficiency and the spatial resolution attainable with the proposed algorithm and the

required CPU time were validated using the cosmic ray data taken by the C 'IS in 2006.

The results of validation showed the efficiency and spatial resolution attainable with the









algorithm are well within the High Level T i.-.-- r requirements and the algorithm's timing

performance is by far superior to all algorithms previously used in C'\ S.









CHAPTER 1
INTRODUCTION

If we knew what it was we were doing,
it would not be called research, would it?
-Albert Einstein

The Standard Model (SM) as a fundamental theory of elementary particles and their

interactions has been so far very successful. But a very important question of this theory

is still unanswered: does the Higgs boson exist? Our dissertation describes work done by

the author, as a member of the Compact Muon Solenoid (C' \S) collaboration.

One of the main goals of C'\ S is to discover Higgs boson. For this we will need good

understanding of our detector and data, and we need to prepare data analysis tools to be

ready to discover new physics we are looking for. The dissertation describes author's work

in both directions.

First, it gives details on commissioning of Cathode Strip C!i i'hlers (CSCs), C'i\S

endcap muon detectors, through analyzing data and validating new muon track hits and

segments reconstruction algorithm, which would be suitable for High Level T i:---. r (HLT).

Note that default algorithm used for this at the time of our analysis were just too slow for

HLT.

Another 1i i .ri part of the dissertation -is our C' \! the SM Higgs boson discovery

potential in H -+ ZZ(*) -- 4 decay channel with M(4p) dependent cuts analysis using

Monte Carlo (M\C) data. Here we describe our analysis, which is much more realistic than

previously existed ones on the topic:


we used calibration from data techniques, as we would do with real data;


full treatment of systematic errors was included and folded into signal-to-background
significance calculations;


dedicated effort was done to generate physics of signal and background processes
properly, including use of dedicated matrix element generators, events re-weighting
with dynamic Next-to-Leading Order (NLO) corrections;









* latest available at the moment full simulation and reconstruction validated software
was used;


* C'\ IS performance was optimized for all allowed SM Higgs boson masses;


* and as a result our analysis is part of the C'\ S Physics Technical Design Report, i.e.
is an official C'\IS strategy for the SM Higgs boson discovery in H -+ ZZ(*) -- 4/
decay channel.









CHAPTER 2
THE STANDARD MODEL HIGGS BOSON

2.1 The Standard Model: Electroweak Symmetry Breaking.

The best known description of the fundamental particles and their interactions1 is

provided by a quantum field theory known as the Standard Model (SM). We will mention

here some of the most important features of the model, not attempting by any means to

compete with many already existing good descriptions of it.

The theory includes three generations of fermions (quarks and leptons, spin

1/2-particles) as well as interaction mediators (spin-1 bosons: 7, Z0, W' and gluons -

mediators of electromagnetic, weak and strong forces) (Tab. 2-1).

Table 2-1: The Standard Model elementary particles.

Particles Spin C!i irge
Fermions Leptons (t) 1/2 (1)
Quarks (u) 1/2 (2)
d -1/3)
Gauge 7 1 0
Bosons ZO, W 1 0, 1
8 gluons 1 0
[Higgs] [Ho] [0] [0]


The description of the SM interactions is implemented by a gauge theory based on

SU(3)c x SU(2)L x U(i)y symmetry. Gauge symmetry provides bosons, which serve

as interaction mediators. Local gauge invariance makes the theory renormalizible and

requires the gauge bosons to be massless. At the same time we know that Z0, W' bosons

have masses.

The solution of the problem in the SM is the Higgs mechanism, which spontaneously

breaks the gauge symmetry. Corresponding scalar potential added to the Lagrangian

generates the vector boson and fermion masses in a gauge invariant way. This remnant



1 gravitational effects not included









scalar field, the Higgs boson, is a part of the physical spectrum. This is the only missing

piece of the SM that is still not confirmed experimentally.

In the 1960s, the theory of Quantum Electrodynamics (QED) was already very

successful, and the theoretical community evolved Fermi's theory of weak interactions

into a Yang-Mills theory [22] based on the symmetry group SU(2)L. The problem with

this approach was that gauge invariance does not allow masses for both the gauge bosons

and the leptons. The observation of Peter Higgs was that the gauge invariance could be

spontaneously broken with the addition of a doublet of complex scalar fields, p, with

Lagrangian


LHiggs (-a,)t(a9) V(Q) (2-1)

where the potential


V()) _-2t + A(0t0)2 (2-2)

is the key to spontaneous symmetry breaking [23, 24]. Having this mechanism for

electroweak symmetry Ioi, i1;iii Glashow [25], Weinberg [26], and Salam [27] proposed a

unified electroweak theory of the leptons. This theory has still a massless photon, allows

for massive W bosons and leptons, predicts the massive, neutral, spin-1 Z boson, and

predicts the massive, neutral scalar Higgs boson. The W' and Z bosons were confirmed at

the CERN SPS by the UA1 and UA2 collaborations [28, 29].

The Higgs boson itself acquires mass through self-coupling in the Higgs potential

V(Q) (Eq. 2-2). The Higgs mass at the tree level is


H = 2Av2 (2-3)

where A is an independent and unknown parameter. mH can not be predicted by the

SM.









At present the SM gives an excellent description of nature. It is a renormalizable

quantum field gauge theory with massive fermions and vector gauge bosons. The only

unknown parameter of the SM is the mass of the Higgs boson.

2.2 The Standard Model Higgs Boson Mass: Current Experimental and
Theoretical Constraints

2.2.1 Theoretical Constraints

Besides allowing the introduction of Z/W-boson masses without breaking renormalizability

and quark/lepton masses within the weak sector of SM with broken P-symmetry, the

introduction of a new scalar particle was motivated by divergences in the scattering of a

longitudinally polarized W bosons in the high energy limit [30, 31]. Without an additional

interaction, the cross-section of that process (Fig. 2-1), would diverge and would violate

unitarity bounds above /s = 1.2 TeV. An interaction with the Higgs boson cancels

those divergences. This mechanism would work only if the Higgs boson is not too heavy,

otherwise, the Higgs boson would not contribute enough to the scattering amplitude

before unitarity is violated. Therefore, the Higgs boson mass should be less than 850 GeV

[31].

800 1


wL wT LwL w w wL 600 mt = 175 GeV
S zV .I / a.(Mz) = 0.118
a 400
not allowed
w1 i w- w W
WjL /VLb ./L W[L 200 allowed

wnot allowed

103 106 10 109 1012 10
Figure 2-1: Divergent (top three) A [GeV]
WW cross-section graphs and their
cancellation (bottom two graphs). Figure 2-2: Theoretical bounds on
the SM Higgs mass as a function of
the cut-off scale.









Stronger bounds on the Higgs boson mass can be derived from the energy scale up to

which the SM should be valid without the necessity of introducing new physics [30, 31].

If the Higgs boson is light, Higgs self-coupling strength A is small and dominant loop

contribution to the Higgs potential comes from top loops (At is large due to the large

mass of the top quark). If the SM is valid as an effective theory up to the scale A, then

these loop contributions have to be summed up to this scale. For the Higgs mechanism

to remain valid, the coupling A must remain positive, otherwise no minimum exists in the

potential and no stable spontaneous symmetry breaking occurs. This places a lower bound

on A, hence on mH, depending on cut-off scale A (Fig. 2-2).

The energy dependence A(Q2) can be derived from the renormalization group

equations [30, 31]. If A is large, the Higgs loop dominates over the top loop. Neglecting

the graph with a top-quark loop, we can write A as


A(Q2) A (24)
1 In(Q2/v2)

For a heavy Higgs boson, A could grow to infinity (Landau-pole). Requiring that

the self-coupling A remains finite for arbitrary values of Q implies A(v) = 0. Since

A(v2) m2/2v2, this would result in the non-interacting trivial theory. If, instead, A

is required to be finite only up to a scale ANp, where the new physics enters, the mass

bound can be written as


2 _87r2V2
M2o < 2- (2-5)
S31nANP/v2
One can read that if there is no new physics up to the Planck scale (A 1019 GeV),

iH must be ~ 160 180 GeV [32] (Fig. 2-2).

2.2.2 Searches at Large Electron-Positron Collider (LEP)

Direct searches for the Higgs boson have been conducted at the LEP experiments

at CERN. No evidence for a signal was observed in a data from e+e- collisions up to the









center of mass energies of 209 GeV. An experimental lower bound is set to mH > 114.4

GeV at the 95'. confidence level [33].

2.2.3 Searches at Tevatron

The search for the SM Higgs particle is continuing at the Tevatron for Higgs masses

up to ~ 130GeV/c2.

The most recent update was presented e.g. at the Moriond conference in March,

2007 and shows joint CDF/DO sensitivity curves together with the SM prediction. The

Tevatron experiments are close to the SM predictions and may add to competition during

early LHC data taking (Fig.2-3 [34]).


5

4


3

2


Figure 2-3: Expected and observed
95' CL cross section ratios for the
combined CDF and DO analyses,
and the expected 95'. CL ratios for
the CDF and DO experiments alone.


300


mH [GeV]

Figure 2-4: The Ax2 curve derived
from high-Q2 precision electroweak
measurements.


2.2.4 Electroweak Precision Measurements

An indirect measurement of mH within the SM framework is possible using the

precision measurements of the fundamental parameters, e.g., mz Fzo m.w. Such

measurements have been performed by several experiments and a global fit to these









electroweak observables with the Higgs boson mass as a free parameter sets limits on mH

[35, 36].

Figure 2-4 [36] shows the Ax2 curve, derived from high-Q2 precision electroweak

measurements, performed at LEP and by SLD, CDF, and DO, as a function of the Higgs

boson mass, assuming the SM to be the correct theory of nature. The preferred value for

its mass, corresponding to the minimum of the curve, is at 76 GeV, with an experimental

uncertainty of +33 and -24 GeV (at ,-'-. confidence level derived from AX2 = 1 for

the black line, thus not taking the theoretical uncertainty shown as the blue band into

account). This result is only affected a little by the low-Q2 results such as the NuTeV

measurement. (While this is not a proof that the Standard-Model Higgs boson actually

exists, it does serve as a guideline in what mass range to look for it. The precision

electroweak measurements tell us that the mass of the Standard-Model Higgs boson is

lower than about 144 GeV (one-sided 95'. confidence level upper limit derived from

Ax2 =2.7 for the blue band, thus including both the experimental and the theoretical

uncertainty). This limit increases to 182 GeV when including the LEP-2 direct search

limit of 114 GeV shown in yellow.)

2.3 The Standard Model Higgs Boson at LHC

The Higgs coupling depends on the mass of the particle. So, in general Higgs boson

production comes through production of the heavy particles and it decays to heavy

particles as well.

2.3.1 Production

The main production channels for the SM Higgs boson are (diagrams are shown at

Fig. 2-5 (left)):

a gluon fusion via top loop

b vector boson (W/Z) fusion

c W/Z associated production

d tt fusion










The production cross section at LHC as a function of the the Higgs boson mass is

shown in Fig. 2-6.


]tb ---.



(a)


1V IV,, Z'



*3 / ^H


(b)



-H

9 Wb--n-C -t
(d)


Figure 2-5: Typical diagrams for
all relevant Higgs boson production
mechanisms at leading order at the
LHC.


IU -- s -- -- -- -- -- -- ---- --, -r ---
0 200 400 60D 800 100


Figure 2-6: Higgs production
cross sections at the LHC for the
various production mechanisms as a
function of the Higgs mass.


Due to the large size of the top Yukawa couplings and the gluon densities, the gluon

fusion, pp -i gg -i H, is the dominant Higgs boson production mechanism for the whole

Higgs mass range.

In our Higgs boson search analysis, the NLO cross sections and branching ratios for

the Higgs boson calculated with the programs HDECAY [37], HIGLU [38] are used, as

well as the NLO cross sections for the background processes, when available2 QCD NLO

corrections known for long time, they increase Higgs boson production cross section by

~ 50 1ii' NNLO QCD corrections are of the order of 211'

2.3.2 Decays

The Higgs decay modes can be divided into two different mass ranges. For MH <

135GeV/c2 the Higgs boson mainly decays into bb and r-r- pairs with branching ratios of




2 The same is true for results in section 2.4


0pp-H+X) [pb|
s= 14 TeV
g-tSH M=l75 GeV
-- CTHQ6MM


'""^ f^HW ^iN -


ggij-4f i


::::-:-::::









about ,".' and s'. respectively (see Fig. 2-7, right plot). Both bb and r+r- Higgs boson

decay modes are (almost) impossible for discovery at hadron collider due to overwhelming

background level.

The decay modes into cc and gluon pairs, with the latter mediated by top and bottom

quark loops, accumulate a branching ratio of up to about 10' but do not p1 l a relevant

role at the LHC (they are even harder to discover than bb and T+-r modes due to huge

QCD background level).

The most important Higgs decays in this mass range at the LHC is the decay

into photon pairs, which is mediated by W, top and bottom quark loops. It reaches

a branching fraction of up to 2 x 10-3. Its importance is in clean environment and

controllable background, which makes it an important discovery channel for small Higgs

boson masses.



10 2 FHif, a W W ............

10 -
--/,: :
1 1 if


1010

2i I I \
-3" -^------------ ^ -. n"3 -- siL-*' ," \- --
O 10
50 10D I) I,'.(I 50 100 XI 5M 100


Figure 2-7: Total decay width and branching ratios of the dominant decay modes of the
SM Higgs boson particle.


For Higgs masses above 135 GeV/c2 the main decay modes are those into WW and

ZZ pairs, where one of the vector bosons is off-shell below the corresponding kinematical

threshold. These are very important discovery channels, in particular H -- ZZ(*) -- 4/1









is called golden mode" for very clean (4 muons) final state, which provides low/medium

integrated luminosity level for discovery of the Higgs boson for all possible Higgs boson

masses in consideration (~ 115 600GeV/c2).

WW/ZZ decay modes dominate over the decay into tt pairs, the branching ratio of

which does not exceed ~ 2n'-. as can be seen from Fig. 2-7 (right plot).

The total decay width of the Higgs boson, shown in Fig. 2-7 (left plot), does not

exceed about 1 GeV/c2 below the WW threshold. For very large Higgs masses the total

decay width grows up to the order of the Higgs mass itself so that the interpretation of

the Higgs boson as a resonance becomes questionable.

2.4 The Standard Model Higgs Boson Searches at CMS Summary: Discovery
Reach

Figure 2-8 [8] shows the integrated luminosity needed for the 5a discovery of the

inclusive Higgs boson production pp -+ H + X with the Higgs boson decay modes H 77,

H -+ ZZ -+ 4U, and H -+ WW 22v three front runners among all possible decay

channels. H -- 77 dominates discovery reach up to nH ~ 130GeV/c2, similarly it's

H -+ WW -- 22v dominates for ~ 150 180 GeV and H -+ ZZ 4 in the rest of the

possible Higgs boson masses.

Figure 2-9 [8] shows the signal significance as a function of the Higgs boson mass for

30 fb-1 of the integrated luminosity for the different Higgs boson production and decay

channels.

2.5 Issues in the Standard Model

In spite of the experimental success of the SM, it is believed that it does not describe

Nature completely, but it is a low energy effective theory of a more fundamental theory.

Following is an overview of some of the SM obstacles.

There is experimental evidence that the neutrinos are massive, contrary to the

assumptions of the SM [39]. During the past years several neutrino experiments like the

SuperKamiokande, K2K, SNO and Kamland [40, 41, 42, 43] have established the presence










c 1 1 1 1 CMS, 30 fb"

SCMS 1


10 i '
e3: r ,LI --. H-*yy cuts
I -- H-ryy opt
o i A. H-ZZ-41
SI H-*HWW--212v
SI I H-yy cuts i i i --qqH, H-vWW-vjj
H-*yy opt | qqH, H-+tn-'+jet
1 I -ZZ-41 qqH, H-yy
I H-WW- 212v 1
| | ||WW- 00 200 300 400 500 6
100 200 300 400 500 600 M, GeV/c'
MH,GeV/c2
Figure 2-9: The signal significance
Figure 2-8: The integrated as a function of the Higgs boson
luminosity needed for the 5a mass for 30 fb-1 of the integrated
discovery of the inclusive Higgs luminosity for the different Higgs
boson production. boson production and decay
channels.

of neutrino oscillations. This is a sign of the neutrino masses: only massive particles

have a time evolution and therefore can oscillate if mass differences between the various

neutrino mass eigenstates exist.

The second obstacle comes from the creation of the Universe and the cosmological

precision measurements. The lack of the antimatter in the visible Universe and the

measured ratio n/nb w 109 [44] places a lower bound on the amount of CP violation,

which is one of the three requirements for the creation of the matter-antimatter

.i-v ii i.! I ry [45]. The SM incorporates CP violation only by the CKM mechanism

[46]. The measured CP violation in the SM is smaller by at least eight orders of

magnitude than the one needed to generate the cosmologically observed matter-antimatter

S 1 1 iii!1, I i'y.

Third, another condition formulated by Andrei Sakharov requires baryon number

violation. Currently, there is no experimental evidence of particle interactions where the

conservation of baryon number is broken perturbatively: this would appear to -ii-:.- -1 that

all observed particle reactions have equal baryon number before and after.









Another observation that cannot be explained within the SM is the requirement

for dark matter [47, 48]. According to the contemporary measurements, only 5' of the

amount of the total energy is stored in an ordinary matter as known by the SM. For the

remaining 95'. of the energy of the Universe there is no explanation in the SM.

Furthermore, if the SM is valid up to an energy scale A, then the size of divergent

loop contributions is Am' ~ A2. On the other hand loop corrections to the fermion

masses are only subject to the logarithmic divergences, so that the overall correction is

of the scale of mass itself and no fine-tuning problem emerges. For the Higgs boson this

means that if the SM is valid up to the Planck scale of Ap = 1019 GeV, then the natural

scale of the Higgs boson mass is Ap while all other particles have natural mass scales

below VEV v. This is the so-called hierarchy problem, which refers to the extremely large

splitting of the weak scale and natural cut-of scale, the Planck scale. In order to achieve

the necessary Higgs mass range of mH < 1 TeV, an unnatural fine-tuning with the relative

precision of mH/Ap > 10-16 has to be applied. This fine-tuning is not explained in the

context of the SM but it can be solved by extensions of the SM.

The SM leaves unexplained why the strong and the electroweak gauge structure is

SU(3)c x SU(2)L x U(1)y with different gauge couplings and fermionic quantum numbers

whose values are not predicted by the model. There have been many efforts to unify the

gauge groups and to have only one coupling at the energy scale of Grand Unification

[49, 50, 51].

There are more questions: why neutrino masses are so small; why there is no CP

violation in QCD, etc.

2.6 Supersymmetry Higgs

Some of the problems mentioned above can be solved by introducing a new symmetry

which relates bosons to fermions. Under such a symmetry, so called Supersymmetry

(SUSY), every SM fermion has a bosonic partner, and every SM boson has a fermionic

partner.









An additional Higgs doublet together with its supersymmetric partner must be

introduced into the SUSY extension of the SM.

Generally broken SUSY has a huge parameter space and therefore has a very

limited predictive power. An example of much more constrained version, with less free

parameters, is the Minimal Supersymmetric Standard Model (\LSSM).

In MSSM the eight real scalar degrees of freedom (that correspond to two complex

Higgs doublets) are three Goldstone bosons absorbed by the W' and Z, and five physical

Higgs bosons: neutral CP-even h and H, neutral CP-odd A, and two charged H'. The

features of the Higgs particles greatly depend on choice of parameters for particular SUSY

models.

MSSM predicts that mass of the lightest Higgs should have mass < 135 GeV [52].









CHAPTER 3
LARGE HADRON COLLIDER

The Large Hadron Collider (LHC) [53] is a 27 km circumference particle accelerator

spanning the Swiss-French boarder. It is a project carried out by CERN (European

Organization of Nuclear Research) [54], which is a large international collaboration

consisting of many countries from all over the world. The LHC is currently under

construction and has the first beams scheduled in 2008. When completed it will

provide proton-proton collisions at the center-of-mass energy 14 TeV as well as heavy

ion collisions.

The LHC consists of a number of accelerator units (Fig. 3-1 [53]). Two linear

accelerators, LINAC2 and LINAC3, will be used for the initial acceleration of protons and

lead ions respectively. The PS (Proton Synchrotron) will be used to provide a low energy

beam (25 GeV) with the final bunch structure. The beams are pre-accelerated using the

SPS (Super PS), and then are injected into the LHC near IP2/IP8 for the beam that

circulates clock wise/counter-clock wise, where particles will be accelerated from 450 GeV

to the nominal energy of 7 TeV (for proton beam).

Inside the LHC accelerator, the particles circulate in opposite directions in two

separate beam pipes. The diameter of the beam is reduced by focusing magnets before

the beam enters every experimental interaction point to achieve high luminosity collisions.

The two beam pipes are surrounded by shells of superconducting coils creating the

magnetic field which guides the beams to follow a circular path. Particles inside the

beam are grouped into approximately cylindrical bunches with a diameter and a length of

approximately 16p and 8 cm at the interaction point respectively. The LHC filling scheme

for proton beam will have a bunch separation of 25 ns.

The LHC will undergo a series of accelerator/detector commissioning stages and pilot

runs at the beginning of its operation in 2008. Regular physics runs will start with low

luminosity runs, planned luminosity is up to 2 1033cm- s-1, with up to about 10fb-1 of



















CERN Accelerators
(not to scale)


0.99999 by her
M


(ass son lQMOw SM6 (1)


LHC Large Hadron Collider
SPS: Suprr Phmon Synchhrrom
AnD: Antiprotnn Decelerator
ISOLDE: sotope Separator OnLine DEvice
PS8 Proton Synchrotron Booster
PS: Proton Synchrotron
LINAC LINear ACcelerator
LEIR Low Encrgy Ion Rng
CNGS: Crm Neutrin to Gran SauN


GnCm SiN (1)
TCta Qs


kMW MYl rb Lk nx
U- -d6 .I.. ItfL... a
D Magleaknthy OWN.14


IStartt e pro s ot h


Figure 3-1: Accelerator complex at CERN.


I0.87 by ere
0.3c ^^byhr









integrated luminosity per year and will go up to ~ 1034m -ls-1 with about 100fb-1 of

integrated luminosity per year.

There are four main experiments at the LHC: two general purpose detectors, C' \!

(Compact Muon Solenoid) [55] and ATLAS [56] (A Toroidal Lhc ApparatuS), and two

with dedicated detectors, ALICE [57] (A Large Ion Collider Experiment) and LHCb (the

Large Hadron Collider Beauty experiment) [58] which will study heavy ion and B-physics

respectively plus relatively small TOTEM experiment [59] for diffraction/forward

physics situated next to C\ IS, which also would allow to have more precise luminosity

measurements. The C\ IS detector is situated approximately 100 m underground at IP5

(Interaction Point in 5th LHC octant), and will be taking data from both the p-p and

Pb-Pb collision runs.









CHAPTER 4
COMPACT MUON SOLENOID (C' \S) DETECTOR

We give a short introduction into the C \!S and its subsystems the following chapter

(see e.g. [7, 8, 60] for more details).
4.1 Introduction

The total proton-proton cross-section s = 14 TeV is roughly 100 mb. At design

luminosity the general-purpose detectors will therefore observe an event rate of approximately

109 inelastic events/s. This leads to a number of experimental challenges. The online

event selection process ( I 1.--.. i") must reduce the approximately 1 billion interactions/s

to no more than about 100 events/s for storage and subsequent analysis. The short time

between bunch crossings, 25 ns, has 1 i ri" implications for the design of the readout and

trigger systems.

At the design luminosity, a mean of about 20 inelastic (hard-core scattering)

collisions will be superimposed on the event of interest. This implies that around 1000

charged particles will emerge from the interaction region every 25 ns. The products of an

interaction under study may be confused with those from other interactions in the same

bunch crossing. This problem clearly becomes more severe when the response time of a

detector element and its electronic signal is longer than 25 ns. The effect of this pile-up

can be reduced by using high-granularity detectors with good time resolution, resulting in

low occupancy. This requires a large number of detector channels. The resulting millions

of detector electronic channels require very good synchronization.

The large flux of particles coming from the interaction region leads to high radiation

levels, requiring radiation-hard detectors and front-end electronics.

The detector requirements for C'\! S [7, 8] to meet the goals of the LHC physics

program can be summarized as follows:


Good muon identification and momentum resolution over a wide range of moment
and angles, good dimuon mass resolution ( t 1 at 100 GeV/c), and the ability to
determine unambiguously the charge of muons with p < 1 TeV/c.









Good charged particle momentum resolution and reconstruction efficiency in the
inner tracker. Efficient triggering and offline J -:1ii-; of T's and b-i, I- requiring pixel
detectors close to the interaction region.


Good electromagnetic energy resolution, good diphoton and dielectron mass
resolution ( 1 at at 100 GeV/c), wide geometric coverage, measurement of the
direction of photons and/or correct localization of the primary interaction vertex, 7r
rejection and efficient photon and lepton isolation at high luminosities.


Good missing-transverse-energy and dijet-mass resolution, requiring hadron
calorimeters with a large hermetic geometric coverage and with fine lateral
segmentation.

The coordinate system adopted by C'\!S has the origin centered at the nominal

collision point inside the experiment, the y-axis pointing vertically upward, and the

x-axis pointing radially inward toward the center of the LHC. Thus, the z-axis points

along the beam direction counter clockwise looking from above. The azimuthal angle

Q is measured from the x-axis in the x-y plane. The polar angle 0 is measured from

the z-axis. Pseudorapidity is defined as q = -ln(tan(0/2)). Thus, the momentum and

energy measured transverse to the beam direction, denoted by PT and ET, respectively,

are computed from the x and y components. The imbalance of energy measured in the

transverse plane is denoted by E "8.

4.2 General Description

An important aspect driving the detector design and layout is the choice of the

magnetic field configuration for the measurement of the momentum of muons. Large

bending power is needed to measure precisely the momentum of charged particles. This

forces a choice of superconducting technology for the magnets. The design configuration

chosen by C'\!S [61] is discussed below.

At the heart of C'\ !S sits a 13-m-long, 5.9 m inner diameter, 4 T superconducting

solenoid (Fig. 4-1). In order to achieve good momentum resolution within a compact

spectrometer without making stringent demands on muon-chamber resolution and

alignment, a high magnetic field was chosen. The return field is large enough to saturate






























Compact Muon Solenoid


Figure 4-1: An exploded view of the C\ \S detector.


1.5 m of iron, allowing 4 muon "stations" to be integrated to ensure robustness and

full geometric coverage. Each muon station consists of several l .-;-i~ of aluminium drift

tubes (DT) in the barrel region and cathode strip chambers (CSCs) in the endcap region,

complemented by resistive plate chambers (RPCs).

The bore of the magnet coil is also large enough to accommodate the inner tracker

and the calorimetry inside. The tracking volume is given by a cylinder of length 5.8

m and diameter 2.6 m. In order to deal with high track multiplicities, C\ IS employs

10 1 .i.-i~ of silicon microstrip detectors, which provide the required granularity and

precision. In addition, 3 l-i.-rs of silicon pixel detectors are placed close to the interaction

region to improve the measurement of the impact parameter of charged-particle tracks,

as well as the position of secondary vertices. The EM calorimeter (ECAL) uses lead

tungstate (PbWO4) crystals with coverage in pseudorapidity up to Ir\ < 3.0. The

scintillation light is detected by silicon avalanche photodiodes (APDs) in the barrel

region and vacuum phototriodes (VPTs) in the endcap region. A preshower system is









installed in front of the endcap ECAL for 7r rejection. The ECAL is surrounded by

a brass/scintillator sampling hadron calorimeter with coverage up to I|q| < 3.0. The

scintillation light is converted by wavelength-shifting (WLS) fibres embedded in the

scintillator tiles and channeled to photodetectors via clear fibres. This light is detected

by novel photodetectors (hybrid photodiodes, or HPDs) that can provide gain and

operate in high axial magnetic fields. This central calorimetry is complemented by a

"tailcatcher" in the barrel region-ensuring that hadronic showers are sampled with nearly

11 hadronic interaction lengths. Coverage up to a pseudorapidity of 5.0 is provided by an

iron/quartz-fibre calorimeter. The Cerenkov light emitted in the quartz fibres is detected

by photomultipliers. The forward calorimeters ensure full geometric coverage for the

measurement of the transverse energy in the event. An even higher forward coverage

is obtained with additional dedicated calorimeters (not shown in the figure) and with

TOTEM.

The overall dimensions of the C'\ S detector are a length of 21.6 m, a diameter of 14.6

m and a total weight of 12 500 tons. The thickness of the detector in radiation lengths

is greater than 25 X0 for the ECAL, and the thickness in interaction lengths varies from

7-11A for HCAL depending on rl.

4.3 The Superconducting Magnet

The compact design of C'\ !S requires a very strong magnetic field in order to induce

enough bending of the charged particle trajectories so that the momentum of the particles

can be measured up to the highest momentum expected at the LHC. The basic design

goal was to be able to reconstruct 1 TeV muons with ~ 10' PT resolution which scales

with 1/B where B is the strength of the magnetic field. In the 4 Tesla field, trajectories

of charged particles with pT > 0.7 GeV reach the ECAL front surface (in the absence of

tracker material), and muons with pT > 4 GeV extend through the muon chambers.

The superconducting magnet for C'\lS has been designed to reach a 4 T field in a

free bore of 6m diameter and 12.5m length with a stored energy of 2.6 GJ at full current.









The flux is returned through a 10000 t yoke comprising of 5 wheels and 2 end caps

composed of three disks each [62]. The distinctive feature of the 220-t cold mass is the

four-li-v r winding made from a stabilized reinforced NbTi conductor. The ratio between

stored energy and cold mass is critically high (11.6KJ/ kg), causing a large mechanical

deformation (0.15 .) during energization, well beyond the values of previous solenoidal

detector magnets.

4.4 Inner Tracking System

The inner tracking system of C' \S is designed to provide a precise and efficient

measurement of the trajectories of charged particles emerging from the LHC collisions

as well as a precise reconstruction of secondary vertices. It surrounds the interaction

point and has a length of 5.8 m and a diameter of 2.5 m. The C'\!S solenoid provides a

homogeneous magnetic field of 4 Tesla over the full volume of the tracker. At the LHC

design luminosity of 1034cm-2s-1 there will be on average about 1000 particles from more

than 20 overlapping proton proton interactions traversing the tracker for each bunch

crossing, i.e. every 25 ns. Therefore a detector technology featuring high granularity

and read-out speed is required, such that the trajectories can be identified reliably and

attributed to the correct bunch crossing. However, these features imply a high power

density of the on detector electronics which in turn requires efficient cooling. This is in

direct conflict with the aim to reduce the amount of material in the tracker to a minimum

in order to limit multiple scattering, bremsst i i1ii.- photon conversion and nuclear

interactions, and a compromise had to be found in this respect. The intense particle flux

will also cause severe radiation damage to the tracking system. The main challenge in the

design of the tracking system was to develop detector components able to operate in this

harsh environment for an expected lifetime of 10 years. These requirements on granularity,

speed and radiation hardness lead to a tracker design entirely based on silicon detector

technology. The C'\ !S tracker is composed of a pixel detector with three barrel lV. ri

at radii between 4.4 cm and 10.2 cm and a silicon strip tracker with 10 barrel detection









l-iv r~ extending outwards to a radius of 1.1 m. Each system is completed by end caps

which consist of 2 disks in the pixel detector and 9 disks in the strip tracker on each side

of the barrel, extending the acceptance of the tracker up to a pseudo-rapidity of IT]1 < 2.5.

With about 200 m2 of active silicon area the C'\!S tracker is the largest silicon tracker ever

built [63].

For single charged particle of transverse moment of 1, 10 and 100 GeV the expected

resolution of transverse momentum depends on pseudorapidity [7, 8] and high momentum

tracks (100 GeV) the transverse momentum resolution is around 1 2'. up to IT\ 1.6,

beyond which it degrades due to the reduced lever arm. At a transverse momentum of 100

GeV multiple scattering in the tracker material accounts for 20 to "3i l' of the transverse

momentum resolution while at lower momentum it is dominated by multiple scattering.

The transverse impact parameter resolution reaches 10pm for high PT tracks, dominated

by the resolution of the first pixel hit, while at lower momentum it is degraded by multiple

scattering (similarly for the longitudinal impact parameter). For muons, the expected

track reconstruction efficiency (a function of pseudo-rapidity) is about 9' ".- over most

of the acceptance. For Ir] w 0 the efficiency decreases slightly due to gaps between

the ladders of the pixel detector at z w 0. At high the efficiency drop is mainly due to

the reduced coverage by the pixel forward disks. For pions and hadrons in general the

efficiency is lower because of interactions with the material in the tracker.

4.5 Electromagnetic Calorimeter (ECAL)

The Electromagnetic Calorimeter of C' \S is a hermetic homogeneous calorimeter

made of 61,200 lead tungstate (PbWO4) ( iv-I ,1J mounted in the central "barrel" part,

closed by 7,324 < i --1 i1 in each of the two end-caps. A preshower detector is placed in

front of the end-caps ( iv-- i- Avalanche photodiodes (APDs) are used as photodetectors

in the barrel and vacuum phototriodes (VPTs) in the end-caps. The use of high density

crystals has allowed the design of a calorimeter which is fast, has fine granularity and

is radiation resistant, all important characteristics in the LHC environment. One of the








driving criteria in the design was the capability to detect the decay to two photons of the
Higgs boson. This capability is enhanced by the good energy resolution provided by a
homogeneous i --1 I calorimeter.
A typical energy resolution for central impact when the energy was reconstructed in a
3 x 3 matrix of crystals was found to be:

(%72 ( 'G 2 (125(MeV))2
S(( ) (41)
Ex V/E/(GeV) E( (41)
The principal aim of the C'\ S Preshower detector is to identify neutral pions in the
endcaps within a fiducial region 1.653 < T]\ < 2.6. It also helps the identification of
electrons against minimum ionizing particles, and improves the position determination of
electrons and photons with high granularity.
4.6 Hadron Calorimeter (HCAL)
The hadron calorimeters are particularly important for the measurement of hadron
jets and neutrinos or exotic particles resulting in apparent missing transverse energy [61].
Hadronic showers have a large lateral shower size, so the degree of granularity
provided by the ECAL is not required for the HCAL. However, moderate transverse
granularity, hermeticity and wide rapidity coverage are essential in measuring the total
missing ET of an event. The HCAL consists of three main parts: Hadron Barrel (HB) and
Hadron Endcap (HE), which extends to ] = 3.0, and Hadron Forward (HF) calorimeters,
which are located around the beam pipe outside the muon system at zl = 10.9 m to
complete the coverage up to r = 5.3. Due to the restricted space available between the
ECAL and the magnetic coil, the HB is only 1 m in thickness whereas the HE is 1.8 m
thick; additional scintillation lv -ri (Hadron Outer calorimeter, HO) are installed just
outside the magnet, using the coil as an absorber, to provide a combined depth of at least
11 absorption lengths.
For gauging the performance of the HCAL, it is usual to look at the jet energy
resolution and the missing transverse energy resolution. The granularity of the sampling in









the 3 parts of the HCAL has been chosen such that the jet energy resolution, as a function

of ET, is similar in all 3 parts. The resolution of the missing transverse energy (EtiS8)

in QCD dijet events with pile-up is given by a(ET"i') w 1.0 EET if energy clustering

corrections are not made, while the average ET"i is given by (Ei"88) 1.25 E/ [7].

4.7 Muon System

The muon system is a tracking device in the outermost region of the C'\ !S detector,

where only muons and neutrinos would have passed through the calorimeters without

depositing a large fraction of their energy. The challenge of the muon system is to provide

fast recognition, efficient reconstruction and online background suppression, which are

required for tri_ -. -- ri-_i; purposes and also, good measurement for higher PT muons. There

are four 1l-. r-i of muon stations in the barrel and the endcaps interleaved with the iron

return yokes, each providing track segments reconstructed from a few distributed hits.

These will later be combined with information from the inner silicon tracking system to

form complete muon tracks.

The muon system employs a combination of three different detectors: Drift Tubes

(DT) in the barrel and Cathode Strip C'!i ilers (CSC) in the endcaps, both assisted by

Resistive Parallel plate C'! ih!ers (RPC).

The DT and CSC detectors are used to obtain a precise position measurement;

background rejection is ensured by the highly redundant design, where each of the four

stations contains 12 DT 1'-. r-i for the barrel and 6 CSC 1'-. r-i in the endcap. The RPCs,

which have a very fast response, provide a time resolution of the order of 1 ns, and are

dedicated for the triggers.

4.7.1 Barrel Muon System

The C'\ !S barrel muon detector is made of four stations forming concentric cylinders

around the beam line: three of them consist of 60 drift chambers each, the fourth, of 70.

The total number of sensitive wires is about 195,000. The choice of a drift chamber as the









tracking detector for the barrel muon system was possible due to the low expected rate

and the relatively low intensity of the local magnetic field.

Within a single station, the DT l1v.-r-i provide a final spacial resolution of 100 pm.

4.7.2 Endcap Muon System

The endcap muon system is described in details in the following chapter.

A final spacial resolution within a single station for CSCs is 80-150 pm depending on

the station.

The CSCs are chosen for the endcaps for their performance in the presence of a high

intensity varying magnetic field, and their ability to handle high particle rates.

4.7.3 Resistive Plate Chambers

Resistive Plate CI'! iihers (RPC) are gaseous parallel-plate detectors that combine

good spatial resolution with a time resolution comparable to that of scintillators. An RPC

is capable of .- -.ii.,-; the time of an ionizing event in much shorter time than the 25 ns

between two consecutive bunch crossings (BX). A total of six l1v.-r-i of RPC chambers will

be embedded in the barrel iron yoke, two located in each of the first and second muon

stations and one in each of the two last stations.

4.8 Muon Trigger

The C'\!S muon trigger and CSC muon trigger in particular will be used in the

following chapters. We will give a short overview here, more details are available in

[64, 65, 7, 8, 60].

Muon Trigger. All three muon systems (DT, CSC, RPC) take part in the trigger.

The DT chambers provide local trigger information: Q and Tl-projections for track

segments. The CSCs deliver 3-dimensional track segments. All chamber types also identify

the bunch crossing at which event took place. The Regional Muon Til-.--. -r consists of the

DT and CSC Track Finders, which join segments to complete tracks and assign physical

parameters to them. In addition, the RPC trigger chambers provide their own track

candidates based on regional hit patterns. The Global Muon T i-:.--- r then combines the









information from the three subdetectors, achieving an improved momentum resolution and

efficiency compared to the stand-alone systems. The initial rapidity coverage of the muon

trigger is |T1 < 2.1 at the startup of LHC. The design coverage is |T1 < 2.4.

CSC local trigger. Muon track segments, also called Local C'!i irged Tracks (LCT),

consisting of positions, angles and bunch crossing information are first determined

separately in the nearly orthogonal anode and cathode views. The trigger electronics

determines the centre of gravity of the charge with a resolution of half a strip width,

between 1.5 and 8 mm, depending on the radius. By demanding that at least four 1-i. -~

are hit, the position of a muon can be determined with a resolution of 0.15 strip widths.

The best two LCTs of each chamber are transmitted to the regional CSC trigger, the

CSC Track Finder, which joins segments to complete tracks. The CSC trigger hardware

consist of several different electronics boards, in particular, ALCT (for anode LCT) and

CLCT (for cathode LCT). T i.-.-. motherboard (TMB) looks for coincidence between

two (ALCT, CLCT), if found, information transmitted to Muon Port Cards (\!PC). Up

to three best candidates then sent from MPC to CSC track finder. If then global tri._. r

provides accept signal, anode and cathode information is passed to Data Acquisition

system (DAQ) and then stored for further offline reconstruction and analysis.









CHAPTER 5
CATHODE STRIP CHAMBERS (CSC)

We give a short introduction into the CSC in the following chapter (see e.g. [65, 7, 8,

60] for more details).

5.1 Introduction

At the time of the LHC start-up, the C\MS Endcap Muon System [65, 7, 8] will

consist of 468 Cathode Strip C('i ihlers (CSCs) arranged in groups as follows: 72 ME1/1,

72 ME1/2, 72 ME1/3, 36 ME2/1, 72 ME2/2, 36 ME3/1, 72 ME3/2, 36 ME4/1 (Figs. 5-1

and 5-2). The de-scoped 72 ME4/2 chambers will not be available during early years

of C\ !S operation. The chambers are trapezoidal and cover either 10 or 20 degrees in

Q; all chambers, except for the ME1/3 ring, overlap and provide contiguous (-coverage.

A muon in a pseudorapidity range 1.2 < Tr1 < 2.4 crosses 3 or 4 CSCs. In the endcap

barrel overlapping range 0.9 < |T1 < 1.2, muons are detected by both the barrel Drift

Tubes (DTs) and endcap CSCs. For pseudorapidity |T1 < 2.0, muons are also detected by

Resistive Plate C('!i ilers (RPCs).

CSCs are multi-wire proportional chambers comprised of six anode wire planes

interleaved between seven cathode panels (Fig. 5-3). Wires run azimuthally and define

track's radial coordinate. Strips are milled on cathode panels and run lengthwise at

constant A0 width. Following the original CSC idea [66], we obtain the muon coordinate

along wires ((-coordinate in C\ !S coordinate system) by interpolating charges induced

on strips (Fig. 5-4). The largest chambers, ME2/2 and ME3/2, are about 3.4 x 1.5 m2 in

size. The overall area covered by the sensitive planes of all chambers is about 5000 m2,

the total gas volume is > 50 m3, and the total number of wires is about 2 million. There

are about 9,000 high voltage channels in the system. The number of cathode strip readout

channels with 12-bit signal digitization is about 220,000, and the number of anode readout

channels is about 180,000.

CSCs provide functions of a precision muon measurement and muon trigger in one

device. They can operate at high rates and in large and non-uniform magnetic fields.









11=0.9
=1.2 I Ma/2/ J| MBOl/4 i MB/1









r ..... ........ I EMs_.






H E E E E E 0 0. 1
rl=,4(B'-- L TI, I ME


Figure 5-1: Quater-view of the
Muon System are highlighted.


C' S Detectors. Cathode Strip C'!i inlers of the Endcap


They do not require tight gas, temperature or pressure control. Additionally, the radial

fan-shaped strip pattern, natural for measurements in the endcap region, can be easily

arranged on cathodes.

The requirements for the C\ IS cathode strip chamber performance are as follows:


Reliable and low maintenance operation for at least 10 years at the full LHC
luminosity, i.e. at estimated random hit rates up to 1 kHz/cm2


At least 9' '. efficiency per chamber for finding track stubs at the 1st Level trigger.


At least 9'"- probability per chamber of identifying correct bunch crossings at the
1st Level trigger. With such an efficiency per chamber and 3-4 CSCs on a muon
track path, a plain ini, Pi i ly rule ensures that the reconstructed muons will be
assigned to a correct bunch crossing number in more than 9'' of cases.


About 2 mm resolution in r6 at the 1st Level tri.-U-r.














































Figure 5-2: The ME2 cathode strip chambers. The outer ring consists of 36 ME2/2
chambers, each spanning 100 in Q; while 18 20-degree ME2/1 chambers form the inner
ring. The chambers overlap to provide contiguous coverage in 4.













47










wire plane (a few wires shown)
cathode plane with strips muon
cathode
.F . . . wires
cathode





induced charge
-cathode with strips
Avalanche wires
cathode

7 trapezoidal panels forming 6 gas gaps


Figure 5-3: A schematic view of a C\l!S
e AFche of igure 5-4: A schematic view of a single
cathode strip chamber made of seven
trapezoidal paes gap illustrating the principle of a CSC
trapezoidal panels. o io
operation.


About 75 pm off-line spatial resolution in r6 for ME1/1 and ME1/2 chambers and
about 150 pm for all others.

5.2 Chamber Mechanical Design

The larger 396 chambers and the smallest 72 ME1/1 chambers have somewhat

different mechanical designs. Below, we describe the design of the larger chambers on an

example of ME2/2 and, then, at the end of this section, summarize the MEl/1-specific

features that distinguish them from all other chambers.

The largest chambers, 144 ME2/2 and ME3/2, are 3.4 m long and 1.5 m wide

(Fig. 5-5). Seven 16.2-mm thick trapezoidal panels are the basis for the mechanical

structure. The panels are made of 12.7-mm thick polycarbonate honeycomb core with two

1.6-mm FR4 skins commercially glued on each side. FR4 is fire-retardant fiberglass/epoxy

material widely used for printed circuit boards. The FR4 skins are copper-clad on their

outer surfaces-the copper lw r-i serve the role of cathodes.

FR4 cathode gap bars are glued to both sides of each other panel (panels 1, 3, 5,

7 in Fig. 5-5) so that when all panels are stacked together, these cathode bars define

six gas gaps of 9.5 mm. To provide an additional support, there are four spacers placed









frame extrusions


panels wires insulation strips
gap bars





Ja.
--. \ _. n, ,E
S\




-- .............- -- ... --- --- "---




\ wire fixation bars blocking capacitors
HV resistors and capacitors holes for tie rod
holes for assembly bolts anode protection boards

Figure 5-5: Mechanical design of the CM'\S Cathode Strip ('!i l hers (exploded view).


between panels along the chamber centerline. When all seven panels are put together,

the entire stack is tighten down with bolts along the chamber perimeter (through holes

in the cathode gap bars) and in 4 points along the chamber centerline (through holes

in the spacers). Such arrangement ensures that panels have no more than 60 cm of

unsupported length. Measurements showed that most of the panels were flat within the

required 300 pm on such spans. This specification arises from the desire to keep gas gain

variations within a factor of two.

One side of six panels has a milled pattern of 80 strips. Strips, being radial, have a

varying pitch from 8.4 mm at the bottom to 16.0 mm at the top. The gap between strips

is about 0.5 mm. The precision of milling is better than 50 pm (rms). Milling is done with

a cutter tilted at 45 degrees to make groove edges smoother (otherwise, sharp edges and

burrs might provoke sparking and discharges).

Anode wires are wound on three so-called ,ii.de" panels (panels 2, 4, 6 in Fig. 5-5);

these panels do not have gap bars. A special winding machine was designed to wind wires









directly on a panel rotating around its long axis at a speed of about 5 turns per minute.

This allows for winding one panel (about 1000 wires on each side) in less than 4 hours.

The wire spacing of about 3.2 mm is defined by combs, threaded rods running the full

panel length and attached to the panel edges during winding. Gold-plated tungsten wires,

50 pm in diameter, are stretched at 250-gram tension (about 71' '. of the elastic limit) and

run their full length up to 1.2 m without any intermediate supports. The electrostatic

stability limit for the longest wires is above 6 kV (the nominal operational point is 3.6

kV). Based on the measurements during production, the wire tension non-uniformity does

not exceed 10' while wire spacing variations are less than 150 pm. Wires found to

fall out of these specs were replaced.

After winding, the wires are first glued and then soldered to anode bars of 4.75 mm

height (half of the gas gap). The anode bars are made of copper-clad FR4 and carry an

electric artwork. An automated soldering machine allowed for soldering at a speed of 3.5 s

per joint. Groups of sixteen wires make one anode readout channel of about 5 cm width.

High Voltage (HV) is distributed to the wire groups on one end and signals are read out

on the other end via 1 nF blocking capacitors.

Each wire plane is sub-divided by spacer bars into 5 independent HV segments,

which allows us to independently regulate or turn off HV on any of the five sections. In

places where the spacer bars are inserted (and prior to their installation), eight wires are

removed. Two gold-plated 200 pm guard wires inserted in place of the first and eighth

removed thin wires to eliminated edge effects. The very first and the last wires in each

plane are also thick. If the edge thin wires are left unguarded, the electric field on them

would be much larger than for the rest of wires, which would provoke discharges. All

in all, such plane segmentation, by virtue of introducing intermediate panel supports

and individual HV control over smaller wire plane sections, makes the overall chamber

performance very robust.










1,000,000 .................. ..... 10

::::::::~~~~- --::^:::::::::|:::: ::::::::: ::::::::;:^ --------- ---------- --------

: 7 : 4 :

10,000


1


1,0-----------------0----------------------------- -------------- ------------------

3.0 3.5 4.0 3.0 3.5 4.0
High Voltage, kV High Voltage, kV


Figure 5-6: CSC gas gain vs high Figure 5-7: ME2/1 chamber singles rate
voltage. vs high voltage (the overall sensitive
area of all six planes in this chamber is
~9.5 m2).


After stacking the panels and tightening bolts (with o-rings), continuous beads of

sealant RTV are applied along the outer seams between panels and gap bars. The o-rings

around bolts and the RTV seal make chambers hermetic. Should a need arise, chambers

can be opened, serviced, and resealed. The gas enters into one of the outer gas gaps via

an inlet in a cathode gap bar, then flows from one plane to another in zigzag manner via

special holes in panels, and finally exits from the last gas gap via an outlet in a gap bar.

The leak rate was measured during production and upon installation of chambers and was

required to be less than 1 of the chamber volume per d4i at overpressure of 7.5 mbar

(<2 cc/min for the largest chambers whose gas volume is about 200 liters).

Side plates made of 3.2-mm thick Al extrusions are attached along the chamber

perimeter. They stiffen the chamber and interconnect top and bottom copper skins to

form a complete Fa-gvdiv cage.

The nominal gas mixture is Ar + CO2 + CF4 .' + 50'-, + 10', The CO2

component is a non-flammable quencher needed to achieve large gas gains, while the









main function of CF4 is to prevent polymerization on wires. The detailed discussion of the

gas optimization for C'\IS CSCs can be found elsewhere [67].

Figure 5-6 shows the chamber gas gain vs. high voltage. The nominal operational

HV point is chosen to be 3.6 kV, which corresponds to gas gain of the order 7 x 104.

Taking into account that a minimum ionizing particle (mip) leaves behind in a gas gap

about 100 electrons, the total charge in an avalanche per mip is about 1 pC. As will be

shown below, at this operational point, the cathode and anode electronics have a very

high efficiency and an adequate signal-to-noise ratio. The operational range of chambers

extends to 3.9 kV. Typically, we start seeing a sharp rise in rate of spurious pulses at

about 3.9-4.0 kV-see Fig. 5-7.

The 72 smallest ME1/1 chambers have variations in the mechanical design with

respect to all other chambers. First, the gas gap is 7 mm, wire diameter is 30 pm, and

wire spacing is 2.5 mm. As a consequence, the nominal high voltage for these chambers is

somewhat lower, 2.9 kV. Secondly, and more importantly, anode wires in ME1/1 chambers

are not azimuthal, but rather rotated by an angle aL = 290 as shown in Fig. 5-8. Unlike

all other CSCs, the ME1/1 chambers are inside the C'\ S solenoid and see its strong and

uniform 4 Tesla axial field. If the wires were not tilted, the ionization electrons drifting

toward wires would be carried -idl, .. li- at the Lorentz angle and become spread along

wires and across strips. The wire tilt compensates for the Lorentz angle so that electrons

drift parallel to strips and the precise measurement of the rw-coordinate remains possible.

5.3 Electronic Design

Figure 5-9 shows a schematic layout of the custom tri --r and readout electronic

boards developed for the Endcap Muon System.

An anode front-end board (AFEB) has one 16-channel amplifier-discriminator ASIC.

The amplifier has a 30 ns shaper (semi-Gaussian with a 2-exponent tail cancellation

designed to suppress the slow signal component associated with a drift of positive ions

away from anode wires), about 7 mV/fC sensitivity, and 1.4 fC noise at a typical 180 pF












XO
-1 1.- Ax= Ay tan (aL


I-Il


Cathode strips


Figure 5-8: Signal in ME1/1 chambers for not tilted and tilted wires.


Muon Port Card (MPC) ----
Clock Control Board (CCB) ---- Trig Motherboard (TMB)
--r-- Q Motherboard (DMB)
Controller (CC) ------

RRoadout Dal3

Peripheral Ct are
/ (on iron disk)


Cathode Front-End Board
(CFEB)


Anode Local Chard
Track Board (ALCT)
LV Distribution Board (LVDB) -


CMS
DAQ


" FED Crate
(in service cavern)


EMU DAQ


....... Anode Front-End Board (AFEB)


Figure 5-9: An overview of all custom-made CSC trigger and readout electronics.
Functionality of various boards is described in the text.


Crate









wire group capacitance for the largest chambers. With the 30-ns shaping time, AFEB sees

about 1"'-. of the total avalanche charge, i.e. about 130 fC on average. A typical chamber

signal as seen at the output of this amplifier is shown in Fig. 5-10. The constant-fraction

discriminator has a threshold nominally set at 20 fC (input equivalent charge) and its

slewing time is less than 3 ns for 60-600 fC signal range. Depending on a chamber size,

there are 12 to 42 AFEBs per chamber. All details on the AFEB design and performance

can be found elsewhere [67].

Every 25 ns in sync with the LHC collisions, all AFEB outputs, 40-ns long step

pulses, are sampled by an FPGA-based Anode Local C'!i irged Track (ALCT) board, one

board per chamber. The recorded yes/no information is stored in FIFO. Upon receiving

a C'\!S-wide "Level-1 trigger .... 'l command (L1A), the recorded information from a

proper time window is extracted and reported to DAQ. The latency of the L1A command

with respect to the time of a collision is 3.2 ps. The width of a record with raw hits

transmitted to DAQ can be set to be as large as 32 bits (1 bit per 25 ns), i.e. 800 ns.

The ALCT board has another important function. Based on the information from

all anode channels, the FPGA code constantly (every 25 ns) searches for patterns

of hits in six planes that would be consistent with muon tracks originating from the

interaction point. For a pattern to be valid, we require that hits from at least 4 planes

were present in the pattern. Figure 5-11 illustrates how patterns are identified among

spurious single-plane hits. Due to a large neutron-induced photon background, we

expect a substantial rate of such single-plane hits. However, these hits, being completely

uncorrelated, would not typically line up to form track-like patterns. Found patterns,

Anode Local C'I irged Tracks, are tri -. --r primitives. They are transmitted further

upstream to the muon Level-1 trigger electronics that builds muon track candidates from

these primitives. The time it takes to form an anode track trigger primitive is 225 ns

(including drift time). The ALCT board can find up to 2 such patterns per each bunch










1 50


.00 ---
E
< 0.75



0.50
... i --


40.75 -.-.-.-.--..- ..............-----4 I .------- ----- I ------ --- ------ -- I------- ~ ~
-0.75
-1 00 ....
-200 -100 0 100 200 300 400 500 600 6 5 4 3 2 1
Time, ns Layers

Figure 5-10: Muon signals as seen at the Figure 5-: A schematic event d
Figure 5-11: A schematic event di pliv
AFEB amplifier output.
showing anode signals in a six-plane
chamber.


crossing, which is well adequate for the expected chamber track occupancy at the nominal

LHC luminosity.

One cathode front-end board serves (6 planes) x (16 strips) = 96 channels and has

six parallel chains of the following chips (Fig. 5-12): 16-channel amplifier-shaper ASIC,

16-channel Switched Capacitor Array ASIC, 12-bit 1-channel ADC, and 16-channel

Comparator ASIC. There are 4 to 5 CFEBs per chamber.

The front-end amplifier-shaper ASIC has 100 ns shaping time and 0.85 mV/fC

sensitivity over the liner range up to 1 V. The equivalent noise level at ~300 pF strip

capacitance is typically 1.5 fC. The shaping is based on semi-Gaussian transfer function

with an overshoot designed to compensate for the 1/t signal tail due to slow drift of

positive ions. After convolution with the current pulse produced in a chamber by a muon,

the amplifier-shaper signal peaks at around 150 ns and has no tails-see Fig. 5-13. With

the 100-ns shaping time, CFEB sees about S'~. of the total avalanche charge, i.e. about

100 fC on average.










3500
SCA ADC to DAQ E Digitized Chamber Pulses

3000
0/ dots: sampled, stored and digitized data
Scomparator > to CLCT curves: scope traces of preomp-shoper output
network
Amplifier-Shaper w
S 2500 .. . .
0
3" 2000 L
E
a, 1750 2 zOmps 2000 --.-------" --.----.- ----
- 1500 Impulse charge inputs _
p 1mip- o .110 fC
S1250 1500 .-- '/ "

O 1000 om ip.
750 1000
500
250 500 .
I i 500 ------.-J


-250
0 100 200 300 400 I l, i
Time (ns) -200 -100 0 100 200 300 400
time (ns)

Figure 5-12: Basic functional diagram of Figure 5-13: Muon hit signals from six
a Cathode Front-End Board. nearby strips. Four curves are actual

oscillograms.


The output from this chip is split in two pathrv--l- One leads to the Switched

Capacitor Array (SCA) ASIC. The SCA chip samples a waveform of each strip signal

every 50 ns in sync with the LHC clock and stores this analog information on its

capacitors. The depth of such analog memory is 96 capacitor cells per each channel,

or 96 x 50 ns = 4.8 ps. Upon receiving the L1A command in 3.2 ps after collision, 8

or 16 consecutive samples from the proper time range in the SCA line of capacitors are

retrieved and one-by-one digitized by the 12-bit flash ADCs. The digital information

is passed on to the data acquisition system via an intermediate digital data buffer. For

the digitization and subsequent readout by DAQ to happen, the L1A signal must be in

a coincidence with the Cathode Local C('i 'ged Track primitive decision to be described

next.













QL -









Threshold



Figure 5-14: A simplified schematic Figure 5-15: CLCT (Cathode Local
illustrating the idea behind the C'!i ,iged Track) is a pattern of half-strip
comparator network. hits consistent with a muon track.

The second amplifier-shaper output goes to the Comparator ASIC. This chip

compares signals on triplets of nearby strips at the time when signals reach the maximum

amplitude. By means of such comparisons, the Comparator ASIC can identify muon

hit location within one half of a strip width, independently of a signal amplitude, an

induced charge shape (as long as it is "bell"-like), and a strip width itself [68]. Figure 5-14

illustrates the basic idea behind the Comparator ASIC algorithm.

Comparator half-strip hits are sent to the Ti i--.- r Mother Board (TMB). Similarly

to the ALCT board, the TMB searches for patterns of half-strip comparator hits that

would be consistent with muon tracks of interest-see Fig. 5-15. Up to two Cathode Local

C!i irged Tracks (CLCTs) per bunch crossing can be found. As in the ALCT pattern

search, for a CLCT pattern to be found, half-strip hits from at least four planes should

be present in it. There is one TMB per chamber. Unlike ALCT boards, TMBs are not on

chambers, but rather in peripheral crates mounted along the outer rim of the endcap iron

disks.









The TMB also matches ALCT and CLCT patterns found within a chamber to make

correlated 2d-LCTs (2d-LCT ALCTxCLCT), up to two per bunch crossing. These

2d-LCTs are sent to the Muon Port Card (\!PC), which serves 9 chambers covering either

60- or 300-sectors in Q. For each bunch crossing, MPC performs a preliminary sorting of

all received correlated 2d-LCTs and finds 3 best candidates-these are then sent further

upstream to the muon Li-trigger electronics.

All raw data are collected by DAQ Mother Boards (DMB). There is one DMB for

each chamber; DMBs are located in peripheral crates as well. The data consists of raw

anode and comparator hits within a time windows up to 32 bunch crossing, ALCT and

CLCT decisions in the same windows, and digitized strip signal waveform (8 or 16 50-ns

time samples). Status of the various electronic boards is also a part of the event record.

The data collected by DMB is passed to DDU (Detector-Dependent Unit), then further to

DCC (Data Concentration Card), and finally to the C'\IS Filter Farm to be processed by

the C'\!S High-Level Ti:.-.-. r software. Event size per chamber is about 5 kBytes.

It is important to note that the CSC readout is intrinsically zero-suppressed. The

anode raw data in a particular chamber are passed upstream only if there was an ALCT

pattern in coincidence with the L1A signal. Likewise, the cathode information, comparator

hits and digitized strip signal waveforms, are passed upstream to DAQ only if there was a

similar CLCTxL1A coincidence. The coincidence window is programmable, but nominally

set at 75 ns, i.e. 1 bunch crossing.

At the ultimate LHC luminosity, on average, we expect to find track stubs in 2

chambers per each L1A signal. With the maximum C'\ IS L1A rate of 100 kHz, the data

flow rate from CSCs to HLT is estimated to be around 1 GB/sec.

Operation of peripheral VME crates is supported by Clock-Control Board (CCB) and

custom Crate Controller (CC). CCB, as its name implies, distributes LHC clock and all

C'\!S control commands (like L1A signals).









The High Voltage system is custom-made and provides channel-by-channel regulated

voltage up to 4.0 kV with about 10 V precision. Currents of less than 10 pA can be

measured with a precision of 100 nA, while the precision for larger currents is about 1

The system can provide more than 100 pA current for individual channels as long as the

average consumption does not exceed 40 /A per channel. The maximum expected current

at the ultimate LHC luminosity for the most loaded HV segment is <10 pA.









CHAPTER 6
FAST ALGORITHM FOR TRACK SEGMENT AND HIT RECONSTRUCTION IN
CATHODE STRIP CHAMBERS

6.1 Introduction

Muon reconstruction in the High Level Ti i--.- i (HLT) starts from finding local track

segments in muon chambers (for the endcap muon system, in six-plane cathode strip

chambers). The found local track segments in all muon chambers and associated with

them reconstructed hits, RecHits, are then used for a standalone muon reconstruction.

An event is accepted by the L2 trigger (the first stage of the HLT event selection),

if there is at least one muon with transverse momentum pr > 19 GeV/c or if there

are at least two muons with pr > 7 GeV/c each (these are luminosity dependent

threshold taken from [64] for 2 1033cm-lsec-1 luminosity). The following stages of

HLT involve calorimeter (L2.5 muon isolation) and reconstruction of tracks in the tracker

(L3-refined momentum and muon isolation). More details on the C' \! High Level

Ti1 i-:. r" can be found elsewhere [64].

This chapter deals with the track segment reconstruction. The four main criteria

qualifying a track segment finding performance are as follows:


CPU time per event and operational code robustness;


efficiency of finding segments associated with muons of interest;


spatial resolution for localizing such segments;


rate of finding secondary segments not directly associated with a muon of interest.

CPU time per event is derived from the overall HLT Event Filter Farm size (4000

CPUs) and the maximum L1 trigger rate of 100 kHz. Given these numbers, one has only

40 ms per event for the whole HLT pro'. ---iv:, which includes the L2 stage (standalone

muon) with 1(I11I weight, L2.5 calorimeterr isolation) with .-:I' ", weight, and L3

(tracks in the tracker) with ~21 1' weight. The weights come from the fact that each









next level trigger sees only a smaller faction of the incoming events. The local segment

reconstruction within a single chamber is among the simplest operations required from

HLT and, therefore, only a very small fraction of the overall 40-ms budget can be allotted

for that. A typical Li-trigger event is expected to have on average three CSCs with track

stubs in them. At the time of writing this chapter, the HLT algorithms have not been

optimized for time performance and, therefore, it is not possible to set a quantitative

specification on the maximum CPU allotment for a local segment reconstruction, but,

clearly, it is hardly going to be much more generous than a few ms per CSC (existing

algorithm were measured to take about 15 ms per event). So, our goal was to de-

velop and validate with real data a new, fast, reliable, efficient, and precise

algorithm for CSCs which would be suitable for HLT needs.

Efficiency of finding muon track segments is desired to be at least 9' I. over the

sensitive chamber area. This will allow for reconstruction of muon tracks in the whole

system with high efficiency and small systematic uncertainties.

Spatial resolution per segment at HLT is not required to be much better than ~0.5

mm. This can be illustrated by the following simple considerations:

> As far as the standalone muon reconstruction is concerned (L2-part of HLT), the
error in a sagitta for the ME1-ME2-ME3 stations due to multiple scattering in 60-cm iron
disks separating them for a muon with pT=20 GeV at Tr=2.0 (p=75 GeV) is about 1 mm.
The error for sagitta based on IP-ME1-ME2 (IP-interaction point) for the same muons is
larger than 3 mm.

> For matching a found standalone muon to tracks in the Si tracker (or wise
versa, matching Si tracks to muon stubs in the muon chambers), one needs to make
an extrapolation through the calorimeters. The extrapolation error due to multiple
scattering for the same benchmark muons (pT=20 GeV at r= 2.0) is about 4 mm.

> Once the tracker hits are included in the muon reconstruction, the muon momentum
resolution is then completely defined by the tracker up to 200 GeV in PT [69]. Therefore,
below the highest expected pr threshold of 19 GeV, the final muon momentum measurement
at HLT is insensitive to the muon chamber resolution at all.

Rate of finding associated secondary segments not directly associated with a muon

of interest is another very important measure of merit for segment finding performance









both for HLT and offline muon reconstruction. Secondary segments can originate from

physical secondary tracks (e.g., jet punchthrough and electromagnetic showers associated

with high-energy muon bremsstrahlung). Combinatorial fakes resulting from all possible

pairings of anode and cathode hits as one forms 2d-hits and/or segments are yet another

source of secondary segments. A single secondary segment, simply by an incorrect

association with a physical soft muon, can result in a dramatic mis-measurement of a

muon momentum and promoting it to a much higher pr. At the tri 'i. -r level, this leads

to larger rates of fake high-pr muons and, even worse, to flattening of the tri -. --r rate vs.

muon pT threshold, thus, resulting in inability to control muon trigger rates. In offline

data ,in l -- one would have to cope with the associated high rates of fake high pr

muons with poorly understood systematic uncertainties.

Development and testing of the proposed algorithm was performed in new C'\ lS

software framework (C'\ SSW), which is beign developed for the future use by C'l\

detector for taking and analyzing data.

Details of our analysis and supporting studies on the new algorithm are being

prepared for publication and can be also found in the author's talks at the internal C'\ !

collaboration meetings of different levels [21].

6.2 Algorithm Principles

Si,--.: -I. 1 by the University of Florida group, the fast track segment finding

algorithm described in this chapter has become known as the UF algorithm. In the

interests of brevity and to distinguish the new algorithm from the others (SK, TC, DF,

ST, see for example [70] for details), we will further refer to it under this nickname. The

UF algorithm is built on the following principles.


First, instead of starting from pairing anode and cathode strip charge clusters in
individual planes to form 2d-RecHits and then trying to build 2d-segments of them,
the UF algorithm starts from finding id-segments, Anode- and CathodeSegments,
and pairs them to build 2d-ProtoSegments. This order saves substantially CPU time
needed to reconstruct complex events with more than just one hit per plane.









Second, having found 2d-ProtoSegments, the UF algorithm then takes only those
strip charge clusters that fall within a close proximity of the found segments and
reconstructs their precise coordinates. To speed up calculations, the reconstruction
is done without accessing any calibrations and/or databases. By default, we use the
ratio-method [71]. The ratio method requires no iterations, is very fast, and delivers
a spatial resolution substantially better than what is required for the HLT purposes.


And finally, the last step is to refine parameters of the found 2d-ProtoSegments and
remove "bad" RecHits. To do that, we make a linear fit of all RecHits originally
associated with each 2d-ProtoSegment. If fit's X2 is not satisfactory, some RecHits
are allowed to be pruned. This is also a linear non-iterative procedure taking very
little time. The final linear fit defines the RefinedSegment parameters.

In short the key principles of the UF algorithm are: first, Id-segments; second,

2d-segments; then, refine RecHits; and, at last, pruning of "bad" RecHits and refinement

of 2d-segment parameters. Certainly, additional l.v.-ir of sophistication can be added to

each of the three steps of the UF algorithm, provided they help improve the performance

and keep the overall computational time in check.

6.2.1 ProtoSegment Building

To define 2d-ProtoSegments, in the simplest implementation of the UF algorithm,

we just use ALCT (Anode Local C' irged Track) and CLCT (Cathode Local Charged

Track)tri -'._ .r primitives for AnodeSegment and CathodeSegment. These trigger primitives

(patterns of raw hits consistent with muons originating from the interaction point) are

already found by the i. i r-frontend" electronics and reported in the data stream together

with all raw data. All details on electronic architecture and functions can be found in

Muon Technical Design Report [65]. Up to two ALCTs (ALCTO and ALCT1) and two

CLCTs (CLCTO and CLCT1) can be found per each bunch crossing. If more than one

primitive is found in either projection, then we combine them combinatorially to form

2d-ProtoSegments, of which we can have 1, 2 or 4 per chambers. The two coordinates

of a 2d-ProtoSegment are ALCT's KeyWireGroup and CLCT's KeyHalfStrip. All in all,

building ProtoSegments this way requires virtually no CPU usage.









It is important to note that whenever two ALCTs and/or two CLCTs are reported

by the frontend electronics, they are alv-1-, the very best two muons candidates from all

possible hit combinations. From physics side, the chances of finding two prompt muons

within one chamber are very small. So, in fact, limiting ourselves to the best two muon

segment candidates is probably more of an advantage than a limitation as long as the

efficiency of finding the segment corresponding to the muon of interest remains high (see

below).

Also worthwhile stressing is that this approach does not introduce any potential

inefficiency as compared to any other algorithm that starts from raw data. By design, the

CSC readout in C'\ S is intrinsically zero-suppressed and one does not get any raw data in

the readout stream unless there was a trigger primitive found [65].

Efficiencies of finding trigger primitives were extensively studied in past on chamber

prototypes operating in a muon beam or cosmic rays, in a muon beam with a superimposed

flux of random hits [72], and with high energy muons accompanied by bremsstrahlung

radiation due to muons passing through an iron slab in front of a chamber [73]. More

recently, these efficiencies were studied in situ with 18 chambers installed in C' \S and

operating in cosmic rays [74]. In all these cases, the efficiency of finding tri r._. r primitives

was measured to be higher than 9,' For example, the latter studies gave an efficiency of

99.90.05'. for finding ALCTxCLCT 2d-patterns for muons passing through the chamber

sensitive volume.

6.2.2 RecHits and Their Coordinates

As was mentioned earlier, the default UF algorithm does not use any calibration

constants and does not access any databases. All information on the internal chamber

geometry needed for the local reconstruction is taken from the chamber drawings.

Below are a few notations and definitions to be used further in the chapter:


Noise associated with a single time sample (first time sample in CFEB readout) is
ai. Three samples added together have a spread of a3.









The averages of the first two SCA samples in CFEB readout, measured on
event-by-event and channel-by-channel basis, define pedestals for the current
event. Below in the section on the validation of the algorithm, we show that using
tabulated pedestals allows one to achieve an even better resolution, but at an
additional CPU cost.


CFEBCluster is defined as a 3x3 matrix = (3 strips)x (3 SCA samples) with the
central sample being a local maximum. For a cluster to be identified as such, the
amplitude of the central strip is required to be larger than kac (k is a configurable
parameter; by default, k=10). For each strip in a cluster, three time samples are
added together (pedestal subtracted). This way the 9-sample information in a
cluster is reduced to three charges: Qi, Q,, Qr (left, central, right). The cluster
charge is Q = Qi + Qc + Qr.


Throughout this chapter, noise, pedestals, and charges are expressed in terms of
ADC counts.

For each ProtoSegment present in the chamber, the following is performed.

Starting from ProtoSegment's coordinates (KeyHalfStrip, KeyWireGroup), the nearest

CFEBCluster in each plane within n strips around the KeyHalfStrip is identified (n is a

configurable parameter; by default, n=5). At the moment, clusters from different planes

are not required to line up in time. Strip signal charges Qi, Qc, Q, for such clusters are

calculated and used to build the ratio r defined as follows:


1 QT, Q1
2 Q, min(Qr, (Qi)
This ratio changes monotonically, but not linearly from -0.5 to 0 to 0.5 as the true

hit coordinate in strip width units changes from -0.5 to 0 to 0.5 (in points -0.5, 0, and

0.5, they coincide). Cluster's local coordinate in strip width units is then calculated via

a function x f(r, w) that corrects for non-linearity between r and x (see Appendix I

for details). The strip width parameter w in this function is derived from the chamber

geometry and ProtoSegment's coordinates (KeyHalfStrip, KeyWireGroup). Note that x,

being measured in strip width units, can be thought of as a Q-coordinate of a hit in units

of strip AQ-width.









The x-coordinate is then given an offset, according to the number of its central strip

and taking into account the 0.25 -1 I.-.- lii.-; of strip patterns in odd/even planes in the

large chambers.

It is important to note that, being built from charge differences, the ratio method is

intrinsically not very sensitive to cross talks variations. Another important feature is that

for hits close to strip edges, the ratio r basically becomes the ratio of two large charges on

nearby strips and has nearly no sensitivity to charge fluctuations on the third strip with

a very small share of the induced charge (and, thus, very little useful information) for

example, this is not the case for the center-of-gravity method.

By default, the UF algorithm assigns errors to reconstructed RecHits x-coordinates

according to the tabulated functions aox(CSC'ir,, HVsegment, x\) for five ranges of x|.

The values for this function were obtained directly from MTCC data. More details are

given further below where we discuss the algorithm validation.

The default option does not attempt to correct the errors in x for the charge in

a cluster. To take into account the charge, a good understanding of charge-dependent

and charge-independent contributions is needed. Note that at nominal gas gain, the

errors for x-coordinates in the area between strips, where the resolution is the best, are

already dominated by charge-independent contributions. Therefore, a charge-dependent

assignment of errors may improve performance somewhat, but the effect is not expected

to be dramatic. These studies are in progress. If they show that taking into account the

charge-dependent component gives better performance without a risk for the algorithm's

robustness, the found corrections can be easily included.

If CFEB data are not present, then the x-coordinate is given by the center of the

nearest half-strip with a comparator response. Again, as in the case with CFEBClusters,

the search is done in the range of n strips around the KeyHalfStrip. The error in this

case is defined as ao = 0.5//2 = 0.144.









RecHits are assigned local y-coordinates according to the center-of-gravity of anode

hits in the corresponding plane and falling within the envelope of the corresponding

AnodeSegment pattern. To calculate RecHit's y-coordinate in ME1/1 chambers with tilted

wires, one also needs to take into account RecHit's x-coordinate.

6.2.3 RefinedSegment

To refine the x-coordinate of the original ProtoSegment, we make a standard

non-iterative weighed linear fit of x-coordinates of all RecHits associated with the

segment. The returned parameters are an intercept segment at local coordinate z=0

(center of the middle chamber panel), with an estimated error on it, slope dx/dz with

an estimated error, and X2/d.o.f. The local xsegment-coordinate is directly related to the

Q-coordinate of the C' \IS coordinate system.

If x2/d.o.f.> X2t, we re-fit the line by leaving out one plane at a time from the fit.

The best fit gives new parameters for the RefinedSegment. Another round of such pruning

is allowed (with the same criteria on X2) as long as the number of remaining hits does not

fall below four, i.e. the final refined segment must have at least four RecHits associated

with it.

For a 2d-RefinedSegment, the local y-coordinate is improved (in comparison to

what one gets by taking plain KeyWireGroup) by taking a simple center of gravity of

actual hits belonging to the AnodeSegment pattern. The local y-coordinate can be easily

converted into the C'l\ S global coordinates r and T1.

Note that the refinement of 2d-segments means just that, refinement; i.e. the earlier

found 2d-segments do not get removed, nor are new segments added in this process.









6.3 Algorithm Validation at MTCC

During August-October 2006, a slice of the C'\ S Detector was used to take cosmic

ray data during the first activation of the C'\ !S solenoid. This important C'\ !S milestone

was named Magnet Test and Cosmic ('!i ,! i,', (\!TCC). Details on the MTCC scope and

the obtained results can be found elsewhere [75]. This section describes the performance of

the UF algorithm as obtained using the MTCC data.

Measurements of the processing time for the default UF algorithm were done on

an Intel P4 2.8 GHz Dual Xeon Server with the C\!iSSW version 1.2.0 as the default

framework. To benchmark the algorithm performance we used the C'\ S Global Run No.

4188. On average, the UF algorithm was found to take ~0.36 ms per matched LCT. This

time includes finding segments, reconstruction/selection of uncorrupted RecHits associated

with these segments, and refining segment parameters.

6.3.1 ProtoSegments

As was mentioned earlier, in the current version of the UF algorithm, the ProtoSegment

is defined as a matched pair of trigger primitive patterns, ALCTxCLCT. Efficiency of

finding such correlated LCTs and their properties were extensively studied elsewhere [74].

Here we just summarize the main results from those studies:


For muons going through a fully sensitive chamber area, the efficiency of finding a
correlated LCT was measured in data to be 99.90.05'.


The highest quality correlated LCT found in a chamber was near the muon position
predicted from external chambers; the distribution spread of less than 2 cm (RMS)
was consistent with multiple scattering of muons.

Therefore, the MTCC data confirm the earlier results showing that ALCT and CLCT

patterns provide a very robust method of identifying muons. In contrast to the earlier

results obtained in beam tests with only a small portion of a chamber illuminated by

muons, the MTCC studies were done with cosmic rays probing nearly the entire area of 8











25000 =550
S40 -
sigma= 5.6
120- 20000-

100
15000 -
80-

60- 10000-

40-
5000-
20-

-25 -20 -15 -10 -5 0 5 10 15 20 25 200 40 800 10001200140016001800200
Sum of three time samples, ADC counts Quick coordinate Landau, ADC counts


Figure 6-1: The noise distribution for Figure 6-2: The three-strip cluster
a three-sample-sum in the absence of a charge distribution for the largest
signal for the largest ME23/2 chambers. ME23/2 chambers.


large chambers operated in situ together with the rest of the C'\ !, detectors participating

in the MTCC.

6.3.2 Spatial Resolution

This sub-section summarizes results of the studies of the spatial resolution attainable

with the ratio method in the context of the CMS Cathode Strip C'i ,inl>ers.

A signal-to-noise ratio is among the key parameters affecting the chamber performance.

As described earlier, a strip charge in the UF algorithm is defined as a three-time-sample

sum. Figure 6-1 shows the three-sample sum distribution in the absence of a signal-the

Gaussian sigma of this distribution is g3~5.5 ADC counts. Figure 6-2 shows the Landau

distribution for 3-strip charge clusters. The average charge is around 550 ADC counts.

Therefore, the MTCC data were taken with the gas gain corresponding to signal-to-noise

ratio of approximately 100:1.

Figure 6-3 shows the single-sample noise level. It has a spread al-3.0 ADC counts.

It is worthwhile noting that a3 > /-3al, which clearly demonstrates the expected noise

correlations between time samples.


























Figure 6-3: The noise distribution for a Figure 6-4: The noise distribution for
single-sample in the absence of a signal a two-sample sum in the absence of a
for the largest ME23/2 chambers. signal for the largest ME23/2 chambers.


Figure 6-4 shows the two-sample sum distribution in an absence of a signal-the

Gaussian sigma of this distribution is a2y4.6 ADC counts. The two-sample sum is of

special interest for further considerations. The UF algorithm assumes two modes of

operation. First, default, mode does not use any calibration constants, including pedestals.

The pedestals for individual channels are evaluated on event-by-event basis from the

first two samples, where signal is not present: ped = samplel + sample2)/2. Then,

such dynamically defined pedestal is subtracted from all three samples used to form a

three-sample-sum charge. This procedure contributes an additional error in determination

of the charge: (a2/2) x 3 = 6.9 ADC counts. Combined with a3 = 5.5, the total error in

charge measurements becomes 8.8 ADC counts, which is 1.6 times worse than 5.5 ADC

counts one could have, if tabulated (calibrated) pedestals were used. This is a substantial

penalty for not using calibrated pedestals. Therefore, we single out pedestals from all

calibration constants and implement a second mode of the UF algorithm that does use

pre-defined pedestals. UF algorithm takes ~0.45 ms per matched LCT with this option

ON.









6.3.2.1 Resolution for high-pr muons


The best chamber resolution matters only for high-pr muons whose track stubs are

nearly perpendicular to cathode strips. Softer muons undergo large multiple scattering

and a good detector resolution is less important. To evaluate the chamber resolution

performance in this context, we select events according to the following criteria:


We require only one ALCT and only one CLCT per chamber. This allowed us to
extract the spatial resolution parameters intrinsic to the chambers themselves and
not obscured by presence of showers.


CLCT patterns are required to be half-strip patterns only (di-strip patterns are
invoked for finding highly inclines muons in the absence of half-strip patterns).


Also, to stay away from edge effects, we require that the ProtoSegment's KeyWireGroup
is at least one wire group away from chamber edges and dead areas separating HV
segments, and KeyHalfStrip is at least 2 strips away from the chamber sides.


We require six charge clusters to be associated with a ProtoSegment, which gets rid
of incomplete tracks.


The resolution is evaluated via residuals between a hit coordinate in the 3rd plane
and the track coordinate in this plane as predicted from a straight line fit of hits in
the remaining 5 planes (1, 2, -, 4, 5, 6). These residuals give a conservative estimate
of the chamber resolution as they include track prediction errors. We do not attempt
to correct for this effect (it can be easily estimated to result in an extra factor as
large as 1.2 for a single-plane resolution in the ;ood" strip areas).


We expect a few percent of 6-electrons per plane. The fraction of dramatically
corrupted 5-plane fits due to 6-electrons in at least one plane is consequently
is 5 times larger. Therefore, to make sure that the reference prediction is not
dramatically compromised, we cut events whose 5-plane fit X2/dof X2/3 > 10.
This removes about "-'. of tracks. Since typical 6-electrons do not have enough
energy to penetrate from plane to plane, this cut does not bias the measurement in
the test plane number 3. It rather insures that the reference prediction for the hit
position in the 3rd plane is self-consistent.

The resolution per plane is analyzed and reconstructed for five regions across a strip

width: 0-0.1 (strip center), 0.1-0.2, 0.2-0.3, 0.3-0.4, and 0.4-0.5 (strip edges). To obtain









the final resolution values for these five strip regions, we perform the procedure described

above iteratively. The convergence is reached in just a few iterations.

This analysis is performed for each chamber HV segment separately and, therefore, at

the end, the chamber resolution is tabulated as a(CSC' Ul, HVsegment, Ix|), where there

are 5 distinct ranges for Ixl: 0-0.1, 0.1-0.2, 0.2-0.3, 0.3-0.4, and 0.4-0.5.

Figure 6-5 shows residuals for these five distinct ranges for the middle part (HV

segment 3) of the largest ME23/2 chambers. The last plot shows a summary for

a single-plane resolution a, vs. i-th Ixl-part of a strip (i=1 for 0< xl<0.1, i=2 for

0.1< x <0.2, ..., i=5 for 0.4< x <0.5).

These aj's allow one to evaluate the overall six-plane (full-chamber) resolution as

follows:


1 3 3
2 2+2 (62)
csc,i i 6-i
Figure 6-6 shows single-plane and full-chamber resolutions for all HV segments for

ME23/2 chambers. For completeness, the results are shown both in units of strip widths

and microns. Figure 6-7 show similar results for another distinct chamber type: ME1/1.

Clearly, the obtained resolutions surpass by far the HLT goal of 0.5 mm or so. To

remind, the presented results do not use any calibration constants. Nor do we correct for

internal plane mis-alignment, which, from the FAST site measurements, is estimated to

contribute about 50 pm RMS per plane [76].
















0. l sigma=0.0503





40 -




A -03 -0.2 -01 -0 0.1 0.2 0.3 0.4 0
3d p lane reaid usl, strip width unit



20- .


0,3<1xl<0.4
S40 sigma=0 0326
20




40
20 -
-4A -3 40.2 -0.1 -0 0.1 0.2 0.3 0.4 05
3id plane reaidusala, trip width unit


0.1
= 0.09


S 0.07
0.06
-0.05
" 0.04


g 0.02
00
0.01


-u1 -.0 -u01 -0 L a 2 ut 3 u 4 cA
3rd plane reaiduals, strip width unit


---~~ ~ 1' I--- I--r *-l---- l- l-

ME23/2 chambers,
3rd HV segment
-



- *-
*


0 0.05 0.1 0.150.2 0.25 0.3 0.35 0.4 0.45 0.5
Track coordinate, strip width units


Figure 6-5: The 3rd-plane residuals for the five distinct xl strip part ranges for the middle
part of the largest ME23/2 chambers. And a summary for a single-plane spatial resolution
as a function of a hit position across a strip.



73


3rd plane residuals, strip width units


o 0,4<|xl<0.5
sigma=0 0269





S_0 -


-Ua -oA






















- -K


A X
V


o i





3 0.1 0.2 0.3 0.4 0.5
Track coordinate, strip width units


10 0.1 0.2 0.3 0.4 0.5
Track coordinate, strip width units


0.1

(1
>a,
cu 0.08
CL
c 0.06
o

"- .
U) 0.04
a,
(D


Figure 6-6: The 3rd-plane residuals o's vs. five ix strip part ranges and overall six-plane
resolution vs. |x|-part of a strip (in units of strip widths and microns). [Notations are:
black empty squares HV segment 1; red squares HV segment 2; triangles up HV
segment 3; triangles down HV segment 4; stars HV segment 5.]


4- -


3


2 -




. I .. I . I '4 . 0
0 0.1 0.2 0.3 0.4 0.5
Track coordinate, strip width units





o -



D -
0








0 0.1 0.2 0.3 0.4 0.5
Track coordinate, strip width units


E 1500
=L
>,
CU


o o
0
cn
O 500
I2


S.


A v V
A

1PS a






















0.1

a
( 0.08

a
0.
o 0.06


M 0.04


0.02


0o



e1500







a



500


40.05


L0.04




0.03

S0.02


U
., .l I , l, .


. .. .. .: : ............










0.1 02 0.3 0.4 0.5
track coordinate, strip part

















I. . . I .


i i i i i l l i l i i 1 i i t i l i i i .
"o 0.1 0.2 os 0.4 o.s
track coordinate, strip part


00 0.1 0.2 0.3 0.4 0.5 0 0 0.1 02 0.3 0.4 0.5


track coordinate, strip part


Figure 6-7: The 3rd-plane residuals a's vs. Ix/wl-part of a strip and overall six-plane
resolution vs. Ix/wl-part of a strip (in units of strip widths and microns). Same notations
as for Fig. 6-6.


E ouu
WU
C ,,

600
a


5 400


. . .. ...




- I. ................ I. I ...........
-


S* . ..


track coordinate, strip part









6.3.2.2 Resolution for high-pr muons with tabulated pedestals

Measuring pedestals dynamically (on event-by-event basis) allows one to not use

calibrations and would help at very high rates when pedestals are expected to float.

However, as it was mentioned earlier, these benefits come with a penalty-the electronic

noise results in errors in the measured pedestals, which makes a significant contribution to

the final errors in charge measurements.

To evaluate the effect quantitatively, we measured pedestals directly from data and

used them in the analysis. The 3rd-plane residuals obtained with tabulated pedestals are

then used to evaluate the full six-plane resolution for ME23/2 chambers. The results are

shown in Figures 6-8, 6-9. The left plot is for dynamically measured pedestals, the right

plot is for tabulated pedestals. The gain in resolution is quite tempting. As long as the

rate of hits in chambers is not too high, one can certainly take advantage of this option,

especially in the offline analysis.

As far as the HLT is concerned, the improvement in resolution, however impressive

it might be, is not critical (the resolution with dynamic pedestals is already better than

needed). Nevertheless, the option of using tabulated pedestals certainly can be used for

the HLT as well, as long as the CPU penalty (yet to be measured) due to the need to

manage about 200K array of constants is acceptable.














5Rn ..


600


400


0 0.1 0.2 0.3 0.4
Track coordinate, strip width


E buu
B


CC
U-

O

"* 400

rE


0.5
units


Figure 6-8: The overall six-plane
chamber resolution vs. five Ixl strip
part ranges for dynamically measured
pedestals. Same notations as for
Fig. 6-6.


E 800
0"
O

Cl)
Bo

- 400

CC
oc


1000


-0.5 0 0.5 1 1.5
Track inclination, radians


I Ix


ta I i a

I,... 1


0.1 0.2 03 0.4 0.5
Track coordinate, strip width units


Figure 6-10: Inclination angles for track
segments reconstructed from di-strip
CLCTs (M\!1: /1 chambers).


Figure 6-11: The spatial resolution
per six-plane chamber for tracks
reconstructed from di-strip CLCTs
(\!1- !/1 chambers) for five Ixl strip part
ranges. Same notations as for Fig. 6-6.


0 0.1 02 0.3 0.4 0.5
Track coordinate, strip width units



Figure 6-9: The overall six-plane
chamber resolution vs. five xl strip
part ranges for calibrated pedestals.
Same notations as for Fig. 6-6.









6.3.2.3 Resolution for low-pr, or highly-inclined, muons

Very low pr tracks, will have a noticeable angle (due to magnetic field and multiple

scattering) of inclination dx/dz. For example, muons with pr~3 GeV/c at r =1.6 (just

barely making into ME1/1 chambers), will be swept -id. i,-i-v by magnetic field B and

would go through ME1/1 chambers at an angle of acB0.35 radians. For other stations

and for muons with larger PT and/or at higher pseudo-rapidities, the angle aB is smaller.

In the current version of CLCT-finding firmware, such "highly-inclined" muons are

captured by CLCT di-strip patterns. The resolution for such segments is going to be

worse due to smearing of the ionization charge along anode wires. But the requirements

on the resolution for soft muons are also less stringent due to much larger multiple

scattering (in comparison to benchmark muons of PT=20 GeV/c that we have used so far).

To evaluate the spatial resolution attainable with the UF algorithm for "highly-inclined"

muons, we selected ProtoSegments based on LCT (di-strip CLCT) xALCT in ME23/2

chambers. This is a conservative estimate as the effect of worsening in ME1/1 chambers

will be smaller. The results are shown in Figures 6-10, 6-11. The left plot shows angles

for track segments reconstructed from di-strip CLCTs. One can see that the range is

much wider than expected for a few GeV pr muons. The right plot is the resolution

per six-plane chamber-clearly, the resolution is still much better than a few-millimeter

resolution needed for such soft muons.









6.3.2.4 Resolution for high-pr muons with charge-dependent sigmas

An additional 1 rv.r of sophistication can be added to the RecHit reconstruction by

adjusting the errors on RecHit coordinates a(CSC',l,.- HVsegment, Ixl) for the actual

charges in clusters. Before implementing such an adjustment, one must perform detailed

studies of the relative interplay of different sources of errors in the hit reconstruction. To

name a few, among these contributions are electronic noise (improvement in resolution

with charge as ~1/Q), 6-electrons (degradation of resolution for too large charges), and

a number of constant terms (place-to-place variations and event-to-event fluctuations, in

induced charge shape, mis-calibrations, mechanical plane misalignment, etc.),

Given that the spatial resolution that we obtain without any of these corrections is

already much better than what is needed for HLT, these corrections, once properly tuned,

would make much more sense for offline analysis than for HLT. Of course, they can be

used at HLT as well.

6.3.3 RefinedSegments

As was outlined in section on the algorithm principles, a 2d-RefinedSegment is

basically a 2d-ProtoSegment, whose parameters have been refined using the precise

coordinates of RecHits. The last touch in building RefinedSegments is pruning of

seemingly "bad" RecHits. This is done by identifying and throwing away one-two "bad"

RecHits that make dramatically bad contributions to a linear fit X2. We derive the X2-cut

criteria from MTCC data as follows.

We select ProtoSegments based on LCT (half-strip CLCT) x ALCT in ME23/2

chambers and suppress complicated events with showers by requiring that there is only

one ProtoSegment per chamber. To stay .,.- li from edge effects, we require that the

ProtoSegment's KeyWireGroup is at least a unit away from the chamber edges and the

borders separating HV segments, and KeyHalfStrip is at least 2 strips away from the

chamber sides. The chamber ME2/3/28 is excluded form further analysis as it has one HV

segment switched off.











E ~25000 -
|0.9
-0.8
-_ --20000-
0.7
0.6 5
15000-
0.5
0.4 10000-
0.3
0.2 5000
0.1

1 2 3 4 5 6 2 3 4 5 6
Number of hits per segment CSC plane


Figure 6-12: Probabilities for different Figure 6-13: L ,-- r charge clusters
numbers of charge clusters found in occupancy in association with ME23/2
association with ME23/2 chamber chambers ProtoSegments.
ProtoSegments.


Figure 6-12 shows how many charge clusters are found on such ProtoSegments. Note

that by the UF algorithm design, only one cluster per plane can be associated with a

given ProtoSegment. Figure 6-13 shows an occupancy of a particular plane from 1st to 6th

with charge clusters.

Figure 6-14 shows X2/dof distribution for a linear fit for events with six, five, and

four clusters associated with a ProtoSegment. The distributions have a long tails due to

6-electrons and other possible sources of hit corruption (e.g., showers).

Figure 6-15 shows a scatter plot of min(xj/dof) vs. X2/dof for events with six-cluster

ProtoSegments. Here, X /dof is obtained for a six-plane fit, min(x /dof) is a minimum

between 6 five-plane fits with one plane dropped from the fit. The choice of cuts for

pruning bad hits is shown by dashed lines. Probability of having 6-hit, 5-hit, and 4-hit

RefinedSegments is then 7-' -, 1;:'. 9'. correspondingly. We do not prune bad hits in

segments with four remaining hits.

Figure 6-16 shows that the occupancy for x l-coordinates for RefinedSegments remains

flat, which demonstrate that the pruning of hits does not bias segments. Figure 6-17












4@00 90
35000 ProtoSegments with 6 hits 9 80
S -- ProtoSegments with 5 hits
30000 ProtoSegments with 4 hits 1 70
E
25000- 60
'550
20000 E
S=40
15000
S30
10000- 20
20 -
5000 10

` 10 20 30 40 50 60 70 80 90 10C 00 10 20 30 40 50 60 70 80 90 100
ProtoSegment track fit X2/ndf X2/ndf for 6-plane fit


Figure 6-14: The X2/dof distributions Figure 6-15: Scatter plot of min(xj/dof)
for a linear fit for ME23/2 chamber vs. x I/dof for six-cluster ME23/2
ProtoSegments with six clusters, five chamber ProtoSegments.
clusters, and four clusters.


shows pools for all retained RecHits (i.e., RecHits associated with RefinedSegments. One

can see an obvious reduction in tails in comparison to distributions shown in Fig. 6-5.

The final optimization of the X2 cuts can be done when a good reference for a whole

segment is available (e.g., in detector Monte Carlo simulation, provided that Monte

Carlo is shown to reproduce data at the adequate level of details). In principle, more

sophisticated additional criteria for pruning can be further empl-o. d- e.g., one can

take into account the charge cluster shape and number of anode hits associated with a

RecHit. These were studied in the past and shown to have some, albeit very limited,

discriminating power against "bad" hits. However, their use for the HLT purposes is

hardly justifiable.

6.4 Conclusions

A new fast algorithm for reconstructing track segments in Cathode Strip ('! i,,hIers

is proposed. The algorithm was validated with the real cosmic ray data taken with 36

CSC chambers operated as a part of C'\ S-wide MTCC test program in the second half





























-0.4 -0.3 -0.2 -0.1 -0 0.1 0.2 0.3 0.4 0.5
RefinedSegment x-coordinate, strip width units


sigma = 1.24










. .


20 -15 -10 -5 0


5 10 15 20
Pool, A(x)/o


Figure 6-16: Occupancy for
|x/w -coordinates of RefinedSegments.


Induced Charge Density


0.05


0
-15 -10
-15 -10


-5 0
x (mm)


Figure 6-17: Pools for all retained
RecHits (i.e., RecHits associated with
RefinedSegments.)
0.5 1


5 10 15


Figure 6-18: Induced charge distribution
calculated according to Gatti for large
and ME1/1 chamber geometries.


0 0.1 0.2 0.3
(XlW)true


Figure 6-19: Ratio r versus a local
coordinate x for large chambers
calculated for variety of strip widths
in the assumption of the Gatti charge
distribution for large chambers.


4000

3000


1000

.5


0.4 0.5









of 2006. The algorithm proved to provide high speed, high efficiency, and good spatial

precision-all well within the High Level Ti 1;-:. r requirements.

6.5 Ratio Method

After charges Qi, Qc, Q, on left, central, and right strips are measured, their ratio is

built as follows:


1 Q- Q-
S= Q Q (6-3)
2 Q, min(Qr, ,Q)

Figure 6-19 shows ratio r as a function of a local coordinate x calculated for the

induced charged distribution according to Gatti et al. [77] (Fig. 6-18). The local

coordinate x is assumed to be in strip units, x=0 corresponds to the strip center, and

x = 0.5 means right/left strip edges. This ratio r is a monotonic, but not linear function

of a hit position across a strip x

We find the conversion function from r to x in two steps.

The first correction is an approximate inversion of the I !, i I function r(x, w).

Figure 6-20 shows a correction that one needs to add to r to obtain the coordinate x.

The points correspond to the I I, l I I" Gatti function. We find that the 1st-order

correction can be parameterized quite well with the following empirical function (w is a

strip width in cm):



g(r, w)= (. ) (6-4)
a/wb+clrl'
where a=0.27 (0.11), b=2.7 (2.9), c=1.25 (1.25) for large (and ME1/1) chambers.

After applying this correction, we get the 1st-order corrected coordinate x1 =

r + g(r, w) is expected to be within 1 of the true coordinate x-see Fig. 6-20. This is

already sufficient for the HLT purposes. However, this correction is purely theoretical and

must be checked against the reality.

Figure 6-21 shows the experimental occupancy distribution dN/dxl. It has an obvious

,., ', which is a manifestation of the fact that the induced charge does not quite follow











w=b Uatl w=b approximaton 2h = 95 mm )
Sw=8 Gattl w=8 approxmaton 00
02 -- = ." .
5000

r 01




S-0.1 -- --.--- 2000-

-0 2 1000
T0 r C1'T100 0

-03 00.05 1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
0 0.1 0.2 0.3 0.4 0.5 1st-order corrected coordinate x,
Ratio r

Figure 6-20: Correction that must be Figure 6-21: An occupancy distribution
added to the measured ratio r to obtain for the 1st-order corrected coordinate
1st
the hit position across a strip.


the I !I I ical" Gatti. Curiously enough, the shape of the wave looks very similar for

all parts of all chambers, i.e. it is very similar for different strip widths. This allows us to

introduce a second-order empirical correction in a strip-width independent manner. This

correction can be derived directly from the shape of the dN/dxl occupancy distribution

and can be parameterized as follows:



X2 + C1Cb + C2b + Co, (6-5)


where b1 = -11, b2 = -6.5, ci 0.070273, c2 -0.072769, and co = -(ci + c2). The

second-order correction x2 x1 is actually quite modest (see Fig. 6-22). However, it does

make the occupancy plots for x2 almost perfectly flat.

The sensitivity of the ratio method to the typical electronics noise, calibration errors,

and cross-talk uncertainties are shown in Figs. 6-24, 6-25, 6-26. The typical values used

to make these plots are the results of pre-installation testing of all 396 large chambers at

so-called Final Assembly and System Tests sites [76].










0.01 --




---- '** 4000

-0.005
5000



-0.015

1000
-0.02
0 0.1 0.2 03 0.4 0.05 0. 15 0.2 0.25 0.3 0.35 0.45 0.5
1 t-order corrected coordinate 2nd-order corrected coordinate x2

Figure 6-22: The second-order correction F r An oc y
Figure 6-23: An occupancy distribution
that must be added to .x' to obtain the ,
that must be added to x to obtain the or the 2nd-order corrected coordinate
hit position across a strip x2.


The typical noise is ~1 of the average cathode cluster charge, when chambers

operate at the nominal gas gain. In addition to the FAST site measurements, the noise

levels were again measured in situ during data taking for all installed chambers and found

to remain unchanged. If one does not use calibration constants, the spread of electronic

channel gains between nearby strips is found to be ~1 The differences in crosstalks

between strips for chambers of the same type are very small; the number used for

making Fig. 6-26 represents the RMS value for all crosstalks collected in one histogram

regardless of the chamber type.

One can see that channel gain calibrations and crosstalks are the least of our

concerns. The noise contribution to the resolution between strips is also very small.

These figures allows one to gauge how electronics performance may contribute to the CSC

spatial resolution when the ratio method is used.


















6%
2h = 9.5 mm I- w=1
QstripQcluster=0.01 w=14
5% -- w=12
w=10
II w=8



20 % --.----------- ------- .---- - =
C \
1 4%, --- ... .................... ..............-......
* 3%





1%


0%
0 0.1 0.2 0.3 0.4 0.5

(x/W)t,ue


0 0.1 0.2 0.3 0.4 0.5

(xiW)true


Figure 6-24: Sensitivity of the ratio method to electronic noise (in of strip width and

mm).


1.0%


C
S0.8%



0.6%



S0.4%
0
"a
-i
E 0.2%



0.0%


0 01 0.2 0.3 0.4 0.5

(X/W)true


0.10
2h = 9.5 mm --w=16
8Gain/Gain=0.01 -w=14
-,-w=12
.0
0.08 ----------------- -------------- ------- ---- -----------

-o-w=6




* 0.04
S0 .0 6 ----- ------ ---- : ----- ------ ------- ------ ------ ------ ------ ------ ----------------






E 0.02
E ) 0 .0 2 -- -- -------...-.. .. .. .





0.00
0 0.1 0.2 0.3 0.4 0.5

(X/W)true


Figure 6-25: Sensitivity of the ratio method to errors in electronic gain calibrations (in

of strip width and mm).









86


2h = 9.5 mm --w=16
8Gain/Gain=0.01 w=14
-, w=12
------------- ----------------- -------------- ---*-w=10
Sw=10
-- w=8


w6




------------ --------
=""' 7





































1.0% 0.10
2h = 9.5 mm --w=16mm 2h = 9.5 mm --w=16 mm
6(crosstalk) = 0.0214mm (crosstalk) = 0.02 w=14 mm
-w=12mm a -, w=12 mm
S 0.8% ........................................
o 0.8% .w=lO mm 0.08 ............ ........ --w=lO mm
S-- w=8 mm -a w=8 mm
Sw=6 mm > --o w=6 mm

S0.6% 0.06






00
.2 0.4% 0.04







0.0% 000
0 0 0 0 0.1 0.2 0.3 0.4 0.51 2 3 0
True Coordinate (x/w) True Coordinate (x/w)


Figure 6-26: Sensitivity of the ratio method to uncertainties in cross talks between strips
(in of strip width and mm).









CHAPTER 7
SEARCH STRATEGY FOR THE STANDARD MODEL HIGGS BOSON IN THE
H -+ ZZ(*) -+ 4/ DECAY CHANNEL USING M(4p)-DEPENDENT CUTS.

7.1 Introduction

The H -+ ZZ(*) -- 4P process is one of the cleanest channels (also known as a

sold-plated" channel) for discovering the Standard Model Higgs boson at the LHC. In

this chapter, we outline a complete analysis strategy for discovering the Standard Model

Higgs boson in the H -+ ZZ(*) -- 4 channel. The explored range of Higgs masses is

115-600 GeV/c2.

The cuts, smooth functions of the four-muon invariant mass M(4p), are such that at

whatever unknown a priori mass the Higgs boson might appear, the signal-to-background

ratio is already optimal to give the best chance of discovering it. This allows one to

avoid a posteriori cut optimization. We give a direct comparison of results obtained with

MA(4p)-dependent (dynamic) and MA(4p)-independent (flat) cuts.

The search for the Higgs boson 4/p resonance-like peak can be done using the

Log-Likelihood Ratio (LLR) [78, 79, 80] built for the entire MA(4p)-distribution, binned

or unbinned, or taking a straightforward counting experiment approach. We give a direct

comparison of the two approaches.

A full treatment of the most important theoretical and instrumental systematic

errors and their effect on the evaluation of the significance of the Higgs boson observation

are presented. To minimize systematic errors, new methods of reconstructing the most

important corrections directly from data were developed. Among them are the muon

reconstruction and isolation cut efficiencies. We also show that by using the measured

Z -- 2/p cross section, or an event count in the sidebands of the M((4p) distributions, one

can substantially reduce a number of theoretical and instrumental systematic errors.

In addition we verify by how much the local excess significance will be effectively

degraded due to the fact that we look for a narrow resonance in a broad range of M(4p)

invariant masses.









The results are obtained with the official full C\ S detector simulation and reconstruction

software [81, 82] and include pile-up events corresponding to an instantaneous luminosity

of 2 x 1033 cm-2 -1

The final results are presented in terms of the required integrated luminosity for

observing the Standard Model Higgs boson at 5a and 3a significance levels and 95'.

CL exclusion limits. Also, we present the significance for a fixed value of an integrated

luminosity equal to 30fb-1 and 95'. CL exclusion contours in the (MH, a) plane for

integrated luminosities of 3, 10, and 30 fb-1.

Previous studies on the search for the Standard Model Higgs boson in the H -

ZZ(*) 4z channel with C'\!S are described in [83, 84, 85]. Another ongoing study

exploring the discovery potential with a different set of mass-independent cuts can be

found elsewhere [86]. The results of the two parallel analyses using the H -- 4e and

H -- 2e2p channels can be found in [87, 88].

In short, we believe our H -- ZZ(*) -- 4p analysis is more realistic than previously

existed ones on the topic:


we used calibration from data techniques, as we would do with real data;


full treatment of systematic errors was included and folded into signal-to-background
significance calculations;


dedicated effort was done to generate physics of signal and background processes
properly, including use of dedicated matrix element generators, events re-weighting
with dynamic Next-to-Leading Order (NLO) corrections;


latest available at the moment full simulation and reconstruction validated software
was used;


C \!S performance was optimized for all allowed SM Higgs boson masses;









and as a result our analysis is part of the CMS Physics Technical Design
Report [7, 8], i.e. is an official CMS strategy for the SM Higgs boson
discovery in H -- ZZ(*) -- 4p decay channel.

Details of our analysis and supporting studies can be found in a list of refereed papers

[1, 2, 3, 4, 5, 6, 7, 8, 9, 10], in all of which the author of this thesis was either the leading

author or one of the leading co-authors. Results of our work presented at 10 international

conferences [11, 12, 13, 14, 15, 16, 17, 18, 19, 20]; the author also gave more than 70 talks

at the internal C'S\ S collaboration meetings of different levels [21].

7.2 Physics Processes and Their Simulation

The Higgs boson event samples for 18 Higgs boson mass points (see Table 7-1) and

the three main background processes, tt, (Z/7*)bb, and Z/7*Z/7*, were simulated using

the full C' \! detector simulation and reconstruction software. Many other plausible

background candidates, bbbb, bbcc, cccc, single-top, Zcc, Wbb, Wcc, fake, and 7r/K decay

muons in QCD, were considered and found to be negligible.

To save CPU time, only events with at least 2p+ and 2p- in the pseudorapidity

range |\ < 2.4 and with pr >3 GeV/c were retained for further analysis. Muons outside

these kinematical limits could not be reconstructed by C'\ S. Additional cuts were applied

to di-muon invariant masses for the Higgs boson samples (m(Z) > 5 GeV/c2) and

for Z/7*Z/7* and (Z/7*)bb samples (m(pp/-) > 5 GeV/c2). (The first P+P- pair in

Z/7*Z/7* and (Z/7*)bb samples was defined as the one with its invariant mass closest

to m(Zo), while the second p+p- pair was made out of the two remaining highest pr

muons of opposite signs.) All analysis cuts on these observables, to be described below,

are much more stringent than these generator-level preselection cuts. The expected

numbers of surviving 4/z events for signal and backgrounds for an integrated luminosity of

L = 30 fb-1 are given in Table 7-1. The M(4p) distribution of events after these cuts is

shown in Figures 7-1 and 7-2.









Table 7-1: The LO/NLO cross sections for various Higgs boson masses and backgrounds,
corresponding number of events with four muons surviving the generator level preselection
cuts (see section 7.2) calculated for L 30fb-1, and the number of simulated events.

Process aLO, pb (NLO, pb 4/p events at L=30 Simulated Events
pp H (mH = 115) 47.7 7.69 10000
pp H (mH = 120) 44.3 13.6 10000
pp H (mH = 130) 38.4 31.1 9000
pp H (mH = 140) 33.7 49.2 10000
pp H (mH = 150) 29.8 54.1 9000
pp H (mH = 160) 26.6 25.6 9000
pp H (mH = 170) 23.9 12.3 10000
pp H (mH = 180) 21.6 28.5 9000
pp H (mH = 190) 19.7 101 10000
pp H (mH = 200) 18.0 109 10000
pp H (mH = 250) 12.4 87.5 10000
pp H (mH = 300) 9.58 72.3 10000
pp H (mH = 350) 9.12 72.6 9000
pp H (mH = 400) 8.81 63.4 9000
pp H (mH = 450) 6.44 45.1 10000
pp H (mH = 500) 4.46 31.8 10000
pp H (mH = 550) -3.07 22.6 10000
pp H (mH = 600) 2.13 16.3 10000
pp tt 840 7000 92236
pp -- (Z/7*)bb -- 2pbb 116 278 8694 124500
pp Z/7*Z/7* 4p 0.113 see text 2622 118000
pp Z/7*Z/7* 2p27 0.157 see text 48.8 10000


7.2.1 Signal: H ZZ(*) 4/

The Higgs boson samples were generated with Pythia 6.225 [89] (LO gluon and

weak-boson fusion), interfaced via Ci\ ilN [90] version 310 (PDF CTEQ5L). Only decay

channels Z 21 (where I stands for e, p, and 7) were considered. Z qq 21 were

not included in the simulation: being very similar to the (Z/7*)bb background, those

events would be suppressed together with the (Z/7*)bb background by our analysis cuts.

QED radiation from the final-state muons is modeled with PHOTOS [91, 92]. Events

were re-weighted to correspond to the total NLO cross-section a(pp H) BR(H -

ZZ) BR(Z 21)2, where a(pp H) and BR(H ZZ) were taken from [93, 94] and

BR(Z -2 21) 0.101 [95].









There exists an additional enhancement to the cross section for H -4 4/ (in

comparison to a(pp H) BR(H -+ ZZ) BR(Z -- 2/)2) due to interference of

permutations of muons originating from different Z's. The corrections are calculated with

CompHEP and presented in Figure 7-3.

The M(4p) distribution for mH = 140 GeV/c2 after event generator-level cuts is

shown in Figures 7-1 and 7-2. The low-mass tail is mostly due to events where muons did

not come directly from Z-decays (e.g., via r-decays) and to internal bremsstrahlung that

also tends to move the 4-muon invariant mass off the peak.

7.2.2 Background: tt

The tt sample was generated with Pythia 6.225 (LO gg -i tt and qq -i tt), interfaced

via Ci\IIN version 110 (PDF CTEQ5L). Only the decay channels t -i Wb -- Ivb

were considered. Events were re-weighted to correspond to the total NLO cross-section

a(pp tt) BR(W -- lv)2, where a(pp -- tt) = 840 pb was taken from [96] and the
branching ratio BR(W Iv) = 0.320 [95].

7.2.3 Background: (Z/7*)bb -+ 2pbb

The (Z/7*)bb --i pp-bb samples were generated with CompHEP 4.2pl [97] matrix

element generator (PDF CTEQ5L, with QCD scales pR = pp = Mz, b-quark mass

mb = 4.85 GeV/c2, and a di-muon mass cut m(ptp-) > 5 GeV/c2), interfaced to PYTHIA
6.225 for showering and hadronization. Included sub-processes were qq/gg Z/7*bb -

p +fibb, where q could be any of the light quarks, (u, d, s, c) (initial states with b quarks
were also considered at the generator level and found to be negligible). No restriction

on b decays was applied. The corresponding CompHEP LO cross section was found to

be 116 pb. To obtain the NLO cross section given in Table 7-1, we calculated the NLO

K-factor using MCFM [98]: KNLO = 2.4 0.3. The conditions for the MCFM NLO and

LO calculations were as follows: CTEQ6, pR = s, nb = 0, M(Z..) > 5 GeV/c2

pr(b) > 5 GeV/c, |b| < 10, M(bb) > 10 GeV/c2.









7.2.4 Backgrounds: qq Z/7*Z/7* -+ 4p and qq Z/7*Z/7* -- 2p27-

These two event samples were generated with CompHEP 4.2pl matrix element

generator (PDF CTEQ5L, with QCD scales pR = F = s, and the q quark could be u,

d, s, c or b). The direct muons from Z/7*-decays were required to have pr > 3 GeV/c

and Irq(p)1 < 2.5. The direct r's from Z/7* decays were required to have pr > 3 GeV/c

and d -. '1-, normally. Both t- and s-channel diagrams were included. The s-channel

diagram, not available in PYTHIA, gives a large peak at M(4p) = Mz, contributing about

10'. to events with 120 < M(4p) < 180 GeV/c2, and can be safely neglected for higher

4/p invariant masses; see [2] for details. The interference between t- and s-channels was

found to be alvi.-, negligible. The CompHEP events were further interfaced to PYTHIA

6.225 for showering and hadronization. The CompHEP LO cross sections for the two

sub-processes were 113 and 157 fb, respectively.

To account for contributions to all the NLO diagrams and to the NNLO gluon

fusion process (gg ZZ, known to contribute 2 211'. with respect to the LO [99]), we

re-weighted events with a M(4p)-dependent K-factor K(M4,) = KNLO(M4,) + 0.2. The

NLO K-factor KNLO(M4,) was obtained with MCFM [98] and is shown in Figure 7-4. All

details on calculation of this M(4p)-dependent K-factor and the dynamical differences

between NLO and LO are summarized elsewhere [6].

The M(4p) distributions after generator-level cuts are shown in Figures 7-1 and 7-2.

The peak at Mz is due to the s-channel contribution. This peak sits on the shoulder of

the enhancement around M(4p) > 100 GeV/c2-this peak corresponds to one of the two

Z's going on-shell in the t-channel. The next bump around 2Mz is due to both Z's going

on-shell.

7.3 Trigger and Offline Muon Reconstruction

Muons have a very clean detection signature resulting in a high trigger efficiency [7].

The inclusive muon tri -1. -r-i based on the selection of a single muon with pr > 19 GeV/c









or di-muons with pr > 7 GeV/c assures an efficiency of practically 1C( for collecting

events that have four high-pr muons.

In order to minimize muon reconstruction systematic uncertainties, we select only

those reconstructed muons that have a transverse momentum pr > 7 GeV/c if they are in

the central pseudorapidity region (Irl < 1.1), or with total momentum p > 13 GeV/c if

they are in the endcaps (Irl > 1.1). Figures 7-5 and 7-6 show the efficiency turn-on curves

- the choice of these cuts is obvious from the figures. These cuts do not dramatically affect

the number of accepted signal events.

We require that all four possible combinations of reconstructed di-muon masses

satisfy m(ptt-) > 12 GeV/c2. As in the previous case, this cut has very little effect on

the Higgs boson events and is primarily intended to suppress poorly simulated hadronic

background contributions originating from charmonium and bottomonium di-muon decays.

The most important characteristic distinguishing the Higgs boson decays from all

backgrounds is the presence of a peak in the four-muon invariant mass distribution.

Figure 7-7 shows such a distribution for MH=150 GeV/c2. A Gaussian fit of the peak

gives a=1.1 GeV/c2. One can see a noticeable tail toward smaller masses mostly due to

internal bremsstrahlung and events with intermediate r-leptons (ZZ -+ 2r2/p -- 4p4v

and ZZ -- 4r -- 4/8v). The four-muon mass detector resolution 6M(4p) as a function of

M(4p) is given in Figure 7-8.

7.4 Higgs Search Strategy

The strategy for searching the Higgs boson involves successive steps:


First, given a distinct localization of the Higgs boson signal as a resonance-like peak
in the invariant mass of four muons, the cuts can be made MA(4p)-dependent. The
cut optimization is described in section 7.4.2. The results with flat, MA(4p)-independent,
cuts are also presented for comparison.


Second, after the cuts are applied, we search for the 4/1 resonance-like peak
over the continuum background. We require the peak to be consistent with the
Standard Model Higgs. We explicitly compare the potential sensitivity of the two









approaches: Log-Likelihood Ratio that takes into account the entire shape of the
M(4p)-distributions for the signal and background as well as a straightforward
counting experiment (at this stage, no systematic errors are yet included; thus,
we call this part of the analysis a potential sensitivity). In the future, more
sophisticated statistical tools can and will be added.


Third, the most important theoretical and instrumental systematic errors are
evaluated (section 7.4.3.1). We propose and analyze the merits of a number of
methods for obtaining various analysis corrections directly from data in order
to minimize our reliance on Monte Carlo simulation (both physics and detector
performance). By doing this, we significantly reduce systematic errors and
uncertainties.


Fourth, systematic uncertainties are included in the evaluation of the significance
of observing the Higgs boson signal (section 7.4.3.2). For the counting experiment
approach, this can be done analytically in a straightforward way. For LLR and other
more sophisticated statistical tools, this can be done only by running a large number
of pseudo-experiments and would also require a knowledge of all correlations across
the data, the M(4p)-spectrum in this case,-these correlations are not yet available.


Finally, should an excess of events consistent with the Higgs boson be observed,
one should be careful in probabilistic interpretation of a local significance. A
considerable over-estimation is possible due to the fact that the range of masses open
for searching a relatively narrow signal is very large. In section 7.4.4, we outline a
straightforward methodology of evaluating the scope of this effect. To be able to do
it right, well defined search assumptions must be set a priori.

7.4.1 Introductory Remarks on Significance

As discussed above, after applying cuts, we will be searching for a possible local

excess of events in the M(4p) invariant mass spectrum. This can be done by using a

likelihood ratio of the probability of observing the data in the case of the signal-plus-background

hypothesis, to the probability of observing the data in the presence of the background-only

hypothesis: Q = P(observablesls + b)/P(observableslb). For the purpose of this study,

the final observable we use for the likelihood ratio is the four-muon invariant mass. In

principle, the list of observables can be extended further, but this requires a substantially

larger sample of Monte Carlo data to be able to take into account all correlations properly.









The likelihood ratio is known to give the best statistical discriminating power between two

hypotheses [78].

The log-likelihood ratio (LLR) can be built for a narrow region in the vicinity of the

peak (counting experiment):


-21n(Q) 2s 2nln + (7 1)

for the entire binned spectrum:

N
-21n(Q) 2 S 2 n In 1 + (7-2)
bins
or the unbinned spectrum:



-21n(Q) 2 S -2 ln + ( (7 3)
events pdfb f /
The significance estimator S = N/2 (Q), ScL for a counting experiment and SL for

the entire spectrum, is known to follow very closely (even for cases with few data) the

one-sided Gaussian probability that one associates with the true significance S:

/+OC 1 x2
P exp(- )dx. (7-4)

Figures 7-9 and 7-10 show that the ScL tracks the true significance S for even small

numbers of background events. Note that other popular quick estimators, S1 = s/I b,

S2 = s/n,, and S12 = 2(/n Vb) also shown in Figures 7-9 and 7-10, do not work as

well for smaller event counts and large values of significance.

The SL estimator, being sensitive to the full shape of the signal and background

distributions, has a leading edge over the simple counting experiment ScL estimator. We

typically observe 5-10(' difference. The ScL estimator, being local, is the natural tool

for optimization of M(4p)-dependent cuts. The convolution of systematic errors into









evaluation of a significance of an event excess is also very transparent and can be done

analytically.

By definition, the significance values we obtain using our Monte Carlo samples are

actually mean values for the expected significance to be measured in real data, should the

Higgs boson exist at a given mass. However, the actual significance to be observed may be

higher or smaller. The spread would be 1 in the limit of infinite data and is somewhat

larger for the typical numbers of signal and background events in our case.

Since the estimators SL and ScL are only estimators, one needs to re-evaluate

their probabilistic properties. For the LLR approach, one needs to run a large number

of pseudo-experiments in order to evaluate the true confidence levels CLb and CLb+s

(probabilities) of observing SL smaller than a particular threshold (or, equivalently,

-21n(Q) larger than a particular threshold), for the background-only and background-plus-signal

hypotheses, respectively. For the counting experiment approach this can be done

analytically: e.g., assuming the background-only hypothesis, the probability of observing

ScL larger than SeL, corresponding to the number of observed events no, is

+0 b k
1 CLb P P(n > nob) = e-b (75)
k=no
For non-integer n, (as is the case for weighted Monte Carlo events), we can use a

smooth interpolation of this function between integers n.

Before we include the systematic errors, it is this probability P that we use in

conjunctions with Eq. 7-4 to define the true counting experiment significance Sc.

Including systematic errors into evaluation of significance is discussed in Section 7.4.3.2.

7.4.2 Optimization of the M(4p)-Dependent Cuts

By taking advantage of the fact that the Higgs boson resonance H -- 4p is relatively

narrow, we use M(4p)-dependent cuts for its search. The analysis steps in this case are as

follows:









First, events with 4 muons (2j+2 -) satisfying pr, p, and M(pp+-) quality cuts
as described in section 7.3 are selected. This ensures that muons are reliably
reconstructed and removes a "contamination" originating from heavy quarkonia
decays.


Second, after reconstructing a four-muon invariant mass, the M(4p)-dependent cuts
are applied. The cuts, being smooth functions of M(4p), are optimized in such a
way that they maximize the ScL significance of the Higgs signal excess at all Higgs
boson mass points.


And, finally, the resulting M(4p) distribution is analysed for the presence of a Higgs
boson resonance. The search can be done using any statistical technique. In this
analysis, we explicitly compare the potential of the LLR built from the M(4p)
distributions and a straightforward counting experiment approach.

To perform the desired M(4p)-dependent cut optimization, we used a recently

developed program GARCON 1 [100]. The counting experiment significance estimator ScL

is the natural tool for such optimization.

For cut optimization, we considered the following muon kinematic variables:


Tracker-based and calorimeter-based isolation energy for all four muons, ordered by
isolation energy.


Transverse momentum pr of all four muons, ordered by pr.





1 GARCON stands for Genetic Algorithm for Rectangular Cuts OptimizatioN. This
program allows one, in an automatic way, to optimize cut values and then verify the
stability of the results, checking effectively a large number of cut sets, which, in a
straightforward approach, would take an astronomical amount of time. In this analysis,
we optimized 18 MH points using 16 cut variables, with the step on each one equal to
0.025 (40 steps) of a typical variable value, which makes 18 (2 16)40 cut set values
to try in a straightforward way. We run the optimization and verification on a 3 105
events. More details on GARCON are available on its web-page and in a dedicated
paper [100]. This program, being relatively new, is already widely used for other C'\ S
analyses [101, 102, 103].









Two invariant masses of di-muons mi( p -) and m2(/ p ), where the first pair of
muons is the one that gives the invariant mass closest to mz.


Parameter describing mismatch of muon displaced vertices.

The first half of the available Monte Carlo simulated data was used for the cut

optimization (no systematic errors were included at this stage). The results for the 18

Higgs mass points were then fit to obtain smooth M(4p)-dependent cuts. It was found

that, given the level of the expected dominant backgrounds (tt, Zbb, ZZ), there are only

three critical discriminating cuts:


Muon isolation cut, both tracker- and calorimeter-based, on the worst isolated muon,
or equivalently one common cut on all four muons (Figs. 7-11 and 7-12). This cut
strongly suppresses tt and Zbb backgrounds. One can see that the cuts gets tighter
as M(4p) gets smaller as the role of Zbb and tt increases (Fig. 7-1).


PT cut on the second lowest pr muon, or equivalently one common cut on the three
highest PT muons (Fig. 7-13). This cut helps to further suppress Zbb background to
the level well below ZZ and reduces the ZZ background at high four-muon invariant
masses. This cut gets more stringent with increasing M(4p), as the transverse
momentum of muons from a high mass Higgs boson tend to be higher than those
from ZZ background.


And, of course, M(4p) window being used for scanning over the background
(Fig. 7-14). It roughly corresponds to the 2a width, where a is the Higgs boson
peak width that includes the detector resolution and the Standard Model Higgs
boson width.

In the next step, we applied the three critical cuts to the second half of the available

Monte Carlo events that were not used for the optimization of the cuts. We tried both the

cuts as they came from the optimization with GARCON and, alternatively, the smooth

cuts shown by the lines in Figures 7-11- 7-14. The results of these comparisons are shown

in Figure 7-19. The observed stability of the results as we switch from the first half of

the sample used for cut optimization to the second half, as well as from "the optimal"

to the smooth cuts, ensures that the cut optimization did not pick peculiar phase space









corners corresponding to statistical flukes. Note that some "optimal" value points are

absent in Figures 7-11-7-14: they are out of the plot-ranges and, in some cases, pushed by

GARCON to the extreme limits. The latter means that this particular cut parameter for

this particular Higgs mass point is not effective, as often is the case, for example, for two

isolation-parameter-based variables, due to their very high degree of correlation.

Figure 7-20 shows the M(4p) invariant mass distribution for the three background

subprocesses and a Higgs boson signal at MH = 150 GeV/c2 after applying the three

smooth M(4p)-dependent cuts. One can see that the tt and Zbb backgrounds are now

suppressed well below the irreducible ZZ background.

Other possible cuts such as invariant masses of the muon pairs, impact point

parameters, kinematical cuts on other muons, and isolation parameters on other muons

do not significantly help to improve the results further, see Figures 7-15-7-18. The cuts on

these observables may still be useful and p1 l a role of !,. i, i l- to suppress possible

unaccounted-for backgrounds related to the beam halo, detector mis-performance, etc.

Additional variables that may help to discriminate H from the dominant ZZ

background have been studied: pr(4p), number of jets and their ET, etc. However, these

variables are driven by the NLO production processes, while our samples were generated

at the Leading Order by Pythia and CompHEP. Therefore, any conclusions that we might

derive from these samples would not be reliable. Some angular distributions built from

muons also have some differences originating from the underlying spin structures, but they

are not sufficiently discriminating to be used for cuts and may be strongly affected by the

NLO diagrams.

Figures 7-21 and 7-22 show the significance ScL and SL at L = 30 fb-1 for the

expected excesses of events for different Higgs boson masses. To emphasize the gain in the

sensitivity achievable with M(4))-dependent cuts, the results for flat cuts, optimized for

MH = 150 GeV/c2, are also superimposed. As expected, one universal set of cuts cannot

deliver the optimal performance for the full range of possible Higgs masses. The gain in









significance can be easily translated into probabilistic terms. For example, the Higgs boson

with MH = 500 GeV/c2 is right at the 5c-discovery threshold for an integrated luminosity

L = 30 fb-1 (Fig. 7-21). The difference in the average expected significance, 5.3 and 4.6,

means in this case that the chances of discovering the Higgs boson with MH = 500 GeV/c2

at L = 30 fb-1 are < !'1 '. for the flat cuts and > I11', for the MA(4p)-dependent cuts.

Figure 7-23 shows the same results as in Figure 7-21 in terms of the luminosity

needed for observing an excess of events over the expected background in the presence of

the Standard Model Higgs boson at 5a significance.

Figure 7-24 gives an idea of how the experimental exclusion limits will map onto the

plane of cross section vs. Higgs boson mass for a few different integrated luminosities.

7.4.3 Systematic Errors

The analysis of the systematic errors can be divided into two distinct stages. First,

one needs to understand the uncertainties in predicting the background. Second, these

uncertainties in the background have to be included in the evaluation of the significance of

an excess of events, should it be observed.

7.4.3.1 Uncertainties in the background

Uncertainties in the signal are not very important for establishing an excess of events

over the background. It is the uncertainties in the background that are of main concern.

After applying the analysis cuts as described earlier, ZZ production is the dominant

irreducible background, with all other processes giving much smaller contributions. This

reduces the analysis of systematic errors to the ZZ -- 4/ process. The main uncertainties

are as follows:

PDF and QCD scale uncertainties;

NLO and NNLO contributions vs LO;

Integrated luminosity;

Ti -.- r efficiency;

Muon reconstruction efficiency;









Muon isolation cut efficiency;

Four-muon mass M(4p) resolution;

Four-muon mass M(4p) absolute scale.

One can try to evaluate/guess the theoretical and detector performance related

uncertainties starting from the first principles. However, the credibility of the detector

performance systematic errors estimated this way is ahv-- shaky, especially during the

earlier stages of the detector operation when the changes in the system are frequent and

hard to monitor; and they must be timely incorporated into the detector Monte Carlo

simulation.

Therefore, we developed methods to evaluate various corrections, such as muon

reconstruction efficiency, muon isolation cut efficiency, M(4p) resolution, and absolute

scale, directly from data in order to minimize our relying on the Monte Carlo simulation,

and, thus, significantly reducing the associated systematic errors.

Moreover, throughout this analysis, we estimate the background around a particular

M(4p) area (signal region) in reference to a measured control sample. Note that this

completely eliminates uncertainties associated with measuring the luminosity and reduces

the sensitivity to PDF and QCD-scales. For the control sample, we use either the inclusive

Z -- 2p process or sidebands of the M(4p) spectrum itself. When we refer to a control

sample, we will use the factor p, defined so that b = p B, where b is the expected number

of background events in the signal window and B is the measured number of events in the

control sample. Particular examples are given below in overview and all the details are

available in referenced below papers we published. In general, selection criteria for Z -- 2P

events were selected as close as possible to selection cuts described above in this analysis

and control samples for ZZ -are events on the left and right side of the signal window.

The PDF and QCD scale uncertainties in the ZZ -- 4p production cross section

were studied at the NLO level using MCFM [98]. Systematic errors associated with PDFs

were estimated by giving la variations to the 20 CTEQ6M parameters. By varying









independently the renormalization and factorization scales by a factor of two up and down

from their default values pR = PF = 2Mz, we found the sensitivity of the ZZ cross

section to the QCD scale uncertainties. All details of these studies can be found in Ref.

[5]. Figure 7-25 shows these PDF and QCD scale uncertainties, added in quadrature,

versus M(4p). The three curves correspond to (a) the absolute predictions (relatively flat,

Jb/b 1 i,' ); (b) the prediction normalized to the measured Z -- 2p cross section (note

that the Z -- 2p cross section can be measured with instrumental systematic errors, not

including luminosity, of less than 2". (CDF results, Phys. Rev. Letter, 94 (2005) 091803))

(6p/p t 1 for M(4p) close to Mz and then steadily increasing toward larger four-muon
invariant masses); and (c) prediction normalized to sidebands of the M(4p) distribution

itself in the range from 100-600 GeV/c2 ((6p/p is at its minimum when the signal window

is at the place where most of the events are).

Beyond-Leading-Order correction uncertainties were estimated as follows. The

M(4p)-dependent K-factor K(M4,) for the ZZ -+ 4p process was evaluated with two

very different programs: MCFM [98] and EffNLO [6]. The latter is a package smoothly

splicing together MadGraph [104] (NLO pp -- 4/ + jet) and Pythia (LO pp -- 4/ +

ISR-jets). The relative difference in K(M4,) is shown in Figure 7-26. The NNLO diagrams

include new processes (we define a process as new if it has a distinctly different initial

state and, therefore, variations of QCD scales do not necessarily give a feel for its relative

importance): gg i ZZ -- 4/ (box-diagram), contributing about (20 8) to the LO

cross section [99] (note that this contribution was calculated without virtual photons in

the propagators) and qq 4/ + qq via Z-bremsstrahlung (not yet calculated) or via

vector-boson fusion (implemented in Pythia, very small). Since the nature of all these

differences/variations is not well understood, we present the final results with and without

these uncertainties included. Certainly, more theoretical work in this area is needed. All

other higher-level diagrams can be considered as corrections to the distinct LO, NLO,

and NNLO processes discussed above. Omission of these higher-order corrections would









manifest itself as a sensitivity of the calculated cross sections to the QCD scale variations

discussed earlier.

Luminosity measurement uncertainties are expected to be 10'. at the time of an

integrated lumunosity of 1 fb-1, 5' at L = 10 fb-1, and :'-. for larger luminosities. When

we estimate the ZZ background events in the signal region via the measured number of

events in the control samples, the luminosity uncertainties largely cancel out.

The muon trigger efficiency, being very close to 1C1iI' due to the presence of four

muons, does not have substantial systematic errors.

The muon reconstruction efficiency can be measured directly from data with an

uncertainty of better that 1 The method uses a data sample based on single-muon

HLT (HLT stands for High Level Tii.-.-. i, the final stage of online filtering after which

the data are recorded on tape) tri r-. --r with pr > 19 GeV/c. This sample will contain

inclusive W, Z, and other processes in the approximate ratio W:Z:others = 10:l:small

[105]. By counting the number of Z -- 2p events in the resonance peak of the invariant

mass distributions built from the HLT muon and all other tracks, the HLT muon and

all other standalone muons and the HLT muon and all other globally reconstructed

muons, one can evaluate the efficiency of finding globally-reconstructed muons with better

than 1 precision. Such a measurement will automatically account for the real detector

performance, including intermittent and smooth variations in time. All details can be

found elsewhere [3], see also Sec. C. The four-muon efficiency therefore will be known with

an absolute error of better than I!'. When we deduce the expected ZZ -4 4P events from

the measured Z -2 2p cross section, this uncertainty partially cancels out and becomes

'' This efficiency remains fairly flat vs M(4p), which makes this error completely

negligible if sidebands are used for evaluating the number of expected background events

in the signal region.

The muon isolation cut is very important as it allows us to suppress otherwise

overwhelming tt and Zbb backgrounds well below the ZZ background. As we apply this









cut, we also cut ZZ (and Higgs) events by t 15 :n This cut is very sensitive to

the underlying event physics, which, unfortunately, is not very well understood and has

substantial uncertainties. As in the case of the muon efficiency, we developed a scheme for

evaluating the muon isolation cut efficiency directly from data. Again, we appeal to the

inclusive Z -- 2/ sample. The Z events have very similar underlying event activity as ZZ

events. We show that, by using random directions in Z events and evaluating the energy

flow in isolation cones around them, one can predict the 4-muon ZZ event losses due to

the muon isolation cut with a systematic error of less than '" [4] (Fig. 7-27).

The uncertainty on the muon pr resolution directly propagates into the four-muon

invariant mass M(4p) reconstruction. This almost does not affect the background

distribution. However, the M(4p) distribution width drives the width of the M(4p)

window that we use for evaluating the signal excess significance at low Higgs boson

masses. Fortunately, even making a mistake in the M(4p) distribution width by as much

as 25' has only a tiny effect on evaluating a significance of an excess of events (Fig. 7-28).

Also, the muon pr resolution is fairly easy to measure from data using the measured J/Q

and Z peak widths with a precision much better than needed.

The uncertainty on the muon pr scale can be similarly calibrated from data using

the measured J/& and Z peaks. The effect of these uncertainties on the number of

background events in a signal window appears only on steep slopes of the M(4p)

distribution. For the steepest part of the M(4p) distribution in the 180 200 GeV/c2

range, we obtain 6b/b 0.16M4,, where 6M4, is in GeV/c2. This implies that to be

able to neglect this effect, one needs to know the momentum scale with a precision of

0.1 GeV/c at pT w 50 GeV/c. This can be easily achieved with just a few hundred

Z -+ 2p events.

Figures 7-29 and 7-30 summarize all the systematic errors on the expected number

of events in the ZZ -- 4p background for the two methods: via referencing to the total









measured Z -- 2p cross section and via referencing to the event count in the sidebands of

the M(4p) spectrum itself.

7.4.3.2 Significance with the background uncertainties included

If the background has uncertainties, which we will express in terms of a probability

density function f(b), the probability of observing at least no events becomes

+o0
P =p(n > nob) f(b) p(n > nlb) f (b) db, (7-6)

which can be again converted into true significance Sc using Eq. 7-4.

We will use a log-normal form of a probability density function for the absolute

systematic errors for expected number of background events bo with a relative uncertainty

6 = Ab/bo:


1 e In2(b/bo)> 1(7 7)
f(b)= exp -1n-e. (7-7)
Sln (k) 21n2(k)

In this equation, k = 1 + 6 and 6 is the sum in quadrature of all the uncertainties.

For relatively small errors, this form of equation gives a Gaussian distribution with

average bo and a -= 6 bo. One advantage of using the log-normal presentation is that

it does not have a tail spilling over into b < 0. Also, and maybe more importantly,

this equation give an intuitively correct representation for very large uncertainties. For

example, such a statement as "we estimate that the background is bo with a factor of 2

uncertainty" probably implies that we assume that the chances for the true background to

be somewhere between bo/2 and 2bo are about 1 -'*, while the chances of being larger than

2bo or smaller than bo/2 are approximately equal-Eq. 7-7 does just that for any value of

k, small or large (k would be equal 2 in this case).









The statistical part of the probability density function f(b) for the background in the

signal region, estimated from the observed event count in a control sample B (b = pB),

can be obtained using B i,, -' theorem and a flat prior:


1 (b/p)Be-b/p
f(b) (- (7-8)
p F(B + 1)
The full probability density function f(b) for the background, estimated using

sidebands when there are uncertainties on the factor p, can be easily obtained by a

convolution of the two equations shown above.

Figure 7-31 gives three curves: the significance vs Higgs mass in the absence of any

systematic errors (both for the plain ScL estimator and the true significance S,) as well as

the significance that includes all uncertainties in the background when it is estimated from

the measured Z -- 2p cross section. All three curves correspond to the total integrated

luminosity of 30 fb- Figure 7-32 shows curves (with and without systematic errors) for

the required luminosity for 5a-discovery, 3a-evidence, and 95'. CL exclusion limit for the

Standard Model Higgs boson.

The comparison between two v--i- of normalization, to the Z -+ 2/ process and

the ZZ -+ 4/z sidebands, is made in terms of the luminosity required for 5a-discovery

(Fig. 7-33). The difference is not dramatic. The true benefit of using two approaches to

estimating background from data is in their complementarity.

Finally, Table 7-2 summarizes the most important results for the M(4p)-dependent

cuts that we presented in this chapter.

7.4.4 Local Significance and Overall Statistical Fluctuation Probability

In searching for a new phenomena in a wide range of parameter phase space (in our

case, we search for a narrow resonance in a very broad range of invariant masses), one

inevitably encounters a well-known problem of overestimating the overall significance

of a "local discovery." The scale of the effect can be quite large (see Appendix, Sec. 7.6

for all details) and one must exercise a caution in evaluating probabilistic interpretation









Table 7-2: Summary of the results: number of signal and background events in a window
used for a counting experiment with the M(4p)-dependent cuts. Systematic error on the
background is normalized to the Z -- 2p process (6KNLO/KNLO is not included); three
different significance without systematic errors included: the SL estimator for the Log
Likelihood Ratio (LLR) built for the full M(4p) spectrum, ScL LLR estimator built for a
counting experiment approach, and the Sc true significance for the counting experiment
approach; the final result for Sc, now including all systematic errors.

Mass Signal Bkgd Syst Error SL ScL Sc Sc
(GeV/c2) s b 8b/b (no syst) (no syst) (no syst) (with syst)
115 2.13 0.92 ;, 1.98 1.75 1.54 1.54
120 4.00 1.15 2.88 2.72 2.57 2.56
130 12.45 2.06 5.97 5.64 5.54 5.52
140 23.22 2.65 9.09 8.45 8.39 8.35
150 28.09 2.42 !, 10.84 9.92 9.87 9.81
160 14.25 3.01 '. 6.04 5.64 5.55 5.53
170 6.32 3.63 '. 3.00 2.73 2.61 2.60
180 14.54 7.10 '. 4.83 4.38 4.30 4.26
190 54.95 17.00 '. 10.85 9.89 9.85 9.59
200 62.78 19.93 '. 11.32 10.48 10.45 10.11
250 54.48 21.62 5% 9.83 9.09 9.05 8.61
300 40.43 13.40 1.' 9.26 8.30 8.25 7.90
350 40.68 10.75 7' 9.50 8.93 8.88 8.47
400 33.59 8.00 9.07 8.36 8.31 7.95
450 24.07 5.68 7.57 7.10 7.03 6.81
500 16.65 5.58 9' 5.81 5.31 5.23 5.05
550 12.10 5.70 9' 4.44 4.04 3.92 3.80
600 8.72 5.04 9' 3.58 3.20 3.10 3.00


of observing an


excess of events at a particular mass. In the case study we consider


in the Appendix, we show that observing a 2j excess would be basically guaranteed,

an observation of a "local 3j" excess would be hardly of any significance (a 15'.

chance), and the -,iin,!i n:e of a local 57-d( ..-' y would actually correspond to a

true statistical significance of a 3.8j. In the same Appendix, we also discuss possible v-i-,

to reduce the scale of the effect.

7.5 Summary

Discovery of the Standard Model Higgs boson in the ;old-plated" decay mode

H ZZ(*) 4p was analyzed in the context of the C' \S Detector. The explored

range of Higgs boson masses was 115-600 GeV/c2. The Monte Carlo samples for









signal and background were generated to represent the NLO cross sections, including

M(4p)-dependent K-factors. To simulate the detector response and reconstruct physics

objects, the full C'\ S Detector simulation and reconstruction software was used. We

explored the Higgs boson discovery potential for different analysis variations, including

the use of M(4p)-dependent and flat cuts, Log Likelihood Ratio based on the full M(4p)

spectrum, and a straightforward counting experiment approach.

A full treatment of the most important theoretical and instrumental systematic

errors and their effect on evaluation of significance of the Higgs boson observation using

mass-dependent cuts and a counting experiment approach were presented. To minimize

systematic errors, a number of methods of reconstructing the necessary corrections directly

from data were developed.

We showed that at a 2 fb-1 of integrated luminosity, we would be able to start

excluding the SM Higgs boson at 95'. CL for MH in vicinity of 200 GeV/c2. By the time

we reach w 30 fb-1, we would exclude the Standard Model Higgs in its four-muon decay

mode in the mass range MH = 120 600 GeV/c2, if indeed it does not exist.

The discoveries at the level of "5o" local significance could be already possible at

S10 fb-1 for MH in the range 140-150 and 190-400 GeV/c2. By the time we reach

S30 fb-1, the discovery range would open up to 130-160 and 180-500 GeV/c2. An

observation of the Higgs boson with the mass MH t 170 GeV/c2 or w 600 GeV/c2 in the

H -+ ZZ(*) -- 4p decay channel would require an integrated luminosity of the order of

100 fb-.

7.6 On the true significance of a local excess of events

In searching for new phenomena in a wide range of possible signal hypotheses (e.g., a

narrow resonance of unknown mass over a broad range background), special care must be

exercised in evaluating the true significance of observing a local excess of events. In past,

this fact was given substantial scrutiny by statisticians (e.g., [106, 107]) and physicists

alike (e.g., [108, 109, 110, 111, 112]). The purpose of this Appendix is to quantify a













SI I II ttbar
|":zbb
> 102 Mzz4mu
aL2 _0 jzztau
(D Emhl40

10






101



10-2 o
0 100 200 300 400 500 600 700
4p invariant mass (GeV)


Figure 7-1: M(4p) distributions after
generator-level cuts for tt, (Z/7*)bb,
Z/7*Z/7*, and m"H 140 GeV/c2 (log
scale).






ro 1.14 1 I 1 .... I I .... I '
1.14

jF 1.12

1.1

E 1.08

S1.06
..c
LI 1.04

1.02
1-

100 200 300 400 500 600
M(4pL) (GeV)


Figure 7-3: Enhancement to the
signal samples' cross sections due
to interference effects not accounted
for at the generator level.


5 450

o 400
(.
- 350

S300

- 250
U-
> 200

150

100

50
I..


CMS, L=30 fb
- ttbar
":zbb
zz4mu
zz2au
] mh140


0 100 200 300 400 500 600 700
4p invariant mass (GeV)


Figure 7-2: Same as Figure 7-1, but on a
linear scale.


Range 30 GeV 750 GeV
sigma NLO = 19.89 fb
sigma LO = 14.73 fb
K factor = 1.3500


100 200 300 400 500 600 700
M(4g) (GeV)


Figure 7-4: The M(4p)-dependent
NLO K-factor KNLO((M4/) for the
ZZ -+ 4/i process evaluated with
MCFM [98].






















0 *, I I I I I
0








Figure 7-5: Global muon
0.68 "


S0.6

0 ,



0 20 40 60 80 100
barrel: pT(Ct) (GeV)


Figure 7-5: Global muon
reconstruction efficiency calculated
from matching reconstructed and
true Monte Carlo muons in the
barrel region vs. pr.


0.8- ~ .'

S0.6
o
% 0.4

0 ,

cii

0 20 40 60 80 100
endcap: p([L) (GeV)


Figure 7-6: Global muon
reconstruction efficiency calculated
from matching reconstructed and
true Monte Carlo muons in the
endcap region vs. momentum.


130 140 150 160
M(4g) (GeV)


Figure 7-7: M(4p) distribution
for m(H)=150 GeV/c2 and the fit
described in the text.


100 200 300 400 500 600 700
M(4pL) (GeV)


Figure 7-8: M(4p) resolution vs.
MH.













SS (Gausslan one-sided tall)
........ Sc,
......... S
Sc,


10-3 \

10-4 % "o

10-5 ''a

10-6

10-7

108 12345678
10 1 2 3 4 5 6 7 8
Threshold Significance S


Figure 7-9: Comparison of different
significance estimators for 1
background event: probability of
measuring significance S > So,
background only case, NB = 1
event.


6


5-


4


3


2.-


smooth
v not corrected


100 200 300 400 500 600
Higgs boson mass (GeV)


Figure 7-11: Dependence of the
tracker-based muon isolation cut on
the least isolated muon versus Higgs
mass.


2 1 ......... seL
1 10-1 .

S10-2 '.




10-5 2 "
10-

10-
10-85 t J I I ,I

10\
10 1 2 3 4 5 6
Threshold Significan


Figure 7-10: Same as Figure 7-9,
but for NB =10 events.


- smooth
not corrected


(,
(9 10

o 9
o
. 8
0
o
0 7


X 6
E


100 200 300 400 500 600
Higgs boson mass (GeV)


Figure 7-12: Dependence of the
calorimeter-based cut on the least
isolated muon versus Higgs mass.


/















70
(,
( 60

o 50

40

e 30
E


0 700
(9

S600
o
S500
o

E 400

E 300

200


100 200 300 400 500 600
Higgs boson mass (GeV)


Figure 7-13: Dependence of the pr
cut on the second-lowest-pr muon
versus Higgs mass.


(D 10
(9
CD

t-*-

> 1
LU



10-1




1n-2


smooth
not corrected


Higgs boson mass (GeV)


Figure 7-14: Dependence of the
M(4p) window cuts versus Higgs
mass.


CMS, L=30 fb-


4 ,,,,,


20 40 60 80 100 120 140
Mass of first pair, M(Z1) (GeV)


Figure 7-15: First muon pair
invariant mass distribution, M(Z1),
after analysis cuts were applied.


CMS, L=30 fb-


4 ,,,,,
',r -iiL


20 40 60 80 100120140160180200
Mass of second pair, M(Z2) (GeV)


Figure 7-16: Second muon pair
invariant mass distribution, M(Z2),
after analysis cuts were applied.


I A
















(D
(9
04
C1 1

"-*-
u(D
LU

10-1






in-2 L


CMS, L=30 fb-


4,,,,
_1_,1_,L


0 20 40 60 80 100120140160180200
Lowest muon PT (GeV)


Figure 7-17: Lowest muon Pr
distribution, after analysis cuts
were applied.










CMS, L=30 fb
U) 1 6 0 optimization
() verification
c3 2 venf smooth cuts (ANALYSIS)


0

8

6

4

2


100 200 300 400 500 600
Higgs boson mass (GeV)


Figure 7-19: ScL vs. Higgs boson
mass.


"N


CMS, L=30 fb-
r 1,.


10-1





1 o-2
0 0.01 0.02 0.03 0.04 0.05
Distance between muons at IP (cm)


Figure 7-18: Maximum distance in
XY-plane between muon impact
point coordinates distribution, after
analysis cuts were applied.


SCMS, L=30 fb-
1ttbar
Zbb
izz4mu
"zz2tau
mh250


4-

2 I


0 100 200 300 400 500 600 700
M(4p) (GeV)


Figure 7-20: M(4p) invariant
mass distribution for the three
background subprocesses and
a Higgs boson signal at MH
150 GeV/c2, after applying cuts on
muon isolation and pr.















S14


- SCL (dynamic cuts)
....... SCL (fixed cuts)


100 200 300 400 500 600
M(4ji) (GeV)


Figure 7-21: Expected excess
significance SeL with L 30 fb-1
for different Higgs boson masses for
M(4p)-dependent and independent
cuts. No systematic errors included.


4-

102


Cn
o
10
E
-j


10


- Dynamic cuts
-...- Constant cuts


- SL (dynamic cuts)
....... SL (fixed cuts)


100 200 300 400 500 600
M(4 )) (GeV)


Figure 7-22: Expected excess
significance SL with L 30 fb-1
for different Higgs boson masses for
M(4p)-dependent and independent
cuts. No systematic errors included.


6 -

5



\ '"'*


.......... Exclusion at 95% CL (3 fb)
.............. Exclusion at 95% CL (10 fb )
........... Exclusion at 95% CL (30 fb )
-- SM Higgsx-section (pp H ZZ 4p)


Figure 7-23: Luminosity required
to reach a 5a event excess for
different Higgs boson masses for
M(4p)-dependent and independent
cuts. No systematic errors included.


100 200 300 400 500 600
MH (GeV)


Figure 7-24: The 95' CL exclusion
contours for the SM Higgs
hypothesis.

















ZZ->4p cross section

Ratio: normalization to Z->2p

Ratio: normalization to sidebands


7:7
./ '
^''


14
/)

12

o 10

- 8
8
0

O 6

O 4



0
E O
V)


300 400 500
M(4g) (GeV)


200 300 400 500 600
M(4p) (GeV)


Figure 7-25: Combined systematic
error on the number of background

events due to PDF and QCD
scale uncertainties for the

o(qq ZZ 4zp) process at NLO.


C I


E0.95


0.9


o 0.85
uJ


0.75


Figure 7-26: Top: the factors

KNLO(M4p) in MCFM and EffNLO
calculations versus M(4p); bottom:
the difference between them.


1 2 3 4 5
Isolation cut (GeV)


Figure 7-27: Muon isolation cut
efficiency for random cone direction
for Z-inclusive (dashed lines) and
for ZZ (solid lines) events.


0
1 1.2 1.4 1.6 1.8 2
True resolution /Assumed MC resolution


Figure 7-28: An example of a
possible bias in evaluating the
significance of an event excess due
to a non-optimal choice of the
signal window width.


4 al
za 0,


MCFM
SEffNLO
.... Average = 1 35


100 200 300 400 500 600
M(4p) (GeV)


100 200


02-
015
005 -
01
-005
-0 1 100


ScL, if we knew the true resolution
.......... Biased ScL if we trust MC


/ --- -
A "


/*


I .
r" /* ,































20 Combined
S ......... dK/K NLO +(gg -ZZ)
SPDFatNLO
18 QCD at NLO
Muon Reconstruction Eficiency
1 ........ Muon Isolation Cut Eficiency
Muon pT resolution: negligible
V 14 Muon pT-scale: negligible
12 Muon Trigger: negligible
O 12 Luminosity: negligible .
10 .
nC 8 .
S6

c 4

t 2 .. ........... ... ...... ..... .
S 0
S 100 200 300 400 500 600
M(4pL) (GeV)


Figure 7-29: Uncertainties in the
number of ZZ -i 4p background
events in the signal region window
at different M(4p). The event
count is referenced to the number of
Z -- 2p events.


- Combined
......... dK/K NLO+(gg -ZZ)
PDF at NLO
QCD at NLO
Muon Reconstruction Efficiency: negligible
Muon Isolation Cut Efficiency: negligible
Muon pT resolution: negligible
Muon pT-scale: negligible
Muon Trigger: negligible
Luminosity: negligible




--l-. *


S 100 200 300 400 500 600
M(4jp) (GeV)


Figure 7-30: Uncertainties in the
number of ZZ -- 4p background
events in the signal region window
at different M(4p). The event
count in signal region, is calculated
from the number of ZZ 4
events in the range 100-700
GeV/c2 (excluding the signal region
window).

























..... ..... S, without systematic errors
12 S, without systematic errors
.......... S, with syst errors (normalization on Z)


S ..... ...... S, without systematic errors
12 S without systematic errors
........... So with syst errors (normalization on Z)


CMS, L=30fb1
H ZZ 4p
NLO x-sections


0 I
100


CMS, L=30fb"1
H ZZ 4
NLO x-sections


0 I
100


200 300 400 500 600
M(4 )) (GeV)


200 300 400 500 600
M(4 )) (GeV)


Figure 7-31: Significance vs. Higgs mass (with and without dF/K contribution).
103 ....... S=5, no systematics 103 ........ S=5, no systematics
' -- S=5, with systematics -- S=5, with systematics
......... S=3, no systematics ......... S=3, no systematics
S- S=3, with systematics -- S=3, with systematics
I Exclusion at 95% CL Exclusion at 95% CL
0 0
102 S 102


E 0 10


10 10


CMS
H ZZ 44
NLO x-sections
I I I II I I 1 1 L
200 300 400 500 600 1100
M(4 ) (GeV)


CMS
H ZZ 4
NLO x-sections


200 300 400 500 600
M(4 ) (GeV)


Figure 7-32: Integrated luminosity needed for 95' .CL exclusion, 3a, and 5o discovery

versus Higgs boson mass (with and without dF/K contribution).


100










1 3.......... No uncertainties .3 ......... No uncertainties
S- With syst errors (normalization on Z) 0 With syst errors (normalization on Z)
4 With stat+syst errors (sidebands) 4 With stat+syst errors (sidebands)

o1 a





10 10

CMS, Significance=5 CMS, Significance=5
H -> ZZ 4 H-> ZZ 4
NLO x-sections NLO x-sections

1 100 200 300 400 500 600 1 100 200 300 400 500 600
M(4jp) (GeV) M(4p) (GeV)

Figure 7-33: Integrated luminosity needed for a 5a discovery of the Higgs boson versus its
mass for (with and without dF/K contribution).


possible scope of this effect on an example of a search for the Standard Model Higgs boson

in the H -+ ZZ(*) -- 4/ decay channel. As the case study, we chose a counting experiment

approach widely used in this volume.

The dashed line in Fig. 7-34 shows the expected 4/z invariant mass distribution for

background at L = 30fb-1 after applying all the m4p-dependent analysis cuts. Using this

distribution, we p1- i,- d out ~ 108 pseudo-experiments; an example is shown in Fig. 7-34.

For each pseudo-experiment, we slid a signal region window across the spectrum looking

for a local event excess over the expectation. The size of the window Am = w(m4)

was optimised and fixed a priori (about 2o) to give close to the best significance for a

resonance with a width corresponding to the experimental SM Higgs boson width oa(m4p).

The step of probing different values of m4p was iiii. -i1 !!y" small (0.05GeV/c2) in

comparison to the Higgs boson width of more than 1GeV/c2. The scanning was performed

in a priori defined range of 115-600GeV/c2.

We used a significance estimator SL = signs) /2 -oln(1 + s/b) 2s, where b is

the expected number of background events, no is the number of observed events, and the

signal is defined as s = n, b). This estimator, based on the Log-Likelihood Ratio, is

known to follow very closely the true Poisson significance, only slightly over-estimating









it in the limit of small statistics [1]. Figure 7-35 presents the results of such a scan for

the pseudo-experiment shown in Fig. 7-34. The maximum value of Set, Smax, and the

corresponding mass of a "Higgs boson ( 'ii l ii. obtained in each pseudo-experiment

were retained for further statistical studies.

After performing 108 pseudo-experiments, the differential probability density function

for S,,m and its corresponding cumulative probability function P(S,,m > S) (Fig. 7-36)

were calculated. From Fig. 7-36, one can see that the frequency of observing some large

values of ScL (solid line) is much higher than its naive interpretation might imply (dashed

line). If desired, the actual probability can be converted to the true significance. The

result of such i' -ii 1 ii ii i- ,I ,i" is presented in Fig. 7-37. One can clearly see that the

required de-rating of significance is not negligible; in fact, it is larger than the effect of

including all theoretical and instrumental systematic errors for this channel. More details

on the various aspects of these studies can be found in [1].

There are v--,i- of reducing the effect. A more detailed analysis of the shape of

the m4, distribution will help somewhat. Using the predicted number of signal events

8 = theory in the significance estimator to begin with and, then, for validating the

statistical consistency of an excess no b with the expectation theory will reduce the

effect further. One can also use a non-flat prior on the Higgs mass as it comes out from

the precision electroweak measurements. Whether one will be able to bring the effect to

a negligible level by using all these additional constraints on the signal hypotheses is yet

to be seen. The purpose of this Appendix is not to give the final quantitative answer,

but rather to assert that these studies must become an integral part of all future search

analyses when multiple signal hypotheses are tried.





























150200250300350400450500 550 600
4a mass, GeV


Figure 7-34: The background pdf and an
example of one pseudo-experiment with
a statistical fluctuation appearing just
like a signal.


> 1
I 10-1
-0
g 10-2
S104-3
C 10-4
a10-5
10-6-
10-7

10-8
10 1 2



Figure 7-36: ScL
density function.


3 4 5 6
ScL


cumulative probability


100 200 300 400 500 600
4a mass, GeV


Figure 7-35: Profile of the ScL scan
corresponding to the pseudo-experiment
example shown on the left. Green
(inner) and yellow (outer) bands denote
lo( and 2 a intervals. Spikes that can
be seen are due to events coming in or
dropping off the trial-window, a feature
of low-statistics searches.

() 6/

S5-

U) 4/

3-

2-

1-

03 4 5
ScL (Observed)


Figure 7-37: Local significance
- ii. in 1.!!i .ii". from an observed
value to the true significance with a
proper probabilistic interpretation.









CHAPTER 8
CONCLUSIONS

The LHC will enable production of the SM Higgs boson in the entire range of its

allowed possible mass (from 114.4CGV/c2 to ~ 1TeV/c2).

An early discovery of the Higgs boson is one of the most important goals of the C'\!S

experiment. This will require a good understanding of both the data and the physics of

the background processes. The research work presented in the dissertation contributes

to both the better understanding of the C'\ !S detector and the development of methods

which could lead to the discovery of the Higgs boson. The following is the summary of the

main results:

Fast algorithm for track segment and hit reconstruction in cathode strip

chambers

A new fast algorithm for reconstructing track segments in Cathode Strip C('! i~hers

is proposed. The algorithm was validated with the real cosmic ray data taken with 36

CSC chambers operated as a part of C'\ !S-wide MTCC test program in the second half

of 2006. The algorithm proved to provide high speed, high efficiency, and good spatial

precision-all well within the High Level Ti i1:.- r requirements.

Search strategy for the Standard Model Higgs boson in the H -+ ZZ(*) -- 4/

decay channel using M(4p)-dependent cuts

Discovery of the Standard Model Higgs boson in the ;old-plated" decay mode

H -+ ZZ(*) 4p was analyzed in the context of the C' \S Detector. The explored

range of Higgs boson masses was 115-600 GeV/c2. The Monte Carlo samples for

signal and background were generated to represent the NLO cross sections, including

M(4p)-dependent K-factors. To simulate the detector response and reconstruct physics

objects, the full C'\MS Detector simulation and reconstruction software was used. We

explored the Higgs boson discovery potential for different analysis variations, including

the use of M(4p)-dependent and flat cuts, Log Likelihood Ratio based on the full M(4p)

spectrum, and a straightforward counting experiment approach.









A full treatment of the most important theoretical and instrumental systematic

errors and their effect on evaluation of significance of the Higgs boson observation using

mass-dependent cuts and a counting experiment approach were presented. To minimize

systematic errors, a number of methods of reconstructing the necessary corrections directly

from data were developed.

We showed that at a 2 fb-1 of integrated luminosity, we would be able to start

excluding the SM Higgs boson at 95'. CL for MH in vicinity of 200 GeV/c2. By the time

we reach w 30 fb-1, we would exclude the Standard Model Higgs in its four-muon decay

mode in the mass range MH = 120 600 GeV/c2, if indeed it does not exist.

The discoveries at the level of "5o" local significance could be already possible at

S10 fb-1 for MH in the range 140-150 and 190-400 GeV/c2. By the time we reach

S30 fb-1, the discovery range would open up to 130-160 and 180-500 GeV/c2. An

observation of the Higgs boson with the mass MH t 170 GeV/c2 or w 600 GeV/c2 in the

H -+ ZZ(*) 4p decay channel would require an integrated luminosity of the order of

100 fb-1

Supporting studies

T ii,, additional topics in "support" of main analyses were studied -results for which

are partly outlined in the set of following appendices.









APPENDIX A
RELATIVE CONTRIBUTIONS OF T- AND S-CHANNELS TO THE ZZ(*) -- 4/
PROCESS


(A. Drozdetskiy et al., "Relative Contributions of t- and s-C! in ., 1- to the ZZ(*) 4/
Process",
C'\!S Note 2006/057)
The qq i ZZ 4/z process is the main irreducible background in searches for
the Higgs boson via its H ZZ 4/z decay mode. PYTHIA [89], an event generator
commonly used for simulation of this process at the LHC, unfortunately is missing the
s-channel contribution (Fig. A-i).
In our detailed study we show that the s-channel subprocess and its interference
with the t-channel cannot be neglected in the context of the backgrounds for the H 4/z
analysis in the area of interest, M4. > 115 GeV if one aims to simulate the ZZ-background
with a precision of 10' or better. This contribution remains non-negligible after all
analysis cuts (Fig. A-i). A spectacular peak appearing at the 2-muon invariant mass
m(4p) ~ mzo due to presence of the s-channel may prove to become a standard candle for
monitoring the level of the ZZ-background.


































Zoy*


U+ 0




C)
0LI



i


ZO/y*


I I



- I








.1~




tj& --I -
I -


CMS SL=W30


Szz4rnu
L..J
5mm40














I I


0 100 200 300 400 500 600 700
4g invariant mass (GeV)


s-channel


Figure A-l: ZZ background: t- and s- channel diagrams and s-channel contribution peak
around Zo mass after pre-selection cuts for fully simulated events.









APPENDIX B
ZZ DISCOVERY WITH < 1FB-1 WITH C \!8 IN ZZ(*) 4p MODE


(A. Drozdetskiy et al., "Observation of the ZZ di-boson production in the ZZ(*) 4/
channel",
C'\iS Note in C'\IS approval process.)
We show that the observation of the first few ZZ 4p events with significance in
excess of 5a should be expected by the time C\ IS integrates luminosity of 0.5-1 fb-l. The
current estimate of the number of background events is < 0.0173Zbb+tt 0.00998Zbb
0.0116tt and is limited by the available MC tt sample statistics. A methodology of
calculating significance in circumstances when no or very few Monte Carlo background
events survive analysis cuts is described. We also demonstrate that a control sample of
41-combinations of wrong flavor and/or charge combination provides a powerful tool for
cross checking that the observed events are not due to some unaccounted backgrounds.









APPENDIX C
MEASURING MUON RECONSTRUCTION EFFICIENCY FROM DATA


(A. Drozdetskiy et al., \!. i-ii,_ig Muon Reconstruction Efficiency from D iI ,
C \!S Note 2006/060)
A method of measuring the global muon reconstruction efficiency e directly from
data was studied. With the data corresponding to an integrated luminosity L 10 fb-1,
the precision of measuring e for muons in the PT range of 10 100 GeV is expected to
be better than 1 potentially much better. The method largely alleviates uncertainties
associated with our ability to monitor and reproduce in Monte Carlo simulation all of
details of the underlying detector performance.
The method uses a data sample based on single-muon HLT (HLT stands for High
Level Ti.-.- r, the final stage of online filtering after which the data are recorded on
tape) trigger with pr > 19 GeV/c. This sample will contain inclusive W, Z, and other
processes in the approximate ratio W:Z:others = 10:1:small [105]. By counting the number
of Z -- 2p events in the resonance peak of the invariant mass distributions built from
the HLT muon and all other tracks, the HLT muon and all other standalone muons and
the HLT muon and all other globally reconstructed muons (Fig. C-l), one can evaluate
the efficiency of finding globally-reconstructed muons with better than 1 precision. The
four-muon efficiency therefore will be known with an absolute error of better than !' .
This efficiency remains fairly flat vs M(4p), which makes this error completely negligible if
sidebands are used for evaluating the number of expected background events in the signal
region.

































> t t E...'* 1054 .. .......... > ., ....... . .._. oi- ...'B


4 1 100



101
10- -1
1002 ... .. ... ..10 1.




0 20 40 60 80100120140160180200 0 20 40 60 80100120140160180200 0 20 40 60 80 100201401601802C
M(HLc tracK) (GeV) M(JHLTvsAM) (GeV) M.HLVGMR) (GE


Figure C-l: Z 2p invariant mass peak: built from the HLT muon and all other tracks,
the HLT muon and all other standalone muons and the HLT muon and all other globally
reconstructed muons.
























128









APPENDIX D
SENSITIVITY OF THE MUON ISOLATION CUT EFFICIENCY TO THE
UNDERLYING EVENT UNCERTAINTY


(A. Drozdetskiy et al., "Sensitivity of the Muon Isolation Cut Efficiency to the Underlying
Event Uncert ,iiii -
C'\S Note 2006/033.)
The isolation cut efficiency per muon due to uncertainties in the considered
underlying event (UE) models vary as much as 5'.- (the efficiency itself and its
uncertainty strongly depend on how tight the isolation cut is). The 4-muon isolation
cut efficiency per event for ZZ -- 4p background is measured to be (78 6) .
To decrease these large uncertainties to a negligible level with respect to other
systematic uncertainties, we propose to calibrate the isolation cut efficiency from data
using Z-inclusive events (Z -- 2p) and the random-cone technique. We show that this
indeed significantly decreases the uncertainties associated with a poor understanding
of the UE physics. There might be ~- "' systematic shift in the 4-muon isolation cut
efficiencies obtained this way. In principle, one could correct for this shift, but it does not
appear to be necessary as this uncertainty is already smaller than other systematic and
statistical errors.









APPENDIX E
GENETIC ALGORITHM FOR RECTANGULAR CUTS OPTIMIZATION


(A. Drozdetskiy et al., "GARCON: Genetic Algorithm for Rectangular Cuts
OptimizatioN. User's manual for version 2.0", hep-ph/0605143,
http://drozdets.home.cern.ch/drozdets/home/genetic/)
Typically HEP analysis has quite a few selection criteria (cuts) to optimize for
example a significance of the -,s1 I!" over "background" events: transverse energy/momenta
cuts, missing transverse energy, angular correlations, isolation and impact parameters, etc.
In such cases simple scan over multi-dimensional cuts space (especially when done on top
of a scan over theoretical predictions parameters space like for SUSY e.g.) leads to CPU
time demand varying from d ,v- to many years... One of the alternative methods, which
solves the issue is to employ a Genetic Algorithm (GA), see e.g. [113, 114, 115].
We wrote a code, GARCON [100], which automatically performs an optimization
and results stability verification effectively trying ~ 1050 cut set parameters/values
permutations for millions of input events in hours time. Examples of analyses are
presented in the Physics TDR [8] and recent papers [1, 101, 102, 103].
In comparison to other automatized optimization methods GARCON output is
transparent to user: it just z-,v what rectangular cut values are optimal and recommended
in an analysis. Interpretation of these cut values is absolutely the same as with eye-balling
cuts when one selects a set of rectangular cut values for each variable in a "classical" way
by ev,
All-in-all it is a simple yet powerful ready-to-use tool with flexible and transparent
optimization and verification parameters setup. It is publicly available along with a paper
on it [100] consisting of an example case study and user's manual.









REFERENCES


[1] A. Drozdetskiy et al., Search strategy for the Standard Model Higgs boson in the
H -+ ZZ(*) 4/ decay channel using M(4p)-dependent cuts, 2006, 2006/122 C\!S
Note.

[2] A. Drozdetskiy et al., Relative contributions of t- and s-channels to the ZZ -+ 4/
process, 2006, 2006/057 C'\ S Note.

[3] A. Drozdetskiy et al., Measuring muon reconstruction efficiency from data, 2006,
2006/060 C\ !S Note.

[4] A. Drozdetskiy et al., Sensitivity of the muon isolation cut efficiency to the
underlying event uncertainties, 2006, 2006/033 C'\ S Note.

[5] P. Bartalini et al., Study of PDF and QCD scale uncertainties in pp -+ ZZ -+ 4/z
events at the LHC, 2006, 2006/068 C'\!S Note.

[6] P. Bartalini et al., NLO vs. LO: kinematical differences for signal and background in
the H -+ ZZ(*) -+ 4P analysis, 2006, 2006/130 C'\IS Note.

[7] M. Della Negra, A. Petrilli, A. Herve, L. Foa, et al., C'l!S physics: technical design
report, volume I, 2006, (C' \S collaboration), in particular see section 9.3.4.

[8] M. Della Negra, A. Petrilli, A. Ball, L. Foa, et al., C'\!S physics technical design
report, volume II, 2007, (C'\IS collaboration), in particular see sections 3.1,
Appendix A.2, Appendix D.

[9] V. Buescher et al., Tevatron-for-LHC report: preparations for discoveries, 2006,
hep-ph/0' .1 1 ;-2

[10] C. Buttar et al., Les houches physics at TeV colliders 2005, standard model and
Higgs working group: summary report, 2005, contributed to Les Houches workshop
on physics at TeV colliders, Les Houches, France. hep-ph/0604120.

[11] A. Drozdetskiy et al., SM Higgs boson: recent developments at C' IS and ATLAS,
2007, (C' IS and ATLAS collaboration), talk given at The XLIInd Rencontres de
Moriond on Electroweak Interactions, La Thuile, Italy, 2007 (will be published in the
conference proceedings).

[12] A. Drozdetskiy et al., C'\MS Detector sensitivity to the Standard Model Higgs boson
in H -+ ZZ(*) -- 4 decay channel, 2007, (C\ IS collaboration), talk given at
American Physical April 2007 meeting.

[13] A. Drozdetskiy and S. Abdullin, GARCON: genetic algorithm for rectangular cuts
optimizatioN, 2007, talk given at The XI International Workshop on Advanced
Computing and Analysis Techniques in Physics Research, Amsterdam, Netherlands,
2007.









[14] A. Drozdetskiy et al., C'\!S detector sensitivity to the Standard Model Higgs boson
in H ZZ(*) 4/ decay channel, 2007, (C'\!S collaboration), talk given at The
Hadron Collider Physics Symposium 2007, Elba, Italy (will be published in the
conference proceedings).

[15] S. Abdullin et al., Search strategy for the standard model Higgs boson in the
H ZZ(*) 4/ decay channel using M(4mu)-dependent cuts, 2007, (C'\!S
collaboration), talk given at The Physics at LHC, Cracow, Poland, 2006 (will be
published in proceedings).

[16] A. Drozdetskiy et al., Sensitivity of the muon isolation cut efficiency to underlying
event uncertainty, 2006, (C'\!S collaboration), talk given at The 2nd HERA LHC
workshop, CERN, Geneva, Switzerland (published in proceedings).

[17] A. Drozdetskiy et al., MCANLO vs PYTHIA for H ZZ(*) z 4/, 2005, (C'\IS
collaboration), talk given at The Les Houches 2005: Physics at TeV Colliders, Les
Houches, France.

[18] A. Drozdetskiy et al., Effect of UE on isolation, 2005, (C'\ S collaboration), talk
given at The Les Houches 2005: Physics at TeV Colliders, Les Houches, France
(published in proceedings).

[19] A. Drozdetskiy et al., The discovery potential of supersymmetry at C' \S within
the mSUGRA model using same-sign di-muons., 2005, (C'\ S collaboration),
talk given at The 40th Rencontres de Moriond on QCD and High Energy
Hadronic Interactions, La Thuile, Aosta Valley, Italy (published in proceedings,
hep-ex/0505034).

[20] A. Drozdetskiy et al., First detailed study on the C \!S SUSY discovery
potential with two same sign muons in the mSUGRA model, 2004, (C' !S
collaboration), talk given at The Physics at LHC, Vienna, Austria (Published in
Czech.J.Phys.55:B249-B256,2005).

[21] A. Drozdetskiy et al., internal C'\ S collaboration meetings of different levels,
2002-2007, see: http://indicosearch.cern.ch/ and http://agenda.cern.ch/search.php.

[22] C.-N. Yang et al., Conservation of isotopic spin and isotopic gauge invariance. Phys.
Rev., 96:191-195, 1954.

[23] P. Higgs et al., Broken symmetries, massless particles and gauge fields. Phys. Lett.,
12:132-133, 1964.

[24] P. Higgs et al., Broken symmetries and the masses of gauge bosons. Phys. Rev.
Lett., 13:508-509, 1964.

[25] S. Glashow, Partial symmetries of weak interactions. Nucl. Phys., 22:579-588, 1961.

[26] S. Weinberg, Phys. Rev. Lett. 19, 1264 1266 (1967).









[27] A. Salam, Elementary particle theory,. Almqvist and Wiksells, Stockholm, 1968.

[28] G. Arnison et al., Experimental observation of lepton pairs of invariant mass around
95- gev/c2 at the cern sps collider. Phys. Lett., B126:398-410, 1983.

[29] G. Arnison et al., Experimental observation of isolated large transverse
energy electrons with associated missing energy at / =s 540-gev. Phys. Lett.,
B122:103-116, 1983.

[30] J. F. Gunion et al., The Higgs hunter's guide, Westview Press, 2001.

[31] M. Spira et al., Electroweak symmetry breaking and Higgs physics, hep-ph/9803257.

[32] T. Hambye et al., Phys. Rev. D 55, 7255 (1997), arXiv:hepph/9610272.

[33] R. Barate et al., LEP working group for Higgs boson searches, Phys. Lett. B 565, 61
(2003), arXiv:hep-ex/0306033.

[34] Tevatron new phenomena & Higgs working group, combined DO and
CDF upper limits on the Standard Model Higgs boson production,
http://tevnphwg.fnal.gov/results/d0conf_5227/, CDF Note 8384, DO Note 5227.

[35] M. Griinewald, Precision tests of the Standard Model, hep-ex/0511018.
[36] LEP Electroweak Working Group. LEPEWWG/2005-01, arXiv:hep-ex/0511027,
http://lepewwg.web.cern.ch/LEPEWWG/ (Status of March 2007).

[37] A. Djouadi, J. Kalinowski, and M. Spira, HDECAY: a program for Higgs boson
decays in the Standard Model and its supersymmetric extension, Comput. Phys.
Commun., vol. 108 (1998), 56-74, hep-ph/9704448.

[38] M. Spira, HIGLU: a program for the calculation of the total Higgs production cross
section at hadron colliders via gluon fusion including QCD corrections, (1995),
hep-ph/9510347.

[39] M. Gonzalez-Garcia et al., Rev. Mod. Phys. 75, 345 (2003), hepph/0202058.
[40] Y. Fukuda et al., Nucl. Instrum. Meth. A 501, 418 (2003).

[41] A. Suzuki et al., Nucl. Instrum. Meth. A 453, 165 (2000), hep-ex/0004024.

[42] A. Suzuki et al., Nucl. Instrum. Meth. A 453, 165 (2000), hep-ex/0004024.

[43] A. Piepke et al., Nucl. Phys. Proc. Suppl. 91, 99 (2001).

[44] E. Kolb et al., The Early universe, Addison-Wesley, Redwood City, USA (1990),
Frontiers in physics, 69.

[45] A. Sakharov, Pisma Zh. Eksp. Teor. Fiz. 5, 32 (1967) (JETP Lett. 5, 24 (1967
SOPUA,34,392-393.1991 UFNAA,161,61-64.1991)).









[46] N. Cabibbo et al., Phys. Rev. Lett. 10, 531 (1963).

[47] C. Bennett et al., Astrophys. J. Suppl. 148, 1 (2003), astro-ph/0302207.
[48] D. Spergel et al., Astrophys. J. Suppl. 148, 175 (2003), astro-ph/0302209.

[49] E. Witten, Nucl. Phys. B 188, 513 (1981).
[50] S. Dimopoulos et al., Nucl. Phys. B 193, 150 (1981).

[51] S. Dimopoulos et al., Phys. Rev. D 24, 1681 (1981).

[52] W. Yao et al., (Particle Data Group Collaboration), J. Phys. G33, 1 (2006).

[53] LHC Project, http://lhc.web.cern.ch/lhc/,
http://lhc-machine-outreach.web.cern.ch/lhc-machine-outreach/lhc-in-pictures.htm.

[54] CERN, http://www.cern.ch/.

[55] C\ !S Collaboration, http://cms.cern.ch/.
[56] ATLAS Collaboration, http://atlas.web.cern.ch/Atlas/index.html.

[57] ALICE Collaboration, http://aliceinfo.cern.ch/.
[58] LHCb collaboration, http://lhcb.web.cern.ch/lhcb/.

[59] TOTEM collaboration, http://totem.web.cern.ch/Totem/.
[60] C'\!S Collaboration, The compact muon solenoid detector at LHC, in preparation,
2007.

[61] C'\!S Collaboration, The Compact Muon Solenoid Technical Proposal,
CERN/LHCC 94-38 (1994) LHCC/P1.

[62] C'\!S Magnet Collaboration, The magnet project: Technical Design report, CERN,
volume CERN/LHCC97-10, C \!S TDR 1, 1997; A. Herve', et al., Status of the
Construction of C'\!S magnet, IEEE Trans. Appl. Superconduct, Vol 14, No 2, pp.
524-547, June 2004; A. Herve', The C'\MS Detector Magnet, IEEE Trans on Appl.
Superconductor, Vol. 10, Nol, pp. 389-394, 2000; F. Kircher, et al., Final Design
of the C'\!S Solenoid Cold Mass, IEEE Trans on Appl. Superconduct., Vol 10, nl,
407-410, March 2000.

[63] C'\!S Collaboration, The Tracker Project Technical Design Report, CERN/LHCC
98-006, C\ !S TDR 5, April 1998; C'\!S Collaboration, Addendum to the C'\iS
Tracker TDR, CERN/LHCC 2000-016, CM\!S TDR 5 Addendum 1, April 2000.

[64] C'\!S Collaboration, C'\MS DAQ and HLT technical design report, CERN/LHCC
2002-26, 2002.









[65] C'\!S Collaboration, The muon project. Technical design report, CERN/LHCC,
97-32, 1997.

[66] G. C'I i'pak et al., High-accuracy, two-dimensional read-out in multiwire
proportional chambers, Nucl. Instrum. Meth. 113:381-385, 1973.

[67] C. Anderson et al., Effect of gas composition on the performance of cathode strip
chambers for the C'\!S Endcap Muon System, C' \! Note 2004/033.

[68] M. Baarmand et al., Spatial resolution attainable with cathode strip chambers at
the tri -r level, Nucl. Instrum. Meth. A425:92-105, 1999.

[69] On 'Global Muon Momentum Resolution' see for example: C'\ S Collaboration,
Physics Technical Design Report, vol.1, CERN/LHCC, 2006-001, 2006.

[70] On 'muon (TC) segment reconstruction algorithm' see for example: C'\ !S
Collaboration, Physics Technical Design Report, vol.1, CERN/LHCC, 2006-001,
2006.

[71] J. Chiba et al., Study of position resolution for cathode readout MWPC with
measurement of induced charge distribution Nucl. Instrum. Meth. 206 (1983), p.451.

[72] D. Acosta et al., Large C'\IS cathode strip chambers: design and performance, Nucl.
Instrum. Meth. 453 (2000) 182-187.

[73] M. Baarmand et al., Spatial resolution attainable with cathode strip chambers at
the tri -r level, Nucl. Instrum. Meth. A425:92-105, 1999.

[74] Y. Pakhotin et al., Performance of endcap muon system of the compact muon
solenoid, American Physical Society Conference, April 14-17, 2007, Jacksonville,
Florida, USA.

[75] C'\ !S collaboration, Summary report on magnet test and cosmic challenge in
preparation, 2007.

[76] V. Barashko et al., Performance validation tests of the cathode strip chambers for
C'\ !S muon system, in Nuclear Science Symposium Conference Record, 2005 IEEE,
Vo. 2, p.827-829, 23-29 Oct. 2005.

[77] E. Gatti et al., Optimum geometry for strip cathodes or grids in mwpc for avalanche
localization along the anode wires, Nucl. Instrum. Meth. 163 (1979) 83.

[78] J. Neyman and E. Pearson, On the use and interpretation of certain test criteria for
purposes of statistical inference, Part I, Biometrika, Vol. 20A, No.1/2 (July 1928),
pp.175-240.

[79] J. Neyman and E. Pearson, On the problem of the most efficient tests of statistical
hypotheses, Philosophical Transactions of the Royal Society of London, Vol. 231
(1933), pp. 289-337.









[80] S. Wilks, The large-sample distribution of the likelihood ratio for testing composite
hypotheses, Annals of Mathematical Statistics, 9:60-62, 1938.

[81] C'\ S collaboration, Object-oriented simulation for C'\ S ,iin ,iS-i and reconstruction,
http://cmsdoc.cern.ch/OSCAR.

[82] C'\ S collaboration, Object-oriented reconstruction for C'\ S analysis,
http://cmsdoc.cern.ch/ORCA.

[83] S. Abdullin et al., Summary of the C'\!S Potential for the Higgs boson discovery,
C'\!S Note 2003/033.

[84] M. Sani et al., Search for the Standard Model Higgs boson in four-muon final state
with C' MS, C' S CR-2004/035, proceeding of Ph,- at LHC, Vienna, Austria, July
2004.

[85] V. Bartsch, Simulation of silicon sensors and study of the Higgs decay H -+ ZZ(*)
4p for C'\!S (LHC), Ph.D. thesis, IEKP-KA/2003-26 University of Karlsruhe (2003).

[86] M. Ald-v-i et al., Search for the Standard Model Higgs boson in the H -+ ZZ(*) 4
decay channel using a mass-independent analysis, C\ !S Note 2006/106.

[87] S. Baffioni et al., Search for the Standard Model Higgs Boson in the four-electron
final state with C'\!S, C'\IS Note 2006/115.

[88] D. Futyan et al., Search for the Standard Model Higgs boson in the two-electron and
two-muon final state with C'M\S, CMS Note 2006/136.

[89] T. Sjostrand, P. Eden, C. Friberg, L. Lonnblad, G. Miu, S. Mrenna, and E. Norrbin,
Comput. Phys. Commun. 135, 238 (2001).

[90] C'\!S collaboration, C'\IS interface for event generators, C'I\I IN,
http://cmsdoc.cern.ch/cmsoo/projects/C\ I Il N.

[91] E. Barberio et al., Computer Phys. Commun. 79 (1994) 291.

[92] E. Barberio et al., Computer Phys. Commun. 66 (1991) 115.

[93] M. Spira et al., HDECAY: a program for Higgs Boson decays in the Standard Model
and its supersymmetric extension, hep-ph/9704448.

[94] M. Spira et al., HIGLU: a program for the calculation of the total Higgs production
cross section at hadron colliders via gluon fusion including QCD corrections,
hep-ph/9510347.

[95] Particle Data Group, Review of Particle Properties, Phys. Lett. B 562 (2004) 1.

[96] F. Maltoni, Theoretical issues and aims at the Tevatron and LHC, proceeding
of The Hadron Collider Physics Symposium, Les Diableretes, Switzerland, July









2005, printed version: Springer Proceedings in Physics, Vol. 108, 2006, ISBN:
3-540-32840-8.

[97] CompHEP collaboration, CompHEP a package for evaluation of Feynman
diagrams and integration over multi-particle phase space. User's manual for version
33, year =.

[98] J. Campbell, W/Z + B anti-B / jets at NLO using the Monte Carlo MCFM,
hep-ph/0105226.

[99] C. Zecher et al., Leptonic signals from off-shell Z boson pairs at hadron colliders,
hep-ph/9404295.

[100] A. Drozdetskiy et al., GARCON: genetic algorithm for rectangular cuts
optimization. User's manual for version 2.0, 2006, hep-ph/0605143,
http://drozdets.home.cern.ch/drozdets/home/genetic/.

[101] D. Acosta et al., Potential to discover supersymmetry in events with muons, jets and
missing energy in pp collisions at s = 14 TeV with the CM\!S detector, C'\S Note
2006/134.

[102] V. Abramov et al., Selection of single top events with the C'\!S detector at LHC,
C'\!S Note 2006/084.

[103] W. Boer et al., Trilepton final state from neutralino-chargino production in
mSUGRA, C'\IS Note 2006/113 (or just PTDR).

[104] F. Maltoni et al., MadEvent: automatic event generation with MadGraph, JHEP
0302 (2003) 027, arXiv:hep-ph/0208156.

[105] C'\!S Collaboration, The Ti i--- r and Data Acquisition project, v.2, p.308, 2002.

[106] E. Lehmann, A Theory of some multiple decision problems, The Annals of
Mathematical Statistics, Vol. 28 1, 1, (1957).

[107] R. O'Neill et al., The present state of multiple comparison methods, Journal of the
Royal Statistical Society. Series B (\. I I dological), Vol. 33 2, 218, (1971).

[108] J. Hauser, Search for new physics in tails of distributions, private communications,
paper in preparation.

[109] B. Knuteson et al., A quasi-model-independent search for new high pr physics at
DO, thesis, 2000.

[110] B. Abbott et al., Search for new physics in e mu X data at DO using sleuth: a
quasi-model-independent search strategy for new physics, DO Collaboration,
Phys.Rev. D62 (2000) 092004, hep-ex/0006011.









[111] B. Abbott et al., A quasi-model-independent search for new physics at large
transverse momentum, DO Collaboration, Phys.Rev. D64 (2001) 012004,
hep-ex/0011067.

[112] C. Finley et al., On the evidence for clustering in the arrival directions of AGASA's
ultrahigh energy cosmic rays, Astroparticle Physics 21 (2004) 359-567.

[113] J. Holland, Adaptation in natural and artificial systems. The University of Michigan
Press, Ann Arbor, 1975.

[114] D. Goldberg, Genetic algorithms in search, optimization and machine learning.
Addison Wesley, 1989.

[115] S. Abdullin, Genetic algorithm for SUSY trigger optimization in C'\!S detector at
LHC, NIM A 502 (2003) 693-695; Genetic algorithm for SUSY tri ..-r optimization,
talk given at the IV Conference LHC D-,v- in Split, October 8-12, 2002.









BIOGRAPHICAL SKETCH

Alexey A. Drozdetskiy was born on March 31, 1978, in the city of Novosibirsk in

Russia. After graduating with honors from high school in 1995, he continued his education

at the Novosibirsk State University (NSU).

In 1999 he received his bachelor of science degree with honors, and in 2001 he

received his master's degree with honors from the same university. While studying in the

NSU, he worked in the Budker Institute in the Spherical Neutral Detector (SND) group

on the VEPP2M, e+e-collider. His work was related to developing new detectors for

X-rays as well as studying a possible upgrade for the SND detector for forward physics at

VEPP2M.

During his period of study at NSU he was awarded 4 years with International Soros

Student Stipend in Physics, 2 years with Special NSU stipend and 8 times with student

conferences diplomas.

In the winter of 2002, Alexey became a graduate student in the Department

of Physics, University of Florida, Gainesville, FL. In the same year he joined C' \S

collaboration and worked with professors Guenakh Mitselmakher, Andrey Korytov and

Darin Acosta on various aspects of experimental particle physics, particularly on the

H -+ ZZ(*) 4p C'\S discovery potential (as well as C'\!S discovery potential for

SUSY in final state with muons and Ep"88). Since 2002 Alexey has been participating in

commissioning of the Endcap Muon System (EMU) of the C'\ !S detector (starting from

electronics/detector tests at UFL site and up to the most recent performance studies and

fast High Level Ti i---. i algorithm development for the EMU, validated with real cosmic

muons data, taken by C'\S detector slice in fall of 2006).

In the period of his study at UF, Alexey was awarded with the UF Presidential

Recognition, UF International Student Academic Award, and 5 Certificates of Achievements.

He is co-author of more than 10 refereed publications, and results of the studies he






140


performed together with colleagues were presented at more than 10 international

conferences.

He is currently based in Geneva, Switzerland, close to LHC and C'\IS experiment

(CERN international laboratory), and gives free Yoga lessons to the community on

voluntary basis.

In 2007, Alexey A. Drozdetskiy graduated from the University of Florida with the

degree of doctor of philosophy.





PAGE 1

1

PAGE 2

2

PAGE 3

3

PAGE 4

Friendshipisunnecessary,likephilosophy,likeart...Ithasnosurvivalvalue;ratheritisoneofthosethingsthatgivevaluetosurvival.{C.S.LewisThisisprobablyoneofthehardestpartsofthedissertation.AndnotbecauseIhavehardtimeinrememberingallthepeoplewithwhomIhaveworked,whohelpedmeandsupportedme,butbecauseno\sensible"limitsfordissertationsizewouldeverbeabletoaccommodatethelistofallofthem.It'ssoforaverysimplereason:forexamplemanypeoplefromtheUniversitywereworkingwithme,weusedsomethingproducedbeforebyothers,ourfamiliesandclosetouspeoplewerelovingandpatientenoughtoallowustoquenchourthirstinresearch,etc...andwecancontinuetoexpandthelistthiswayonandon.Somesayittakesjustafewpeopletoconnectanyoneofuswithanyoneelsethroughcommonwork,friends,etc.Soit'sclearwhatanimpossibletaskitistolistonafewpagesallthepeople.AndIwouldliketothankall,whohelped,supportedandworkedwithme,andinuencedmeknowinglyorunknowingly:Irememberallinmymindandmyheart.Andifpaperisnotlargeenough,theyare.I'mgratefultoall,includingUniversityofFlorida(UFL,Gainesville,USA)professors,postdocs,students,andnotless-allthewonderfulocemanagersandpersonal;membersoftheCMScollaboration(forcountlessworkhoursandout-of-workmeetingsanddiscussions,helpandsupport);peoplefromNovosibirskStateUniversity(Russia)andBudkerInstitute(Novosibirsk,Russia);allteachersandfriendsinmylife.Still,thereareafewpeopleIwouldliketothankalittlemorethanothers:GuenakhMitselmakherandAndreyKorytov-mysupervisors,foralltheirexperienceandknowledgetheypassedontome,bothonphysicsandnotlessimportantlyonlifeingeneral.Ithankthemforalwaysgivingmethefreedomtogetmyowngrasponatopic.AndmaybeevenmoreimportantlyIwouldliketothankthemfortheirconstantkindand 4

PAGE 5

5

PAGE 6

page ACKNOWLEDGMENTS ................................. 4 LISTOFTABLES ..................................... 9 LISTOFFIGURES .................................... 10 ABSTRACT ........................................ 16 CHAPTER 1INTRODUCTION .................................. 18 2THESTANDARDMODELHIGGSBOSON .................... 20 2.1TheStandardModel:ElectroweakSymmetryBreaking. ........... 20 2.2TheStandardModelHiggsBosonMass:CurrentExperimentalandTheoreticalConstraints ................................... 22 2.2.1TheoreticalConstraints ......................... 22 2.2.2SearchesatLargeElectron-PositronCollider(LEP) ......... 23 2.2.3SearchesatTevatron .......................... 24 2.2.4ElectroweakPrecisionMeasurements ................. 24 2.3TheStandardModelHiggsBosonatLHC .................. 25 2.3.1Production ................................ 25 2.3.2Decays .................................. 26 2.4TheStandardModelHiggsBosonSearchesatCMSSummary:DiscoveryReach ...................................... 28 2.5IssuesintheStandardModel ......................... 28 2.6SupersymmetryHiggs ............................. 30 3LARGEHADRONCOLLIDER ........................... 32 4COMPACTMUONSOLENOID(CMS)DETECTOR .............. 35 4.1Introduction ................................... 35 4.2GeneralDescription ............................... 36 4.3TheSuperconductingMagnet ......................... 38 4.4InnerTrackingSystem ............................. 39 4.5ElectromagneticCalorimeter(ECAL) ..................... 40 4.6HadronCalorimeter(HCAL) .......................... 41 4.7MuonSystem .................................. 42 4.7.1BarrelMuonSystem .......................... 42 4.7.2EndcapMuonSystem .......................... 43 4.7.3ResistivePlateChambers ........................ 43 4.8MuonTrigger .................................. 43 6

PAGE 7

...................... 45 5.1Introduction ................................... 45 5.2ChamberMechanicalDesign .......................... 48 5.3ElectronicDesign ................................ 52 6FASTALGORITHMFORTRACKSEGMENTANDHITRECONSTRUCTIONINCATHODESTRIPCHAMBERS ........................ 60 6.1Introduction ................................... 60 6.2AlgorithmPrinciples .............................. 62 6.2.1ProtoSegmentBuilding ......................... 63 6.2.2RecHitsandTheirCoordinates .................... 64 6.2.3RenedSegment ............................. 67 6.3AlgorithmValidationatMTCC ........................ 68 6.3.1ProtoSegments .............................. 68 6.3.2SpatialResolution ............................ 69 6.3.2.1Resolutionforhigh-pTmuons ................ 71 6.3.2.2Resolutionforhigh-pTmuonswithtabulatedpedestals .. 76 6.3.2.3Resolutionforlow-pT,orhighly-inclined,muons ...... 78 6.3.2.4Resolutionforhigh-pTmuonswithcharge-dependentsigmas 79 6.3.3RenedSegments ............................ 79 6.4Conclusions ................................... 81 6.5RatioMethod .................................. 83 7SEARCHSTRATEGYFORTHESTANDARDMODELHIGGSBOSONINTHEH!ZZ()!4DECAYCHANNELUSINGM(4)-DEPENDENTCUTS. 88 7.1Introduction ................................... 88 7.2PhysicsProcessesandTheirSimulation .................... 90 7.2.1Signal:H!ZZ()!4 91 7.2.2Background:tt ............................. 92 7.2.3Background:(Z=)bb!2bb 92 7.2.4Backgrounds:qq!Z=Z=!4andqq!Z=Z=!22 93 7.3TriggerandOineMuonReconstruction ................... 93 7.4HiggsSearchStrategy ............................. 94 7.4.1IntroductoryRemarksonSignicance ................. 95 7.4.2OptimizationoftheM(4)-DependentCuts ............. 97 7.4.3SystematicErrors ............................ 101 7.4.3.1Uncertaintiesinthebackground ............... 101 7.4.3.2Signicancewiththebackgrounduncertaintiesincluded .. 106 7.4.4LocalSignicanceandOverallStatisticalFluctuationProbability .. 107 7.5Summary .................................... 108 7.6Onthetruesignicanceofalocalexcessofevents .............. 109 7

PAGE 8

................................... 122 APPENDIX ARELATIVECONTRIBUTIONSOFT-ANDS-CHANNELSTOTHEZZ()!4PROCESS ..................................... 124 BZZDISCOVERYWITH<1FB1WITHCMSINZZ()!4MODE ..... 126 CMEASURINGMUONRECONSTRUCTIONEFFICIENCYFROMDATA ... 127 DSENSITIVITYOFTHEMUONISOLATIONCUTEFFICIENCYTOTHEUNDERLYINGEVENTUNCERTAINTY ..................... 129 EGENETICALGORITHMFORRECTANGULARCUTSOPTIMIZATION .. 130 REFERENCES ....................................... 131 BIOGRAPHICALSKETCH ................................ 139 8

PAGE 9

Table page 2-1TheStandardModelelementaryparticles. ..................... 20 7-1TheLO/NLOcrosssectionsforvariousHiggsbosonmassesandbackgrounds,correspondingnumberofeventswithfourmuonssurvivingthegeneratorlevelpreselectioncuts(seesection 7.2 )calculatedforL=30fb1,andthenumberofsimulatedevents. .................................. 91 7-2Summaryoftheresults:numberofsignalandbackgroundeventsinawindowusedforacountingexperimentwiththeM(4)-dependentcuts.SystematicerroronthebackgroundisnormalizedtotheZ!2process(KNLO=KNLOisnotincluded);threedierentsignicanceswithoutsystematicerrorsincluded:theSLestimatorfortheLogLikelihoodRatio(LLR)builtforthefullM(4)spectrum,ScLLLRestimatorbuiltforacountingexperimentapproach,andtheSctruesignicanceforthecountingexperimentapproach;thenalresultforSc,nowincludingallsystematicerrors. ..................... 108 9

PAGE 10

Figure page 2-1Divergent(topthree)WWcross-sectiongraphsandtheircancellation(bottomtwographs). ...................................... 22 2-2TheoreticalboundsontheSMHiggsmassasafunctionofthecut-oscale. .. 22 2-3Expectedandobserved95%CLcrosssectionratiosforthecombinedCDFandD0analyses,andtheexpected95%CLratiosfortheCDFandD0experimentsalone. ......................................... 24 2-4The2curvederivedfromhigh-Q2precisionelectroweakmeasurements. ... 24 2-5TypicaldiagramsforallrelevantHiggsbosonproductionmechanismsatleadingorderattheLHC. ................................... 26 2-6HiggsproductioncrosssectionsattheLHCforthevariousproductionmechanismsasafunctionoftheHiggsmass. ........................... 26 2-7TotaldecaywidthandbranchingratiosofthedominantdecaymodesoftheSMHiggsbosonparticle. ............................... 27 2-8Theintegratedluminosityneededforthe5discoveryoftheinclusiveHiggsbosonproduction. ................................... 29 2-9ThesignalsignicanceasafunctionoftheHiggsbosonmassfor30fb1oftheintegratedluminosityforthedierentHiggsbosonproductionanddecaychannels. 29 3-1AcceleratorcomplexatCERN. ........................... 33 4-1AnexplodedviewoftheCMSdetector. ....................... 37 5-1Quater-viewoftheCMSDetectors.CathodeStripChambersoftheEndcapMuonSystemarehighlighted. ............................ 46 5-2TheME2cathodestripchambers.Theouterringconsistsof36ME2/2chambers,eachspanning10in;while1820-degreeME2/1chambersformtheinnerring.Thechambersoverlaptoprovidecontiguouscoveragein. ............ 47 5-3AschematicviewofaCMScathodestripchambermadeofseventrapezoidalpanels. ......................................... 48 5-4AschematicviewofasinglegapillustratingtheprincipleofaCSCoperation. 48 5-5MechanicaldesignoftheCMSCathodeStripChambers(explodedview). .... 49 5-6CSCgasgainvshighvoltage. ............................ 51 5-7ME2/1chambersinglesratevshighvoltage(theoverallsensitiveareaofallsixplanesinthischamberis9.5m2). ......................... 51 10

PAGE 11

............. 53 5-9Anoverviewofallcustom-madeCSCtriggerandreadoutelectronics.Functionalityofvariousboardsisdescribedinthetext. ...................... 53 5-10MuonsignalsasseenattheAFEBamplieroutput. ............... 55 5-11Aschematiceventdisplayshowinganodesignalsinasix-planechamber. .... 55 5-12BasicfunctionaldiagramofaCathodeFront-EndBoard. ............. 56 5-13Muonhitsignalsfromsixnearbystrips.Fourcurvesareactualoscillograms. .. 56 5-14Asimpliedschematicillustratingtheideabehindthecomparatornetwork. .. 57 5-15CLCT(CathodeLocalChargedTrack)isapatternofhalf-striphitsconsistentwithamuontrack. .................................. 57 6-1Thenoisedistributionforathree-sample-sumintheabsenceofasignalforthelargestME23/2chambers. .............................. 69 6-2Thethree-stripclusterchargedistributionforthelargestME23/2chambers. .. 69 6-3Thenoisedistributionforasingle-sampleintheabsenceofasignalforthelargestME23/2chambers. .................................. 70 6-4Thenoisedistributionforatwo-samplesumintheabsenceofasignalforthelargestME23/2chambers. .............................. 70 6-5The3rd-planeresidualsforthevedistinctjxjstrippartrangesforthemiddlepartofthelargestME23/2chambers.Andasummaryforasingle-planespatialresolutionasafunctionofahitpositionacrossastrip. .............. 73 6-6The3rd-planeresiduals'svs.vejxjstrippartrangesandoverallsix-planeresolutionvs.jxj-partofastrip(inunitsofstripwidthsandmicrons).[Notationsare:blackemptysquares-HVsegment1;redsquares-HVsegment2;trianglesup-HVsegment3;trianglesdown-HVsegment4;stars-HVsegment5.] ... 74 6-7The3rd-planeresiduals'svs.jx=wj-partofastripandoverallsix-planeresolutionvs.jx=wj-partofastrip(inunitsofstripwidthsandmicrons).SamenotationsasforFig. 6-6 .................................... 75 6-8Theoverallsix-planechamberresolutionvs.vejxjstrippartrangesfordynamicallymeasuredpedestals.SamenotationsasforFig. 6-6 ................ 77 6-9Theoverallsix-planechamberresolutionvs.vejxjstrippartrangesforcalibratedpedestals.SamenotationsasforFig. 6-6 ...................... 77 6-10Inclinationanglesfortracksegmentsreconstructedfromdi-stripCLCTs(ME1/1chambers). ....................................... 77 11

PAGE 12

6-6 ........................................ 77 6-12ProbabilitiesfordierentnumbersofchargeclustersfoundinassociationwithME23/2chamberProtoSegments. .......................... 80 6-13LayerchargeclustersoccupancyinassociationwithME23/2chambersProtoSegments. 80 6-14The2/dofdistributionsforalineartforME23/2chamberProtoSegmentswithsixclusters,veclusters,andfourclusters. .................. 81 6-15Scatterplotofmin(25/dof)vs.26/dofforsix-clusterME23/2chamberProtoSegments. 81 6-16Occupancyforjx=wj-coordinatesofRenedSegments. ............... 82 6-17PoolsforallretainedRecHits(i.e.,RecHitsassociatedwithRenedSegments.) 82 6-18InducedchargedistributioncalculatedaccordingtoGattiforlargeandME1/1chambergeometries. ................................. 82 6-19RatiorversusalocalcoordinatexforlargechamberscalculatedforvarietyofstripwidthsintheassumptionoftheGattichargedistributionforlargechambers. 82 6-20Correctionthatmustbeaddedtothemeasuredratiortoobtainthehitpositionacrossastrip. .................................... 84 6-21Anoccupancydistributionforthe1st-ordercorrectedcoordinatex1st. ...... 84 6-22Thesecond-ordercorrectionthatmustbeaddedtox1toobtainthehitpositionacrossastripx2. ................................... 85 6-23Anoccupancydistributionforthe2nd-ordercorrectedcoordinatex2. ...... 85 6-24Sensitivityoftheratiomethodtoelectronicnoise(in%ofstripwidthandmm). 86 6-25Sensitivityoftheratiomethodtoerrorsinelectronicgaincalibrations(in%ofstripwidthandmm). ................................ 86 6-26Sensitivityoftheratiomethodtouncertaintiesincrosstalksbetweenstrips(in%ofstripwidthandmm). .............................. 87 7-1M(4)distributionsaftergenerator-levelcutsfortt,(Z=)bb,Z=Z=,andmH=140GeV=c2(logscale). ............................ 110 7-2SameasFigure 7-1 ,butonalinearscale. ...................... 110 7-3Enhancementtothesignalsamples'crosssectionsduetointerferenceeectsnotaccountedforatthegeneratorlevel. ...................... 110 12

PAGE 13

98 ]. ............................. 110 7-5GlobalmuonreconstructioneciencycalculatedfrommatchingreconstructedandtrueMonteCarlomuonsinthebarrelregionvs.pT. ............. 111 7-6GlobalmuonreconstructioneciencycalculatedfrommatchingreconstructedandtrueMonteCarlomuonsintheendcapregionvs.momentum. ........ 111 7-7M(4)distributionform(H)=150GeV=c2andthetdescribedinthetext. ... 111 7-8M(4)resolutionvs.MH. .............................. 111 7-9Comparisonofdierentsignicanceestimatorsfor1backgroundevent:probabilityofmeasuringsignicanceS>S0,backgroundonlycase,NB=1event. ..... 112 7-10SameasFigure 7-9 ,butforNB=10events. .................... 112 7-11Dependenceofthetracker-basedmuonisolationcutontheleastisolatedmuonversusHiggsmass. .................................. 112 7-12Dependenceofthecalorimeter-basedcutontheleastisolatedmuonversusHiggsmass. .......................................... 112 7-13DependenceofthepTcutonthesecond-lowest-pTmuonversusHiggsmass. .. 113 7-14DependenceoftheM(4)windowcutsversusHiggsmass. ............ 113 7-15Firstmuonpairinvariantmassdistribution,M(Z1),afteranalysiscutswereapplied. ........................................ 113 7-16Secondmuonpairinvariantmassdistribution,M(Z2),afteranalysiscutswereapplied. ........................................ 113 7-17LowestmuonpTdistribution,afteranalysiscutswereapplied. .......... 114 7-18MaximumdistanceinXY-planebetweenmuonimpactpointcoordinatesdistribution,afteranalysiscutswereapplied. ........................... 114 7-19ScLvs.Higgsbosonmass. .............................. 114 7-20M(4)invariantmassdistributionforthethreebackgroundsubprocessesandaHiggsbosonsignalatMH=150GeV=c2,afterapplyingcutsonmuonisolationandpT. ........................................ 114 7-21ExpectedexcesssignicanceScLwithL=30fb1fordierentHiggsbosonmassesforM(4)-dependentandindependentcuts.Nosystematicerrorsincluded. 115 7-22ExpectedexcesssignicanceSLwithL=30fb1fordierentHiggsbosonmassesforM(4)-dependentandindependentcuts.Nosystematicerrorsincluded. 115 13

PAGE 14

... 115 7-24The95%CLexclusioncontoursfortheSMHiggshypothesis. .......... 115 7-25CombinedsystematicerroronthenumberofbackgroundeventsduetoPDFandQCDscaleuncertaintiesforthe(qq!ZZ!4)processatNLO. ..... 116 7-26Top:thefactorsKNLO(M4)inMCFMandENLOcalculationsversusM(4);bottom:thedierencebetweenthem. ........................ 116 7-27MuonisolationcuteciencyforrandomconedirectionforZ-inclusive(dashedlines)andforZZ(solidlines)events. ........................ 116 7-28Anexampleofapossiblebiasinevaluatingthesignicanceofaneventexcessduetoanon-optimalchoiceofthesignalwindowwidth. ............. 116 7-29UncertaintiesinthenumberofZZ!4backgroundeventsinthesignalregionwindowatdierentM(4).TheeventcountisreferencedtothenumberofZ!2events. ....................................... 117 7-30UncertaintiesinthenumberofZZ!4backgroundeventsinthesignalregionwindowatdierentM(4).Theeventcountinsignalregion,iscalculatedfromthenumberofZZ!4eventsintherange100-700GeV=c2(excludingthesignalregionwindow). ................................ 117 7-31Signicancevs.Higgsmass(withandwithoutdF/Kcontribution). ....... 118 7-32Integratedluminosityneededfor95%CLexclusion,3,and5discoveryversusHiggsbosonmass(withandwithoutdF/Kcontribution). ............. 118 7-33Integratedluminosityneededfora5discoveryoftheHiggsbosonversusitsmassfor(withandwithoutdF/Kcontribution). .................. 119 7-34Thebackgroundpdfandanexampleofonepseudo-experimentwithastatisticaluctuationappearingjustlikeasignal. ....................... 121 7-35ProleoftheScLscancorrespondingtothepseudo-experimentexampleshownontheleft.Green(inner)andyellow(outer)bandsdenote1and2intervals.Spikesthatcanbeseenareduetoeventscominginordroppingothetrial-window,afeatureoflow-statisticssearches. ......................... 121 7-36ScLcumulativeprobabilitydensityfunction. .................... 121 7-37Localsignicance\renormalisation"fromanobservedvaluetothetruesignicancewithaproperprobabilisticinterpretation. ..................... 121 A-1ZZbackground:t-ands-channeldiagramsands-channelcontributionpeakaroundZ0massafterpre-selectioncutsforfullysimulatedevents. ........ 125 14

PAGE 15

............................ 128 15

PAGE 16

16

PAGE 17

17

PAGE 18

Ifweknewwhatitwasweweredoing,itwouldnotbecalledresearch,wouldit?{AlbertEinsteinTheStandardModel(SM)asafundamentaltheoryofelementaryparticlesandtheirinteractionshasbeensofarverysuccessful.Butaveryimportantquestionofthistheoryisstillunanswered:doestheHiggsbosonexist?Ourdissertationdescribesworkdonebytheauthor,asamemberoftheCompactMuonSolenoid(CMS)collaboration.OneofthemaingoalsofCMSistodiscoverHiggsboson.Forthiswewillneedgoodunderstandingofourdetectoranddata,andweneedtopreparedataanalysistoolstobereadytodiscovernewphysicswearelookingfor.Thedissertationdescribesauthor'sworkinbothdirections.First,itgivesdetailsoncommissioningofCathodeStripChambers(CSCs),CMSendcapmuondetectors,throughanalyzingdataandvalidatingnewmuontrackhitsandsegmentsreconstructionalgorithm,whichwouldbesuitableforHighLevelTrigger(HLT).NotethatdefaultalgorithmusedforthisatthetimeofouranalysiswerejusttooslowforHLT.Anothermajorpartofthedissertation{isourCMStheSMHiggsbosondiscoverypotentialinH!ZZ()!4decaychannelwithM(4)dependentcutsanalysisusingMonteCarlo(MC)data.Herewedescribeouranalysis,whichismuchmorerealisticthanpreviouslyexistedonesonthetopic: 18

PAGE 19

19

PAGE 20

2-1 ). Table2-1: TheStandardModelelementaryparticles. Particles Spin Charge Fermions Leptons 0 Bosons 0,1 8gluons 1 0 [Higgs] [H0] [0] [0] ThedescriptionoftheSMinteractionsisimplementedbyagaugetheorybasedonSU(3)CSU(2)LU(1)Ysymmetry.Gaugesymmetryprovidesbosons,whichserveasinteractionmediators.Localgaugeinvariancemakesthetheoryrenormalizibleandrequiresthegaugebosonstobemassless.AtthesametimeweknowthatZ0;Wbosonshavemasses.ThesolutionoftheproblemintheSMistheHiggsmechanism,whichspontaneouslybreaksthegaugesymmetry.CorrespondingscalarpotentialaddedtotheLagrangiangeneratesthevectorbosonandfermionmassesinagaugeinvariantway.Thisremnant 20

PAGE 21

22 ]basedonthesymmetrygroupSU(2)L.Theproblemwiththisapproachwasthatgaugeinvariancedoesnotallowmassesforboththegaugebosonsandtheleptons.TheobservationofPeterHiggswasthatthegaugeinvariancecouldbespontaneouslybrokenwiththeadditionofadoubletofcomplexscalarelds,,withLagrangian 23 24 ].Havingthismechanismforelectroweaksymmetrybreaking,Glashow[ 25 ],Weinberg[ 26 ],andSalam[ 27 ]proposedauniedelectroweaktheoryoftheleptons.Thistheoryhasstillamasslessphoton,allowsformassiveWbosonsandleptons,predictsthemassive,neutral,spin-1Zboson,andpredictsthemassive,neutralscalarHiggsboson.TheWandZbosonswereconrmedattheCERNSPSbytheUA1andUA2collaborations[ 28 29 ].TheHiggsbosonitselfacquiresmassthroughself-couplingintheHiggspotentialV()(Eq. 2{2 ).TheHiggsmassatthetreelevelis 21

PAGE 22

2.2.1TheoreticalConstraintsBesidesallowingtheintroductionofZ/W-bosonmasseswithoutbreakingrenormalizabilityandquark/leptonmasseswithintheweaksectorofSMwithbrokenP-symmetry,theintroductionofanewscalarparticlewasmotivatedbydivergencesinthescatteringofalongitudinallypolarizedWbosonsinthehighenergylimit[ 30 31 ].Withoutanadditionalinteraction,thecross-sectionofthatprocess(Fig. 2-1 ),woulddivergeandwouldviolateunitarityboundsabovep 31 ]. Divergent(topthree)WWcross-sectiongraphsandtheircancellation(bottomtwographs). TheoreticalboundsontheSMHiggsmassasafunctionofthecut-oscale. 22

PAGE 23

30 31 ].IftheHiggsbosonislight,Higgsself-couplingstrengthissmallanddominantloopcontributiontotheHiggspotentialcomesfromtoploops(tislargeduetothelargemassofthetopquark).IftheSMisvalidasaneectivetheoryuptothescale,thentheseloopcontributionshavetobesummeduptothisscale.FortheHiggsmechanismtoremainvalid,thecouplingmustremainpositive,otherwisenominimumexistsinthepotentialandnostablespontaneoussymmetrybreakingoccurs.Thisplacesalowerboundon,henceonmH,dependingoncut-oscale(Fig. 2-2 ).Theenergydependence(Q2)canbederivedfromtherenormalizationgroupequations[ 30 31 ].Ifislarge,theHiggsloopdominatesoverthetoploop.Neglectingthegraphwithatop-quarkloop,wecanwriteas 13(2) 82ln(Q2=2)(2{4)ForaheavyHiggsboson,couldgrowtoinnity(Landau-pole).Requiringthattheself-couplingremainsniteforarbitraryvaluesofQimplies()=0.Since(2)=m2H=22,thiswouldresultinthenon-interactingtrivialtheory.If,instead,isrequiredtobeniteonlyuptoascaleNP,wherethenewphysicsenters,themassboundcanbewrittenas 32 ](Fig. 2-2 ). 23

PAGE 24

33 ]. 2-3 [ 34 ]). Expectedandobserved95%CLcrosssectionratiosforthecombinedCDFandD0analyses,andtheexpected95%CLratiosfortheCDFandD0experimentsalone. The2curvederivedfromhigh-Q2precisionelectroweakmeasurements. 24

PAGE 25

35 36 ].Figure 2-4 [ 36 ]showsthe2curve,derivedfromhigh-Q2precisionelectroweakmeasurements,performedatLEPandbySLD,CDF,andD0,asafunctionoftheHiggsbosonmass,assumingtheSMtobethecorrecttheoryofnature.Thepreferredvalueforitsmass,correspondingtotheminimumofthecurve,isat76GeV,withanexperimentaluncertaintyof+33and-24GeV(at68%condencelevelderivedfrom2=1fortheblackline,thusnottakingthetheoreticaluncertaintyshownasthebluebandintoaccount).Thisresultisonlyaectedalittlebythelow-Q2resultssuchastheNuTeVmeasurement.(WhilethisisnotaproofthattheStandard-ModelHiggsbosonactuallyexists,itdoesserveasaguidelineinwhatmassrangetolookforit.TheprecisionelectroweakmeasurementstellusthatthemassoftheStandard-ModelHiggsbosonislowerthanabout144GeV(one-sided95%condencelevelupperlimitderivedfrom2=2:7fortheblueband,thusincludingboththeexperimentalandthetheoreticaluncertainty).Thislimitincreasesto182GeVwhenincludingtheLEP-2directsearchlimitof114GeVshowninyellow.) 2-5 (left)): a gluonfusionviatoploop b vectorboson(W/Z)fusion c W/Zassociatedproduction d ttfusion 25

PAGE 26

2-6 TypicaldiagramsforallrelevantHiggsbosonproductionmechanismsatleadingorderattheLHC. HiggsproductioncrosssectionsattheLHCforthevariousproductionmechanismsasafunctionoftheHiggsmass. DuetothelargesizeofthetopYukawacouplingsandthegluondensities,thegluonfusion,pp!gg!H,isthedominantHiggsbosonproductionmechanismforthewholeHiggsmassrange.InourHiggsbosonsearchanalysis,theNLOcrosssectionsandbranchingratiosfortheHiggsbosoncalculatedwiththeprogramsHDECAY[ 37 ],HIGLU[ 38 ]areused,aswellastheNLOcrosssectionsforthebackgroundprocesses,whenavailable 2.4 26

PAGE 27

2-7 ,rightplot).Bothbband+Higgsbosondecaymodesare(almost)impossiblefordiscoveryathadroncolliderduetooverwhelmingbackgroundlevel.Thedecaymodesintoccandgluonpairs,withthelattermediatedbytopandbottomquarkloops,accumulateabranchingratioofuptoabout10%,butdonotplayarelevantroleattheLHC(theyareevenhardertodiscoverthanbband+modesduetohugeQCDbackgroundlevel).ThemostimportantHiggsdecaysinthismassrangeattheLHCisthedecayintophotonpairs,whichismediatedbyW,topandbottomquarkloops.Itreachesabranchingfractionofupto2103.Itsimportanceisincleanenvironmentandcontrollablebackground,whichmakesitanimportantdiscoverychannelforsmallHiggsbosonmasses. TotaldecaywidthandbranchingratiosofthedominantdecaymodesoftheSMHiggsbosonparticle. ForHiggsmassesabove135GeV=c2themaindecaymodesarethoseintoWWandZZpairs,whereoneofthevectorbosonsiso-shellbelowthecorrespondingkinematicalthreshold.Theseareveryimportantdiscoverychannels,inparticularH!ZZ()!4

PAGE 28

2-7 (rightplot).ThetotaldecaywidthoftheHiggsboson,showninFig. 2-7 (leftplot),doesnotexceedabout1GeV=c2belowtheWWthreshold.ForverylargeHiggsmassesthetotaldecaywidthgrowsuptotheorderoftheHiggsmassitselfsothattheinterpretationoftheHiggsbosonasaresonancebecomesquestionable. 2-8 [ 8 ]showstheintegratedluminosityneededforthe5discoveryoftheinclusiveHiggsbosonproductionpp!H+XwiththeHiggsbosondecaymodesH!,H!ZZ!4`,andH!WW!2`2{threefrontrunnersamongallpossibledecaychannels.H!dominatesdiscoveryreachuptomH130GeV=c2,similarlyit'sH!WW!2`2dominatesfor150180GeVandH!ZZ!4`{intherestofthepossibleHiggsbosonmasses.Figure 2-9 [ 8 ]showsthesignalsignicanceasafunctionoftheHiggsbosonmassfor30fb1oftheintegratedluminosityforthedierentHiggsbosonproductionanddecaychannels. 39 ].DuringthepastyearsseveralneutrinoexperimentsliketheSuperKamiokande,K2K,SNOandKamland[ 40 41 42 43 ]haveestablishedthepresence 28

PAGE 29

Theintegratedluminosityneededforthe5discoveryoftheinclusiveHiggsbosonproduction. ThesignalsignicanceasafunctionoftheHiggsbosonmassfor30fb1oftheintegratedluminosityforthedierentHiggsbosonproductionanddecaychannels. ofneutrinooscillations.Thisisasignoftheneutrinomasses:onlymassiveparticleshaveatimeevolutionandthereforecanoscillateifmassdierencesbetweenthevariousneutrinomasseigenstatesexist.ThesecondobstaclecomesfromthecreationoftheUniverseandthecosmologicalprecisionmeasurements.ThelackoftheantimatterinthevisibleUniverseandthemeasuredration=nb109[ 44 ]placesalowerboundontheamountofCPviolation,whichisoneofthethreerequirementsforthecreationofthematter-antimatterasymmetry[ 45 ].TheSMincorporatesCPviolationonlybytheCKMmechanism[ 46 ].ThemeasuredCPviolationintheSMissmallerbyatleasteightordersofmagnitudethantheoneneededtogeneratethecosmologicallyobservedmatter-antimatterasymmetry.Third,anotherconditionformulatedbyAndreiSakharovrequiresbaryonnumberviolation.Currently,thereisnoexperimentalevidenceofparticleinteractionswheretheconservationofbaryonnumberisbrokenperturbatively:thiswouldappeartosuggestthatallobservedparticlereactionshaveequalbaryonnumberbeforeandafter. 29

PAGE 30

47 48 ].Accordingtothecontemporarymeasurements,only5%oftheamountofthetotalenergyisstoredinanordinarymatterasknownbytheSM.Fortheremaining95%oftheenergyoftheUniversethereisnoexplanationintheSM.Furthermore,iftheSMisvaliduptoanenergyscale,thenthesizeofdivergentloopcontributionsism2H2.Ontheotherhandloopcorrectionstothefermionmassesareonlysubjecttothelogarithmicdivergences,sothattheoverallcorrectionisofthescaleofmassitselfandnone-tuningproblememerges.FortheHiggsbosonthismeansthatiftheSMisvaliduptothePlanckscaleofP=1019GeV,thenthenaturalscaleoftheHiggsbosonmassisPwhileallotherparticleshavenaturalmassscalesbelowVEV.Thisistheso-calledhierarchyproblem,whichreferstotheextremelylargesplittingoftheweakscaleandnaturalcut-ofscale,thePlanckscale.InordertoachievethenecessaryHiggsmassrangeofmH<1TeV,anunnaturalne-tuningwiththerelativeprecisionofmH=P>1016hastobeapplied.Thisne-tuningisnotexplainedinthecontextoftheSMbutitcanbesolvedbyextensionsoftheSM.TheSMleavesunexplainedwhythestrongandtheelectroweakgaugestructureisSU(3)CSU(2)LU(1)Ywithdierentgaugecouplingsandfermionicquantumnumberswhosevaluesarenotpredictedbythemodel.TherehavebeenmanyeortstounifythegaugegroupsandtohaveonlyonecouplingattheenergyscaleofGrandUnication[ 49 50 51 ].Therearemorequestions:whyneutrinomassesaresosmall;whythereisnoCPviolationinQCD,etc. 30

PAGE 31

52 ]. 31

PAGE 32

53 ]isa27kmcircumferenceparticleacceleratorspanningtheSwiss-Frenchboarder.ItisaprojectcarriedoutbyCERN(EuropeanOrganizationofNuclearResearch)[ 54 ],whichisalargeinternationalcollaborationconsistingofmanycountriesfromallovertheworld.TheLHCiscurrentlyunderconstructionandhastherstbeamsscheduledin2008.Whencompleteditwillprovideproton-protoncollisionsatthecenter-of-massenergy14TeVaswellasheavyioncollisions.TheLHCconsistsofanumberofacceleratorunits(Fig. 3-1 [ 53 ]).Twolinearaccelerators,LINAC2andLINAC3,willbeusedfortheinitialaccelerationofprotonsandleadionsrespectively.ThePS(ProtonSynchrotron)willbeusedtoprovidealowenergybeam(25GeV)withthenalbunchstructure.Thebeamsarepre-acceleratedusingtheSPS(SuperPS),andthenareinjectedintotheLHCnearIP2/IP8forthebeamthatcirculatesclockwise/counter-clockwise,whereparticleswillbeacceleratedfrom450GeVtothenominalenergyof7TeV(forprotonbeam).InsidetheLHCaccelerator,theparticlescirculateinoppositedirectionsintwoseparatebeampipes.Thediameterofthebeamisreducedbyfocusingmagnetsbeforethebeamenterseveryexperimentalinteractionpointtoachievehighluminositycollisions.Thetwobeampipesaresurroundedbyshellsofsuperconductingcoilscreatingthemagneticeldwhichguidesthebeamstofollowacircularpath.Particlesinsidethebeamaregroupedintoapproximatelycylindricalbuncheswithadiameterandalengthofapproximately16and8cmattheinteractionpointrespectively.TheLHCllingschemeforprotonbeamwillhaveabunchseparationof25ns.TheLHCwillundergoaseriesofaccelerator/detectorcommissioningstagesandpilotrunsatthebeginningofitsoperationin2008.Regularphysicsrunswillstartwithlowluminosityruns,plannedluminosityisupto21033cm1s1,withuptoabout10fb1of 32

PAGE 33

AcceleratorcomplexatCERN. 33

PAGE 34

55 ]andATLAS[ 56 ](AToroidalLhcApparatuS),andtwowithdedicateddetectors,ALICE[ 57 ](ALargeIonColliderExperiment)andLHCb(theLargeHadronColliderBeautyexperiment)[ 58 ]whichwillstudyheavyionandB-physicsrespectivelyplusrelativelysmallTOTEMexperiment[ 59 ]fordiraction/forwardphysicssituatednexttoCMS,whichalsowouldallowtohavemorepreciseluminositymeasurements.TheCMSdetectorissituatedapproximately100mundergroundatIP5(InteractionPointin5thLHCoctant),andwillbetakingdatafromboththep-pandPb-Pbcollisionruns. 34

PAGE 35

7 8 60 ]formoredetails). 7 8 ]tomeetthegoalsoftheLHCphysicsprogramcanbesummarizedasfollows: 35

PAGE 36

61 ]isdiscussedbelow.AttheheartofCMSsitsa13-m-long,5.9minnerdiameter,4Tsuperconductingsolenoid(Fig. 4-1 ).Inordertoachievegoodmomentumresolutionwithinacompactspectrometerwithoutmakingstringentdemandsonmuon-chamberresolutionandalignment,ahighmagneticeldwaschosen.Thereturneldislargeenoughtosaturate 36

PAGE 37

AnexplodedviewoftheCMSdetector. 1.5mofiron,allowing4muon\stations"tobeintegratedtoensurerobustnessandfullgeometriccoverage.Eachmuonstationconsistsofseverallayersofaluminiumdrifttubes(DT)inthebarrelregionandcathodestripchambers(CSCs)intheendcapregion,complementedbyresistiveplatechambers(RPCs).Theboreofthemagnetcoilisalsolargeenoughtoaccommodatetheinnertrackerandthecalorimetryinside.Thetrackingvolumeisgivenbyacylinderoflength5.8manddiameter2.6m.Inordertodealwithhightrackmultiplicities,CMSemploys10layersofsiliconmicrostripdetectors,whichprovidetherequiredgranularityandprecision.Inaddition,3layersofsiliconpixeldetectorsareplacedclosetotheinteractionregiontoimprovethemeasurementoftheimpactparameterofcharged-particletracks,aswellasthepositionofsecondaryvertices.TheEMcalorimeter(ECAL)usesleadtungstate(PbWO4)crystalswithcoverageinpseudorapidityuptojj<3:0.Thescintillationlightisdetectedbysiliconavalanchephotodiodes(APDs)inthebarrelregionandvacuumphototriodes(VPTs)intheendcapregion.Apreshowersystemis 37

PAGE 38

38

PAGE 39

62 ].Thedistinctivefeatureofthe220-tcoldmassisthefour-layerwindingmadefromastabilizedreinforcedNbTiconductor.Theratiobetweenstoredenergyandcoldmassiscriticallyhigh(11.6KJ/kg),causingalargemechanicaldeformation(0.15%)duringenergization,wellbeyondthevaluesofprevioussolenoidaldetectormagnets. 39

PAGE 40

63 ].Forsinglechargedparticleoftransversemomentaof1,10and100GeVtheexpectedresolutionoftransversemomentumdependsonpseudorapidity[ 7 8 ]andhighmomentumtracks(100GeV)thetransversemomentumresolutionisaround1-2%uptojj1:6,beyondwhichitdegradesduetothereducedleverarm.Atatransversemomentumof100GeVmultiplescatteringinthetrackermaterialaccountsfor20to30%ofthetransversemomentumresolutionwhileatlowermomentumitisdominatedbymultiplescattering.Thetransverseimpactparameterresolutionreaches10mforhighPTtracks,dominatedbytheresolutionoftherstpixelhit,whileatlowermomentumitisdegradedbymultiplescattering(similarlyforthelongitudinalimpactparameter).Formuons,theexpectedtrackreconstructioneciency(afunctionofpseudo-rapidity)isabout99%overmostoftheacceptance.Forjj0theeciencydecreasesslightlyduetogapsbetweentheladdersofthepixeldetectoratz0.Athightheeciencydropismainlyduetothereducedcoveragebythepixelforwarddisks.Forpionsandhadronsingeneraltheeciencyislowerbecauseofinteractionswiththematerialinthetracker. 40

PAGE 41

E2=2:8% 61 ].Hadronicshowershavealargelateralshowersize,sothedegreeofgranularityprovidedbytheECALisnotrequiredfortheHCAL.However,moderatetransversegranularity,hermeticityandwiderapiditycoverageareessentialinmeasuringthetotalmissingETofanevent.TheHCALconsistsofthreemainparts:HadronBarrel(HB)andHadronEndcap(HE),whichextendsto=3:0,andHadronForward(HF)calorimeters,whicharelocatedaroundthebeampipeoutsidethemuonsystematjzj=10:9mtocompletethecoverageupto=5:3.DuetotherestrictedspaceavailablebetweentheECALandthemagneticcoil,theHBisonly1minthicknesswhereastheHEis1.8mthick;additionalscintillationlayers(HadronOutercalorimeter,HO)areinstalledjustoutsidethemagnet,usingthecoilasanabsorber,toprovideacombineddepthofatleast11absorptionlengths.ForgaugingtheperformanceoftheHCAL,itisusualtolookatthejetenergyresolutionandthemissingtransverseenergyresolution.Thegranularityofthesamplingin 41

PAGE 42

7 ]. 42

PAGE 43

64 65 7 8 60 ].MuonTrigger.Allthreemuonsystems(DT,CSC,RPC)takepartinthetrigger.TheDTchambersprovidelocaltriggerinformation:and-projectionsfortracksegments.TheCSCsdeliver3-dimensionaltracksegments.Allchambertypesalsoidentifythebunchcrossingatwhicheventtookplace.TheRegionalMuonTriggerconsistsoftheDTandCSCTrackFinders,whichjoinsegmentstocompletetracksandassignphysicalparameterstothem.Inaddition,theRPCtriggerchambersprovidetheirowntrackcandidatesbasedonregionalhitpatterns.TheGlobalMuonTriggerthencombinesthe 43

PAGE 44

44

PAGE 45

65 7 8 60 ]formoredetails). 65 7 8 ]willconsistof468CathodeStripChambers(CSCs)arrangedingroupsasfollows:72ME1/1,72ME1/2,72ME1/3,36ME2/1,72ME2/2,36ME3/1,72ME3/2,36ME4/1(Figs. 5-1 and 5-2 ).Thede-scoped72ME4/2chamberswillnotbeavailableduringearlyyearsofCMSoperation.Thechambersaretrapezoidalandcovereither10or20degreesin;allchambers,exceptfortheME1/3ring,overlapandprovidecontiguous-coverage.Amuoninapseudorapidityrange1:250m3,andthetotalnumberofwiresisabout2million.Thereareabout9,000highvoltagechannelsinthesystem.Thenumberofcathodestripreadoutchannelswith12-bitsignaldigitizationisabout220,000,andthenumberofanodereadoutchannelsisabout180,000.CSCsprovidefunctionsofaprecisionmuonmeasurementandmuontriggerinonedevice.Theycanoperateathighratesandinlargeandnon-uniformmagneticelds. 45

PAGE 46

Quater-viewoftheCMSDetectors.CathodeStripChambersoftheEndcapMuonSystemarehighlighted. Theydonotrequiretightgas,temperatureorpressurecontrol.Additionally,theradialfan-shapedstrippattern,naturalformeasurementsintheendcapregion,canbeeasilyarrangedoncathodes.TherequirementsfortheCMScathodestripchamberperformanceareasfollows: 46

PAGE 47

TheME2cathodestripchambers.Theouterringconsistsof36ME2/2chambers,eachspanning10in;while1820-degreeME2/1chambersformtheinnerring.Thechambersoverlaptoprovidecontiguouscoveragein. 47

PAGE 48

AschematicviewofaCMScathodestripchambermadeofseventrapezoidalpanels. Figure5-4: AschematicviewofasinglegapillustratingtheprincipleofaCSCoperation. 5-5 ).Seven16.2-mmthicktrapezoidalpanelsarethebasisforthemechanicalstructure.Thepanelsaremadeof12.7-mmthickpolycarbonatehoneycombcorewithtwo1.6-mmFR4skinscommerciallygluedoneachside.FR4isre-retardantberglass/epoxymaterialwidelyusedforprintedcircuitboards.TheFR4skinsarecopper-cladontheiroutersurfaces|thecopperlayersservetheroleofcathodes.FR4cathodegapbarsaregluedtobothsidesofeachotherpanel(panels1,3,5,7inFig. 5-5 )sothatwhenallpanelsarestackedtogether,thesecathodebarsdenesixgasgapsof9.5mm.Toprovideanadditionalsupport,therearefourspacersplaced 48

PAGE 49

MechanicaldesignoftheCMSCathodeStripChambers(explodedview). betweenpanelsalongthechambercenterline.Whenallsevenpanelsareputtogether,theentirestackistightendownwithboltsalongthechamberperimeter(throughholesinthecathodegapbars)andin4pointsalongthechambercenterline(throughholesinthespacers).Sucharrangementensuresthatpanelshavenomorethan60cmofunsupportedlength.Measurementsshowedthatmostofthepanelswereatwithintherequired300monsuchspans.Thisspecicationarisesfromthedesiretokeepgasgainvariationswithinafactoroftwo.Onesideofsixpanelshasamilledpatternof80strips.Strips,beingradial,haveavaryingpitchfrom8.4mmatthebottomto16.0mmatthetop.Thegapbetweenstripsisabout0.5mm.Theprecisionofmillingisbetterthan50m(rms).Millingisdonewithacuttertiltedat45degreestomakegrooveedgessmoother(otherwise,sharpedgesandburrsmightprovokesparkinganddischarges).Anodewiresarewoundonthreeso-called\anode"panels(panels2,4,6inFig. 5-5 );thesepanelsdonothavegapbars.Aspecialwindingmachinewasdesignedtowindwires 49

PAGE 50

50

PAGE 51

CSCgasgainvshighvoltage. Figure5-7: ME2/1chambersinglesratevshighvoltage(theoverallsensitiveareaofallsixplanesinthischamberis9.5m2). Afterstackingthepanelsandtighteningbolts(witho-rings),continuousbeadsofsealantRTVareappliedalongtheouterseamsbetweenpanelsandgapbars.Theo-ringsaroundboltsandtheRTVsealmakechambershermetic.Shouldaneedarise,chamberscanbeopened,serviced,andresealed.Thegasentersintooneoftheoutergasgapsviaaninletinacathodegapbar,thenowsfromoneplanetoanotherinzigzagmannerviaspecialholesinpanels,andnallyexitsfromthelastgasgapviaanoutletinagapbar.Theleakratewasmeasuredduringproductionanduponinstallationofchambersandwasrequiredtobelessthan1%ofthechambervolumeperdayatoverpressureof7.5mbar(<2cc/minforthelargestchamberswhosegasvolumeisabout200liters).Sideplatesmadeof3.2-mmthickAlextrusionsareattachedalongthechamberperimeter.TheystienthechamberandinterconnecttopandbottomcopperskinstoformacompleteFaradaycage.ThenominalgasmixtureisAr+CO2+CF4=40%+50%+10%.TheCO2componentisanon-ammablequencherneededtoachievelargegasgains,whilethe 51

PAGE 52

67 ].Figure 5-6 showsthechambergasgainvs.highvoltage.ThenominaloperationalHVpointischosentobe3.6kV,whichcorrespondstogasgainoftheorder7104.Takingintoaccountthataminimumionizingparticle(mip)leavesbehindinagasgapabout100electrons,thetotalchargeinanavalanchepermipisabout1pC.Aswillbeshownbelow,atthisoperationalpoint,thecathodeandanodeelectronicshaveaveryhigheciencyandanadequatesignal-to-noiseratio.Theoperationalrangeofchambersextendsto3.9kV.Typically,westartseeingasharpriseinrateofspuriouspulsesatabout3.9-4.0kV|seeFig. 5-7 .The72smallestME1/1chambershavevariationsinthemechanicaldesignwithrespecttoallotherchambers.First,thegasgapis7mm,wirediameteris30m,andwirespacingis2.5mm.Asaconsequence,thenominalhighvoltageforthesechambersissomewhatlower,2.9kV.Secondly,andmoreimportantly,anodewiresinME1/1chambersarenotazimuthal,butratherrotatedbyanangleL=29asshowninFig. 5-8 .UnlikeallotherCSCs,theME1/1chambersareinsidetheCMSsolenoidandseeitsstronganduniform4Teslaaxialeld.Ifthewireswerenottilted,theionizationelectronsdriftingtowardwireswouldbecarriedsidewaysattheLorentzangleandbecomespreadalongwiresandacrossstrips.ThewiretiltcompensatesfortheLorentzanglesothatelectronsdriftparalleltostripsandtheprecisemeasurementofther-coordinateremainspossible. 5-9 showsaschematiclayoutofthecustomtriggerandreadoutelectronicboardsdevelopedfortheEndcapMuonSystem.Ananodefront-endboard(AFEB)hasone16-channelamplier-discriminatorASIC.Theamplierhasa30nsshaper(semi-Gaussianwitha2-exponenttailcancellationdesignedtosuppresstheslowsignalcomponentassociatedwithadriftofpositiveionsawayfromanodewires),about7mV/fCsensitivity,and1.4fCnoiseatatypical180pF 52

PAGE 53

SignalinME1/1chambersfornottiltedandtiltedwires. Figure5-9: Anoverviewofallcustom-madeCSCtriggerandreadoutelectronics.Functionalityofvariousboardsisdescribedinthetext. 53

PAGE 54

5-10 .Theconstant-fractiondiscriminatorhasathresholdnominallysetat20fC(inputequivalentcharge)anditsslewingtimeislessthan3nsfor60-600fCsignalrange.Dependingonachambersize,thereare12to42AFEBsperchamber.AlldetailsontheAFEBdesignandperformancecanbefoundelsewhere[ 67 ].Every25nsinsyncwiththeLHCcollisions,allAFEBoutputs,40-nslongsteppulses,aresampledbyanFPGA-basedAnodeLocalChargedTrack(ALCT)board,oneboardperchamber.Therecordedyes/noinformationisstoredinFIFO.UponreceivingaCMS-wide"Level-1triggeraccept"command(L1A),therecordedinformationfromapropertimewindowisextractedandreportedtoDAQ.ThelatencyoftheL1Acommandwithrespecttothetimeofacollisionis3.2s.ThewidthofarecordwithrawhitstransmittedtoDAQcanbesettobeaslargeas32bits(1bitper25ns),i.e.800ns.TheALCTboardhasanotherimportantfunction.Basedontheinformationfromallanodechannels,theFPGAcodeconstantly(every25ns)searchesforpatternsofhitsinsixplanesthatwouldbeconsistentwithmuontracksoriginatingfromtheinteractionpoint.Forapatterntobevalid,werequirethathitsfromatleast4planeswerepresentinthepattern.Figure 5-11 illustrateshowpatternsareidentiedamongspurioussingle-planehits.Duetoalargeneutron-inducedphotonbackground,weexpectasubstantialrateofsuchsingle-planehits.However,thesehits,beingcompletelyuncorrelated,wouldnottypicallylineuptoformtrack-likepatterns.Foundpatterns,AnodeLocalChargedTracks,aretriggerprimitives.TheyaretransmittedfurtherupstreamtothemuonLevel-1triggerelectronicsthatbuildsmuontrackcandidatesfromtheseprimitives.Thetimeittakestoformananodetracktriggerprimitiveis225ns(includingdrifttime).TheALCTboardcanndupto2suchpatternspereachbunch 54

PAGE 55

MuonsignalsasseenattheAFEBamplieroutput. Figure5-11: Aschematiceventdisplayshowinganodesignalsinasix-planechamber. crossing,whichiswelladequatefortheexpectedchambertrackoccupancyatthenominalLHCluminosity.Onecathodefront-endboardserves(6planes)(16strips)=96channelsandhassixparallelchainsofthefollowingchips(Fig. 5-12 ):16-channelamplier-shaperASIC,16-channelSwitchedCapacitorArrayASIC,12-bit1-channelADC,and16-channelComparatorASIC.Thereare4to5CFEBsperchamber.Thefront-endamplier-shaperASIChas100nsshapingtimeand0.85mV/fCsensitivityoverthelinerrangeupto1V.Theequivalentnoiselevelat300pFstripcapacitanceistypically1.5fC.Theshapingisbasedonsemi-Gaussiantransferfunctionwithanovershootdesignedtocompensateforthe1=tsignaltailduetoslowdriftofpositiveions.Afterconvolutionwiththecurrentpulseproducedinachamberbyamuon,theamplier-shapersignalpeaksataround150nsandhasnotails|seeFig. 5-13 .Withthe100-nsshapingtime,CFEBseesabout8%ofthetotalavalanchecharge,i.e.about100fConaverage. 55

PAGE 56

BasicfunctionaldiagramofaCathodeFront-EndBoard. Figure5-13: Muonhitsignalsfromsixnearbystrips.Fourcurvesareactualoscillograms. Theoutputfromthischipissplitintwopathways.OneleadstotheSwitchedCapacitorArray(SCA)ASIC.TheSCAchipsamplesawaveformofeachstripsignalevery50nsinsyncwiththeLHCclockandstoresthisanaloginformationonitscapacitors.Thedepthofsuchanalogmemoryis96capacitorcellspereachchannel,or9650ns=4.8s.UponreceivingtheL1Acommandin3.2saftercollision,8or16consecutivesamplesfromthepropertimerangeintheSCAlineofcapacitorsareretrievedandone-by-onedigitizedbythe12-bitashADCs.Thedigitalinformationispassedontothedataacquisitionsystemviaanintermediatedigitaldatabuer.ForthedigitizationandsubsequentreadoutbyDAQtohappen,theL1AsignalmustbeinacoincidencewiththeCathodeLocalChargedTrackprimitivedecisiontobedescribednext. 56

PAGE 57

Asimpliedschematicillustratingtheideabehindthecomparatornetwork. Figure5-15: CLCT(CathodeLocalChargedTrack)isapatternofhalf-striphitsconsistentwithamuontrack. Thesecondamplier-shaperoutputgoestotheComparatorASIC.Thischipcomparessignalsontripletsofnearbystripsatthetimewhensignalsreachthemaximumamplitude.Bymeansofsuchcomparisons,theComparatorASICcanidentifymuonhitlocationwithinonehalfofastripwidth,independentlyofasignalamplitude,aninducedchargeshape(aslongasitis\bell"-like),andastripwidthitself[ 68 ].Figure 5-14 illustratesthebasicideabehindtheComparatorASICalgorithm.Comparatorhalf-striphitsaresenttotheTriggerMotherBoard(TMB).SimilarlytotheALCTboard,theTMBsearchesforpatternsofhalf-stripcomparatorhitsthatwouldbeconsistentwithmuontracksofinterest|seeFig. 5-15 .UptotwoCathodeLocalChargedTracks(CLCTs)perbunchcrossingcanbefound.AsintheALCTpatternsearch,foraCLCTpatterntobefound,half-striphitsfromatleastfourplanesshouldbepresentinit.ThereisoneTMBperchamber.UnlikeALCTboards,TMBsarenotonchambers,butratherinperipheralcratesmountedalongtheouterrimoftheendcapirondisks. 57

PAGE 58

58

PAGE 59

59

PAGE 60

64 ]for21033cm1sec1luminosity).ThefollowingstagesofHLTinvolvecalorimeter(L2.5|muonisolation)andreconstructionoftracksinthetracker(L3|renedmomentumandmuonisolation).MoredetailsontheCMSHighLevelTriggercanbefoundelsewhere[ 64 ].Thischapterdealswiththetracksegmentreconstruction.Thefourmaincriteriaqualifyingatracksegmentndingperformanceareasfollows: 60

PAGE 61

69 ].Therefore,belowthehighestexpectedpTthresholdof19GeV,thenalmuonmomentummeasurementatHLTisinsensitivetothemuonchamberresolutionatall.Rateofndingassociatedsecondarysegmentsnotdirectlyassociatedwithamuonofinterestisanotherveryimportantmeasureofmeritforsegmentndingperformance 61

PAGE 62

21 ]. 70 ]fordetails),wewillfurtherrefertoitunderthisnickname.TheUFalgorithmisbuiltonthefollowingprinciples. 62

PAGE 63

71 ].Theratiomethodrequiresnoiterations,isveryfast,anddeliversaspatialresolutionsubstantiallybetterthanwhatisrequiredfortheHLTpurposes. 65 ].UptotwoALCTs(ALCT0andALCT1)andtwoCLCTs(CLCT0andCLCT1)canbefoundpereachbunchcrossing.Ifmorethanoneprimitiveisfoundineitherprojection,thenwecombinethemcombinatoriallytoform2d-ProtoSegments,ofwhichwecanhave1,2or4perchambers.Thetwocoordinatesofa2d-ProtoSegmentareALCT'sKeyWireGroupandCLCT'sKeyHalfStrip.Allinall,buildingProtoSegmentsthiswayrequiresvirtuallynoCPUusage. 63

PAGE 64

65 ].Ecienciesofndingtriggerprimitiveswereextensivelystudiedinpastonchamberprototypesoperatinginamuonbeamorcosmicrays,inamuonbeamwithasuperimposeduxofrandomhits[ 72 ],andwithhighenergymuonsaccompaniedbybremsstrahlungradiationduetomuonspassingthroughanironslabinfrontofachamber[ 73 ].Morerecently,theseeciencieswerestudiedinsituwith18chambersinstalledinCMSandoperatingincosmicrays[ 74 ].Inallthesecases,theeciencyofndingtriggerprimitiveswasmeasuredtobehigherthan99%.Forexample,thelatterstudiesgaveaneciencyof99.90.05%forndingALCTCLCT2d-patternsformuonspassingthroughthechambersensitivevolume. 64

PAGE 65

2QrQl 65

PAGE 66

66

PAGE 67

67

PAGE 68

75 ].ThissectiondescribestheperformanceoftheUFalgorithmasobtainedusingtheMTCCdata.MeasurementsoftheprocessingtimeforthedefaultUFalgorithmweredoneonanIntelP42.8GHzDualXeonServerwiththeCMSSWversion1.2.0asthedefaultframework.TobenchmarkthealgorithmperformanceweusedtheCMSGlobalRunNo.4188.Onaverage,theUFalgorithmwasfoundtotake0.36mspermatchedLCT.Thistimeincludesndingsegments,reconstruction/selectionofuncorruptedRecHitsassociatedwiththesesegments,andreningsegmentparameters. 74 ].Herewejustsummarizethemainresultsfromthosestudies: 68

PAGE 69

Thenoisedistributionforathree-sample-sumintheabsenceofasignalforthelargestME23/2chambers. Figure6-2: Thethree-stripclusterchargedistributionforthelargestME23/2chambers. largechambersoperatedinsitutogetherwiththerestoftheCMSdetectorsparticipatingintheMTCC. 6-1 showsthethree-samplesumdistributionintheabsenceofasignal|theGaussiansigmaofthisdistributionis35.5ADCcounts.Figure 6-2 showstheLandaudistributionfor3-stripchargeclusters.Theaveragechargeisaround550ADCcounts.Therefore,theMTCCdataweretakenwiththegasgaincorrespondingtosignal-to-noiseratioofapproximately100:1.Figure 6-3 showsthesingle-samplenoiselevel.Ithasaspread13.0ADCcounts.Itisworthwhilenotingthat3>p 69

PAGE 70

Thenoisedistributionforasingle-sampleintheabsenceofasignalforthelargestME23/2chambers. Figure6-4: Thenoisedistributionforatwo-samplesumintheabsenceofasignalforthelargestME23/2chambers. Figure 6-4 showsthetwo-samplesumdistributioninanabsenceofasignal|theGaussiansigmaofthisdistributionis24.6ADCcounts.Thetwo-samplesumisofspecialinterestforfurtherconsiderations.TheUFalgorithmassumestwomodesofoperation.First,default,modedoesnotuseanycalibrationconstants,includingpedestals.Thepedestalsforindividualchannelsareevaluatedonevent-by-eventbasisfromthersttwosamples,wheresignalisnotpresent:ped=(sample1+sample2)=2.Then,suchdynamicallydenedpedestalissubtractedfromallthreesamplesusedtoformathree-sample-sumcharge.Thisprocedurecontributesanadditionalerrorindeterminationofthecharge:(2=2)3=6:9ADCcounts.Combinedwith3=5:5,thetotalerrorinchargemeasurementsbecomes8.8ADCcounts,whichis1.6timesworsethan5.5ADCcountsonecouldhave,iftabulated(calibrated)pedestalswereused.Thisisasubstantialpenaltyfornotusingcalibratedpedestals.Therefore,wesingleoutpedestalsfromallcalibrationconstantsandimplementasecondmodeoftheUFalgorithmthatdoesusepre-denedpedestals.UFalgorithmtakes0.45mspermatchedLCTwiththisoptionON. 70

PAGE 71

71

PAGE 72

6-5 showsresidualsforthesevedistinctrangesforthemiddlepart(HVsegment3)ofthelargestME23/2chambers.Thelastplotshowsasummaryforasingle-planeresolutionivs.i-thjxj-partofastrip(i=1for0
PAGE 73

The3rd-planeresidualsforthevedistinctjxjstrippartrangesforthemiddlepartofthelargestME23/2chambers.Andasummaryforasingle-planespatialresolutionasafunctionofahitpositionacrossastrip. 73

PAGE 74

The3rd-planeresiduals'svs.vejxjstrippartrangesandoverallsix-planeresolutionvs.jxj-partofastrip(inunitsofstripwidthsandmicrons).[Notationsare:blackemptysquares-HVsegment1;redsquares-HVsegment2;trianglesup-HVsegment3;trianglesdown-HVsegment4;stars-HVsegment5.] 74

PAGE 75

The3rd-planeresiduals'svs.jx=wj-partofastripandoverallsix-planeresolutionvs.jx=wj-partofastrip(inunitsofstripwidthsandmicrons).SamenotationsasforFig. 6-6 75

PAGE 76

6-8 6-9 .Theleftplotisfordynamicallymeasuredpedestals,therightplotisfortabulatedpedestals.Thegaininresolutionisquitetempting.Aslongastherateofhitsinchambersisnottoohigh,onecancertainlytakeadvantageofthisoption,especiallyintheoineanalysis.AsfarastheHLTisconcerned,theimprovementinresolution,howeverimpressiveitmightbe,isnotcritical(theresolutionwithdynamicpedestalsisalreadybetterthanneeded).Nevertheless,theoptionofusingtabulatedpedestalscertainlycanbeusedfortheHLTaswell,aslongastheCPUpenalty(yettobemeasured)duetotheneedtomanageabout200Karrayofconstantsisacceptable. 76

PAGE 77

Theoverallsix-planechamberresolutionvs.vejxjstrippartrangesfordynamicallymeasuredpedestals.SamenotationsasforFig. 6-6 Figure6-9: Theoverallsix-planechamberresolutionvs.vejxjstrippartrangesforcalibratedpedestals.SamenotationsasforFig. 6-6 Figure6-10: Inclinationanglesfortracksegmentsreconstructedfromdi-stripCLCTs(ME1/1chambers). Figure6-11: Thespatialresolutionpersix-planechamberfortracksreconstructedfromdi-stripCLCTs(ME1/1chambers)forvejxjstrippartranges.SamenotationsasforFig. 6-6 77

PAGE 78

6-10 6-11 .Theleftplotshowsanglesfortracksegmentsreconstructedfromdi-stripCLCTs.OnecanseethattherangeismuchwiderthanexpectedforafewGeVpTmuons.Therightplotistheresolutionpersix-planechamber|clearly,theresolutionisstillmuchbetterthanafew-millimeterresolutionneededforsuchsoftmuons. 78

PAGE 79

79

PAGE 80

ProbabilitiesfordierentnumbersofchargeclustersfoundinassociationwithME23/2chamberProtoSegments. Figure6-13: LayerchargeclustersoccupancyinassociationwithME23/2chambersProtoSegments. Figure 6-12 showshowmanychargeclustersarefoundonsuchProtoSegments.NotethatbytheUFalgorithmdesign,onlyoneclusterperplanecanbeassociatedwithagivenProtoSegment.Figure 6-13 showsanoccupancyofaparticularplanefrom1stto6thwithchargeclusters.Figure 6-14 shows2/dofdistributionforalineartforeventswithsix,ve,andfourclustersassociatedwithaProtoSegment.Thedistributionshavealongtailsdueto-electronsandotherpossiblesourcesofhitcorruption(e.g.,showers).Figure 6-15 showsascatterplotofmin(25/dof)vs.26/dofforeventswithsix-clusterProtoSegments.Here,26/dofisobtainedforasix-planet,min(25/dof)isaminimumbetween6ve-planetswithoneplanedroppedfromthet.Thechoiceofcutsforpruningbadhitsisshownbydashedlines.Probabilityofhaving6-hit,5-hit,and4-hitRenedSegmentsisthen78%,13%,9%correspondingly.Wedonotprunebadhitsinsegmentswithfourremaininghits.Figure 6-16 showsthattheoccupancyforjxj-coordinatesforRenedSegmentsremainsat,whichdemonstratethatthepruningofhitsdoesnotbiassegments.Figure 6-17 80

PAGE 81

The2/dofdistributionsforalineartforME23/2chamberProtoSegmentswithsixclusters,veclusters,andfourclusters. Figure6-15: Scatterplotofmin(25/dof)vs.26/dofforsix-clusterME23/2chamberProtoSegments. showspoolsforallretainedRecHits(i.e.,RecHitsassociatedwithRenedSegments.OnecanseeanobviousreductionintailsincomparisontodistributionsshowninFig. 6-5 .Thenaloptimizationofthe2cutscanbedonewhenagoodreferenceforawholesegmentisavailable(e.g.,indetectorMonteCarlosimulation,providedthatMonteCarloisshowntoreproducedataattheadequatelevelofdetails).Inprinciple,moresophisticatedadditionalcriteriaforpruningcanbefurtheremployed:e.g.,onecantakeintoaccountthechargeclustershapeandnumberofanodehitsassociatedwithaRecHit.Thesewerestudiedinthepastandshowntohavesome,albeitverylimited,discriminatingpoweragainst\bad"hits.However,theirusefortheHLTpurposesishardlyjustiable. 81

PAGE 82

Occupancyforjx=wj-coordinatesofRenedSegments. Figure6-17: PoolsforallretainedRecHits(i.e.,RecHitsassociatedwithRenedSegments.) Figure6-18: InducedchargedistributioncalculatedaccordingtoGattiforlargeandME1/1chambergeometries. Figure6-19: RatiorversusalocalcoordinatexforlargechamberscalculatedforvarietyofstripwidthsintheassumptionoftheGattichargedistributionforlargechambers. 82

PAGE 83

2QrQl 6-19 showsratiorasafunctionofalocalcoordinatexcalculatedfortheinducedchargeddistributionaccordingtoGattietal.[ 77 ](Fig. 6-18 ).Thelocalcoordinatexisassumedtobeinstripunits,x=0correspondstothestripcenter,andx=0.5meansright/leftstripedges.Thisratiorisamonotonic,butnotlinearfunctionofahitpositionacrossastripxWendtheconversionfunctionfromrtoxintwosteps.Therstcorrectionisanapproximateinversionofthe\theoretical"functionr(x;w).Figure 6-20 showsacorrectionthatoneneedstoaddtortoobtainthecoordinatex.Thepointscorrespondtothe\theoretical"Gattifunction.Wendthatthe1st-ordercorrectioncanbeparameterizedquitewellwiththefollowingempiricalfunction(wisastripwidthincm): 6-20 .ThisisalreadysucientfortheHLTpurposes.However,thiscorrectionispurelytheoreticalandmustbecheckedagainstthereality.Figure 6-21 showstheexperimentaloccupancydistributiondN=dx1.Ithasanobvious\wave",whichisamanifestationofthefactthattheinducedchargedoesnotquitefollow 83

PAGE 84

Correctionthatmustbeaddedtothemeasuredratiortoobtainthehitpositionacrossastrip. Figure6-21: Anoccupancydistributionforthe1st-ordercorrectedcoordinatex1st. the\theoretical"Gatti.Curiouslyenough,theshapeofthewavelooksverysimilarforallpartsofallchambers,i.e.itisverysimilarfordierentstripwidths.Thisallowsustointroduceasecond-orderempiricalcorrectioninastrip-widthindependentmanner.ThiscorrectioncanbederiveddirectlyfromtheshapeofthedN=dx1occupancydistributionandcanbeparameterizedasfollows: 6-22 ).However,itdoesmaketheoccupancyplotsforx2almostperfectlyat.Thesensitivityoftheratiomethodtothetypicalelectronicsnoise,calibrationerrors,andcross-talkuncertaintiesareshowninFigs. 6-24 6-25 6-26 .Thetypicalvaluesusedtomaketheseplotsaretheresultsofpre-installationtestingofall396largechambersatso-calledFinalAssemblyandSystemTestssites[ 76 ]. 84

PAGE 85

Thesecond-ordercorrectionthatmustbeaddedtox1toobtainthehitpositionacrossastripx2. Figure6-23: Anoccupancydistributionforthe2nd-ordercorrectedcoordinatex2. Thetypicalnoiseis1%oftheaveragecathodeclustercharge,whenchambersoperateatthenominalgasgain.InadditiontotheFASTsitemeasurements,thenoiselevelswereagainmeasuredinsituduringdatatakingforallinstalledchambersandfoundtoremainunchanged.Ifonedoesnotusecalibrationconstants,thespreadofelectronicchannelgainsbetweennearbystripsisfoundtobe1%.Thedierencesincrosstalksbetweenstripsforchambersofthesametypeareverysmall;the2%numberusedformakingFig. 6-26 representstheRMSvalueforallcrosstalkscollectedinonehistogramregardlessofthechambertype.Onecanseethatchannelgaincalibrationsandcrosstalksaretheleastofourconcerns.Thenoisecontributiontotheresolutionbetweenstripsisalsoverysmall.TheseguresallowsonetogaugehowelectronicsperformancemaycontributetotheCSCspatialresolutionwhentheratiomethodisused. 85

PAGE 86

Sensitivityoftheratiomethodtoelectronicnoise(in%ofstripwidthandmm). Figure6-25: Sensitivityoftheratiomethodtoerrorsinelectronicgaincalibrations(in%ofstripwidthandmm). 86

PAGE 87

Sensitivityoftheratiomethodtouncertaintiesincrosstalksbetweenstrips(in%ofstripwidthandmm). 87

PAGE 88

78 79 80 ]builtfortheentireM(4)-distribution,binnedorunbinned,ortakingastraightforwardcountingexperimentapproach.Wegiveadirectcomparisonofthetwoapproaches.AfulltreatmentofthemostimportanttheoreticalandinstrumentalsystematicerrorsandtheireectontheevaluationofthesignicanceoftheHiggsbosonobservationarepresented.Tominimizesystematicerrors,newmethodsofreconstructingthemostimportantcorrectionsdirectlyfromdataweredeveloped.Amongthemarethemuonreconstructionandisolationcuteciencies.WealsoshowthatbyusingthemeasuredZ!2crosssection,oraneventcountinthesidebandsoftheM(4)distributions,onecansubstantiallyreduceanumberoftheoreticalandinstrumentalsystematicerrors.InadditionweverifybyhowmuchthelocalexcesssignicancewillbeeectivelydegradedduetothefactthatwelookforanarrowresonanceinabroadrangeofM(4)invariantmasses. 88

PAGE 89

81 82 ]andincludepile-upeventscorrespondingtoaninstantaneousluminosityof21033cm2s1.ThenalresultsarepresentedintermsoftherequiredintegratedluminosityforobservingtheStandardModelHiggsbosonat5and3signicancelevelsand95%CLexclusionlimits.Also,wepresentthesignicanceforaxedvalueofanintegratedluminosityequalto30fb1and95%CLexclusioncontoursinthe(MH,)planeforintegratedluminositiesof3,10,and30fb1.PreviousstudiesonthesearchfortheStandardModelHiggsbosonintheH!ZZ()!4channelwithCMSaredescribedin[ 83 84 85 ].Anotherongoingstudyexploringthediscoverypotentialwithadierentsetofmass-independentcutscanbefoundelsewhere[ 86 ].TheresultsofthetwoparallelanalysesusingtheH!4eandH!2e2channelscanbefoundin[ 87 88 ].Inshort,webelieveourH!ZZ()!4analysisismorerealisticthanpreviouslyexistedonesonthetopic: 89

PAGE 90

7 8 ],i.e.isanocialCMSstrategyfortheSMHiggsbosondiscoveryinH!ZZ()!4decaychannel.Detailsofouranalysisandsupportingstudiescanbefoundinalistofrefereedpapers[ 1 2 3 4 5 6 7 8 9 10 ],inallofwhichtheauthorofthisthesiswaseithertheleadingauthororoneoftheleadingco-authors.Resultsofourworkpresentedat10internationalconferences[ 11 12 13 14 15 16 17 18 19 20 ];theauthoralsogavemorethan70talksattheinternalCMScollaborationmeetingsofdierentlevels[ 21 ]. 7-1 )andthethreemainbackgroundprocesses,tt,(Z=)bb,andZ=Z=,weresimulatedusingthefullCMSdetectorsimulationandreconstructionsoftware.Manyotherplausiblebackgroundcandidates,bbbb,bbcc,cccc,single-top,Zcc,Wbb,Wcc,fake,and=KdecaymuonsinQCD,wereconsideredandfoundtobenegligible.TosaveCPUtime,onlyeventswithatleast2+and2inthepseudorapidityrangejj<2:4andwithpT>3GeV=cwereretainedforfurtheranalysis.MuonsoutsidethesekinematicallimitscouldnotbereconstructedbyCMS.Additionalcutswereappliedtodi-muoninvariantmassesfortheHiggsbosonsamples(m(Z)>5GeV=c2)andforZ=Z=and(Z=)bbsamples(m(+)>5GeV=c2).(Therst+pairinZ=Z=and(Z=)bbsampleswasdenedastheonewithitsinvariantmassclosesttom(Z0),whilethesecond+pairwasmadeoutofthetworemaininghighestpTmuonsofoppositesigns.)Allanalysiscutsontheseobservables,tobedescribedbelow,aremuchmorestringentthanthesegenerator-levelpreselectioncuts.Theexpectednumbersofsurviving4eventsforsignalandbackgroundsforanintegratedluminosityofL=30fb1aregiveninTable 7-1 .TheM(4)distributionofeventsafterthesecutsisshowninFigures 7-1 and 7-2 90

PAGE 91

TheLO/NLOcrosssectionsforvariousHiggsbosonmassesandbackgrounds,correspondingnumberofeventswithfourmuonssurvivingthegeneratorlevelpreselectioncuts(seesection 7.2 )calculatedforL=30fb1,andthenumberofsimulatedevents. 4eventsatL=30 SimulatedEvents 47.7 7.69 10000 44.3 13.6 10000 38.4 31.1 9000 33.7 49.2 10000 29.8 54.1 9000 26.6 25.6 9000 23.9 12.3 10000 21.6 28.5 9000 19.7 101 10000 18.0 109 10000 12.4 87.5 10000 9.58 72.3 10000 9.12 72.6 9000 8.81 63.4 9000 6.44 45.1 10000 4.46 31.8 10000 3.07 22.6 10000 2.13 16.3 10000 840 7000 92236 278 8694 124500 seetext 2622 118000 seetext 48.8 10000 89 ](LOgluonandweak-bosonfusion),interfacedviaCMKIN[ 90 ]version310(PDFCTEQ5L).OnlydecaychannelsZ!2l(wherelstandsfore,,and)wereconsidered.Z!qq!2lwerenotincludedinthesimulation:beingverysimilartothe(Z=)bbbackground,thoseeventswouldbesuppressedtogetherwiththe(Z=)bbbackgroundbyouranalysiscuts.QEDradiationfromthenal-statemuonsismodeledwithPHOTOS[ 91 92 ].Eventswerere-weightedtocorrespondtothetotalNLOcross-section(pp!H)BR(H!ZZ)BR(Z!2l)2,where(pp!H)andBR(H!ZZ)weretakenfrom[ 93 94 ]andBR(Z!2l)=0:101[ 95 ]. 91

PAGE 92

7-3 .TheM(4)distributionformH=140GeV=c2aftereventgenerator-levelcutsisshowninFigures 7-1 and 7-2 .Thelow-masstailismostlyduetoeventswheremuonsdidnotcomedirectlyfromZ-decays(e.g.,via-decays)andtointernalbremsstrahlungthatalsotendstomovethe4-muoninvariantmassothepeak. 96 ]andthebranchingratioBR(W!l)=0:320[ 95 ]. 97 ]matrixelementgenerator(PDFCTEQ5L,withQCDscalesR=F=MZ,b-quarkmassmb=4:85GeV=c2,andadi-muonmasscutm(+)>5GeV=c2),interfacedtoPYTHIA6.225forshoweringandhadronization.Includedsub-processeswereqq=gg!Z=bb!+bb,whereqcouldbeanyofthelightquarks,(u,d,s,c)(initialstateswithbquarkswerealsoconsideredatthegeneratorlevelandfoundtobenegligible).Norestrictiononbdecayswasapplied.ThecorrespondingCompHEPLOcrosssectionwasfoundtobe116pb.ToobtaintheNLOcrosssectiongiveninTable 7-1 ,wecalculatedtheNLOK-factorusingMCFM[ 98 ]:KNLO=2:40:3.TheconditionsfortheMCFMNLOandLOcalculationswereasfollows:CTEQ6,2R=2F=^s,mb=0,M(Zres)>5GeV=c2,pT(b)>5GeV=c,jbj<10,M(bb)>10GeV=c2. 92

PAGE 93

2 ]fordetails.Theinterferencebetweent-ands-channelswasfoundtobealwaysnegligible.TheCompHEPeventswerefurtherinterfacedtoPYTHIA6.225forshoweringandhadronization.TheCompHEPLOcrosssectionsforthetwosub-processeswere113and157fb,respectively.ToaccountforcontributionstoalltheNLOdiagramsandtotheNNLOgluonfusionprocess(gg!ZZ,knowntocontribute20%withrespecttotheLO[ 99 ]),were-weightedeventswithaM(4)-dependentK-factorK(M4)=KNLO(M4)+0:2.TheNLOK-factorKNLO(M4)wasobtainedwithMCFM[ 98 ]andisshowninFigure 7-4 .AlldetailsoncalculationofthisM(4)-dependentK-factorandthedynamicaldierencesbetweenNLOandLOaresummarizedelsewhere[ 6 ].TheM(4)distributionsaftergenerator-levelcutsareshowninFigures 7-1 and 7-2 .ThepeakatMZisduetothes-channelcontribution.ThispeaksitsontheshoulderoftheenhancementaroundM(4)>100GeV=c2|thispeakcorrespondstooneofthetwoZ'sgoingon-shellinthet-channel.Thenextbumparound2MZisduetobothZ'sgoingon-shell. 7 ].TheinclusivemuontriggersbasedontheselectionofasinglemuonwithpT>19GeV=c

PAGE 94

7-5 and 7-6 showtheeciencyturn-oncurves-thechoiceofthesecutsisobviousfromthegures.Thesecutsdonotdramaticallyaectthenumberofacceptedsignalevents.Werequirethatallfourpossiblecombinationsofreconstructeddi-muonmassessatisfym(+)>12GeV=c2.Asinthepreviouscase,thiscuthasverylittleeectontheHiggsbosoneventsandisprimarilyintendedtosuppresspoorlysimulatedhadronicbackgroundcontributionsoriginatingfromcharmoniumandbottomoniumdi-muondecays.ThemostimportantcharacteristicdistinguishingtheHiggsbosondecaysfromallbackgroundsisthepresenceofapeakinthefour-muoninvariantmassdistribution.Figure 7-7 showssuchadistributionforMH=150GeV=c2.AGaussiantofthepeakgives=1.1GeV=c2.Onecanseeanoticeabletailtowardsmallermasses-mostlyduetointernalbremsstrahlungandeventswithintermediate-leptons(ZZ!22!44andZZ!4!48).Thefour-muonmassdetectorresolutionM(4)asafunctionofM(4)isgiveninFigure 7-8 7.4.2 .Theresultswithat,M(4)-independent,cutsarealsopresentedforcomparison. 94

PAGE 95

7.4.3.1 ).WeproposeandanalyzethemeritsofanumberofmethodsforobtainingvariousanalysiscorrectionsdirectlyfromdatainordertominimizeourrelianceonMonteCarlosimulation(bothphysicsanddetectorperformance).Bydoingthis,wesignicantlyreducesystematicerrorsanduncertainties. 7.4.3.2 ).Forthecountingexperimentapproach,thiscanbedoneanalyticallyinastraightforwardway.ForLLRandothermoresophisticatedstatisticaltools,thiscanbedoneonlybyrunningalargenumberofpseudo-experimentsandwouldalsorequireaknowledgeofallcorrelationsacrossthedata,theM(4)-spectruminthiscase,|thesecorrelationsarenotyetavailable. 7.4.4 ,weoutlineastraightforwardmethodologyofevaluatingthescopeofthiseect.Tobeabletodoitright,welldenedsearchassumptionsmustbesetapriori. 95

PAGE 96

78 ].Thelog-likelihoodratio(LLR)canbebuiltforanarrowregioninthevicinityofthepeak(countingexperiment): b;(7{1)fortheentirebinnedspectrum: 7-9 and 7-10 showthattheScLtracksthetruesignicanceSforevensmallnumbersofbackgroundevents.Notethatotherpopularquickestimators,S1=s=p 7-9 and 7-10 ,donotworkaswellforsmallereventcountsandlargevaluesofsignicance.TheSLestimator,beingsensitivetothefullshapeofthesignalandbackgrounddistributions,hasaleadingedgeoverthesimplecountingexperimentScLestimator.Wetypicallyobserve5-10%dierence.TheScLestimator,beinglocal,isthenaturaltoolforoptimizationofM(4)-dependentcuts.Theconvolutionofsystematicerrorsinto 96

PAGE 97

1CLb=P=p(nnojb)=+1Xk=nobk 7{4 todenethetruecountingexperimentsignicanceSc.IncludingsystematicerrorsintoevaluationofsignicanceisdiscussedinSection 7.4.3.2 97

PAGE 98

7.3 areselected.Thisensuresthatmuonsarereliablyreconstructedandremovesa\contamination"originatingfromheavyquarkoniadecays. 100 ].ThecountingexperimentsignicanceestimatorScListhenaturaltoolforsuchoptimization.Forcutoptimization,weconsideredthefollowingmuonkinematicvariables: 100 ].Thisprogram,beingrelativelynew,isalreadywidelyusedforotherCMSanalyses[ 101 102 103 ]. 98

PAGE 99

7-11 and 7-12 ).ThiscutstronglysuppressesttandZbbbackgrounds.OnecanseethatthecutsgetstighterasM(4)getssmallerastheroleofZbbandttincreases(Fig. 7-1 ). 7-13 ).ThiscuthelpstofurthersuppressZbbbackgroundtothelevelwellbelowZZandreducestheZZbackgroundathighfour-muoninvariantmasses.ThiscutgetsmorestringentwithincreasingM(4),asthetransversemomentumofmuonsfromahighmassHiggsbosontendtobehigherthanthosefromZZbackground. 7-14 ).Itroughlycorrespondstothe2width,whereistheHiggsbosonpeakwidththatincludesthedetectorresolutionandtheStandardModelHiggsbosonwidth.Inthenextstep,weappliedthethreecriticalcutstothesecondhalfoftheavailableMonteCarloeventsthatwerenotusedfortheoptimizationofthecuts.WetriedboththecutsastheycamefromtheoptimizationwithGARCONand,alternatively,thesmoothcutsshownbythelinesinFigures 7-11 7-14 .TheresultsofthesecomparisonsareshowninFigure 7-19 .Theobservedstabilityoftheresultsasweswitchfromthersthalfofthesampleusedforcutoptimizationtothesecondhalf,aswellasfrom\theoptimal"tothesmoothcuts,ensuresthatthecutoptimizationdidnotpickpeculiarphasespace 99

PAGE 100

7-11 7-14 :theyareoutoftheplot-rangesand,insomecases,pushedbyGARCONtotheextremelimits.ThelattermeansthatthisparticularcutparameterforthisparticularHiggsmasspointisnoteective,asoftenisthecase,forexample,fortwoisolation-parameter-basedvariables,duetotheirveryhighdegreeofcorrelation.Figure 7-20 showstheM(4)invariantmassdistributionforthethreebackgroundsubprocessesandaHiggsbosonsignalatMH=150GeV=c2afterapplyingthethreesmoothM(4)-dependentcuts.OnecanseethatthettandZbbbackgroundsarenowsuppressedwellbelowtheirreducibleZZbackground.Otherpossiblecutssuchasinvariantmassesofthemuonpairs,impactpointparameters,kinematicalcutsonothermuons,andisolationparametersonothermuonsdonotsignicantlyhelptoimprovetheresultsfurther,seeFigures 7-15 7-18 .Thecutsontheseobservablesmaystillbeusefulandplayaroleof\safeguards"tosuppresspossibleunaccounted-forbackgroundsrelatedtothebeamhalo,detectormis-performance,etc.AdditionalvariablesthatmayhelptodiscriminateHfromthedominantZZbackgroundhavebeenstudied:pT(4),numberofjetsandtheirET,etc.However,thesevariablesaredrivenbytheNLOproductionprocesses,whileoursamplesweregeneratedattheLeadingOrderbyPythiaandCompHEP.Therefore,anyconclusionsthatwemightderivefromthesesampleswouldnotbereliable.Someangulardistributionsbuiltfrommuonsalsohavesomedierencesoriginatingfromtheunderlyingspinstructures,buttheyarenotsucientlydiscriminatingtobeusedforcutsandmaybestronglyaectedbytheNLOdiagrams.Figures 7-21 and 7-22 showthesignicancesScLandSLatL=30fb1fortheexpectedexcessesofeventsfordierentHiggsbosonmasses.ToemphasizethegaininthesensitivityachievablewithM(4)-dependentcuts,theresultsforatcuts,optimizedforMH=150GeV=c2,arealsosuperimposed.Asexpected,oneuniversalsetofcutscannotdelivertheoptimalperformanceforthefullrangeofpossibleHiggsmasses.Thegainin 100

PAGE 101

7-21 ).Thedierenceintheaverageexpectedsignicances,5.3and4.6,meansinthiscasethatthechancesofdiscoveringtheHiggsbosonwithMH=500GeV=c2atL=30fb1are<40%fortheatcutsand>60%fortheM(4)-dependentcuts.Figure 7-23 showsthesameresultsasinFigure 7-21 intermsoftheluminosityneededforobservinganexcessofeventsovertheexpectedbackgroundinthepresenceoftheStandardModelHiggsbosonat5signicance.Figure 7-24 givesanideaofhowtheexperimentalexclusionlimitswillmapontotheplaneofcrosssectionvs.Higgsbosonmassforafewdierentintegratedluminosities. 101

PAGE 102

98 ].SystematicerrorsassociatedwithPDFswereestimatedbygiving1variationstothe20CTEQ6Mparameters.Byvarying 102

PAGE 103

5 ].Figure 7-25 showsthesePDFandQCDscaleuncertainties,addedinquadrature,versusM(4).Thethreecurvescorrespondto(a)theabsolutepredictions(relativelyat,b=b6%);(b)thepredictionnormalizedtothemeasuredZ!2crosssection(notethattheZ!2crosssectioncanbemeasuredwithinstrumentalsystematicerrors,notincludingluminosity,oflessthan2%(CDFresults,Phys.Rev.Letter,94(2005)091803))(=1%forM(4)closetoMZandthensteadilyincreasingtowardlargerfour-muoninvariantmasses);and(c)predictionnormalizedtosidebandsoftheM(4)distributionitselfintherangefrom100-600GeV=c2((=isatitsminimumwhenthesignalwindowisattheplacewheremostoftheeventsare).Beyond-Leading-Ordercorrectionuncertaintieswereestimatedasfollows.TheM(4)-dependentK-factorK(M4)fortheZZ!4processwasevaluatedwithtwoverydierentprograms:MCFM[ 98 ]andENLO[ 6 ].ThelatterisapackagesmoothlysplicingtogetherMadGraph[ 104 ](NLOpp!4+jet)andPythia(LOpp!4+ISR-jets).TherelativedierenceinK(M4)isshowninFigure 7-26 .TheNNLOdiagramsincludenewprocesses(wedeneaprocessasnewifithasadistinctlydierentinitialstateand,therefore,variationsofQCDscalesdonotnecessarilygiveafeelforitsrelativeimportance):gg!ZZ!4(box-diagram),contributingabout(208)%totheLOcrosssection[ 99 ](notethatthiscontributionwascalculatedwithoutvirtualphotonsinthepropagators)andqq!4+qqviaZ-bremsstrahlung(notyetcalculated)orviavector-bosonfusion(implementedinPythia,verysmall).Sincethenatureofallthesedierences/variationsisnotwellunderstood,wepresentthenalresultswithandwithouttheseuncertaintiesincluded.Certainly,moretheoreticalworkinthisareaisneeded.Allotherhigher-leveldiagramscanbeconsideredascorrectionstothedistinctLO,NLO,andNNLOprocessesdiscussedabove.Omissionofthesehigher-ordercorrectionswould 103

PAGE 104

105 ].BycountingthenumberofZ!2eventsintheresonancepeakoftheinvariantmassdistributionsbuiltfromtheHLTmuonandallothertracks,theHLTmuonandallotherstandalonemuonsandtheHLTmuonandallothergloballyreconstructedmuons,onecanevaluatetheeciencyofndingglobally-reconstructedmuonswithbetterthan1%precision.Suchameasurementwillautomaticallyaccountfortherealdetectorperformance,includingintermittentandsmoothvariationsintime.Alldetailscanbefoundelsewhere[ 3 ],seealsoSec. C .Thefour-muoneciencythereforewillbeknownwithanabsoluteerrorofbetterthan4%.WhenwededucetheexpectedZZ!4eventsfromthemeasuredZ!2crosssection,thisuncertaintypartiallycancelsoutandbecomes2%.ThiseciencyremainsfairlyatvsM(4),whichmakesthiserrorcompletelynegligibleifsidebandsareusedforevaluatingthenumberofexpectedbackgroundeventsinthesignalregion.ThemuonisolationcutisveryimportantasitallowsustosuppressotherwiseoverwhelmingttandZbbbackgroundswellbelowtheZZbackground.Asweapplythis 104

PAGE 105

4 ](Fig. 7-27 ).TheuncertaintyonthemuonpTresolutiondirectlypropagatesintothefour-muoninvariantmassM(4)reconstruction.Thisalmostdoesnotaectthebackgrounddistribution.However,theM(4)distributionwidthdrivesthewidthoftheM(4)windowthatweuseforevaluatingthesignalexcesssignicanceatlowHiggsbosonmasses.Fortunately,evenmakingamistakeintheM(4)distributionwidthbyasmuchas25%hasonlyatinyeectonevaluatingasignicanceofanexcessofevents(Fig. 7-28 ).Also,themuonpTresolutionisfairlyeasytomeasurefromdatausingthemeasuredJ=andZpeakwidthswithaprecisionmuchbetterthanneeded.TheuncertaintyonthemuonpTscalecanbesimilarlycalibratedfromdatausingthemeasuredJ=andZpeaks.TheeectoftheseuncertaintiesonthenumberofbackgroundeventsinasignalwindowappearsonlyonsteepslopesoftheM(4)distribution.ForthesteepestpartoftheM(4)distributioninthe180200GeV=c2range,weobtainb=b0:1M4,whereM4isinGeV=c2.Thisimpliesthattobeabletoneglectthiseect,oneneedstoknowthemomentumscalewithaprecisionof0:1GeV=catpT50GeV=c.ThiscanbeeasilyachievedwithjustafewhundredZ!2events.Figures 7-29 and 7-30 summarizeallthesystematicerrorsontheexpectednumberofeventsintheZZ!4backgroundforthetwomethods:viareferencingtothetotal 105

PAGE 106

7{4 .Wewillusealog-normalformofaprobabilitydensityfunctionfortheabsolutesystematicerrorsforexpectednumberofbackgroundeventsb0witharelativeuncertainty=b=b0: 2ln2(k)1 7{7 doesjustthatforanyvalueofk,smallorlarge(kwouldbeequal2inthiscase). 106

PAGE 107

7-31 givesthreecurves:thesignicancevsHiggsmassintheabsenceofanysystematicerrors(bothfortheplainScLestimatorandthetruesignicanceSc)aswellasthesignicancethatincludesalluncertaintiesinthebackgroundwhenitisestimatedfromthemeasuredZ!2crosssection.Allthreecurvescorrespondtothetotalintegratedluminosityof30fb1.Figure 7-32 showscurves(withandwithoutsystematicerrors)fortherequiredluminosityfor5-discovery,3-evidence,and95%CLexclusionlimitfortheStandardModelHiggsboson.Thecomparisonbetweentwowaysofnormalization,totheZ!2processandtheZZ!4sidebands,ismadeintermsoftheluminosityrequiredfor5-discovery(Fig. 7-33 ).Thedierenceisnotdramatic.Thetruebenetofusingtwoapproachestoestimatingbackgroundfromdataisintheircomplementarity.Finally,Table 7-2 summarizesthemostimportantresultsfortheM(4)-dependentcutsthatwepresentedinthischapter. 7.6 foralldetails)andonemustexerciseacautioninevaluatingprobabilisticinterpretation 107

PAGE 108

Summaryoftheresults:numberofsignalandbackgroundeventsinawindowusedforacountingexperimentwiththeM(4)-dependentcuts.SystematicerroronthebackgroundisnormalizedtotheZ!2process(KNLO=KNLOisnotincluded);threedierentsignicanceswithoutsystematicerrorsincluded:theSLestimatorfortheLogLikelihoodRatio(LLR)builtforthefullM(4)spectrum,ScLLLRestimatorbuiltforacountingexperimentapproach,andtheSctruesignicanceforthecountingexperimentapproach;thenalresultforSc,nowincludingallsystematicerrors. Signal Bkgd SystError b b=b (nosyst) (nosyst) (withsyst) 115 2.13 0.92 3% 1.98 1.75 1.54 1.54 120 4.00 1.15 4% 2.88 2.72 2.57 2.56 130 12.45 2.06 3% 5.97 5.64 5.54 5.52 140 23.22 2.65 3% 9.09 8.45 8.39 8.35 150 28.09 2.42 4% 10.84 9.92 9.87 9.81 160 14.25 3.01 4% 6.04 5.64 5.55 5.53 170 6.32 3.63 4% 3.00 2.73 2.61 2.60 180 14.54 7.10 4% 4.83 4.38 4.30 4.26 190 54.95 17.00 4% 10.85 9.89 9.85 9.59 200 62.78 19.93 4% 11.32 10.48 10.45 10.11 250 54.48 21.62 5% 9.83 9.09 9.05 8.61 300 40.43 13.40 6% 9.26 8.30 8.25 7.90 350 40.68 10.75 7% 9.50 8.93 8.88 8.47 400 33.59 8.00 8% 9.07 8.36 8.31 7.95 450 24.07 5.68 8% 7.57 7.10 7.03 6.81 500 16.65 5.58 9% 5.81 5.31 5.23 5.05 550 12.10 5.70 9% 4.44 4.04 3.92 3.80 600 8.72 5.04 9% 3.58 3.20 3.10 3.00 108

PAGE 109

106 107 ])andphysicistsalike(e.g.,[ 108 109 110 111 112 ]).ThepurposeofthisAppendixistoquantifya 109

PAGE 110

SameasFigure 7-1 ,butonalinearscale. Enhancementtothesignalsamples'crosssectionsduetointerferenceeectsnotaccountedforatthegeneratorlevel. TheM(4)-dependentNLOK-factorKNLO(M4)fortheZZ!4processevaluatedwithMCFM[ 98 ]. 110

PAGE 111

GlobalmuonreconstructioneciencycalculatedfrommatchingreconstructedandtrueMonteCarlomuonsinthebarrelregionvs.pT. GlobalmuonreconstructioneciencycalculatedfrommatchingreconstructedandtrueMonteCarlomuonsintheendcapregionvs.momentum. 111

PAGE 112

Comparisonofdierentsignicanceestimatorsfor1backgroundevent:probabilityofmeasuringsignicanceS>S0,backgroundonlycase,NB=1event. SameasFigure 7-9 ,butforNB=10events. Dependenceofthetracker-basedmuonisolationcutontheleastisolatedmuonversusHiggsmass. Dependenceofthecalorimeter-basedcutontheleastisolatedmuonversusHiggsmass. 112

PAGE 113

DependenceofthepTcutonthesecond-lowest-pTmuonversusHiggsmass. DependenceoftheM(4)windowcutsversusHiggsmass. Firstmuonpairinvariantmassdistribution,M(Z1),afteranalysiscutswereapplied. Secondmuonpairinvariantmassdistribution,M(Z2),afteranalysiscutswereapplied. 113

PAGE 114

LowestmuonpTdistribution,afteranalysiscutswereapplied. MaximumdistanceinXY-planebetweenmuonimpactpointcoordinatesdistribution,afteranalysiscutswereapplied. 114

PAGE 115

ExpectedexcesssignicanceScLwithL=30fb1fordierentHiggsbosonmassesforM(4)-dependentandindependentcuts.Nosystematicerrorsincluded. ExpectedexcesssignicanceSLwithL=30fb1fordierentHiggsbosonmassesforM(4)-dependentandindependentcuts.Nosystematicerrorsincluded. Luminosityrequiredtoreacha5eventexcessfordierentHiggsbosonmassesforM(4)-dependentandindependentcuts.Nosystematicerrorsincluded. The95%CLexclusioncontoursfortheSMHiggshypothesis. 115

PAGE 116

CombinedsystematicerroronthenumberofbackgroundeventsduetoPDFandQCDscaleuncertaintiesforthe(qq!ZZ!4)processatNLO. Top:thefactorsKNLO(M4)inMCFMandENLOcalculationsversusM(4);bottom:thedierencebetweenthem. MuonisolationcuteciencyforrandomconedirectionforZ-inclusive(dashedlines)andforZZ(solidlines)events. Anexampleofapossiblebiasinevaluatingthesignicanceofaneventexcessduetoanon-optimalchoiceofthesignalwindowwidth. 116

PAGE 117

UncertaintiesinthenumberofZZ!4backgroundeventsinthesignalregionwindowatdierentM(4).TheeventcountisreferencedtothenumberofZ!2events. UncertaintiesinthenumberofZZ!4backgroundeventsinthesignalregionwindowatdierentM(4).Theeventcountinsignalregion,iscalculatedfromthenumberofZZ!4eventsintherange100-700GeV=c2(excludingthesignalregionwindow). 117

PAGE 118

Signicancevs.Higgsmass(withandwithoutdF/Kcontribution). Integratedluminosityneededfor95%CLexclusion,3,and5discoveryversusHiggsbosonmass(withandwithoutdF/Kcontribution). 118

PAGE 119

Integratedluminosityneededfora5discoveryoftheHiggsbosonversusitsmassfor(withandwithoutdF/Kcontribution). possiblescopeofthiseectonanexampleofasearchfortheStandardModelHiggsbosonintheH!ZZ()!4decaychannel.Asthecasestudy,wechoseacountingexperimentapproachwidelyusedinthisvolume.ThedashedlineinFig. 7-34 showstheexpected4invariantmassdistributionforbackgroundatL=30fb1afterapplyingallthem4-dependentanalysiscuts.Usingthisdistribution,weplayedout108pseudo-experiments;anexampleisshowninFig. 7-34 .Foreachpseudo-experiment,weslidasignalregionwindowacrossthespectrumlookingforalocaleventexcessovertheexpectation.Thesizeofthewindowm=w(m4)wasoptimisedandxedapriori(about2)togiveclosetothebestsignicanceforaresonancewithawidthcorrespondingtotheexperimentalSMHiggsbosonwidth(m4).Thestepofprobingdierentvaluesofm4was\innitesimally"small(0.05GeV=c2)incomparisontotheHiggsbosonwidthofmorethan1GeV=c2.Thescanningwasperformedinaprioridenedrangeof115-600GeV=c2.WeusedasignicanceestimatorScL=sign(s)p 119

PAGE 120

1 ].Figure 7-35 presentstheresultsofsuchascanforthepseudo-experimentshowninFig. 7-34 .ThemaximumvalueofScL,Smax,andthecorrespondingmassofa\Higgsbosoncandidate"obtainedineachpseudo-experimentwereretainedforfurtherstatisticalstudies.Afterperforming108pseudo-experiments,thedierentialprobabilitydensityfunctionforSmaxanditscorrespondingcumulativeprobabilityfunctionP(Smax>S)(Fig. 7-36 )werecalculated.FromFig. 7-36 ,onecanseethatthefrequencyofobservingsomelargevaluesofScL(solidline)ismuchhigherthanitsnaiveinterpretationmightimply(dashedline).Ifdesired,theactualprobabilitycanbeconvertedtothetruesignicance.Theresultofsuch\renormalisation"ispresentedinFig. 7-37 .Onecanclearlyseethattherequiredde-ratingofsignicanceisnotnegligible;infact,itislargerthantheeectofincludingalltheoreticalandinstrumentalsystematicerrorsforthischannel.Moredetailsonthevariousaspectsofthesestudiescanbefoundin[ 1 ].Therearewaysofreducingtheeect.Amoredetailedanalysisoftheshapeofthem4distributionwillhelpsomewhat.Usingthepredictednumberofsignaleventss=stheoryinthesignicanceestimatortobeginwithand,then,forvalidatingthestatisticalconsistencyofanexcessnobwiththeexpectationstheorywillreducetheeectfurther.Onecanalsouseanon-atpriorontheHiggsmassasitcomesoutfromtheprecisionelectroweakmeasurements.Whetheronewillbeabletobringtheeecttoanegligiblelevelbyusingalltheseadditionalconstraintsonthesignalhypothesesisyettobeseen.ThepurposeofthisAppendixisnottogivethenalquantitativeanswer,butrathertoassertthatthesestudiesmustbecomeanintegralpartofallfuturesearchanalyseswhenmultiplesignalhypothesesaretried. 120

PAGE 121

Thebackgroundpdfandanexampleofonepseudo-experimentwithastatisticaluctuationappearingjustlikeasignal. Figure7-35: ProleoftheScLscancorrespondingtothepseudo-experimentexampleshownontheleft.Green(inner)andyellow(outer)bandsdenote1and2intervals.Spikesthatcanbeseenareduetoeventscominginordroppingothetrial-window,afeatureoflow-statisticssearches. Figure7-36: Figure7-37: Localsignicance\renormalisation"fromanobservedvaluetothetruesignicancewithaproperprobabilisticinterpretation. 121

PAGE 122

122

PAGE 123

123

PAGE 124

(A.Drozdetskiyetal.,\RelativeContributionsoft-ands-ChannelstotheZZ()!4Process",CMSNote2006/057)Theqq!ZZ!4processisthemainirreduciblebackgroundinsearchesfortheHiggsbosonviaitsH!ZZ!4decaymode.PYTHIA[ 89 ],aneventgeneratorcommonlyusedforsimulationofthisprocessattheLHC,unfortunatelyismissingthes-channelcontribution(Fig. A-1 ).Inourdetailedstudyweshowthatthes-channelsubprocessanditsinterferencewiththet-channelcannotbeneglectedinthecontextofthebackgroundsfortheH!4analysisintheareaofinterest,M4>115GeVifoneaimstosimulatetheZZ-backgroundwithaprecisionof10%orbetter.Thiscontributionremainsnon-negligibleafterallanalysiscuts(Fig. A-1 ).Aspectacularpeakappearingatthe2-muoninvariantmassm(4)mZ0duetopresenceofthes-channelmayprovetobecomeastandardcandleformonitoringtheleveloftheZZ-background. 124

PAGE 125

ZZbackground:t-ands-channeldiagramsands-channelcontributionpeakaroundZ0massafterpre-selectioncutsforfullysimulatedevents. 125

PAGE 126

(A.Drozdetskiyetal.,\ObservationoftheZZdi-bosonproductionintheZZ()!4channel",CMSNoteinCMSapprovalprocess.)WeshowthattheobservationoftherstfewZZ!4eventswithsignicanceinexcessof5shouldbeexpectedbythetimeCMSintegratesluminosityof0.5-1fb1.Thecurrentestimateofthenumberofbackgroundeventsis0:0173Zbb+tt0:00998Zbb0:0116ttandislimitedbytheavailableMCttsamplestatistics.AmethodologyofcalculatingsignicanceincircumstanceswhennoorveryfewMonteCarlobackgroundeventssurviveanalysiscutsisdescribed.Wealsodemonstratethatacontrolsampleof4l-combinationsofwrongavorand/orchargecombinationprovidesapowerfultoolforcrosscheckingthattheobservedeventsarenotduetosomeunaccountedbackgrounds. 126

PAGE 127

(A.Drozdetskiyetal.,\MeasuringMuonReconstructionEciencyfromData",CMSNote2006/060)Amethodofmeasuringtheglobalmuonreconstructioneciencydirectlyfromdatawasstudied.WiththedatacorrespondingtoanintegratedluminosityL=10fb1,theprecisionofmeasuringformuonsinthePTrangeof10100GeVisexpectedtobebetterthan1%,potentiallymuchbetter.ThemethodlargelyalleviatesuncertaintiesassociatedwithourabilitytomonitorandreproduceinMonteCarlosimulationallofdetailsoftheunderlyingdetectorperformance.Themethodusesadatasamplebasedonsingle-muonHLT(HLTstandsforHighLevelTrigger,thenalstageofonlinelteringafterwhichthedataarerecordedontape)triggerwithpT>19GeV=c.ThissamplewillcontaininclusiveW,Z,andotherprocessesintheapproximateratioW:Z:others=10:1:small[ 105 ].BycountingthenumberofZ!2eventsintheresonancepeakoftheinvariantmassdistributionsbuiltfromtheHLTmuonandallothertracks,theHLTmuonandallotherstandalonemuonsandtheHLTmuonandallothergloballyreconstructedmuons(Fig. C-1 ),onecanevaluatetheeciencyofndingglobally-reconstructedmuonswithbetterthan1%precision.Thefour-muoneciencythereforewillbeknownwithanabsoluteerrorofbetterthan4%.ThiseciencyremainsfairlyatvsM(4),whichmakesthiserrorcompletelynegligibleifsidebandsareusedforevaluatingthenumberofexpectedbackgroundeventsinthesignalregion. 127

PAGE 128

128

PAGE 129

(A.Drozdetskiyetal.,\SensitivityoftheMuonIsolationCutEciencytotheUnderlyingEventUncertainties",CMSNote2006/033.)Theisolationcuteciencypermuonduetouncertaintiesintheconsideredunderlyingevent(UE)modelsvaryasmuchas5%(theeciencyitselfanditsuncertaintystronglydependonhowtighttheisolationcutis).The4-muonisolationcuteciencypereventforZZ!4backgroundismeasuredtobe(786)%.Todecreasetheselargeuncertaintiestoanegligiblelevelwithrespecttoothersystematicuncertainties,weproposetocalibratetheisolationcuteciencyfromdatausingZ-inclusiveevents(Z!2)andtherandom-conetechnique.WeshowthatthisindeedsignicantlydecreasestheuncertaintiesassociatedwithapoorunderstandingoftheUEphysics.Theremightbe2%systematicshiftinthe4-muonisolationcutecienciesobtainedthisway.Inprinciple,onecouldcorrectforthisshift,butitdoesnotappeartobenecessaryasthisuncertaintyisalreadysmallerthanothersystematicandstatisticalerrors. 129

PAGE 130

(A.Drozdetskiyetal.,\GARCON:GeneticAlgorithmforRectangularCutsOptimizatioN.User'smanualforversion2.0",hep-ph/0605143,http://drozdets.home.cern.ch/drozdets/home/genetic/)TypicallyHEPanalysishasquiteafewselectioncriteria(cuts)tooptimizeforexampleasignicanceofthe\signal"over\background"events:transverseenergy/momentacuts,missingtransverseenergy,angularcorrelations,isolationandimpactparameters,etc.Insuchcasessimplescanovermulti-dimensionalcutsspace(especiallywhendoneontopofascanovertheoreticalpredictionsparametersspacelikeforSUSYe.g.)leadstoCPUtimedemandvaryingfromdaystomanyyears...Oneofthealternativemethods,whichsolvestheissueistoemployaGeneticAlgorithm(GA),seee.g.[ 113 114 115 ].Wewroteacode,GARCON[ 100 ],whichautomaticallyperformsanoptimizationandresultsstabilityvericationeectivelytrying1050cutsetparameters/valuespermutationsformillionsofinputeventsinhourstime.ExamplesofanalysesarepresentedinthePhysicsTDR[ 8 ]andrecentpapers[ 1 101 102 103 ].IncomparisontootherautomatizedoptimizationmethodsGARCONoutputistransparenttouser:itjustsayswhatrectangularcutvaluesareoptimalandrecommendedinananalysis.Interpretationofthesecutvaluesisabsolutelythesameaswitheye-ballingcutswhenoneselectsasetofrectangularcutvaluesforeachvariableina\classical"waybyeye.All-in-allitisasimpleyetpowerfulready-to-usetoolwithexibleandtransparentoptimizationandvericationparameterssetup.Itispubliclyavailablealongwithapaperonit[ 100 ]consistingofanexamplecasestudyanduser'smanual. 130

PAGE 131

[1] A.Drozdetskiyetal.,SearchstrategyfortheStandardModelHiggsbosonintheH!ZZ()!4decaychannelusingM(4)-dependentcuts,2006,2006/122CMSNote. [2] A.Drozdetskiyetal.,Relativecontributionsoft-ands-channelstotheZZ!4process,2006,2006/057CMSNote. [3] A.Drozdetskiyetal.,Measuringmuonreconstructioneciencyfromdata,2006,2006/060CMSNote. [4] A.Drozdetskiyetal.,Sensitivityofthemuonisolationcuteciencytotheunderlyingeventuncertainties,2006,2006/033CMSNote. [5] P.Bartalinietal.,StudyofPDFandQCDscaleuncertaintiesinpp!ZZ!4eventsattheLHC,2006,2006/068CMSNote. [6] P.Bartalinietal.,NLOvs.LO:kinematicaldierencesforsignalandbackgroundintheH!ZZ()!4analysis,2006,2006/130CMSNote. [7] M.DellaNegra,A.Petrilli,A.Herve,L.Foa,etal.,CMSphysics:technicaldesignreport,volumeI,2006,(CMScollaboration),inparticularseesection9.3.4. [8] M.DellaNegra,A.Petrilli,A.Ball,L.Foa,etal.,CMSphysicstechnicaldesignreport,volumeII,2007,(CMScollaboration),inparticularseesections3.1,AppendixA.2,AppendixD. [9] V.Buescheretal.,Tevatron-for-LHCreport:preparationsfordiscoveries,2006,hep-ph/0608322. [10] C.Buttaretal.,LeshouchesphysicsatTeVcolliders2005,standardmodelandHiggsworkinggroup:summaryreport,2005,contributedtoLesHouchesworkshoponphysicsatTeVcolliders,LesHouches,France.hep-ph/0604120. [11] A.Drozdetskiyetal.,SMHiggsboson:recentdevelopmentsatCMSandATLAS,2007,(CMSandATLAScollaboration),talkgivenatTheXLIIndRencontresdeMoriondonElectroweakInteractions,LaThuile,Italy,2007(willbepublishedintheconferenceproceedings). [12] A.Drozdetskiyetal.,CMSDetectorsensitivitytotheStandardModelHiggsbosoninH!ZZ()!4decaychannel,2007,(CMScollaboration),talkgivenatAmericanPhysicalApril2007meeting. [13] A.DrozdetskiyandS.Abdullin,GARCON:geneticalgorithmforrectangularcutsoptimizatioN,2007,talkgivenatTheXIInternationalWorkshoponAdvancedComputingandAnalysisTechniquesinPhysicsResearch,Amsterdam,Netherlands,2007. 131

PAGE 132

[14] A.Drozdetskiyetal.,CMSdetectorsensitivitytotheStandardModelHiggsbosoninH!ZZ()!4decaychannel,2007,(CMScollaboration),talkgivenatTheHadronColliderPhysicsSymposium2007,Elba,Italy(willbepublishedintheconferenceproceedings). [15] S.Abdullinetal.,SearchstrategyforthestandardmodelHiggsbosonintheH!ZZ()!4decaychannelusingM(4mu)-dependentcuts,2007,(CMScollaboration),talkgivenatThePhysicsatLHC,Cracow,Poland,2006(willbepublishedinproceedings). [16] A.Drozdetskiyetal.,Sensitivityofthemuonisolationcuteciencytounderlyingeventuncertainty,2006,(CMScollaboration),talkgivenatThe2ndHERA-LHCworkshop,CERN,Geneva,Switzerland(publishedinproceedings). [17] A.Drozdetskiyetal.,MC@NLOvsPYTHIAforH!ZZ()!4,2005,(CMScollaboration),talkgivenatTheLesHouches2005:PhysicsatTeVColliders,LesHouches,France. [18] A.Drozdetskiyetal.,EectofUEonisolation,2005,(CMScollaboration),talkgivenatTheLesHouches2005:PhysicsatTeVColliders,LesHouches,France(publishedinproceedings). [19] A.Drozdetskiyetal.,ThediscoverypotentialofsupersymmetryatCMSwithinthemSUGRAmodelusingsame-signdi-muons.,2005,(CMScollaboration),talkgivenatThe40thRencontresdeMoriondonQCDandHighEnergyHadronicInteractions,LaThuile,AostaValley,Italy(publishedinproceedings,hep-ex/0505034). [20] A.Drozdetskiyetal.,FirstdetailedstudyontheCMSSUSYdiscoverypotentialwithtwosamesignmuonsinthemSUGRAmodel,2004,(CMScollaboration),talkgivenatThePhysicsatLHC,Vienna,Austria(PublishedinCzech.J.Phys.55:B249-B256,2005). [21] A.Drozdetskiyetal.,internalCMScollaborationmeetingsofdierentlevels,2002-2007,see:http://indicosearch.cern.ch/andhttp://agenda.cern.ch/search.php. [22] C.-N.Yangetal.,Conservationofisotopicspinandisotopicgaugeinvariance.Phys.Rev.,96:191-195,1954. [23] P.Higgsetal.,Brokensymmetries,masslessparticlesandgaugeelds.Phys.Lett.,12:132-133,1964. [24] P.Higgsetal.,Brokensymmetriesandthemassesofgaugebosons.Phys.Rev.Lett.,13:508-509,1964. [25] S.Glashow,Partialsymmetriesofweakinteractions.Nucl.Phys.,22:579-588,1961. [26] S.Weinberg,Phys.Rev.Lett.19,1264-1266(1967).

PAGE 133

[27] A.Salam,Elementaryparticletheory,.AlmqvistandWiksells,Stockholm,1968. [28] G.Arnisonetal.,Experimentalobservationofleptonpairsofinvariantmassaround95-gev/c2atthecernspscollider.Phys.Lett.,B126:398-410,1983. [29] G.Arnisonetal.,Experimentalobservationofisolatedlargetransverseenergyelectronswithassociatedmissingenergyatp [30] J.F.Gunionetal.,TheHiggshunter'sguide,WestviewPress,2001. [31] M.Spiraetal.,ElectroweaksymmetrybreakingandHiggsphysics,hep-ph/9803257. [32] T.Hambyeetal.,Phys.Rev.D55,7255(1997),arXiv:hepph/9610272. [33] R.Barateetal.,LEPworkinggroupforHiggsbosonsearches,Phys.Lett.B565,61(2003),arXiv:hep-ex/0306033. [34] Tevatronnewphenomena&Higgsworkinggroup,combinedD0andCDFupperlimitsontheStandardModelHiggsbosonproduction,http://tevnphwg.fnal.gov/results/d0conf 5227/,CDFNote8384,D0Note5227. [35] M.Grunewald,PrecisiontestsoftheStandardModel,hep-ex/0511018. [36] LEPElectroweakWorkingGroup.LEPEWWG/2005-01,arXiv:hep-ex/0511027,http://lepewwg.web.cern.ch/LEPEWWG/(StatusofMarch2007). [37] A.Djouadi,J.Kalinowski,andM.Spira,HDECAY:aprogramforHiggsbosondecaysintheStandardModelanditssupersymmetricextension,Comput.Phys.Commun.,vol.108(1998),56-74,hep-ph/9704448. [38] M.Spira,HIGLU:aprogramforthecalculationofthetotalHiggsproductioncrosssectionathadroncollidersviagluonfusionincludingQCDcorrections,(1995),hep-ph/9510347. [39] M.Gonzalez-Garciaetal.,Rev.Mod.Phys.75,345(2003),hepph/0202058. [40] Y.Fukudaetal.,Nucl.Instrum.Meth.A501,418(2003). [41] A.Suzukietal.,Nucl.Instrum.Meth.A453,165(2000),hep-ex/0004024. [42] A.Suzukietal.,Nucl.Instrum.Meth.A453,165(2000),hep-ex/0004024. [43] A.Piepkeetal.,Nucl.Phys.Proc.Suppl.91,99(2001). [44] E.Kolbetal.,TheEarlyuniverse,Addison-Wesley,RedwoodCity,USA(1990),Frontiersinphysics,69. [45] A.Sakharov,PismaZh.Eksp.Teor.Fiz.5,32(1967)(JETPLett.5,24(1967SOPUA,34,392-393.1991UFNAA,161,61-64.1991)).

PAGE 134

[46] N.Cabibboetal.,Phys.Rev.Lett.10,531(1963). [47] C.Bennettetal.,Astrophys.J.Suppl.148,1(2003),astro-ph/0302207. [48] D.Spergeletal.,Astrophys.J.Suppl.148,175(2003),astro-ph/0302209. [49] E.Witten,Nucl.Phys.B188,513(1981). [50] S.Dimopoulosetal.,Nucl.Phys.B193,150(1981). [51] S.Dimopoulosetal.,Phys.Rev.D24,1681(1981). [52] W.Yaoetal.,(ParticleDataGroupCollaboration),J.Phys.G33,1(2006). [53] LHCProject,http://lhc.web.cern.ch/lhc/,http://lhc-machine-outreach.web.cern.ch/lhc-machine-outreach/lhc in pictures.htm. [54] CERN,http://www.cern.ch/. [55] CMSCollaboration,http://cms.cern.ch/. [56] ATLASCollaboration,http://atlas.web.cern.ch/Atlas/index.html. [57] ALICECollaboration,http://aliceinfo.cern.ch/. [58] LHCbcollaboration,http://lhcb.web.cern.ch/lhcb/. [59] TOTEMcollaboration,http://totem.web.cern.ch/Totem/. [60] CMSCollaboration,ThecompactmuonsolenoiddetectoratLHC,inpreparation,2007. [61] CMSCollaboration,TheCompactMuonSolenoidTechnicalProposal,CERN/LHCC94-38(1994)LHCC/P1. [62] CMSMagnetCollaboration,Themagnetproject:TechnicalDesignreport,CERN,volumeCERN/LHCC97-10,CMSTDR1,1997;A.Herve',etal.,StatusoftheConstructionofCMSmagnet,IEEETrans.Appl.Superconduct,Vol14,No2,pp.524-547,June2004;A.Herve',TheCMSDetectorMagnet,IEEETransonAppl.Superconductor,Vol.10,No1,pp.389-394,2000;F.Kircher,etal.,FinalDesignoftheCMSSolenoidColdMass,IEEETransonAppl.Superconduct.,Vol10,n1,407-410,March2000. [63] CMSCollaboration,TheTrackerProjectTechnicalDesignReport,CERN/LHCC98-006,CMSTDR5,April1998;CMSCollaboration,AddendumtotheCMSTrackerTDR,CERN/LHCC2000-016,CMSTDR5Addendum1,April2000. [64] CMSCollaboration,CMSDAQandHLTtechnicaldesignreport,CERN/LHCC2002-26,2002.

PAGE 135

[65] CMSCollaboration,Themuonproject.Technicaldesignreport,CERN/LHCC,97-32,1997. [66] G.Charpaketal.,High-accuracy,two-dimensionalread-outinmultiwireproportionalchambers,Nucl.Instrum.Meth.113:381-385,1973. [67] C.Andersonetal.,EectofgascompositionontheperformanceofcathodestripchambersfortheCMSEndcapMuonSystem,CMSNote2004/033. [68] M.Baarmandetal.,Spatialresolutionattainablewithcathodestripchambersatthetriggerlevel,Nucl.Instrum.Meth.A425:92-105,1999. [69] On'GlobalMuonMomentumResolution'seeforexample:CMSCollaboration,PhysicsTechnicalDesignReport,vol.1,CERN/LHCC,2006-001,2006. [70] On'muon(TC)segmentreconstructionalgorithm'seeforexample:CMSCollaboration,PhysicsTechnicalDesignReport,vol.1,CERN/LHCC,2006-001,2006. [71] J.Chibaetal.,StudyofpositionresolutionforcathodereadoutMWPCwithmeasurementofinducedchargedistributionNucl.Instrum.Meth.206(1983),p.451. [72] D.Acostaetal.,LargeCMScathodestripchambers:designandperformance,Nucl.Instrum.Meth.453(2000)182-187. [73] M.Baarmandetal.,Spatialresolutionattainablewithcathodestripchambersatthetriggerlevel,Nucl.Instrum.Meth.A425:92-105,1999. [74] Y.Pakhotinetal.,Performanceofendcapmuonsystemofthecompactmuonsolenoid,AmericanPhysicalSocietyConference,April14-17,2007,Jacksonville,Florida,USA. [75] CMScollaboration,Summaryreportonmagnettestandcosmicchallengeinpreparation,2007. [76] V.Barashkoetal.,PerformancevalidationtestsofthecathodestripchambersforCMSmuonsystem,inNuclearScienceSymposiumConferenceRecord,2005IEEE,Vo.2,p.827-829,23-29Oct.2005. [77] E.Gattietal.,Optimumgeometryforstripcathodesorgridsinmwpcforavalanchelocalizationalongtheanodewires,Nucl.Instrum.Meth.163(1979)83. [78] J.NeymanandE.Pearson,Ontheuseandinterpretationofcertaintestcriteriaforpurposesofstatisticalinference,PartI,Biometrika,Vol.20A,No.1/2(July1928),pp.175-240. [79] J.NeymanandE.Pearson,Ontheproblemofthemostecienttestsofstatisticalhypotheses,PhilosophicalTransactionsoftheRoyalSocietyofLondon,Vol.231(1933),pp.289-337.

PAGE 136

[80] S.Wilks,Thelarge-sampledistributionofthelikelihoodratiofortestingcompositehypotheses,AnnalsofMathematicalStatistics,9:60-62,1938. [81] CMScollaboration,Object-orientedsimulationforCMSanalysisandreconstruction,http://cmsdoc.cern.ch/OSCAR. [82] CMScollaboration,Object-orientedreconstructionforCMSanalysis,http://cmsdoc.cern.ch/ORCA. [83] S.Abdullinetal.,SummaryoftheCMSPotentialfortheHiggsbosondiscovery,CMSNote2003/033. [84] M.Sanietal.,SearchfortheStandardModelHiggsbosoninfour-muonnalstatewithCMS,CMSCR-2004/035,proceedingofPhysicsatLHC,Vienna,Austria,July2004. [85] V.Bartsch,SimulationofsiliconsensorsandstudyoftheHiggsdecayH!ZZ()!4forCMS(LHC),Ph.D.thesis,IEKP-KA/2003-26UniversityofKarlsruhe(2003). [86] M.Aldayaetal.,SearchfortheStandardModelHiggsbosonintheH!ZZ()!4decaychannelusingamass-independentanalysis,CMSNote2006/106. [87] S.Baonietal.,SearchfortheStandardModelHiggsBosoninthefour-electronnalstatewithCMS,CMSNote2006/115. [88] D.Futyanetal.,SearchfortheStandardModelHiggsbosoninthetwo-electronandtwo-muonnalstatewithCMS,CMSNote2006/136. [89] T.Sjostrand,P.Eden,C.Friberg,L.Lonnblad,G.Miu,S.Mrenna,andE.Norrbin,Comput.Phys.Commun.135,238(2001). [90] CMScollaboration,CMSinterfaceforeventgenerators,CMKIN,http://cmsdoc.cern.ch/cmsoo/projects/CMKIN. [91] E.Barberioetal.,ComputerPhys.Commun.79(1994)291. [92] E.Barberioetal.,ComputerPhys.Commun.66(1991)115. [93] M.Spiraetal.,HDECAY:aprogramforHiggsBosondecaysintheStandardModelanditssupersymmetricextension,hep-ph/9704448. [94] M.Spiraetal.,HIGLU:aprogramforthecalculationofthetotalHiggsproductioncrosssectionathadroncollidersviagluonfusionincludingQCDcorrections,hep-ph/9510347. [95] ParticleDataGroup,ReviewofParticleProperties,Phys.Lett.B562(2004)1. [96] F.Maltoni,TheoreticalissuesandaimsattheTevatronandLHC,proceedingofTheHadronColliderPhysicsSymposium,LesDiableretes,Switzerland,July

PAGE 137

2005,printedversion:SpringerProceedingsinPhysics,Vol.108,2006,ISBN:3-540-32840-8. [97] CompHEPcollaboration,CompHEP-apackageforevaluationofFeynmandiagramsandintegrationovermulti-particlephasespace.User'smanualforversion33,year=. [98] J.Campbell,W/Z+Banti-B/jetsatNLOusingtheMonteCarloMCFM,hep-ph/0105226. [99] C.Zecheretal.,Leptonicsignalsfromo-shellZbosonpairsathadroncolliders,hep-ph/9404295. [100] A.Drozdetskiyetal.,GARCON:geneticalgorithmforrectangularcutsoptimization.User'smanualforversion2.0,2006,hep-ph/0605143,http://drozdets.home.cern.ch/drozdets/home/genetic/. [101] D.Acostaetal.,Potentialtodiscoversupersymmetryineventswithmuons,jetsandmissingenergyinppcollisionsatp [102] V.Abramovetal.,SelectionofsingletopeventswiththeCMSdetectoratLHC,CMSNote2006/084. [103] W.Boeretal.,Trileptonnalstatefromneutralino-charginoproductioninmSUGRA,CMSNote2006/113(orjustPTDR). [104] F.Maltonietal.,MadEvent:automaticeventgenerationwithMadGraph,JHEP0302(2003)027,arXiv:hep-ph/0208156. [105] CMSCollaboration,TheTriggerandDataAcquisitionproject,v.2,p.308,2002. [106] E.Lehmann,ATheoryofsomemultipledecisionproblems,TheAnnalsofMathematicalStatistics,Vol.281,1,(1957). [107] R.O'Neilletal.,Thepresentstateofmultiplecomparisonmethods,JournaloftheRoyalStatisticalSociety.SeriesB(Methodological),Vol.332,218,(1971). [108] J.Hauser,Searchfornewphysicsintailsofdistributions,privatecommunications,paperinpreparation. [109] B.Knutesonetal.,Aquasi-model-independentsearchfornewhighpTphysicsatD0,thesis,2000. [110] B.Abbottetal.,SearchfornewphysicsinemuXdataatD0usingsleuth:aquasi-model-independentsearchstrategyfornewphysics,D0Collaboration,Phys.Rev.D62(2000)092004,hep-ex/0006011.

PAGE 138

[111] B.Abbottetal.,Aquasi-model-independentsearchfornewphysicsatlargetransversemomentum,D0Collaboration,Phys.Rev.D64(2001)012004,hep-ex/0011067. [112] C.Finleyetal.,OntheevidenceforclusteringinthearrivaldirectionsofAGASA'sultrahighenergycosmicrays,AstroparticlePhysics21(2004)359-567. [113] J.Holland,Adaptationinnaturalandarticialsystems.TheUniversityofMichiganPress,AnnArbor,1975. [114] D.Goldberg,Geneticalgorithmsinsearch,optimizationandmachinelearning.AddisonWesley,1989. [115] S.Abdullin,GeneticalgorithmforSUSYtriggeroptimizationinCMSdetectoratLHC,NIMA502(2003)693-695;GeneticalgorithmforSUSYtriggeroptimization,talkgivenattheIVConferenceLHCDaysinSplit,October8-12,2002.

PAGE 139

AlexeyA.DrozdetskiywasbornonMarch31,1978,inthecityofNovosibirskinRussia.Aftergraduatingwithhonorsfromhighschoolin1995,hecontinuedhiseducationattheNovosibirskStateUniversity(NSU).In1999hereceivedhisbachelorofsciencedegreewithhonors,andin2001hereceivedhismaster'sdegreewithhonorsfromthesameuniversity.WhilestudyingintheNSU,heworkedintheBudkerInstituteintheSphericalNeutralDetector(SND)groupontheVEPP2M,e+e-collider.HisworkwasrelatedtodevelopingnewdetectorsforX-raysaswellasstudyingapossibleupgradefortheSNDdetectorforforwardphysicsatVEPP2M.DuringhisperiodofstudyatNSUhewasawarded4yearswithInternationalSorosStudentStipendinPhysics,2yearswithSpecialNSUstipendand8timeswithstudentconferencesdiplomas.Inthewinterof2002,AlexeybecameagraduatestudentintheDepartmentofPhysics,UniversityofFlorida,Gainesville,FL.InthesameyearhejoinedCMScollaborationandworkedwithprofessorsGuenakhMitselmakher,AndreyKorytovandDarinAcostaonvariousaspectsofexperimentalparticlephysics,particularlyontheH!ZZ()!4CMSdiscoverypotential(aswellasCMSdiscoverypotentialforSUSYinnalstatewithmuonsandEmissT).Since2002AlexeyhasbeenparticipatingincommissioningoftheEndcapMuonSystem(EMU)oftheCMSdetector(startingfromelectronics/detectortestsatUFLsiteanduptothemostrecentperformancestudiesandfastHighLevelTriggeralgorithmdevelopmentfortheEMU,validatedwithrealcosmicmuonsdata,takenbyCMSdetectorsliceinfallof2006).IntheperiodofhisstudyatUF,AlexeywasawardedwiththeUFPresidentialRecognition,UFInternationalStudentAcademicAward,and5CerticatesofAchievements.Heisco-authorofmorethan10refereedpublications,andresultsofthestudieshe 139

PAGE 140

performedtogetherwithcolleagueswerepresentedatmorethan10internationalconferences.HeiscurrentlybasedinGeneva,Switzerland,closetoLHCandCMSexperiment(CERNinternationallaboratory),andgivesfreeYogalessonstothecommunityonvoluntarybasis.In2007,AlexeyA.DrozdetskiygraduatedfromtheUniversityofFloridawiththedegreeofdoctorofphilosophy.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101112_AAAAFK INGEST_TIME 2010-11-13T00:54:15Z PACKAGE UFE0021068_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 1495 DFID F20101112_AADFEL ORIGIN DEPOSITOR PATH drozdetski_a_Page_076.txt GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
bcf48b5a5c162eb5070ccbd0dc3e1ecc
SHA-1
01e9832df1a8b53f4ec526489c1d421f32f05f0d
50819 F20101112_AADEZF drozdetski_a_Page_028.pro
b2e5094bd2d1d4474e164af1c9722a44
fa3c26ed70e466b378d06996353876f8065a262d
2577 F20101112_AADFDX drozdetski_a_Page_061.txt
142ee7117a70d4ce9e0e5c384f996abb
1c4b3e571f7de9fcf22f0b968af73a9e3450e945
25760 F20101112_AADEYR drozdetski_a_Page_005.pro
1e156bd00122d171ca7737ebd85a5e2e
dbb2ba8004a866c6c6619b76f41109f1ea5fbd91
2217 F20101112_AADFFA drozdetski_a_Page_098.txt
6f491b7cdf48bd80ae46bc93f2144070
6f5431fe17f45d2c9e0155181eb54e9df1c51d9a
1629 F20101112_AADFEM drozdetski_a_Page_078.txt
f08c38263c0964dc46970211994e968a
818b5612a5b577dd42e5be23cb189005641f15a7
57163 F20101112_AADEZG drozdetski_a_Page_029.pro
ce64de2b4de8bbac9ee2a23fa40f0dd2
0884f949cffbbd052f88e68966e419b3a4d5c7d5
2323 F20101112_AADFDY drozdetski_a_Page_062.txt
c797a58fc0c7e9ef9f3749446f743e25
e78e13d76ded45e9499073fba25f3363c6f628c3
45949 F20101112_AADEYS drozdetski_a_Page_006.pro
81c0033d97e1306fece26a743f1c7069
387f43595d68ae201706f0645b291d9e180ded86
2360 F20101112_AADFFB drozdetski_a_Page_099.txt
7018bea31158b7aed92fa3546eb7441e
fc8c41fb076b67228fecfbf8f2e34e18e8b5fbdf
2146 F20101112_AADFEN drozdetski_a_Page_079.txt
22e6eee05f23163f38a06475eaff404a
975f210426d4026380adcae709d3dca657d91e92
53316 F20101112_AADEZH drozdetski_a_Page_030.pro
f54c7baf52d8c3cce052c2f615d879a3
85369a4ec57b3b38c6b36ce9d43857f18e7e2ae8
2473 F20101112_AADFDZ drozdetski_a_Page_063.txt
94b6374416d9debcb940b10a6c187ebc
81d4c3250632484fe36ace987952feffec970fcc
66969 F20101112_AADEYT drozdetski_a_Page_007.pro
a182366d3547162a367b47e0177e3713
2a9e00118503b9fde018f34f1d471dd9ec8864a8
502 F20101112_AADECA drozdetski_a_Page_001.txt
d4e57f2a50df647fd943369a50690ff0
369a1e2bfd223708464537cb4851e35a7e9f92cc
1952 F20101112_AADFFC drozdetski_a_Page_101.txt
554e7dc64e1bbb436b7eca422ae472f2
cb70fdcfb0632352ffd7abab696911472ae50cbb
2096 F20101112_AADFEO drozdetski_a_Page_080.txt
1db3ee6e54abf428ba4606ce1a1a5772
f168a9c2dc8d4e44bf82f1c749babfd1c6445605
20591 F20101112_AADEZI drozdetski_a_Page_031.pro
0f56205b8e9e2fca51e415f8ccd1516b
1aeaeca3ea8b66b110710408e764f1744b5b73ad
72060 F20101112_AADEYU drozdetski_a_Page_011.pro
0e37e82094447719dd4bdeae29325f97
45007d850fa122bbed3e7976bfe0eb6045452103
24293 F20101112_AADECB drozdetski_a_Page_022.QC.jpg
4b4b0dbfe1451a5b3fc331014044a811
82a6cdf371080c1b20c6fbafec708f88b1a3dae8
2215 F20101112_AADFFD drozdetski_a_Page_102.txt
60e41269c0934af1c7e2f5b882026b2b
3334f71aa0b5209b0c719259200f1ed613a613f9
2422 F20101112_AADFEP drozdetski_a_Page_081.txt
754e47bc2a9e25c565defd036bca6917
1ed93d88e8caa7d22c85c4280d7a92f73c261412
21450 F20101112_AADEZJ drozdetski_a_Page_034.pro
f303a02af79466dedd66faafb0015057
ae81ea3c3ef2042fe651815f22f8f44ed225b76d
74005 F20101112_AADEYV drozdetski_a_Page_014.pro
b3609893504ff469169abeff43c86b97
e3bcee35c8b8dc26ecfbe17e7921a3ecbac08cd1
17831 F20101112_AADECC drozdetski_a_Page_047.QC.jpg
02de9eb618f8c8e01227035965d07a9d
b6eaa413b208a69f477f80e012e58b8de942888a
2410 F20101112_AADFFE drozdetski_a_Page_103.txt
abd8a155cd2f4e74a6165c5e85f15221
8adda473a5c59dafaaaeec5d6f91f0feaa0aebc2
987 F20101112_AADFEQ drozdetski_a_Page_082.txt
9b030fbccea58591640e40ab86fc7be5
bf8842495ae73f52ae5a4e737aea14d87b28878e
56988 F20101112_AADEZK drozdetski_a_Page_036.pro
4c549d5609657217e6e515ff3475bf7f
06dffdb75f29030042d8a7b3f6bb6442cdf449b8
6200 F20101112_AADEYW drozdetski_a_Page_015.pro
16af9949dc14d3ac45d16d3d357d85f9
476c548f6cb0059a70b045cc557dcf1acb597364
60138 F20101112_AADECD drozdetski_a_Page_131.pro
6bac600d788650eca4a666f34ea17642
bdc6fb0fab60cb4120e8c4d73825f7d263d58190
2257 F20101112_AADFFF drozdetski_a_Page_104.txt
7624d50095b41953103ad64e3801821e
4dc67d9f259962075e179f05fa3c64f623773057
1980 F20101112_AADFER drozdetski_a_Page_084.txt
584e32547371ada54351add74e04f6e9
5e388777bb1f4da8f453461ff059ef653127ad57
33734 F20101112_AADEZL drozdetski_a_Page_037.pro
603c389bf6a2267f85de2c0c34dc2586
c8691015a459b89224c94d33d0d499ef750af444
46676 F20101112_AADEYX drozdetski_a_Page_016.pro
6bd39a9faf570022ce00cafa342ece2e
e232d0ded2565314b2aaa63dabd1a0c9cda2cc08
3372 F20101112_AADECE drozdetski_a_Page_009thm.jpg
1295993a9906620e7bbabe56aae305e4
68a15fd0f2a23a06fc116624df73d57805950f25
2219 F20101112_AADFFG drozdetski_a_Page_105.txt
ea0b5932008d990210d5336feda8f424
c917c3e7c4830f95098d2354b6226aae25486585
1797 F20101112_AADFES drozdetski_a_Page_085.txt
3cc48d7449bb775f09d0ade96a8a311d
ce6c747db5b53686bf8d76d79a59176065a358b0
57206 F20101112_AADEZM drozdetski_a_Page_038.pro
7cf2b4ca35f71be2e51efb17a536a7d2
e8414e9b3b8675c6e30b44799b7d700d803abd34
4621 F20101112_AADEYY drozdetski_a_Page_017.pro
2a6d7ce11fe664e408e60e07fe5b3720
c369383a8218cd86ea83257b9f8eb4d60cda18d0
25271604 F20101112_AADECF drozdetski_a_Page_055.tif
065babda77e11ba22a11c836e9a9e5fa
1ef9c01129046f6e0b06f457851f0f87cfceb864
1883 F20101112_AADFFH drozdetski_a_Page_106.txt
51df29c0a35a0f635b9998abe18189ef
a0ae61fbe853e0d8baca2dd4af06442da754c5b8
1323 F20101112_AADFET drozdetski_a_Page_087.txt
0cbf3498ab5d7686c0d848a351b77986
b32b2fdf4dc5b506db960f71e65cf02b0f07a126
59734 F20101112_AADEZN drozdetski_a_Page_039.pro
c52d62bd73ee9c3d91787b701f23c405
a31f63f82f3ea2b3758e524cb3810af73c476204
40429 F20101112_AADEYZ drozdetski_a_Page_020.pro
e31057ca33aabacbd7ef8e01341d6b17
f8a4548c2ae3bb6f5e07bad21bb25eac69c793ef
1801 F20101112_AADFEU drozdetski_a_Page_089.txt
665c26429c5e0bd7be21a6fbf3e64aec
35e9d11c194e6d77490c6458c2634ce6dab6fe24
59698 F20101112_AADEZO drozdetski_a_Page_040.pro
06d3b8c4a802a8a425741d9cbce3ba96
03445ba692f98945e50658b93288b7a5430d1eb7
48433 F20101112_AADECG drozdetski_a_Page_018.pro
6909a701bd576b27bd3fe869826754f5
27aba119b2b36427622b33bf3e26ecc55157d954
2223 F20101112_AADFFI drozdetski_a_Page_107.txt
601d1bd51d7b8ddb0583fccbeb658515
0fb3df094be70363d692810e79b427ee0f9906e9
2778 F20101112_AADFEV drozdetski_a_Page_091.txt
ce05c94118be784f871889cd5ab7ce77
7891070fca2c79630e3d36cf7560f4be00a3e044
54196 F20101112_AADEZP drozdetski_a_Page_041.pro
e9651b2fe2ba6b6d65d3f4fd97caa107
f65ff289c93b76b929e27da313958d66fb961fa0
F20101112_AADECH drozdetski_a_Page_109.tif
ed218d073933c8f8b27a15e3da89c12c
e1a3426a75294a4636cbaeddbfbc42be2dcd81b0
3013 F20101112_AADFFJ drozdetski_a_Page_108.txt
15df17faa1e8281f4f434d302b7ccc1f
d01453f3f8f58bfb2d7d4cea3601a0bad3827c35
2113 F20101112_AADFEW drozdetski_a_Page_093.txt
913b3ea2798d3231884add68c0d7ef14
2443831daaa539528f8f368d283ae6246b041355
37995 F20101112_AADEZQ drozdetski_a_Page_044.pro
0b7c052c30e5c399699d29105769cae6
226d4e04f44de4285f4a0d5cb90b8c47b3e5163d
2220 F20101112_AADECI drozdetski_a_Page_045.txt
c8f72c9ca8310bb71f4faa6427125e8a
02c9d2f80393c743449f08887dc45f148d4f895c
1389 F20101112_AADFFK drozdetski_a_Page_110.txt
7896a4db890e89beaa5368f290d14121
1446f089b521005fe551e846ed796a8ee987499c
1051983 F20101112_AADEBV drozdetski_a_Page_035.jp2
fae846df4fc2e6b432cac76108aeaf4d
430e257c13554a1b1fa388ca72918cc513dbb361
2279 F20101112_AADFEX drozdetski_a_Page_094.txt
6975e0386543c91819474c7caa11562d
8959e5989707d3b19d4df306f83fcbd1ae32ed96
53906 F20101112_AADEZR drozdetski_a_Page_045.pro
3adcedee5170a7c0889f16ace65940e3
4c1fb225275f07507d2c530c2cc88a6a3f459e51
2345 F20101112_AADFGA drozdetski_a_Page_136.txt
6263ee4c71557c605baac67124591daf
b593f4ed4829787db83668aefb0fd6fb8a58d69b
F20101112_AADECJ drozdetski_a_Page_030.txt
5e81a99cdffa93ae9d2dd7b54292553a
d5833316c1458cabeaf11ce3f4ecf30bb8efd521
1347 F20101112_AADFFL drozdetski_a_Page_112.txt
f6b5120504162d0a3fc4367ecdcaf1ea
85cb99dba8c0832555347a4608668c7c32ee141a
861 F20101112_AADFGB drozdetski_a_Page_138.txt
f063d39aa00434f6fa2b8e5fe5af4ab7
8626f7be50d08faca40ae577c6716af67fd8e26d
1051982 F20101112_AADEBW drozdetski_a_Page_012.jp2
c0dc8c88b14b12a8284738ad6a27d790
46b44e5c94cf5ac693dabd3ae4199438873c08cf
2733 F20101112_AADFEY drozdetski_a_Page_095.txt
62681dce67dbeb7b415a23863b4b5e61
8a54143dbeefec7e5b9524fad5e2a888b7e9bd7a
6297 F20101112_AADEZS drozdetski_a_Page_047.pro
1dbeabe4ef7c564df0601915b5f6d540
d11f5fceeb99a99a936e4d1c192e0062414fce77
67467 F20101112_AADECK drozdetski_a_Page_010.pro
7274fc7a1c6f279262447ba55577f60f
d0c4fc6bc0bb410783f11096d2e6441a4ac355a4
1309 F20101112_AADFFM drozdetski_a_Page_114.txt
bb6aac44893b9181d5adfa0f1da518a9
a135482bb032d393805589da30b507a7c0ba9aab
470 F20101112_AADFGC drozdetski_a_Page_140.txt
c9bd9e0cf254b62fb33454149d9552ca
d013534baf70e1bd3530371eeed51b4054565635
851 F20101112_AADEBX drozdetski_a_Page_128.txt
70892033ac9b108577e11871f5fd6f74
d1342d76bd3d2d725a1935a281eda551487a0bf1
1866 F20101112_AADFEZ drozdetski_a_Page_096.txt
f6a02f0fa98d14d83791496d3669b575
a0ec1e536c35a3dcdd937e34aca7c490c1e000c6
45277 F20101112_AADEZT drozdetski_a_Page_048.pro
25fb47aa2cd1f7edce8d829156d2c283
573ce8ccab84ceb681cf31a159cd7cd1e8d646a8
5630 F20101112_AADEDA drozdetski_a_Page_020thm.jpg
6201864727bd83bbf1bb3535c3c21e81
15b0e9393617d5961734dbc6618b85dcd1533ed9
F20101112_AADECL drozdetski_a_Page_062.tif
c38cbc01a31c08cb85feb36bc4df67b1
e1ea82233bea1cf5a4790b49297ae431c6bfd8e9
1328 F20101112_AADFFN drozdetski_a_Page_115.txt
a3363c4d84b72c35f8e3b68a33147e73
d0484edd33a0cd8bec6ccbc750ee44310fd44ac7
2316 F20101112_AADFGD drozdetski_a_Page_001thm.jpg
72533aa742af637c5baaf53f410c821a
e61dc0d4b772eeceba7e10af630fbd34e1b33d4a
1053954 F20101112_AADEBY drozdetski_a_Page_138.tif
d6547859189d2e67c8ec2dd5f131418c
e786435124ac555a802a1f757635b5539d2a9b53
40964 F20101112_AADEZU drozdetski_a_Page_049.pro
0f8bff29a809d3f50982a42bf5590499
1bf02be9eee521310dcc35d0751322412c6a53df
192 F20101112_AADEDB drozdetski_a_Page_003.txt
f28c11400395f167d409474bbddfe987
59bcf6794b878dd16797d4f9c44fde0e080b2822
27541 F20101112_AADECM drozdetski_a_Page_054.QC.jpg
892695353758f7295146bc0137078949
ac62bc6a9c9458e566cf8ad81cf5784b33a981d7
1600 F20101112_AADFFO drozdetski_a_Page_117.txt
45e4623e4c78134a6a8399b08c702720
4f1b2f70643911b6a52ff9298eb0424e4f4bd7e3
4343183 F20101112_AADFGE drozdetski_a.pdf
f7360ea131f29191f2fb988ebf20c63b
666762632869991b30acc354f1113476475e718c
2278 F20101112_AADEBZ drozdetski_a_Page_054.txt
ac195cfe8db0a152997f6005a4966330
8007fd6f51e61c6c092c6ba64ae1d016e761210b
56485 F20101112_AADEZV drozdetski_a_Page_050.pro
52720d912bced521ebcf94ea38294220
816e1f21c2c5d292e6deee35ab901a9494c05f73
1390 F20101112_AADEDC drozdetski_a_Page_002thm.jpg
c466aac8348084a5c29d12c3af4c3199
71a621bff0758996f130137de2bec4f831b2e783
9543 F20101112_AADECN drozdetski_a_Page_001.pro
4d86e0c57ae5d053e19911dcfb7ba894
395a2dcda11a0750623fa0ee5c0431e19d8359ef
F20101112_AADFFP drozdetski_a_Page_118.txt
948feb5eaf517aa4402fc403c762ea64
31894c1c41269c3d17552953767226db218f8f33
7797 F20101112_AADFGF drozdetski_a_Page_001.QC.jpg
516652c5db40bae360aa1166066645ea
e34cfe8e3073a4a4402b4b3a86484c67ced3ab74
45601 F20101112_AADEZW drozdetski_a_Page_051.pro
6e2dc1a6dd3e5fddde7e7a616c14400c
c180d41ca0f821795699d38cc31307600dffff75
14954 F20101112_AADEDD drozdetski_a_Page_086.QC.jpg
de875770217dbfa9eee46a583c4c7e89
37a21459e2c59119ae64b9e85885ad87135b62ab
F20101112_AADECO drozdetski_a_Page_058.tif
5c0e19c2b0ab90a6690c46d4b60102d9
41132ae2300759092c334872c95c74f0e4ed4051
2025 F20101112_AADFFQ drozdetski_a_Page_120.txt
ac55868acca68f734f615a4184f96d9d
a6e24334626d2ff09b18b02ddb74609f6252e831
3453 F20101112_AADFGG drozdetski_a_Page_002.QC.jpg
d839f6c73aa941f7a4e11efefcf7b1e2
3d6ad857f2e75fa67e3dcdf49cbb8938b03481fe
57484 F20101112_AADEZX drozdetski_a_Page_052.pro
da4424a63808233bd9d0d62b8239db8e
d9094f58271c17528a97d126cf0925c208c60f61
28168 F20101112_AADEDE drozdetski_a_Page_011.QC.jpg
5009593b7ab20b365c869f0f07c22cab
a5ebbe2aa07f9f1ae42e40c0fe93df68eb131d9f
114670 F20101112_AADECP drozdetski_a_Page_065.jp2
9289be4749a44373b0e74e01e405404f
139899668b15749a5230b18c7f1c516f1b71e5ab
1139 F20101112_AADFFR drozdetski_a_Page_121.txt
508b7a7907bd9a0ba18219ebe10b01f4
fa83712734569f2d068d09c92b26d5c77a9e2972
3821 F20101112_AADFGH drozdetski_a_Page_003.QC.jpg
aa19f6745e5b17ac7332cbb409b1458d
7af0f5d51fae8d44761136ef5c5236e95a729fc6
18442 F20101112_AADEZY drozdetski_a_Page_053.pro
a149591063d7e94652fc69bd07614567
18a62d953a84c79c730d88557c5a58aeffd40a35
76113 F20101112_AADEDF drozdetski_a_Page_136.jpg
4a4a04a4351ea0b3242ae98d02fa9afb
61d0ff44558ed2b546d5b0bad751aca5db78b632
F20101112_AADECQ drozdetski_a_Page_026.tif
ae6da7173a1bf13bba46125f50a705ce
4710dbf879a5bc2802da7106eccb6f8f73582749
2126 F20101112_AADFFS drozdetski_a_Page_122.txt
a1420e04e04d20a6017bb0679f8bce3d
455b84ca0c9172b5c8db425a14b4edaddbc7c4b4
23251 F20101112_AADFGI drozdetski_a_Page_004.QC.jpg
74117c2b3803d9eecab1dac0ead3c92d
9bb82336b07ff79e2dbb6549472cb8075212cc4f
57967 F20101112_AADEZZ drozdetski_a_Page_054.pro
cc98160263e4dae0225cfca9dff1794f
862ae5e40b652d64f7265dc7f21277b02da89b0c
3127 F20101112_AADEDG drozdetski_a_Page_087thm.jpg
482532b4533d65b28cca6ce3045dbcf5
e31378d7c2ec534302648703a1f786c882eb5b57
970 F20101112_AADECR drozdetski_a_Page_111.txt
5e1d48c28f4db905c33109fb0f434386
a37fb8a0f8516435aa20fc174d30a43a87e37bc8
1366 F20101112_AADFFT drozdetski_a_Page_123.txt
bf636c966d45c75dbd3f5c5a65ec149c
bc15a625c5958257c73406a7186ba5dd9cb119db
1036 F20101112_AADECS drozdetski_a_Page_126.txt
f85acd8e7423bac2c62a74546941080b
aa7b24070853d710ff89caca22574dbb5f81629e
1232 F20101112_AADFFU drozdetski_a_Page_124.txt
83cb43ad9699d2da67299175f2116127
e4808b25d522ee3ed3a5d4438739e589a1a4cdef
12827 F20101112_AADFGJ drozdetski_a_Page_005.QC.jpg
35ca11dc077c3e738f504ffd01712fcf
023c3c3fb9b77c4edcee0fb16f513752f3534a09
2005 F20101112_AADEDH drozdetski_a_Page_016.txt
da0ba354bfc0e7b917395366dbe09295
1e7e64c39005ca8a096280d8dd6b5df3bce937bf
969096 F20101112_AADECT drozdetski_a_Page_046.jp2
3743fc3a46b5959fa910bb74436df9da
382bb75158f1d6c94c35a8d76dacdd69040b4612
596 F20101112_AADFFV drozdetski_a_Page_125.txt
82206ce7d78b4f8cfabbd6a7b8144564
782d19516bafad3044c48fcae7ea407675940a95
3598 F20101112_AADFGK drozdetski_a_Page_005thm.jpg
e8bd52c25e5202fcdaac8ee1821ccf22
da82fcb99ab3b119540d595d1f22de3933755285
F20101112_AADEDI drozdetski_a_Page_100.tif
7057e1e17c88e784aeb1024177f3ac80
aea0c9a1496f8e21a9dcf734752cb7a86ff2a2d1
F20101112_AADECU drozdetski_a_Page_065.tif
fa90a2c3a0046608e0f96792a117da41
dd440d49e860ef958e1668ef3b29d5f7d060744e
1733 F20101112_AADFFW drozdetski_a_Page_127.txt
97e8e0484efd81adf7384634eeda8c7c
3a64322f3a369eb9dffa178c16f2e8ebd4bfc9c0
1690 F20101112_AADFHA drozdetski_a_Page_017thm.jpg
ce116f84be3b05d739f7688e4ccb7280
d5cb416fe8a47d1dea69b9fd681935a060bc2965
21419 F20101112_AADFGL drozdetski_a_Page_006.QC.jpg
2540b104925f634c74a77d61c16dc855
5da1f1a08ea86b2359b5bc3a5999893c5833f826
40869 F20101112_AADEDJ drozdetski_a_Page_078.pro
9c9bc930d45fe8557e94452014ce6aa9
4dc61a33f0cd4ebae7db0b79b0b475b2a325fbc9
25605 F20101112_AADECV drozdetski_a_Page_099.QC.jpg
4341fdfc6e12f6b1ab13f74d06ecfbe0
f252fe96bafe731f957045a9b8aa8d83e485174c
1364 F20101112_AADFFX drozdetski_a_Page_129.txt
a4b163bb8d95a31ddc185bd7a8d30e62
d2375e5fbb7fbfdda17d72caa738635d3feb2a5b
5835 F20101112_AADFHB drozdetski_a_Page_018thm.jpg
442473d2f3f05c773f10e1021e203bd2
932880becee6ebb020adbea17688e023f513458a
25975 F20101112_AADFGM drozdetski_a_Page_007.QC.jpg
0697647d1db25282df2427084c570e55
647346d71bfee574b1c5aae3311e59e4903a7c9d
25884 F20101112_AADEDK drozdetski_a_Page_028.QC.jpg
72336e86033abf9d66ab548305260a3f
5fce0f0380f23b02564284d1b71567dcbd9b4658
1051984 F20101112_AADECW drozdetski_a_Page_069.jp2
0889b5d84c690e48961be17aa39c0e0d
1f37b3fe4d4d388eac17f95ef0a4e10047febeef
2548 F20101112_AADFFY drozdetski_a_Page_132.txt
3c775a3131ecc80cd6d5464d56895550
54f4d07c10a545081e3ede5724447490adda6d61
2077 F20101112_AADFHC drozdetski_a_Page_019thm.jpg
8bfded595116ffedae9b61487b8e4556
738abbd3b8779a42ff8f8aa96a0c2146dd515702
6201 F20101112_AADFGN drozdetski_a_Page_007thm.jpg
8a95c61af75c398010ec53858e4cd0a1
d999bf4988e9cc82111deeb762279e4de1a490b4
20749 F20101112_AADEEA drozdetski_a_Page_084.QC.jpg
701090155f81e162e564e82b0eeb1a40
8c422d6b0e7655cf946ab8698276c8fe22bcef3c
486307 F20101112_AADEDL drozdetski_a_Page_031.jp2
1602fd148e117ddc3819ed22c4e14bb8
ac9439032eb4eb2c459aeb634991fdb4f73d5ac2
7270 F20101112_AADECX drozdetski_a_Page_061thm.jpg
82b9e4e18f93ee686e611f5630f783bd
de20abe5f00e2a5971322a3415f4e6bd871cfe73
2038 F20101112_AADFFZ drozdetski_a_Page_134.txt
08acc4142cba5f1583580fe9077d7624
dcac84ef226261cab502197ec54da2798e33116f
20038 F20101112_AADFHD drozdetski_a_Page_020.QC.jpg
0155ec155d24af0ccd4b3af519cbe48b
af464790ee8acf77958f4b6eadb8234596480648
3133 F20101112_AADFGO drozdetski_a_Page_008thm.jpg
273c74895e76a8c7a05ceeabc58c8d34
16c9f25fdb2dbd3d314f8c17ddc1d0ef9f6e2879
75570 F20101112_AADEEB drozdetski_a_Page_079.jpg
b44b830383dbe8d289fc7380668e5a1d
6f80df84f75281c0350db2f9cbf8187ddced7f2f
16432 F20101112_AADEDM drozdetski_a_Page_114.QC.jpg
93b67b7037fc3b3bfa3382fc389d1b6c
df56c224b064eb67da90907b6da2ce0ecc5030fb
6360 F20101112_AADECY drozdetski_a_Page_132thm.jpg
061df522baf754d9efacc20c774ad603
591894e67c39b4aa8fe57044d8488256ed35ef58
19995 F20101112_AADFHE drozdetski_a_Page_021.QC.jpg
52fe6823bd35986c22330ed00b43ee84
90ca04fd64fb57e0c331f402aa36e3a912a09d87
27492 F20101112_AADFGP drozdetski_a_Page_010.QC.jpg
6fbf6fa8003480dc114af648f81a8f2a
bc0e2259d321a8d54ee18135dc6c32d4c27e3480
1919 F20101112_AADEEC drozdetski_a_Page_130.txt
37bd3c44f90100dab94d45dd6e2b609a
0aee40da35a22cdeaab942d6d2e2c5a916491358
20053 F20101112_AADEDN drozdetski_a_Page_016.QC.jpg
c8a9e74ca7059c32d6e1b019ab907267
d9166f6ebe51d5d1c72b6717c2df864ec1afe771
F20101112_AADECZ drozdetski_a_Page_011.tif
a3a569a1197922d6f4b4507952aa3c56
5effaf86a1cc024104e670cf95a09b9c55a2eda8
5400 F20101112_AADFHF drozdetski_a_Page_021thm.jpg
ef0816ad1c41c2b4eb3420905b365ab8
a6b88c2d8b29ca2bcbe310e00abf49f2c00c395f
6788 F20101112_AADFGQ drozdetski_a_Page_010thm.jpg
5383cb131d1e1696e1279a0b6466f08f
777219b92ed265fdbac53951e43eebbe3037b6ae
127091 F20101112_AADEED drozdetski_a_Page_071.jp2
b636445ee1c1b02872b45a0898169552
dbc7f7f24dfa64fa72851d8d39a366eddc374c57
6700 F20101112_AADEDO drozdetski_a_Page_064thm.jpg
e9272334404651122bcb55825b949d21
cea9c63bcb08d63639d3cc9f8a804c8b6f9c2b8f
6507 F20101112_AADFHG drozdetski_a_Page_022thm.jpg
ba9320be23cd67ba84b1ab6d44d4e302
d40c50e81598694b48f0a80a75c7283fd5ead8e3
7184 F20101112_AADFGR drozdetski_a_Page_011thm.jpg
621be0d8da6aaf6cffa33d010c6ceaa1
ab34bf585c820e9def311577ff7a11c5554c7968
45864 F20101112_AADEDP drozdetski_a_Page_111.jpg
8e2145c401e3329454dd3f7e0b4d1ee3
8f8a6efa4ccbab83ab51924e41da4a3eee1bf23b
F20101112_AADEEE drozdetski_a_Page_061.tif
eacea3f2188392e800834a73e0a93ebf
319348faede8138ccb3feeb92ae76822697fdc27
23335 F20101112_AADFHH drozdetski_a_Page_023.QC.jpg
ae4edcdab3c1a64d0331ce4079fbc6e5
06c06b32698c0bc91293a25f3a7ffdc392e14939
28538 F20101112_AADFGS drozdetski_a_Page_012.QC.jpg
e27d42608c9eb04c8178a8832dd918cd
e75a0dd398ab8380f46b86f337452980cc99064a
18763 F20101112_AADEDQ drozdetski_a_Page_085.QC.jpg
9369f2a12c72fbfbd0e6efa649037d9d
6f2cd49db0249eacac9af46e34834f45ddadb27f
5978 F20101112_AADEEF drozdetski_a_Page_137thm.jpg
d41788086738c1b83384d94da4b36ebb
2b8374163dfaadde16aab0ba9067e6cdcb8d65a5
6081 F20101112_AADFHI drozdetski_a_Page_023thm.jpg
65ccb3be968654c673acbfc40bd548dc
5735a24315a342d66793d661551f965f005dead8
6699 F20101112_AADFGT drozdetski_a_Page_012thm.jpg
828c9ae6d5b63d40b322da263127c8dc
9dfe70fbe4789a581ac45cebd09473e1c2eba4b2
F20101112_AADEDR drozdetski_a_Page_022.tif
e5727ccbec915bd5bb56e4d608092a82
a6a9f402f181a163318b695974b4862369a01d2d
88325 F20101112_AADEEG drozdetski_a_Page_062.jpg
a354f68519963eab9c58dfa5eae592a8
24ed9daf837bb76941220a920101d30bf2b645ce
22072 F20101112_AADFHJ drozdetski_a_Page_024.QC.jpg
ddae5e6ff7bc01d2d4bfac08100cb2df
67d9d9922ab95980fc939930fa5585c40fb96d19
27238 F20101112_AADFGU drozdetski_a_Page_013.QC.jpg
73027f71ce824334c64b31b0fd2b1bab
2765d5ddf29f24060aaae5040352246d48776c55
3508 F20101112_AADEDS drozdetski_a_Page_034thm.jpg
6db6a7e0d67d24aba5239bddff0aa01d
82485d6c747ff04daf1b4414f21e27cf6f871c59
F20101112_AADEEH drozdetski_a_Page_060.tif
4b93308dc37b7e086131f132fd8a5193
e06e1fd2c14ccc0a3144aac38e7d73e28b42bb2a
6958 F20101112_AADFGV drozdetski_a_Page_013thm.jpg
d5d7d5f09fb77d66445dfb5514cac2ac
87375462c0c47e4d6e76c5e29329a7d37ca61f4a
16750 F20101112_AADEDT drozdetski_a_Page_075.QC.jpg
e2cb0aa0e63aebf695a2f9c76de873f2
fd81f1384da2b4509e06d7daf2db94af26b3da00
5962 F20101112_AADFHK drozdetski_a_Page_024thm.jpg
6404af31482a9fbafce53b9ee49c65c3
57a29781c21f12438a3ae0e1eea577ee9b16a91f
7199 F20101112_AADFGW drozdetski_a_Page_014thm.jpg
9fd39426f892017b83d7f860b4f85112
96a40aaf47a4aa1744a496b693f7c156cca7872a
F20101112_AADEDU drozdetski_a_Page_040.tif
e599ae48fb1df8c2ec6a32af753ab357
2850287a3a0c6c5d0f88c7b2b8352a9ce2284af4
260 F20101112_AADEEI drozdetski_a_Page_015.txt
778bc8f76d14ea8ab5f84f7f1ed150be
9d1407e0d764b39d3bca46456291186ea1c1fde4
12905 F20101112_AADFIA drozdetski_a_Page_034.QC.jpg
74db3fc78605e5c55795ce3632e4a503
b38fd7dc40771000ea05107b7724fe09e97cc8c0
22854 F20101112_AADFHL drozdetski_a_Page_025.QC.jpg
5fea6bf658faf6341a26a0e19fc4aa22
9c2df059d0d0144580e6a53d32397cf2c141dfda
5127 F20101112_AADFGX drozdetski_a_Page_015.QC.jpg
3288f378a6d784fb80f60222a0b8835f
fb6852613b22ec1d4245f529a8a8a5c34d57f51f
9822 F20101112_AADEDV drozdetski_a_Page_125.pro
a70cc8d185f6af65f25e1ae283516c08
5af510a68eb4fddb7c6f8b4b21620e3e66cb651e
1008006 F20101112_AADEEJ drozdetski_a_Page_025.jp2
6b2e81e62bd1ddf111785cb75c481960
2fe5e28c8f28a3e2f2005154517ccec4987771be
6591 F20101112_AADFIB drozdetski_a_Page_035thm.jpg
0482c3fb9d5b972d8a1597f6de9e20d5
48aad384f22c68c036f963af91f7e01c2835f5b8
5930 F20101112_AADFHM drozdetski_a_Page_025thm.jpg
eb2415fa495c27c56f6febb5364af2f7
7eda59237207491387ec6a2c5bf9082d72b2d3f6
F20101112_AADFGY drozdetski_a_Page_015thm.jpg
ebf92d407e13b4ddbf7e02aed1a97974
f1c37b4db5e2f4582567e4292149f34a88823626
61625 F20101112_AADEDW drozdetski_a_Page_026.jpg
c47644627bd31708451a79c62938506f
e899a96b77a6c1f8f2557060544bd166491561cc
48592 F20101112_AADEEK drozdetski_a_Page_060.pro
839b661a6c3e0e7f3d674d7ccff31723
98dca8578edbfe0cac8603565649262e635e2120
25286 F20101112_AADFIC drozdetski_a_Page_036.QC.jpg
0acc94cadf0976a87c73556ce5988159
98c63687c80b0145d9ca8c726e7e33f61144fc14
19442 F20101112_AADFHN drozdetski_a_Page_026.QC.jpg
0f7c2cdd03e7d9d6fc9059793e553dcd
59d9c0e3e9d1353d4d3456c11ec8c67398f027c1
4518 F20101112_AADFGZ drozdetski_a_Page_017.QC.jpg
5a972a9b82c29cc4b45c510d737d5444
cc04c8d7c57442a067261d3d0d74933f47ffb3ad
12818 F20101112_AADEDX drozdetski_a_Page_017.jp2
81d33c9a9ea91c8b56240ea1743b5a21
51eb6acfdf2b310032cf64481d7cf8941a2cee6f
130343 F20101112_AADEFA drozdetski_a_Page_132.jp2
105b509ebfbdd1fe4a8b50f444b9b794
43bae22fa6f127aed02d07d205872d433306664b
1051971 F20101112_AADEEL drozdetski_a_Page_090.jp2
5210a36120d4d15913ab822ac0225dd3
5e2a7e689074aecbf0707cf5b581b62c90457de3
6751 F20101112_AADFID drozdetski_a_Page_036thm.jpg
110ee8f909112daa662a2218921c1a74
64515c5ba88415bf4299d1fb484b8290d7d28c16
5545 F20101112_AADFHO drozdetski_a_Page_026thm.jpg
3e1460fa4ee496d8ddc7b45cd01db5af
cb9e6bfb9806d61794c9bc0b75efa9c04d422626
2364 F20101112_AADEDY drozdetski_a_Page_039.txt
512f1502c8567701166aa974466de48b
421cb34f98dfe5bf55115a5670ca540785baa3b1
4490 F20101112_AADEFB drozdetski_a_Page_127thm.jpg
11846371273c7c5245303838aefb7bf0
497a2ff020572903384f870aec256bb16df1e986
20796 F20101112_AADEEM drozdetski_a_Page_018.QC.jpg
a28dacccda5694ff97ee09d14ea54199
d033a9f87b330ea0810114e7b7ce043a4e95152b
26834 F20101112_AADFIE drozdetski_a_Page_037.QC.jpg
b52cb9f46cee8a0a10065d1c46f4f28a
a171826a258f20597208db8b35bc324a7ad442ff
20973 F20101112_AADFHP drozdetski_a_Page_027.QC.jpg
23ed3eba8c780fd52781fd9a8e3b33c1
ff48b711023042738fcf2a28f44205d01c36b850
4896 F20101112_AADEDZ drozdetski_a_Page_072thm.jpg
9609ae2d97d528c83f768e182d17422e
d1ef809f4bdb91c1e87ed0caaff47535ee637521
78747 F20101112_AADEFC drozdetski_a_Page_006.jpg
a7a61215c71217e13e92264f7d2ee1ea
569f767a840bfb2e5594a5f3881033c017518f3c
5063 F20101112_AADEEN drozdetski_a_Page_116thm.jpg
f4b38dda2b08941fbcfe478410198a0a
73f07b0122e593d749663672495bbd3aad102a39
6854 F20101112_AADFIF drozdetski_a_Page_037thm.jpg
29acc5013deacfd6d19a7888de791c02
b05e374c8e524af7b4b3ea64c44f1866b271373f
5604 F20101112_AADFHQ drozdetski_a_Page_027thm.jpg
370c00ce29e287f700c15bb86b75ce42
72c139f8090484d45a3ff861599ebef466531857
86365 F20101112_AADEFD drozdetski_a_Page_109.jpg
7bb38bbe3f4e0136b3baef2aa77cf1e7
5a11b5960d8d78fd44b0387462bb2470fe6f709a
22789 F20101112_AADEEO drozdetski_a_Page_122.QC.jpg
09d93fb270c0cb5a6a5140ce15c68f3f
102aa8a5289fe459b07effd5d60628bad69d2ab0
25154 F20101112_AADFIG drozdetski_a_Page_038.QC.jpg
bab0e5e487e6cc914d8afd4a1cae42b1
f5cccbaac85fd03123b82e6a7c3a60978d240fbd
6337 F20101112_AADFHR drozdetski_a_Page_029thm.jpg
2225156bf1aa1c7724dcd70e83064d64
5d2056b251eb62bff7846a9be0b5f31386907475
2174 F20101112_AADEFE drozdetski_a_Page_137.txt
c9b94f870fd4c11c9d9becb6c359057c
b4adb586b0d744c81b9748697bccd73eb42c481c
73444 F20101112_AADEEP drozdetski_a_Page_134.jpg
c5a14eac8fafa0e646092f5781279319
578dce822b9285f8df3414ff0dc4874db1fe9021
6332 F20101112_AADFIH drozdetski_a_Page_038thm.jpg
8a7ba3ad5cff2b40b5b10b9b97bc32c4
a9864e9c11ac1fc23f4c69de7608a40107966341
26138 F20101112_AADFHS drozdetski_a_Page_030.QC.jpg
6b8988dcc3447af0aa1f1f5885cbb2c5
8065e5b58d944ecf74be7c479819b08e8dd2ac39
F20101112_AADEFF drozdetski_a_Page_119.jp2
d75cb8819d6352eca9c7c3cd361f6abb
e0b2db03305d5662fd9b020325d2ecc672fa2944
2128 F20101112_AADEEQ drozdetski_a_Page_022.txt
347299de0db08792055a0ae62045bfdf
f0b1ac015f64e23c8a4e00560d4e7274eb04be40
28362 F20101112_AADFII drozdetski_a_Page_039.QC.jpg
841572a2bb164a46fdaceea72b7db669
c9230c27b38b7fd1026c31db26e4003da8f2386f
6493 F20101112_AADFHT drozdetski_a_Page_030thm.jpg
94e3053ac8ad8fa602937efa5081166c
e8312f2ff59a4406a464da8ff3118e8dfdc66d8f
23457 F20101112_AADEFG drozdetski_a_Page_029.QC.jpg
313a16fd39dfffe84505b6d4eb2d3376
8165974df589a41a78ad9e0d4d7509c08904b6c6
F20101112_AADEER drozdetski_a_Page_028.tif
ac01ba99701c61be746503e8e992ec68
4cb6373c4e03bb98826495dcaa6b7f2b04beb442
28022 F20101112_AADFIJ drozdetski_a_Page_040.QC.jpg
eb3f8772717efbc54a6ff3923f2ca118
9108990174a4dbbab4e26a2269d3680078fa3aab
11906 F20101112_AADFHU drozdetski_a_Page_031.QC.jpg
b750075aef3f7de9ec1a5708bf54faff
49e208b708139c239bb0dcdcce9b88b673b5e160
2228 F20101112_AADEFH drozdetski_a_Page_140thm.jpg
90f075cdf704157a2f61fdeb86440347
0fb5e404d0b03610df40dd1840f605b8a3b54b92
17963 F20101112_AADEES drozdetski_a_Page_116.QC.jpg
601c39d088dac766dc1a82ff1d09a016
75921cc078f8744f3daa292de6519b03b84f0f7c
6813 F20101112_AADFIK drozdetski_a_Page_040thm.jpg
bb52dbd27a7a7fefe175413df09cceab
242a31a68f4cce7e8faac2660ce68e28199fabc6
3498 F20101112_AADFHV drozdetski_a_Page_031thm.jpg
a11906942fa27fb1fc251fdb1e2bb795
e6f9a90d8d9e8efa6701bb8098e44ea3b904519e
83766 F20101112_AADEFI drozdetski_a_Page_090.jpg
c33f620acab09e742a8821c5cf66ca8f
9253886dbdb9fbc03210dec09929eb228ac7ea94
5550 F20101112_AADEET drozdetski_a_Page_055thm.jpg
a0ecafa15f14976af5e5f232ed4c78b8
05a147daf15dc9b5209951a34cfd2cd58e3730d2
26094 F20101112_AADFHW drozdetski_a_Page_032.QC.jpg
a5e4f0b070aadd739f2cc84f1f6bebe0
fa20159193039e38c460148391221047da56701c
6598 F20101112_AADEEU drozdetski_a_Page_094thm.jpg
0f46c37013b32929829d9404a6de5e92
3f45529016ff8401856c40ef31d4a90459ef5e75
21008 F20101112_AADFJA drozdetski_a_Page_051.QC.jpg
6042a6d71d144d13772fdc28621318e6
8ff4c2981f2689982397edcf4fd90aeb03e173ea
25241 F20101112_AADFIL drozdetski_a_Page_041.QC.jpg
adea43a4fa455fab167285f9edceed58
2a31bb54d265b674b5fe39dc07aca9ce16bc0bd9
6784 F20101112_AADFHX drozdetski_a_Page_032thm.jpg
97b68306ebaeacb46b9c587db3425d11
4f393c53bcd64f6d31587ad5cf72afcef36ada94
55978 F20101112_AADEFJ drozdetski_a_Page_102.pro
87b5d11a7bb973a5bc880f37b47b7118
095f3f45d6048b15476f7dcffdf4b4f8e5653b00
F20101112_AADEEV drozdetski_a_Page_087.tif
cf4c7c0233ac33386d2e727d0d5dbca8
751aede3b6857673479ceb2c56217ce43d4f42b2
5827 F20101112_AADFJB drozdetski_a_Page_051thm.jpg
dcc3d7ba6cfd69d6408df1ca92ca2f0e
de96a3c3d6f5a9366f7e6d3c9aca03bfb7a2db6b
6604 F20101112_AADFIM drozdetski_a_Page_041thm.jpg
9a29c7620b0dfd5cc1ecc1714b0e1e5e
f950dc8e01c7319d8cbabc5e17bb2d65f4f4dec8
12845 F20101112_AADFHY drozdetski_a_Page_033.QC.jpg
5c5e49beea570d5266fd9b7ec3170fd9
79de5e329f828e6a80f0884b84a537cdc8cea242
36701 F20101112_AADEFK drozdetski_a_Page_057.pro
5b50c6b7ad8e83ad38766728815f8987
8c767a78b98c948a28292ca73cbe81f791ecad69
10081 F20101112_AADEEW drozdetski_a_Page_008.QC.jpg
d71013e6852ce08923eeed5f245b266b
d68733d7a76abae1bbf19e308b08215c5e2b1bee
17191 F20101112_AADFJC drozdetski_a_Page_053.QC.jpg
66db52012401f2c6725425c0d4fa72a8
06b05206a2491fa2b8358b68c78467e1c1731c1a
25828 F20101112_AADFIN drozdetski_a_Page_042.QC.jpg
757d84ad73cccf8ebd4264e4fad62a41
ff0d2129acc1f8dbce000abfe2debbb0ccaee498
4405 F20101112_AADFHZ drozdetski_a_Page_033thm.jpg
6840dbd8858c89fab4479922cc90a73d
3cf40cd484d6e13a412d0f9c196ec638cecce919
6089 F20101112_AADEGA drozdetski_a_Page_120thm.jpg
505a10ef04f33d3543e580fab41b5661
c6c93968eb8ec4228a5d94be8ac5e797161a5cc3
1051973 F20101112_AADEFL drozdetski_a_Page_022.jp2
d04839c904add736e781c2f0a5dd5173
4da06f7ff9052fbb236e7deb637c4858e1bbb55d
6599 F20101112_AADEEX drozdetski_a_Page_102thm.jpg
b2e02ed40cb7f247cb2a38536706787e
69666e295584922b1ee5919977d86df1708faa17
5461 F20101112_AADFJD drozdetski_a_Page_053thm.jpg
f98e94922aad416f07f5a60ba319b012
5100c1010489cfbff399f6cbbbd77358b2605a7b
6510 F20101112_AADFIO drozdetski_a_Page_042thm.jpg
44aae7f8605dcdc75032549fecd8f883
9a43aa9069b6e38538f18bb14dbd0451f4bed637
F20101112_AADEGB drozdetski_a_Page_023.tif
ee1add1e805076dad23351b07dc4042e
c57448bf0acfb720c0673894319fdc387b6ef766
53551 F20101112_AADEFM drozdetski_a_Page_044.jpg
33fef30cc860aa74197e92667bdf6ac3
4091a6b5fac46af300a6def250452b51414eef7d
51569 F20101112_AADEEY drozdetski_a_Page_043.pro
231991ba1d64e338eebe007b956914e5
cbc21d8d19ed61fb868a0e5843a0a54deae5c9de
6790 F20101112_AADFJE drozdetski_a_Page_054thm.jpg
248fa808a5a1e7476e6c735a0f155660
670f90aa7986275c1239c5e50fbe8495ba7262d8
6561 F20101112_AADFIP drozdetski_a_Page_043thm.jpg
0a2f0502d7bf9b94426c2c03d91990db
c1b5c03f43e5ea17a984de05be8a4adafa64fd98
18586 F20101112_AADEGC drozdetski_a_Page_072.QC.jpg
bf7fb0c8638c65a1e73b320bf5da2025
5f9488fa11f2e9f969de2898bf3173c05c127237
21686 F20101112_AADEFN drozdetski_a_Page_137.QC.jpg
69032e5dfa4e158f5bf58e145571194a
f1a0ada9c52462a0a9ba81a7ca3fcf97aedb6323
24270 F20101112_AADEEZ drozdetski_a_Page_132.QC.jpg
fc4ed1c1d1dd5e27c2a6ae1e330470a7
ab512cc3b53baf15c15a8d1a82c1ea80a57ae209
19885 F20101112_AADFJF drozdetski_a_Page_055.QC.jpg
a490e905d2923d5b0eddb5abaf2adaeb
25fb50011b7f29030540e74fa1d5bc9baa4d29f6
17514 F20101112_AADFIQ drozdetski_a_Page_044.QC.jpg
52fd78effea7dcb7e54c3967fde2e1e8
2fb6f138d37246055c99a404c627354a0899ae65
2236 F20101112_AADEGD drozdetski_a_Page_109.txt
b307409ed2fe14b49d0b5c8cedd9e107
fa5249860d5cd67e710aa3de2f9c8e940f637e55
1050538 F20101112_AADEFO drozdetski_a_Page_023.jp2
f6cf01bf285fb8704196f9b4724ffeed
938724a05c159d28228d1ebedd199d49b50181e7
21143 F20101112_AADFJG drozdetski_a_Page_056.QC.jpg
4fe48a4b550b8cef766a9d1cfa522edd
f4fe97537706a498abd6dc5496abdac462241391
4598 F20101112_AADFIR drozdetski_a_Page_044thm.jpg
ea18352ce0397015c9b3eca69cd29ca3
660269cf29ec47844c2126b5867c295aecf95c4d
932101 F20101112_AADEGE drozdetski_a_Page_127.jp2
1217ba594764fb7ca90317fdb0bd8d61
2ce8832bc90c13e0a070d9e369edf399c18a3ab7
2297 F20101112_AADEFP drozdetski_a_Page_052.txt
a8177440fc83cf875f9bdd059c4b87a8
58ad24f315f5e233c7c0b5a92932837c28cfee2c
5489 F20101112_AADFJH drozdetski_a_Page_056thm.jpg
7c162259d849e00549bce30143f08a62
36b89357d0c868355588c594f87c79a20bd5509b
25859 F20101112_AADFIS drozdetski_a_Page_045.QC.jpg
980450c5f69ef47a1dbad45a9e59d3f9
aadafc06a8a4425d4289f059b1fdd4a9c3e95d51
38912 F20101112_AADEGF drozdetski_a_Page_026.pro
11ba6a2b3bb4790e250e513d6f04fe0d
d03e9715200d255e206432ab52bffad9707b787f
F20101112_AADEFQ drozdetski_a_Page_044.tif
f81ce05a068ab2cea27d8d1cde705ae1
f047600acbf93f14c277bc1ca4c28480473ff7e7
22062 F20101112_AADFJI drozdetski_a_Page_057.QC.jpg
863659c2ffb17150381b231e26192f13
130bd58b867c8b87a5d7791db7d73eefb5e7c6b2
19199 F20101112_AADFIT drozdetski_a_Page_046.QC.jpg
ecdc65e4e407c9c885abe3fa561b58f0
7a6eb54a1d86f89d21179e5d9552c04238a060f6
103904 F20101112_AADEGG drozdetski_a_Page_018.jp2
a65959ce69b4f033d4fc6d847008bc39
21fc3b9c0c3d0b4f0d495706493323280e900b9a
66737 F20101112_AADEFR drozdetski_a_Page_016.jpg
7d2976682e078921300d1538ffe35229
0d6bf8915486107206116890f513b02d47f2244e
5990 F20101112_AADFJJ drozdetski_a_Page_057thm.jpg
7ce74e194d85c16c4263ab58d376ab61
4d919a595f815007182ffd584ca57919998007a7
5554 F20101112_AADFIU drozdetski_a_Page_046thm.jpg
80b36aecac7c2d18d5b315f59c49e9ca
74b6d7ed0e83d17f86a7168fcceca9b16229cc47
F20101112_AADEGH drozdetski_a_Page_134.tif
b713fca8c4d8d5f40c421c48a578fcdb
39ace3149da6917c49004b236a742fe135422b01
F20101112_AADEFS drozdetski_a_Page_013.tif
1dc70400c7eaec32400521a183de7c25
05924577aa63fd0228a5b32f9ff94e6c93b07950
24146 F20101112_AADFJK drozdetski_a_Page_058.QC.jpg
6f6e12416c1b86f08f72f34fa2855af3
72c250b1195b4fe7fec6e1a9b5f726505eb59410
5200 F20101112_AADFIV drozdetski_a_Page_047thm.jpg
59fc7edcc88495c8140293327833fa5c
dc907b2b4002a46aa3c137a5e799282736652bd8
10284 F20101112_AADEGI drozdetski_a_Page_138.QC.jpg
fafec52bbcbc58add7b761944c1a336d
54d0e20e3fff46230f5dc9dece93678978af58b4
F20101112_AADEFT drozdetski_a_Page_115.tif
2944dceb0ef393d944e3760b9af817d4
2cd55c3200843bf4cd79a2ea0fb7130177ef8568
6044 F20101112_AADFJL drozdetski_a_Page_058thm.jpg
7a04621b77d7b1d3a9d126382f039a14
017070a1a092f9734e3124b709e387b5bc1be8f7
6123 F20101112_AADFIW drozdetski_a_Page_048thm.jpg
8ce6bf0784f4c3fda00ad5822a07cfab
49cc11656322b199b5d3a02553d7bfd37fe38e61
116204 F20101112_AADEGJ drozdetski_a_Page_079.jp2
68bae0faa9d108214945d2d9e912c70d
538a0e4f6ba385ae145bf73b9d216833fac769c7
1051978 F20101112_AADEFU drozdetski_a_Page_010.jp2
7554195062e47aa1b9c255589cd6d90d
2289a5c839bca9f592ed200402338c31d3803e6c
6547 F20101112_AADFKA drozdetski_a_Page_068thm.jpg
add9b0c371298b1ee4804985812333b3
cdaa5f4427ba1e973d295600ffee85a97a66c151
5805 F20101112_AADFIX drozdetski_a_Page_049thm.jpg
0b779d4309ae8ff0200613a3da9c0e1e
630b9cff8fce5c99ade4ee43e553de712d5460d3
5639 F20101112_AADEFV drozdetski_a_Page_016thm.jpg
c926d091ef2088e424fd9ed49c98c816
986beebf860cf8a89abb61c245913da7f0bd9acd
21735 F20101112_AADFKB drozdetski_a_Page_069.QC.jpg
82cfd3533a5bfcfea3be397e020e5885
d3d0a46bd834eb1ef3eddc0f536fbed467332294
7993 F20101112_AADFJM drozdetski_a_Page_059.QC.jpg
f8d10a8c09f3968b6e5118f209542e77
039c5f302d6f7e128c1e3be2ca4857e52282b2bb
24024 F20101112_AADFIY drozdetski_a_Page_050.QC.jpg
3727c997e441d3f184f9a2d5ede113dd
4dba8ab2e636155ced14aa284d6ce0a1de0a2932
2311 F20101112_AADEGK drozdetski_a_Page_133.txt
c55eed101c50fb6932d3cb8f5ba6089f
6cc174146e29d3b194220cdf50775722b86d06be
F20101112_AADEFW drozdetski_a_Page_030.tif
acdb5234d81b18bb080c034d11ec601c
1d4546728768d248e75751cc8744913f939a596c
5909 F20101112_AADFKC drozdetski_a_Page_069thm.jpg
2c81d9893881bd4cb644c088aa8cabdd
5d6a34ad323a5b82add8895fd1aca4740c6d4cb7
2492 F20101112_AADFJN drozdetski_a_Page_059thm.jpg
c43ab051b60ffc898d3e8e9aad5cf8c3
2e2de79e64b6a59f504dc1411ac34da9572cd6f6
6065 F20101112_AADFIZ drozdetski_a_Page_050thm.jpg
7ee05de90b17658ec5da9f1675f97ead
6fc140cbf1e10ea7627a0c3431391fadc153eb5b
F20101112_AADEGL drozdetski_a_Page_057.tif
be3d13ac1fd4690a5969b4ede629ed0e
fc9aa963760627f39a60508d863f80d72792f85a
17664 F20101112_AADEFX drozdetski_a_Page_110.QC.jpg
3d7894dedc543969cd959207c602d45a
4973b1b9ce7f44d5eb3f01353e651cc61c508853
58477 F20101112_AADEHA drozdetski_a_Page_099.pro
eb8aa6c3b6cf8515e591b46d1af56484
8431e8da8fbeacd24c66ac85a383f3b88cd74925
22369 F20101112_AADFKD drozdetski_a_Page_070.QC.jpg
bcd55948b7c3065bf826ed3fa759c32a
e97fbad5b4169008940cedc69dfcca4e6b1bd50c
24248 F20101112_AADFJO drozdetski_a_Page_060.QC.jpg
610b4be5c76137ff2a4767f0e991369d
d2dd80522617514e5a8cec6ecb6fa2e37b21cd2c
F20101112_AADEGM drozdetski_a_Page_084.tif
3dfda84955893f5c1839f97eca894ff1
0c7b9850730bb91d525163c35307506f5f6b1343
6333 F20101112_AADEFY drozdetski_a_Page_060thm.jpg
109450b09db2f50f851f4a53458648d8
6f3a7a4078849ff22424b20485df978c7a20937b
9676 F20101112_AADEHB drozdetski_a_Page_019.pro
24fa7335d2df251a59981028e6dad132
57f3ef7b9c8bed4f6d7c9307a0d835971c72eddb
6130 F20101112_AADFKE drozdetski_a_Page_070thm.jpg
5c98c7cb98371a956f74cddce670e699
ebc14a9392c4fb9f57b76a327256d189ca9e854f
29422 F20101112_AADFJP drozdetski_a_Page_061.QC.jpg
9cfd0f94005ddfaeba249ac95c84d9ef
7f984d1d1599a23433e2bce420075702b78c667c
24942 F20101112_AADEGN drozdetski_a_Page_035.QC.jpg
1e783a3c67c0b004df59509568a35cac
ea1b4336fa9550aa1073d0922062bbba9f01b966
807139 F20101112_AADEFZ drozdetski_a_Page_072.jp2
68f84c51f67b1298c4832f09833e560a
67c308e9b89116c3ff2502f804463e8ff8592e93
93995 F20101112_AADEHC drozdetski_a_Page_063.jpg
b14e8cfe33ff340873e031287f3da83c
7b5de6416a90f7ef5fe4af2567a320bfa1663d52
23356 F20101112_AADFKF drozdetski_a_Page_071.QC.jpg
a438b2ed911335f3de00992f74af44bc
8e2634fe7d0a7327752f9b9cfbb28d60cc3ee5ad
27041 F20101112_AADFJQ drozdetski_a_Page_062.QC.jpg
d05f0387aef8d3cf47d28773ecfffaa6
d064404560b8ef6e1ed22043c9ffc51e4811e713
58794 F20101112_AADEGO drozdetski_a_Page_136.pro
f46b5dfff537fd67e4a810ff21987831
f7077327d28ddb008ea0422e3d2938d19fdbe75a
2069 F20101112_AADEHD drozdetski_a_Page_139.txt
ea17ef4d9fe8befdfd05e8cb38e19531
d608cbe09e225882139ad05d288f17f7db3841f4
6240 F20101112_AADFKG drozdetski_a_Page_071thm.jpg
dab94536cea209f09db91c78f3a35efc
2ba2e0617dd74da715bf934b844c880100791f5e
6829 F20101112_AADFJR drozdetski_a_Page_062thm.jpg
5d3bee07677cf16d4884ded0a02f0f13
32f95e191794325c775f61e446413b416c731c21
42265 F20101112_AADEGP drozdetski_a_Page_127.pro
2bc67f3bafb44fb564d8e3a2230ab0a0
ab56dfa0efc7b6475b8bd168f3acf93e2d213c50
1580 F20101112_AADEHE drozdetski_a_Page_116.txt
7200403dfcb4d268f7d2d8c1926f9678
e4022ef8ad79cf9e9c3db6ea3237f91012c4dbbd
16277 F20101112_AADFKH drozdetski_a_Page_074.QC.jpg
b8949283078c4f1a2689d8b16eac8a71
bbb845932b521bac61fe3772cc253fca3990ccac
27685 F20101112_AADFJS drozdetski_a_Page_063.QC.jpg
7d51dff0050489f9179bcf366364e125
259ef65499e4601b16461fc40dcddef8af60e801
F20101112_AADEGQ drozdetski_a_Page_048.tif
882c4aecb58ed78b621880ea9e4a58bc
76fad954a56e9ae9935090d8d45c5852378cfd25
6836 F20101112_AADEHF drozdetski_a_Page_028thm.jpg
a051f1bbd652e70e744aaff83a6cef5e
67fc7cc65d84e8cc8a71f7435a8a0e257a10efac
4981 F20101112_AADFKI drozdetski_a_Page_074thm.jpg
21b82d26d1b85a7164680fd8fd845c12
e4f3e9c7dd26b28a9334e2698f11da0127089212
6740 F20101112_AADFJT drozdetski_a_Page_063thm.jpg
ae8bec1db942ce8f20f87f590c3809bb
126a55633b684c77d83769af2633bd5dfaeb55c2
F20101112_AADEGR drozdetski_a_Page_041.tif
1dcb70cccf4711610be65d41ac46cdfb
ac05d1c5b4b58e5541024c8ade0b2bdc856c123e
24255 F20101112_AADEHG drozdetski_a_Page_101.QC.jpg
d4fa459cf6e3070b234f02aa2a231514
5a76e215fe097471200229f6d9ac1afdf62683d4
4603 F20101112_AADFKJ drozdetski_a_Page_075thm.jpg
7bbd8bb649608c8bc9674e57adff2645
f7bc60768c3197470d5ae45020a3b69e7002d7cb
26507 F20101112_AADFJU drozdetski_a_Page_064.QC.jpg
adc42852cb57262bdb15b87957559901
9a9a2e5e584410fc6d6528cb45a18e564efbc84c
F20101112_AADEGS drozdetski_a_Page_051.tif
e8559edabcc217c12a96a6a991f1a71e
4d9372857a182d538a88cd215bce4f8b619385d8
F20101112_AADEHH drozdetski_a_Page_018.tif
19964a618de6e8e3f7ecebbbf60f30ed
1c31482ed86df60065059d0c28f9f44683f285d5
18727 F20101112_AADFKK drozdetski_a_Page_076.QC.jpg
ce357cd3794c0dbff24f4249eaf82717
51c923b1e64c10994ed420c0d970e0e07c0ebce9
5725 F20101112_AADFJV drozdetski_a_Page_065thm.jpg
f57dda4af1a012917439e01cb2a2e0c4
0c16a3e49c9f847e84e2fb40b598fd6a895fbf77
127308 F20101112_AADEGT drozdetski_a_Page_135.jp2
d9d4d73fc7226281ec28a49f8bef5b28
1900c68c2ce83adfcab36b0e57ce972fabdd7c5f
70773 F20101112_AADEHI drozdetski_a_Page_048.jpg
b4e6de8643c2b434092f2ada5d87dc70
56b51ce6ec79dabed58d9ce730b6431d197d82e5
5075 F20101112_AADFKL drozdetski_a_Page_076thm.jpg
567602418c1693d0bfb81568c56f65d0
a636ffaf0ab75ab9271680f9b07bb88c8a35d156
23544 F20101112_AADFJW drozdetski_a_Page_066.QC.jpg
cb241ac27a52e12b188c20e436f2b492
cdbb2e8f58bc468cf2ed48a60e0aef2fbfb0ee81
97452 F20101112_AADEGU drozdetski_a_Page_012.jpg
d1db7778263fd2beb15483ad56365ab9
bc7d54cc3d9298330445b431e5873767e4e09312
57951 F20101112_AADEHJ drozdetski_a_Page_062.pro
159e12405fed85e360bc40b680bcf068
a5eef215150737d309f8ebdcecbf31980c471261
9437 F20101112_AADFLA drozdetski_a_Page_087.QC.jpg
6e05adeb9cf751acd1c1a90c90ca90e3
05c739f54dce39393c57aa398393fc482758f7ee
16618 F20101112_AADFKM drozdetski_a_Page_077.QC.jpg
43fe5d549931f2acdb46a0af3ddae916
eb54747eeebc1337985cfee51db1f374a5b7c31b
5846 F20101112_AADFJX drozdetski_a_Page_066thm.jpg
20c595db67f0e5cb04971f45fd5beeb6
8145740cf77870cb5464fc416fabaf19c61a1624
54340 F20101112_AADEGV drozdetski_a_Page_079.pro
53c2791a15f57a8506bfe779ae2c19fe
09e3f28ed8aadd394c1bf5e51cd3dee53ae56261
2210 F20101112_AADEHK drozdetski_a_Page_041.txt
058ef46f6c0423860606189e9ac0cd1a
538b657f6f348711bbb6fe9bc1e911ee77109504
21628 F20101112_AADFLB drozdetski_a_Page_089.QC.jpg
0666f0ead88265be33021aa4ed63cc99
bd901217ad5afc3c679c3e8492dac94b4a80a750
F20101112_AADFJY drozdetski_a_Page_067thm.jpg
f3080837f6a8aac56e1c8b670508f135
167264acefa562695de030fb24f10f1b91b8a63a
75550 F20101112_AADEGW drozdetski_a_Page_050.jpg
60baa8e3f427ce18fb934a0692e1a0bc
6f2c29e48893a3aad55d330c7fe8a7d3c89633f4
6025 F20101112_AADFLC drozdetski_a_Page_089thm.jpg
251d6cd272003b710460199e9e8939b7
a3c904bdf92bd2e8ee8ea6c4d66951ce7b96e40e
5139 F20101112_AADFKN drozdetski_a_Page_077thm.jpg
12f98c92818465fb94631b748ccffab9
14a0ccab43dd98c2273e8524916f36c758f8530d
24456 F20101112_AADFJZ drozdetski_a_Page_068.QC.jpg
63c7f22c273cb7d3f8fed0aef1306404
a96932f26db078c7d313395a39335e2b09900320
1524 F20101112_AADEGX drozdetski_a_Page_057.txt
49da83daf6ec3ca75bcaaabea3c6fbbd
31670858dc150cbdcf3f77c8e14f629f43b5b230
118794 F20101112_AADEIA drozdetski_a_Page_038.jp2
ff3beecff0afe0f0160d183f9ed9c0c4
c3efe3810f42ffc9dab30bcab6d78de85c3c61f5
F20101112_AADEHL drozdetski_a_Page_085.jp2
b1d6bb54fe76aa1953399e30f184a69f
145fb6b234fa87cd357242f5277f122a4f6a4a6e
26180 F20101112_AADFLD drozdetski_a_Page_090.QC.jpg
34aeb3c0dd9e215b0589d95cc44d4abb
62813d7c6d817050b73ddbf6c7b4182fe352272d
5252 F20101112_AADFKO drozdetski_a_Page_078thm.jpg
33ac212cc4334bf20936f538d213c585
a0764d246d1179e19804c83f43eceb24927562f0
506187 F20101112_AADEGY drozdetski_a_Page_034.jp2
2fd1cdeffe3b41948eaea18541bf503c
8da1fe56639e82ebf83f6a836a2b36d7c0214ef0
63889 F20101112_AADEIB drozdetski_a_Page_106.jpg
7300004ce4c90022d7e9b3950243822b
6952199a6775944cbf196ff490ed406fd7dc97bf
F20101112_AADEHM drozdetski_a_Page_128.tif
b5a8605696c2174b9bb89718b106da43
afb39d393e6fa99599e4302271e4fbec0c70c096
6559 F20101112_AADFLE drozdetski_a_Page_090thm.jpg
5e604ec70e6cb7ce8023bec8a706871b
01f7f2454e6ca0bdf216db46b5913ebd553fddec
24062 F20101112_AADFKP drozdetski_a_Page_079.QC.jpg
7316538c62dc978d5e5f3147478fe757
f646385def6a9cc41adb1d602e7eb3647c1c2b81
785667 F20101112_AADEGZ drozdetski_a_Page_113.jp2
86739f702b18bb3022c7a0d32fecce86
fbf1e90ab9f6447794779ba6b6ce0eeff929638d
2179 F20101112_AADEIC drozdetski_a_Page_018.txt
e97d3e4faf5152bb48877977ba4f0a08
165eefa591ef7c66438e2000fd71a22fd63d897c
2001 F20101112_AADEHN drozdetski_a_Page_020.txt
99fd529dff23f9d2dd5def085f9492fe
4fadd0e8c1a36a42043b1aaa40025fc8464587fc
26300 F20101112_AADFLF drozdetski_a_Page_091.QC.jpg
c7292dbba156880255d790b88f79cfa7
40b5459f22800a272f7f693ac0d0d0a75640b114
6078 F20101112_AADFKQ drozdetski_a_Page_079thm.jpg
227d2eb81a44d83ed27e3b63fda43b81
4771a4ba95faa6008f89eb34f75f89104e0ddf92
F20101112_AADEID drozdetski_a_Page_085.tif
eacb132548d36d115dc55038339cec46
5c9707748b905540081b728d9cd9c904d73844e3
6541 F20101112_AADEHO drozdetski_a_Page_091thm.jpg
73a3dcf31e8151238e544d80f5086a07
b755374ce2b7e10c549edc14778a6380455925ac
25490 F20101112_AADFLG drozdetski_a_Page_092.QC.jpg
e415d4429ebe4b99aadc6b40b918b4b7
42f986ad4765e9ed4a0d7472074e0024c42adf90
5995 F20101112_AADFKR drozdetski_a_Page_080thm.jpg
7389cabedabdb4ba7bad8b7c30e19732
13fcab00d0c11c853a03a7e9696a087ddea2a297
53349 F20101112_AADEIE drozdetski_a_Page_137.pro
f5c571bacc2656ce282dd5702e81c9a6
aac4c1f33617eb7d2bd9b5e34053e37ac3890575
5137 F20101112_AADEHP drozdetski_a_Page_110thm.jpg
8c973cdfe67f2491537142ebbde0b7a2
e237f6b33fdbefaf4197ad8cd176e3f3ae54c47e
6475 F20101112_AADFLH drozdetski_a_Page_092thm.jpg
d5fdbed17b01af5090dad36b888d2862
834a4e4fdf19083fadf8d6590e1e72f15ee3df97
5744 F20101112_AADFKS drozdetski_a_Page_081thm.jpg
71542e567c528a0cf149ddb065e47040
be6264435f65fe4076a50cefd493eade3af614e0
68698 F20101112_AADEIF drozdetski_a_Page_012.pro
03e1882a02b0d1d1b3a43dc756a4b098
fdc3bdebbe667c2a8ea4d1150eeec4ac46fb45e7
2860 F20101112_AADEHQ drozdetski_a_Page_007.txt
62527a08e2b707f1032adf9d1084ba30
dcda941dbb7097905358739ce6efd9dd35384910
26343 F20101112_AADFLI drozdetski_a_Page_093.QC.jpg
4e1a0717abeee23e70608962b6f36a8e
7fb3dd427cc63587813a35b119d81feef6f74ece
16973 F20101112_AADFKT drozdetski_a_Page_082.QC.jpg
f6eef8b95a8d21fd743dad1f7d9c8b57
f060014768e5511e5f8c266138666add033cc676
70237 F20101112_AADEIG drozdetski_a_Page_139.jpg
5efb8c678dd814dd8a0e40359a120ebe
b1bb3508ea1377fb7d8afe95c919ef31963f33e5
46523 F20101112_AADEHR drozdetski_a_Page_080.pro
28a7a679d0209166555dd7720a3fc9fe
de172dac0d67cad7a71ede4b338c24e4e9056329
6459 F20101112_AADFLJ drozdetski_a_Page_093thm.jpg
36cae24793d868cdb57681fec4767da3
0b8ddb52cc9b6a54f0c5fe5dee03c346c7183cf1
4900 F20101112_AADFKU drozdetski_a_Page_082thm.jpg
f85f2ce6782a9cf44f9cdbdc85f64261
464d34f4c0bf9c6c221e1ab7cb11fa6bc327a81c
55588 F20101112_AADEIH drozdetski_a_Page_053.jpg
7e8fe0676e6837cafe27e16fb4f4b24b
34864de2cd18475cb65ad285cc62d890507670ac
61869 F20101112_AADEHS drozdetski_a_Page_063.pro
ec9acd18733b4ba086da3bc2d7560cea
e7238d3ae117cbfe4a65a3263c9b12485c50c751
25433 F20101112_AADFLK drozdetski_a_Page_094.QC.jpg
00199d5055e6abe90f7826baa4ddea92
7b2301a38089b6ae68e90590f16f000605e926bf
22306 F20101112_AADFKV drozdetski_a_Page_083.QC.jpg
f025adf0a0a34678c31a73ca561f029f
b8f268ee8cc4da367d31bff1cf9f0a7ee40b2525
21260 F20101112_AADEII drozdetski_a_Page_049.QC.jpg
b2c17851912f69be9446e94fc144e49f
01aba3c798d41b65bcf318588694378110bb13e7
44434 F20101112_AADEHT drozdetski_a_Page_025.pro
b15bb63d510a6685b663194f184b8884
d6883c82151e218b2816e55ed26f868c371435ab
27920 F20101112_AADFLL drozdetski_a_Page_095.QC.jpg
63ef6ee1946b953bb6220d5ae92df78e
d79b3afae97828e19cbc0c0efe60757b76ae768b
5952 F20101112_AADFKW drozdetski_a_Page_083thm.jpg
f8d5f03a1b0a7dd18cd921f875742751
7fbf582f5bc7407f36f8d22ed67d57b2b592d117
32718 F20101112_AADEIJ drozdetski_a_Page_046.pro
42c4243d5ec9d2fed07cd0c824a6a224
60afbea2841c528696589719a4021e90943741d4
818411 F20101112_AADEHU drozdetski_a_Page_075.jp2
1b3038f6038935e64f1ec67f86bf0c76
3d7e0fafd09337c71e27014f1432626e7768806b
20106 F20101112_AADFMA drozdetski_a_Page_106.QC.jpg
4b69ba14bc01314433410adb470d33e1
42736faae8e8ff1169585dbc152fb0f75594380f
18909 F20101112_AADFLM drozdetski_a_Page_096.QC.jpg
6455e1d8c1d90343c2c04de9aa1a5879
f778e3136186e3061ae3768a0d9d8a2b3a180224
5718 F20101112_AADFKX drozdetski_a_Page_084thm.jpg
db9dece0149ea7c4741be74dd0af8c1d
d81902a444e8049fdcb2b3fa2768b316e187697e
22708 F20101112_AADEIK drozdetski_a_Page_131.QC.jpg
605307b4a44cb14ef46ed1a727d80402
6d234f046c2cbe743937e266658d5103b29ff367
2173 F20101112_AADEHV drozdetski_a_Page_088.txt
aebe5af6d585f6fd6fef636c3bbd9c3b
990f6ea27fae63c719beb205b9dd26671c355090
5589 F20101112_AADFMB drozdetski_a_Page_106thm.jpg
6130e6368ff6a401848396ea551508b4
02d09c371107d7733749852f5781886e64a0ade5
5284 F20101112_AADFLN drozdetski_a_Page_096thm.jpg
a481c1110378f4951f82ac488d0ed53f
1e82b134cf8d767a994615ef8af902914d0fcc1d
5467 F20101112_AADFKY drozdetski_a_Page_085thm.jpg
b432fec4e4360ab989fa5e7ef8d8de29
974dba2ba8d3e8657503f1ebf223885bf1f1e299
2152 F20101112_AADEIL drozdetski_a_Page_097.txt
fde66321ba660a2be3ad3935c60ca611
44de6497201f919e9e97bd5a201ba198f9d8289a
22211 F20101112_AADEHW drozdetski_a_Page_065.QC.jpg
f8d28bcab8e77b53f0c48dbcd85dc591
f056279e858dc5e546ad9b1ad1d05e6e2b1e0d13
25227 F20101112_AADFMC drozdetski_a_Page_107.QC.jpg
6aaf05e02c418df11c801c81bfc36715
fe7c6edd9a7e1de0172b949cce7360a1d8638233
4974 F20101112_AADFKZ drozdetski_a_Page_086thm.jpg
7a3aa5c194e73d3ca1a9f6909358271d
a89640985a3da6a8ba4a1c8461211c74c8aa8946
2803 F20101112_AADEJA drozdetski_a_Page_010.txt
9dae9cc8338179854d5e18ae51de9b7c
aa1ebb836788315ac81f7db1678c2fd815ee9e8c
20849 F20101112_AADEHX drozdetski_a_Page_081.QC.jpg
3900f73570a76e8cc7d6c2ccd6343277
1ade3b91c7f5b03a3e263f3b0a3dc27541d8b700
6544 F20101112_AADFMD drozdetski_a_Page_107thm.jpg
c01ae667ab7882cf97501f173c905c00
cbd0cc1677efbedd4488ce705516e21dc1ff86d0
24858 F20101112_AADFLO drozdetski_a_Page_097.QC.jpg
c6e166c163781119fb68dafb0e920dd8
55f580a618f94316ab673f0f8a108b4db2c14fba
2920 F20101112_AADEIM drozdetski_a_Page_011.txt
6c9043fa40f9801fdf9a9ecd4c0adc11
590dfe4951229a985642cb4f1d0927518730099b
69477 F20101112_AADEHY drozdetski_a_Page_089.jpg
1841de8d9af766703558e83bc2dab0d6
def755819ea8986684c0e9ad556e23e70fcf7bd1
125011 F20101112_AADEJB drozdetski_a_Page_133.jp2
b326aae3a6d6b1a563bf9681d99c031f
c843b5c199a2818479e134ead85b7c114fea34c3
25374 F20101112_AADFME drozdetski_a_Page_108.QC.jpg
3546e8f5b5d8ea459925f639b1a6339b
36d478dfa3cacbd0ae37dced72fc720e4f9d6e1f
6302 F20101112_AADFLP drozdetski_a_Page_097thm.jpg
2c3b57aa778b516c2058cfc8f74eb576
bc39e2960ce130c5f12ce753edcf6f10818861f8
52606 F20101112_AADEIN drozdetski_a_Page_042.pro
ed8c37bd714c5c990504f89a2ee5933d
79addd3427f035f94cd19362f1677b09c967b684
46937 F20101112_AADEHZ drozdetski_a_Page_138.jp2
f17dae82f34eb224451ed13fb6aa8b0c
46ee3cf6cd2fda76e1b941c3d5d017ee80d5a35e
57632 F20101112_AADEJC drozdetski_a_Page_005.jp2
832210f73a8f0c1558f72340015a6572
b43a6c5f2ccd57135195ee9dd50fdd1a9d7291f5
6188 F20101112_AADFMF drozdetski_a_Page_108thm.jpg
3aaf08082c358da49562785596f6911a
b7bd47eb0dc9b2be4b8cb54160b700e4b308312f
23072 F20101112_AADFLQ drozdetski_a_Page_098.QC.jpg
46f7b69ea1b8260576f12a58dded0047
b05343cd2fbb6c6a4cbcd21d4937564471d66123
784705 F20101112_AADEIO drozdetski_a_Page_008.jp2
c16cc584222e7779696ce5c73af7b0a0
a43a1cc65d85efe082baef07c1b41a4ddb7ccbd8
F20101112_AADEJD drozdetski_a_Page_021.tif
18db49e02bd0e0485003fc6255ec1032
7076d574aa4ade7e94be39aad57ae58e9f0c2a8f
26981 F20101112_AADFMG drozdetski_a_Page_109.QC.jpg
08dca67d54d05d6fa7a776ae1986a133
efeb7c3ad1b914e1dbc4bf58fd7446c2fa83690a
6789 F20101112_AADFLR drozdetski_a_Page_099thm.jpg
fb3f890ba46d44b937690b1f6dccb780
df63cb90a7a3d6ae5e879ad4c9d471629d8ac1ef
27966 F20101112_AADEIP drozdetski_a_Page_009.pro
c3eb19918bc4a4d8ceb3ee040685233f
53624bfde697c232c2d697951455c80a268a6624
16041 F20101112_AADEJE drozdetski_a_Page_008.pro
b1bfe8fcf8ed33e3ada10958c22e5c4d
18e8d46c2c18572847590db3aeb0973763cf0ac5
6919 F20101112_AADFMH drozdetski_a_Page_109thm.jpg
d1cc7b47508ea100e37cec0adbae45f3
537af8bcdc3dc152f8af65f3380e6abcd31eb8b6
7011 F20101112_AADFLS drozdetski_a_Page_100thm.jpg
6856d1cab1c35f18a35795818739d342
a5a9454592dfd28eb80da08a28581954e1b22759
2170 F20101112_AADEIQ drozdetski_a_Page_090.txt
5bf4ff3100b95f5af694d76e2378cddd
9ed7df0e195fc505d2706da18080ffe49f0c55fe
21675 F20101112_AADEJF drozdetski_a_Page_048.QC.jpg
cdd45b5f21611056c7743465eac5363d
9de02017230ee16dd52285aa8eb78af3feb27fb5
13981 F20101112_AADFMI drozdetski_a_Page_111.QC.jpg
47a91523193c097a7ff6a9cb0d133b0a
342da3cc163088166e2ad9aeaeb100fd7857dd69
6195 F20101112_AADFLT drozdetski_a_Page_101thm.jpg
fe6ad26d5e302ea26d8ed85600ebb6ad
37fa72e9924b786584d8030fc08746563b7b5376
F20101112_AADEIR drozdetski_a_Page_003.tif
4c181f8c219e8074a2435bf2de8044c2
bd3ecab08b0acd653d49b1e75833d4a9496406e4
4730 F20101112_AADEJG drozdetski_a_Page_113thm.jpg
c09a705f96b162defbbcb86dd7e09c3d
137fe33858631c25d6cc63a21adf27977c0bea79
4371 F20101112_AADFMJ drozdetski_a_Page_111thm.jpg
d65438ca945f7d4383cf55c2cede1825
45b53d139e0f0a248c827c939e8a834c20e79cba
26897 F20101112_AADFLU drozdetski_a_Page_102.QC.jpg
f7e07ccd08d6fbae87fe395dff0d6567
890418498cc57997eded789b11029308ee6b6fa3
26791 F20101112_AADEIS drozdetski_a_Page_052.QC.jpg
65eea524131f162c80a1ee6f295c6db1
72a858a37a664863fa3e8d5aa2017baa7e8fb7b5
19051 F20101112_AADEJH drozdetski_a_Page_067.QC.jpg
012bcdfc0fcb916667f8323ddf44e077
ae044e92c0a42d98cdceaeb1121eb6909a7c7c86
16760 F20101112_AADFMK drozdetski_a_Page_112.QC.jpg
299627e94b351d18dbc1231a24a85f68
777e88e477d98a8c9b33f67b6c72cdf423a7d0a9
29344 F20101112_AADFLV drozdetski_a_Page_103.QC.jpg
d3b65963c79640137918b411993da7fb
82aa6eb779ae6f976e84356a419b89bd02947fc3
F20101112_AADEIT drozdetski_a_Page_050.tif
a49edb60dfcd8406ae9cb77cc6d899c7
161d21b4fdccff5eb3af3f2571be72b1a0239062
20800 F20101112_AADEJI drozdetski_a_Page_078.QC.jpg
e5147b47372ed90bc96feb7316dcc2dd
d8bb88c6c07286dae7d9646faa3535dd8cec8e55
5032 F20101112_AADFML drozdetski_a_Page_112thm.jpg
4cb049acf6899aa4279e52db54202af2
3f8f0195827f890d2cbaf585fd31a6ab0556c066
26657 F20101112_AADFLW drozdetski_a_Page_104.QC.jpg
4942d93835fd2cf9ebf0c334064eab0f
c9f242fdd594130cd7649451314f7afb0dbca80d
115948 F20101112_AADEIU drozdetski_a_Page_058.jp2
90b1d88fd5c2eef92564d16a50f47166
1c290354ced109009eecd6664e74847fa6109202
31366 F20101112_AADEJJ drozdetski_a_Page_059.jp2
ff72e35ae900e34ff5c2e848bc59f768
b3ee107baa2e3ad024fd3103017a2fe33ca81d07
3477 F20101112_AADFNA drozdetski_a_Page_128thm.jpg
d5e93128e24b3316842ec9da5ca9cc8c
11218672557c884be8a45d6c8481402202d880b4
4914 F20101112_AADFMM drozdetski_a_Page_114thm.jpg
2fef49e8db34f312c0c68e8982c9c584
7007ab3d1d2866c7109827ecd8f87e820ac26177
6775 F20101112_AADFLX drozdetski_a_Page_104thm.jpg
f0390853429f2a8c00702f6893acee83
e8bfc50c95c338011142366c49f450b4a3da60ca
2195 F20101112_AADEIV drozdetski_a_Page_066.txt
940ed298d4bf0135274fb5bab2d26b50
f12cf7ed08bd2ffdcad336bb68f6bd6dcf78a8b0
2424 F20101112_AADEJK drozdetski_a_Page_131.txt
f2d3cf9392ddfc1f1ccc3c6acec77a4b
3142e06aac69235e308afc72ff9320e2ecaad6cc
12623 F20101112_AADFNB drozdetski_a_Page_129.QC.jpg
6e17e249c82a29877f81ae49f210f55b
83b34ea0a954f475a45da5f34c1bafba7924df22
18532 F20101112_AADFMN drozdetski_a_Page_115.QC.jpg
396139a42edd0d0d07741b83ee99473b
618b8c9592deef167ca69df9f4c192dbe99344a3
26714 F20101112_AADFLY drozdetski_a_Page_105.QC.jpg
8869aa46bdcb1b3534fba9234a039c2c
a0018a8de44b7d78b8f749eefb0f15234b0aa3b2
859 F20101112_AADEIW drozdetski_a_Page_034.txt
d8675a9d1938e99fd270f1f7003e6ea2
1ff5b085e0db7a9a9c744d57bbf877a7408c23d5
2440 F20101112_AADEJL drozdetski_a_Page_135.txt
1dbea9ea4e917efc1f34fa073d01c416
1516bae91d2bc71f9efcc975736c8a43861b8469
3563 F20101112_AADFNC drozdetski_a_Page_129thm.jpg
0d11bb7d89360643558443ff83b53f8f
8a8ef809f24d2cf0c566da543028ffd2021686f0
14084 F20101112_AADFMO drozdetski_a_Page_117.QC.jpg
240042bb924c1ed135fbda6910d84cab
b715ef71b9cea44f105a9d4d51ac9b3d16eb5f30
6609 F20101112_AADFLZ drozdetski_a_Page_105thm.jpg
533afcbbd7ec4dd51fcead7be093cbf9
d4e025117362718a9077209f445c9c1aa780bcc8
23595 F20101112_AADEIX drozdetski_a_Page_087.pro
f504601c7c56cc0a1e56004e4e25f00a
cd4b0414ffca70b1e388ba7985578bb2e35a99db
71843 F20101112_AADEKA drozdetski_a_Page_122.jpg
f7963746acba8dfe2a8c3c3800cadff8
a2a314ca5f5bc830e9353cd312faf9b801f19d10
4918 F20101112_AADEJM drozdetski_a_Page_121thm.jpg
51f83b85d8c5d784a3c61458c9f00fbd
11bda7a697af0c6d1087781093c98e22475fb118
19960 F20101112_AADFND drozdetski_a_Page_130.QC.jpg
68fb992f3862815c25a6678e102b1824
ac996e6cc38b3a031b6b86982280cbdd4f3eae9e
28356 F20101112_AADEIY drozdetski_a_Page_100.QC.jpg
b5a3a3b8ddd7e076d431547ce3831d53
391da9ba8930c9cf2441a87fa6a3d39d7f1fcdc0
F20101112_AADEKB drozdetski_a_Page_082.tif
f26720cb26eec6984a3162842c79f4d1
89a23df62a3cf8a44e99d3c5fc91a3f37e9d72e6
4979 F20101112_AADFNE drozdetski_a_Page_130thm.jpg
97f67a0fa31e33e6a92b064ce1985a72
e3c767c163c9dcbd006c1cf3b75c420cc7ef57ed
3962 F20101112_AADFMP drozdetski_a_Page_117thm.jpg
60e1ceb85a24efa13797c97b489fc812
fb2dc70541d755dd4939e8ffa5579540a9319fff
56114 F20101112_AADEIZ drozdetski_a_Page_075.jpg
afba46e28131b8f916039d57e9ae9028
f3338cffab5e9128cbe179eed5d0ee6a0cab3afd
63267 F20101112_AADEKC drozdetski_a_Page_132.pro
813f0251d30ad66aecbf73141c3387a4
7908f35d7207e6c16504072859e6960cd28e8cf8
F20101112_AADEJN drozdetski_a_Page_068.tif
b3847323816f32f3ceba0e74f9695f93
a981d6f604ef2b1cc72515353292a7dd5c070d81
23816 F20101112_AADFNF drozdetski_a_Page_133.QC.jpg
6456b43e9cb65a5fc9a147b11ce7de37
dd44cfd54242f10397a785009b95574fa4160537
16611 F20101112_AADFMQ drozdetski_a_Page_118.QC.jpg
891bd991531a7dadc98d7f92e6c99220
9f49cf19b3d91c66ffc4dfe99cfe6d7285469243
F20101112_AADEKD drozdetski_a_Page_127.tif
82f976e5b1cffbd0ea63feda39e4319a
6d3db38f6cdedb6ce52fb4de8a7ffb3a0db0faeb
1051981 F20101112_AADEJO drozdetski_a_Page_011.jp2
2dea94fa6c5e2aec40a9b6665052c024
3a3c90af94b530103b3f1baa7611da3c4f03a5dc
6346 F20101112_AADFNG drozdetski_a_Page_133thm.jpg
41c291a48f889e5e8ba5fb73ab143a4a
82f8e5c90edac56cd217d75ec614633d11988f6f
5002 F20101112_AADFMR drozdetski_a_Page_118thm.jpg
ecc4d5f2d56dd9602ff1f27ba2dafdb8
b0739412ff678d0f04ae655172abe54e9f6e872f
F20101112_AADEKE drozdetski_a_Page_001.tif
3fca307f05698c1163976895b3cd2f02
d9916857c87759b45ff1791d43db3a7579336d41
23522 F20101112_AADEJP drozdetski_a_Page_126.pro
ffa554773aca723c17b9a523a084d071
df5230acd7f985dab3246d976ca9f50a24cbe96e
21937 F20101112_AADFNH drozdetski_a_Page_134.QC.jpg
df20549ce0de475e75f353025fe4d52e
ade990e6b2cb88b45e5ff8e578511769dc2da0bb
6571 F20101112_AADFMS drozdetski_a_Page_119thm.jpg
9b0b25f28ecd026a1966b7967454f599
c0f04b80c7392d3673592bc9de84ad7d0d4a5a41
6668 F20101112_AADEKF drozdetski_a_Page_088thm.jpg
2af5f51e9dc98a0410cfa22cc29d28cb
200e150c320635fdcf3261d342d2b7440922a39d
38545 F20101112_AADEJQ drozdetski_a_Page_085.pro
d12487ca2bfa149efa4514ecd7cf215a
28ed1f817047c29a6689ab44f40b01d7bdccf8e7
5709 F20101112_AADFNI drozdetski_a_Page_134thm.jpg
587ac503d1899320ac5cdb31f685734b
ca14c81eb6acc94b275f8d8447e0a75b46bd303a
23651 F20101112_AADFMT drozdetski_a_Page_120.QC.jpg
7324394ed69c8db649d53d46c39b6ce4
e8e1a32efd740e8c7776f841a2a60f0c1faeb894
1173 F20101112_AADEKG drozdetski_a_Page_077.txt
2b18deb042c7cf99f7ee18eebdbd380e
58b353debf1c67d834a860a459e58532556f7ef2
9869 F20101112_AADEJR drozdetski_a_Page_003.jp2
5caa802581b82c8b2113d552bcbb8b0b
41950db213c559169bcb1a02a33afbb74c06c058
23596 F20101112_AADFNJ drozdetski_a_Page_135.QC.jpg
544ee18dd3d39196b445d42fde20fe81
28ff6ce147ef236cbc1628aa2499753da633f52b
16438 F20101112_AADFMU drozdetski_a_Page_121.QC.jpg
b5d31bbc2db9ff7f2a8e4397ae91e504
6594fe9ffe758dc20457faaf15c0a1a6368e3acb
5920 F20101112_AADEKH drozdetski_a_Page_004thm.jpg
f6f5c5cd11318e8dda9c3a098910da1a
0469ba8ef0f978ad8adb2e1b3098210b6ee3baba
F20101112_AADEJS drozdetski_a_Page_107.tif
eba9a4b4572fc0cb4aa465d41e5b6a9a
983913aa757d3fb4ffc5b53ab1d84596defa69b2
6390 F20101112_AADFNK drozdetski_a_Page_135thm.jpg
c0cb7d1295ba6dfb5964888362fc60d8
e141b62fbaec54732feb98d198b3d2f60146b788
16021 F20101112_AADFMV drozdetski_a_Page_123.QC.jpg
d46e2087648a3c2ebd8b40840d030ab5
2b6a0ff8efeeb723076ae86373ad0f538c554b98
78721 F20101112_AADEKI drozdetski_a_Page_029.jpg
d028ff45b495c7db3d5b77a12098dec0
75530d5b22d833c67a6e9df84bf82b6cb5c583f3
25510 F20101112_AADEJT drozdetski_a_Page_088.QC.jpg
61719b3b1ee38bda181d3defd6528548
93058e35f64e1067a38188572b6d6720483960bd
23101 F20101112_AADFNL drozdetski_a_Page_136.QC.jpg
00f0a1d52c2efbff8bd497f89679c6b8
bbf6e1dc25384bed893967bc945fc86074a81516
3676 F20101112_AADFMW drozdetski_a_Page_124thm.jpg
be1279dc78099508d74f2bd6a54e21c3
8d70bc3c341045bd30ea241aeabb6b307d1aa2eb
6147 F20101112_AADEKJ drozdetski_a_Page_122thm.jpg
2d5a75d382c11442d7d0034513725034
5b376a42b2839eef217e14291f2d3c89234c54d3
1821 F20101112_AADEJU drozdetski_a_Page_086.txt
b753a1bdb9efc32d6940b9151dbfc375
d9ce47092080c2ea8b1e3ff40a3e0d46e84a053c
6246 F20101112_AADFNM drozdetski_a_Page_136thm.jpg
77271fc7965564b5d0a940a5e3f607ea
e5ed3a5fc283cc34262691812cf52cfdaa166c45
10217 F20101112_AADFMX drozdetski_a_Page_125.QC.jpg
18102fa2513836c41fb6393e3878a339
6e879def87b2bfc24a3bd46e87fa1b8df4609548
3457 F20101112_AADEKK drozdetski_a_Page_125thm.jpg
d884f332fefba7a519d1b7a42fecf78f
cb313bb54560db814032c008dd704ec7f60129a6
2134 F20101112_AADEJV drozdetski_a_Page_083.txt
c8f2d8ac49c3c0c8dcdcecb75c41e94d
81397892aba2de76c2693e792bec068ccba0ca97
3066 F20101112_AADFNN drozdetski_a_Page_138thm.jpg
6b7569755749a3fd5b82d282f8dcde81
d47ef1cfb013812c7802ef09b3accca625f8d007
3097 F20101112_AADFMY drozdetski_a_Page_126thm.jpg
ca8e8c6bbe2efd4a4ed3378a3212e2d5
b774d538bb9e2f8046be63533bbb3c69caf15b69
7150 F20101112_AADEKL drozdetski_a_Page_103thm.jpg
fb1610b604c8d793874e4f9b1b0b0f59
5627af2a7d288d0490294160cc209e0049199cf9
6726 F20101112_AADEJW drozdetski_a_Page_045thm.jpg
40ce943f25764067a457c088f43a5118
886b754c075d0b43ef82d531b674f4b603ea6add
22721 F20101112_AADFNO drozdetski_a_Page_139.QC.jpg
8b30e637f15f46a8ffe921edb96a0901
c6917bd37e1518728453ef335602f343e0f5c70f
12383 F20101112_AADFMZ drozdetski_a_Page_128.QC.jpg
8c4a14ae666f81b77e5e1b9d3ba71551
535518dc19473c05f759afe7cbb3ff638af98985
1051974 F20101112_AADELA drozdetski_a_Page_120.jp2
2c455700dab574c4697c0b355e343231
5a37291bc598bda0cfed7613c95d2ed23ba37f3f
30696 F20101112_AADEKM drozdetski_a_Page_087.jpg
aee2b122b21c27e356d3d1da010db299
19e82192974975aa21f238b041af2a86859b057d
1051934 F20101112_AADEJX drozdetski_a_Page_028.jp2
cad7cb4f13805ab8052e6d76302a0a56
e0b44a10c0a32021bf6501531b1d9398bc57ec27
5914 F20101112_AADFNP drozdetski_a_Page_139thm.jpg
f3a2f3af8775e20189df9b7dc3440093
3c5b2d7117472d31fcee4c19ce5bdc82851506ec
18083 F20101112_AADELB drozdetski_a_Page_127.QC.jpg
9ece46fc407f95e23bb180bc5f1981d2
c1fb1408acbea19185eb2311263cf61533064c95
988 F20101112_AADEKN drozdetski_a_Page_113.txt
1a38cf7c0e0f7e649257a9423348ce0b
fc7f29f74d4eb2738a7e0f0447499d6993d1f422
5622 F20101112_AADEJY drozdetski_a_Page_073thm.jpg
97413949800afc5848e3025b1ff000e8
a8f732c98d409c1a50d3cbebd104a34ab2a48649
15591 F20101112_AADELC drozdetski_a_Page_113.QC.jpg
f562665057a5b7b577b15105740e660a
8467044e82ccf75b0eacfb899ea912361c1fa111
6900 F20101112_AADEJZ drozdetski_a_Page_095thm.jpg
e8863c76e6b6f68f93b41caaec7c1406
98c333f6e792042949194d9ca5233e4bb2e763ab
6913 F20101112_AADFNQ drozdetski_a_Page_140.QC.jpg
7091a9e67ba4ebb8433b89ca5d885628
1c85f77d46b26938b9402abbba814f57aba7a5a0
2517 F20101112_AADELD drozdetski_a_Page_119.txt
ddc0f514613fe76f70a6b21cf86606f4
44bf1d9d0ceee446973b6be7916acc21d442e6cd
F20101112_AADEKO drozdetski_a_Page_053.tif
c0a6dae2c44d85f9d7c045f1fff7b47d
93e1f5400a91d754548e99ae9cbd6598aa989ca1
164854 F20101112_AADFNR UFE0021068_00001.mets FULL
f158b029e0db301195dc91be5f348827
55124df843364260e4d2354c1db8037c6af9edb5
61607 F20101112_AADELE drozdetski_a_Page_103.pro
2475e5d90929bafd4198924364a33ee7
0a46640fff74f0556ae03b3acb0448673e3d44c1
76759 F20101112_AADEKP drozdetski_a_Page_101.jpg
de601381d8485ea25d1014d4b652a236
f279ee7941e18ce3c9cf3c691961ad09698fb103
2203 F20101112_AADELF drozdetski_a_Page_032.txt
d3264e892197982e6d834a5e029ea22a
bcf6b893b4ec9fd6c36d9f32448957d4d598f462
1051932 F20101112_AADEKQ drozdetski_a_Page_043.jp2
ecd5d7e86af9b647f0a8b0cd5a7691b1
b9e37cc8fd5fe449c19efb4e63b86a7ae6b080bc
74421 F20101112_AADELG drozdetski_a_Page_023.jpg
7aee94bba04d4ce549435b8d17dfffd8
5823e5bebf6e506b835d4d4e61e21b7f8dc74683
4191 F20101112_AADEKR drozdetski_a_Page_123thm.jpg
720e7fddd17050e8d451359ac8fd0571
1e57932a55296305844129daad0a6788cea46f35
12163 F20101112_AADELH drozdetski_a_Page_009.QC.jpg
2e0de8b211f09958e3322f1f49e0be8c
16e42156c635f712e3a4d2c7c5a34dd1f7adccc1
1997 F20101112_AADEKS drozdetski_a_Page_092.txt
d97de108cdb7c949378b936a9b77c97a
107efb7e7dff72c3e3b221201a6cd820b0f117fd
835334 F20101112_AADELI drozdetski_a_Page_096.jp2
deb24bf63613746f94685f77d569cf8b
9a5d64f05d10522c504538bd22975b468b3355a3
5584 F20101112_AADEKT drozdetski_a_Page_006thm.jpg
6306f2c0f1a39cda71927d5e533e8317
6c5f67de4cc2886ae49a190714bbf4571482199b
F20101112_AADELJ drozdetski_a_Page_089.tif
0590055233dcc8579f366fbd3edb2d8e
2a138f1dc49e79e3add2b991760a06beb360be0f
F20101112_AADEKU drozdetski_a_Page_080.QC.jpg
48773d0b016df49472e4ea25e01e46fa
c859a6b6601c46a495b3dee6d1d30030778eaf2f
2112 F20101112_AADELK drozdetski_a_Page_051.txt
fb0fc1533863c216cd311994279cfafd
091c2388c537e7a10d3c0733017682442680e21f
14135 F20101112_AADEKV drozdetski_a_Page_033.pro
84389b742831526bdce073273eaeb3bc
11ee89e7d6bb27cc6ee96026226e2e3d389588ba
13001 F20101112_AADELL drozdetski_a_Page_124.QC.jpg
95d618d75b3f7e97e8cbe90c38de5952
0d49d53995b053e2983a080db5b98f0f92e6ffbb
74357 F20101112_AADEKW drozdetski_a_Page_130.jpg
8f17f272439575ed310bd76172dbe9ee
1438a3433c578492142685899c669f15cf66ef4e
531 F20101112_AADELM drozdetski_a_Page_059.txt
a28e90ec40cd7d52e7a7487f4517bee4
79044ae050417ac6de2d0811c031f913a284c002
97524 F20101112_AADEKX drozdetski_a_Page_061.jpg
6b55f7695314f540b94576a8935f4da5
0fdea9567aa49a01d5a98fef61d97e89aee7f492
1051980 F20101112_AADEMA drozdetski_a_Page_103.jp2
195f11e3b3b588ac498bfb7d507610c4
e3de3777505b0319fbe421991991e5c8acf0da41
6888 F20101112_AADELN drozdetski_a_Page_052thm.jpg
a42f1c82feafaba78abeb7e7f92ccb7f
eb241e9c88974f958d58a457963594b4498cd168
41798 F20101112_AADEKY drozdetski_a_Page_128.jpg
2ab5b7ab444efa844123f72e66ab21ef
43c6087ee712207bf4cf1892719c2ca6d317ca6f
18695 F20101112_AADEMB drozdetski_a_Page_073.QC.jpg
44e08e042443ba3bd49828cd39def5e6
34d236827cda88f989dff57618806a9bac8b3ea4
24852 F20101112_AADELO drozdetski_a_Page_043.QC.jpg
a4bf4edfc8bb8d1820f645d0fd91e43e
a4d223295bf51d6ff21ccba845d1708866e57460
22113 F20101112_AADEKZ drozdetski_a_Page_082.pro
4ef8c2691fe145d0abdf17ec21902f39
24187a56c08c26a5d6eaaf9824caf71322c7679e
60941 F20101112_AADEMC drozdetski_a_Page_135.pro
b52b87423369eed952cde6fbb80524a0
471644916c49e1eeb510f11d6c3395cee796e81a
2333 F20101112_AADEMD drozdetski_a_Page_100.txt
dc52aece660c4622d7c2eea407274ca5
e0d112467d73b4687b5dd0e6bdc3e9fd3e6e56f8
F20101112_AADELP drozdetski_a_Page_024.jp2
8f05d2e8c7968842b53d45965a2a11e2
a138f8c5bd4edc6273be643f5f90c4fe6cf32db8
82761 F20101112_AADEME drozdetski_a_Page_092.jpg
aca767af1c0fdb62db4e8497715ecc69
ba714eb379fe6779036d138ccbc672cdc2ce76c8
6341 F20101112_AADELQ drozdetski_a_Page_019.QC.jpg
436e15546e9409c2301ada746db319b2
5cccedf3d366076d14d0e327fe7df5eee0feea8c
F20101112_AADEMF drozdetski_a_Page_025.tif
e9ff632e40d7c69a1b106958364b329c
24947ed16ee87c61535f67e284a0cbde6261c157
5370 F20101112_AADELR drozdetski_a_Page_115thm.jpg
4acb97050ba7da4dd23d2bbe72c2afc0
f61806bbe1f894f5cc57be29c2d94e7f4f9473ed
F20101112_AADEMG drozdetski_a_Page_037.tif
f7a6019cc7255cc1f03791fe47e4e172
ba6bf0b6848fb5217b71dea3ffc51fcb94542b1a
114596 F20101112_AADELS drozdetski_a_Page_066.jp2
03101272d80fbe887dbccd6ef2797a42
ad481a5ad2ee62749fc43bbc79eb3e0355ddabbd
29745 F20101112_AADEMH drozdetski_a_Page_014.QC.jpg
98a332d468f935e1b8ac993e4a276343
82dea993bc2d79f88cdb50c2d33a94056c856fe6
1051877 F20101112_AADELT drozdetski_a_Page_130.jp2
724d81fd6bc1952d8a753529faa69b34
81661ab73c990fe3fa6e3753f7426f6e98dcb5ff
53495 F20101112_AADEMI drozdetski_a_Page_035.pro
119a221fb06a5b07f04d2f1543f9d837
d520fc07efd4eb4283d512dddd826206d066d342
54077 F20101112_AADELU drozdetski_a_Page_032.pro
871813a5fed926c1368ec1b64a97299b
b7bff6249edff9d434d7d0b254fc60ab1292935c
88616 F20101112_AADEMJ drozdetski_a_Page_054.jpg
0bb7d81c7a420c95f333cea1e343f076
49019d299eb3523771bdc46d99434c2cb709896d
F20101112_AADELV drozdetski_a_Page_020.tif
5a08e64c3e663a92846a7cbc87939cf9
5844546d1337ef74c66f446df4467193915b2676
10086 F20101112_AADEMK drozdetski_a_Page_126.QC.jpg
75bffcc77dd09bca7cbe3cea7a24b472
93b2d896e17522cc60d95e743b4e99e57e67d1d6
7048 F20101112_AADELW drozdetski_a_Page_039thm.jpg
33ad3b8fea13de6f6ebcd61d5ae50c6c
e7a5ea6f05239f7d2e4d1e5b6e059a134e1fa605
1051957 F20101112_AADEML drozdetski_a_Page_105.jp2
651dc46579399b6336b0b1139da4afcc
27d37fd40d0b9d9238f18dc0b06a71de53ac53b0
42241 F20101112_AADELX drozdetski_a_Page_106.pro
0c7a38894873d81b8ddaa564a722e960
1cca45d818aaad602602b99420a3ada3114d4d9a
26073 F20101112_AADENA drozdetski_a_Page_001.jpg
5780ac8d4c34bc2e742dafd76be9fd3e
ab30194b36d5afab07b3ce511f8a1301c412592c
47224 F20101112_AADEMM drozdetski_a_Page_083.pro
4fc58a48a2c594ccea19214d8778c04b
8558e53c75711a6cb092b63cc06bcdc9cfae1144
16600 F20101112_AADELY drozdetski_a_Page_015.jpg
0c6aff2d5e9080930a90a0e8197e73e7
f3bf411e14a5323812f9c93ab815fbb6d64c9ccf
10467 F20101112_AADENB drozdetski_a_Page_002.jpg
56f9c0581237831e0542c0bc8926465e
87f5f97749d824d4b75dbc1929342184abf2ba7e
68477 F20101112_AADEMN drozdetski_a_Page_013.pro
b9250bb1239825ac3dc58576c1a21d7f
96869347ee8d88d0bfcdec6adf003f1e1187b5b4
11863 F20101112_AADENC drozdetski_a_Page_003.jpg
7cc79a9196b8aa456a844e908cb2ebce
0f4e6de4810f5fda5ae1943ff28b9a283c3e3d3e
F20101112_AADEMO drozdetski_a_Page_029.tif
84c04907f5278da6c48b0e4a38c48112
603dfef16f8f3f36741cf75ef7e71da74d58bc93
5916 F20101112_AADELZ drozdetski_a_Page_098thm.jpg
fe5d94c75a8b19707af525e80695583e
19cb19bb56cb7a2afce8169312814f0eaa86f442
73899 F20101112_AADEND drozdetski_a_Page_004.jpg
f0f9db734dce47cf1cfe71074409cd5b
32818c67a0c9e0ee27e9ebaf0600d1a229a2f5b9
110595 F20101112_AADEMP drozdetski_a_Page_139.jp2
f3b223bdc7bfdde25926d2cb8c6bf996
a18889304a953fcad82b4a527e90a55c4f38deeb
39501 F20101112_AADENE drozdetski_a_Page_005.jpg
24251850c10cb70aa406efe936394800
0ee9c659f9e749a5890caa82d3b05b709a288d40
97063 F20101112_AADENF drozdetski_a_Page_007.jpg
25711b4637cbc71cc3b8b21bfc0a628e
368f4252602f6055a1c08100f5081ba68975ea5a
89296 F20101112_AADEMQ drozdetski_a_Page_091.jpg
27e45847c75eec41fab5e93512b85407
f0d8a4fef97835c1836a7e511f7ec3650f419c0f
33563 F20101112_AADENG drozdetski_a_Page_008.jpg
81a2ab75cefe787f2b083565aa55740e
62c6e3e0fefee38a32af81cd8b3fb4007926a8d7
25138 F20101112_AADEMR drozdetski_a_Page_119.QC.jpg
cc579e8e948261f12bd999b692dc36f8
d77466df7c388e009f61beac3d16f49c1416cc79
44324 F20101112_AADENH drozdetski_a_Page_009.jpg
78c08e8e4a0befbb07e9691732f13d77
b5903be536bcff09ac307fbd3e951f7a82eeba0e
6088 F20101112_AADEMS drozdetski_a_Page_131thm.jpg
b294bf62ae777a467962510d34f3b45d
582b86fddaa28e4a8e2f92ab36f1de78380c1d58
95411 F20101112_AADENI drozdetski_a_Page_010.jpg
0a8bb534bf0e0ac3f2137eb290b478ea
6be10cce8d9eaabd4a75582ce2b785e95e935580
71544 F20101112_AADEMT drozdetski_a_Page_137.jpg
baaefc2cc0c41e505f1c81ced26b8593
d2984f0ef2e5c551d4da9918fa2dad1fc36cf448
100052 F20101112_AADENJ drozdetski_a_Page_011.jpg
c41b7e8af3c01d77d650bc23b8024cbc
d57552d5f8bee62668b81035f25cc656f0cf981d
1551 F20101112_AADEMU drozdetski_a_Page_003thm.jpg
bf592b07a1b9e88effe898190253ef5b
cdc28d1baf394ddf3d48c96d95c4233a57060164
93353 F20101112_AADENK drozdetski_a_Page_013.jpg
5c6f6c6a98c7eed5fc06a1b86bd30837
fd605842d30cbd8d89da6321f087cd16dcd19143
1051965 F20101112_AADEMV drozdetski_a_Page_100.jp2
9a6c9f4b29f1de546f53cea75fc7aef6
d8d36827fb049839e14ec08f20789a57348fb63a
105072 F20101112_AADENL drozdetski_a_Page_014.jpg
9935470b6b64b15832bac27bf3b5546b
c11a334c94d0897eb1760eb9f41b72d5e9672525
632 F20101112_AADEMW drozdetski_a_Page_033.txt
acea86cfba7741c9d5afd0496c64f3fc
d081328dd7a93da1a9e1a669d221e880a658aed5
13495 F20101112_AADENM drozdetski_a_Page_017.jpg
717f132676d2b1ecfa509ea84c6eb938
a2a0ad97ab588ecb3a763b03dcd5ef0d0fe90272
213674 F20101112_AADEMX UFE0021068_00001.xml
f7374987b88ff7ba01ae5798c4887bfd
198f5aafbfbb4bf1e1801f2dc9bc4d2c8d59e769
41556 F20101112_AADEOA drozdetski_a_Page_034.jpg
982bba2d53e0a6e0753812e011cb9ecc
f0a8da3b184d7eb6de68769f2997b079fafbbd8e
66558 F20101112_AADENN drozdetski_a_Page_018.jpg
c8298f328c7e0900fb25d1f717e5748e
038931d57dda03b63dafc58e8afc176740d795fd
78825 F20101112_AADEOB drozdetski_a_Page_035.jpg
573e1295ec9643955d7b2495df435fe3
01df48a602b99779d7adf56e1085d7ae7b53199e
19930 F20101112_AADENO drozdetski_a_Page_019.jpg
e6ca7e10584ff708b42afdf671778158
7a4acda63d88ef1b0512bf4d0afbf15fc0f7b964
84460 F20101112_AADEOC drozdetski_a_Page_036.jpg
6fa31cd2196b1b737baccde34b35da4e
5abab65b48f17963f90cb554508fb7a75e02a424
68648 F20101112_AADENP drozdetski_a_Page_020.jpg
45376e4bbff385a53909ee4406d7ab6c
9cd65524ed6f5b810ba58ed27e54190ad1a9b553
94858 F20101112_AADEOD drozdetski_a_Page_037.jpg
7f5a2624724095e3262e8a80e9108db8
79d69f5ac6ce1b5c5b1862e50180c1d416766d5f
63447 F20101112_AADENQ drozdetski_a_Page_021.jpg
6eba9f52fdf2903f7b7597861df6bb1e
1e952ad6d6e4ac932e3bbaa3e58a0572e6103811
76972 F20101112_AADEOE drozdetski_a_Page_038.jpg
3bfc72bcb8fa1069715f58c16d5bbc3c
50aa066159e1db274d156bb3cd025c4d97da7108
89473 F20101112_AADEOF drozdetski_a_Page_039.jpg
9f015ba169af1fa7bcba15d87e3c1c49
10a6e88ef80937388705c23762f5ca4a73c1b416
75254 F20101112_AADENR drozdetski_a_Page_022.jpg
721c3a0f9a6a3610f8d480f88e5deceb
d112af007ca1b73e1116a357ff937ed439209e36
89201 F20101112_AADEOG drozdetski_a_Page_040.jpg
3fcbe1bab7b22bdcac52cd064f57741e
05a4bb2940fe533ef49bc888838d47775186c3ac
70132 F20101112_AADENS drozdetski_a_Page_024.jpg
55ef1b56c88ec796e0d2fd8ad9058e25
90bf566a2b656f2b291beddc7840a52c14de9799
81738 F20101112_AADEOH drozdetski_a_Page_041.jpg
ca4f47c89f1c5e86246ac8da278a8057
7894f12ef7d8a2fd13bd4df89df4002e248e89a6
72213 F20101112_AADENT drozdetski_a_Page_025.jpg
6d25b5f835cff31c6ed2156770986e62
5b7b79332611a8dceb115a77c3e490eef02040b9
80734 F20101112_AADEOI drozdetski_a_Page_042.jpg
110c2bd20d6c398b3ec35fb049918388
1fde67187c3a521a935599720189f9c327c9ea1e
68237 F20101112_AADENU drozdetski_a_Page_027.jpg
c6fa24cf40bbf4f2360be2e28b224836
f1bbed4e8cb25021ca228d66221c6b157f8f9ad2
75768 F20101112_AADEOJ drozdetski_a_Page_043.jpg
2db2536a00d7a423c4ef2afea60983e4
eb7160f1efda9344c04e9375d58af18f6b9d465e
81641 F20101112_AADENV drozdetski_a_Page_028.jpg
28a3712e8140e37423744085bc08c035
0e71d430928b41bdf9ddca3f2f1ef75c9aaf509e
84625 F20101112_AADEOK drozdetski_a_Page_045.jpg
e100d1822a21cc48d2652a8ebd96606f
1f0ae99720d1905aa194c3d2ad76cb70c1df4965
81500 F20101112_AADENW drozdetski_a_Page_030.jpg
0a056e67c44084a1ca872b7582ac0f2c
935064b45b563ae070f7b3c8010a2220fc9545d2
63201 F20101112_AADEOL drozdetski_a_Page_046.jpg
641a846dbd3f25431a61f09d40bd02bb
70cc3fd92e54f9012756cdae1e9710b63ac98af9
40118 F20101112_AADENX drozdetski_a_Page_031.jpg
f23eba8b676250c702bd4c1e87261b6e
be8a04ce0d4fdea81eaef71bea825f5b85baa7e9
78163 F20101112_AADEPA drozdetski_a_Page_068.jpg
e319d9fbd0f9c004e77141fb4d333f39
19758a62f39683c006a8b19ecfe726c9a9024ae6
64941 F20101112_AADEOM drozdetski_a_Page_047.jpg
d6306144c7bb2920dd1e528ac70d9049
ca857258d1682b697b7e83c2dbed5088806326eb
82946 F20101112_AADENY drozdetski_a_Page_032.jpg
fbd4975130751a130d2b78fabff59257
f39472d168a4ef8efb2632ab05350882d4163d1b
71584 F20101112_AADEPB drozdetski_a_Page_069.jpg
aafa9412a2aba7b9ca8535dc1dfdf6ba
4f3f840d06edd08b18d2f626a7832b160d9495a3
71582 F20101112_AADEON drozdetski_a_Page_049.jpg
ef6f3f3bed823744cb0c433155020bb9
e013be8e38097c809dcba4c50dc153e24b2765a5
38818 F20101112_AADENZ drozdetski_a_Page_033.jpg
44036b4c227427c18867bbe2b90f0a0e
fa2040b9bdea18b20ba04a9caf0c2d2c298a196e
72899 F20101112_AADEPC drozdetski_a_Page_070.jpg
55185621d160098dfe1079f4b032103b
426f5918bf64e8b168eafb2381349fbc5c7eba47
72822 F20101112_AADEOO drozdetski_a_Page_051.jpg
b03f5558eeaba03ac111f6daf4bf2821
0d99f96a976ee09028da17fffbe7730b32d58ebd
81913 F20101112_AADEPD drozdetski_a_Page_071.jpg
32b1dbd4df79a82f4b508b7dc5388e8e
d370f480cfb2e8a5efd1e4af5274511c505a2761
86228 F20101112_AADEOP drozdetski_a_Page_052.jpg
4e443cf872fe20e903e0f7546704c328
f6d7ff1cbf655062a2f966b4cf6581840970f744
57256 F20101112_AADEPE drozdetski_a_Page_072.jpg
dd3db267216a01356a50a66b08506677
f7ed552ca6ae050fd34ed0f3bb97d0916e6aa130
66633 F20101112_AADEOQ drozdetski_a_Page_055.jpg
331f3ff3660205d69de422139e61a887
05696f297593e4c57d2ce0ac20ec3168135c9f88
60430 F20101112_AADEPF drozdetski_a_Page_073.jpg
f920c619d77f3ddb59a9dae371a796cb
50c98c257b19946e8812f541a770508d2e26f0c7
67599 F20101112_AADEOR drozdetski_a_Page_056.jpg
62efbf8890a7e1ee54ac4b71dd0109c9
aff9b53c769b2a0774e25c22329f04080863211f
54946 F20101112_AADEPG drozdetski_a_Page_074.jpg
f394cc3be7c46a8536c5810141e59e62
da55f99513fe05d3a9fd891e14a09de43824ac16
59737 F20101112_AADEPH drozdetski_a_Page_076.jpg
1d76aa72dc2f80bdace8e3805a274b09
a6d91085966dc096b184830b1711f4bf3f05421b
69456 F20101112_AADEOS drozdetski_a_Page_057.jpg
60a52e6ad1dcef902fdd78f843edd044
66e2abd029af19e316966b7e55359a2e281b35cc
52679 F20101112_AADEPI drozdetski_a_Page_077.jpg
09075ffd26ac03ff743db61053859d1a
90e2f573e8f3e40b7b63e2ca407c775afc4f11bf
75310 F20101112_AADEOT drozdetski_a_Page_058.jpg
bc3518be2800d41ce3c2710295d51100
d27aa96364af4588057fbf259057c9fc6150c5ae
66248 F20101112_AADEPJ drozdetski_a_Page_078.jpg
2327f71c5716a93f65637e2273d3d76e
41a4ed2aea76a0fa433599b74842deeeb6b0b911
24277 F20101112_AADEOU drozdetski_a_Page_059.jpg
da00a2b11501191a7689cb559ada5d05
3d4911d7e1eb23270339739bd11fd3b0427dcf18
69770 F20101112_AADEPK drozdetski_a_Page_080.jpg
a9d02909a02d83ffc1d46fd4bd2c712c
793b8dca0572fca75bc2342658dbd832835e7f20
77861 F20101112_AADEOV drozdetski_a_Page_060.jpg
1f8365d804f6fcc098b505aa4dce0978
39490b08318ccac1ba87a0118404140192428802
70795 F20101112_AADEPL drozdetski_a_Page_081.jpg
fa343d86e2611517b427e8f7b0ff5aa7
338c4e28880809cb8fb3f778ecab479cfe2dcadd
84319 F20101112_AADEOW drozdetski_a_Page_064.jpg
8aa6978f7b3cb93a07b552bf6ac10523
03370028437a547aa8ea4d84e9addc9ad87c0cf1
85629 F20101112_AADEQA drozdetski_a_Page_102.jpg
6d406e249c7cc2d7479126678ac5d63d
c9e5a5d90c73ec6d7701794259da19d4b1a7a091
59872 F20101112_AADEPM drozdetski_a_Page_082.jpg
bd4485270b6a05704f9397bd8cf3a97a
adf82bb892d95d366baa508a2d3f8de83cf51852
73738 F20101112_AADEOX drozdetski_a_Page_065.jpg
63208e04cf86a6f765688fc8eaa3dc3c
819d562d2a2b43bca7740fa2bd07d67bbf5206b4
94172 F20101112_AADEQB drozdetski_a_Page_103.jpg
02a83c5258727050a8f4875c77cc2e5e
d993b98e217d81f4ee7ce66b2fa00705b2b6134c
71644 F20101112_AADEPN drozdetski_a_Page_083.jpg
873f91f2e2454a4709835556e30c19e4
1a7270d681f67f45caf49700aaa9b8f2c2fef6ee
73427 F20101112_AADEOY drozdetski_a_Page_066.jpg
e41db3a7752a9adba01da0f261322d81
d6ee0a68a522b7a2057795e5075835bba6d5f907
86014 F20101112_AADEQC drozdetski_a_Page_104.jpg
ef81edb65a3ea6229fe647eecab6c164
1d4ee75a167b8ed1e45ba945bc425aa6e6295689
70466 F20101112_AADEPO drozdetski_a_Page_084.jpg
54f9b323af247af43d56771b736b4ae0
54d1d6dffa48179f8d8fc7345bc051749d7ad62b
59861 F20101112_AADEOZ drozdetski_a_Page_067.jpg
3a6dc0269bfb83952deeb7f0aac74804
1607f4ca9f4e71038acbca6ff208fc3f27229cd6
85746 F20101112_AADEQD drozdetski_a_Page_105.jpg
4ac16454a8d4f9ff25e0f0f17e67aab0
f65c19e8b08791147239e0feefd3df3e52357fbb
64394 F20101112_AADEPP drozdetski_a_Page_085.jpg
26fdccbb015eb617a1735079518bd203
4a72201af6e6e3c40f9c183c56c155297a48dd7b
82081 F20101112_AADEQE drozdetski_a_Page_107.jpg
54a3ecfaf7ee46b75bdf0fa4d9df22e7
83b2df33db543c47eb23a6fd6528197ee5df6142
52125 F20101112_AADEPQ drozdetski_a_Page_086.jpg
b39911267fd282de8d40a9bb4f2aa3da
213f7546f9aef167de463f84c6773560792a0623
87330 F20101112_AADEQF drozdetski_a_Page_108.jpg
b950bbb9b95961931223e0aebe8b87b9
5b464c4b4757ce28bb1a7aa6d3aa9b4202999855
82740 F20101112_AADEPR drozdetski_a_Page_088.jpg
116df5c192e1e105cb8a85c290aea3e2
c1bd7c786902e9c70247e3f06fbd9e5f96d7e5b8
60281 F20101112_AADEQG drozdetski_a_Page_110.jpg
dc18d081f19417e224062a346015a28f
c0b8000108cb6e4dc946f66f98219808bc1b536f
85128 F20101112_AADEPS drozdetski_a_Page_093.jpg
7cbbf6cce7c388471c57ffc738c73327
c12f78bc45525c7fc00b0b0efa4263d3f47f5a70
54565 F20101112_AADEQH drozdetski_a_Page_112.jpg
7e479bcef8cacce27c85d07194a84e57
c80e3473b5258c6dc2dbf2b19baa82fe1b7a4b91
53485 F20101112_AADEQI drozdetski_a_Page_113.jpg
0a69a26a5929a6bfc6e6ff7b61276a86
53df031e5afa835588deeab7cfad07b020869f23
83322 F20101112_AADEPT drozdetski_a_Page_094.jpg
89d07fa63f1a718fa7ca687484f7749b
15d9a0f40d6209cbf02b8fd75570d31cab02078b
55551 F20101112_AADEQJ drozdetski_a_Page_114.jpg
890943e7dfec1467843231e6f690eff9
f0f43d7911463082f9826d9b1a35ec79306df70e
100738 F20101112_AADEPU drozdetski_a_Page_095.jpg
93d2fec06611c16ab304a183cd7f0fbc
803f3d9efe15006481aea14a316ad17a8b1d7a5d
62089 F20101112_AADEQK drozdetski_a_Page_115.jpg
4c5918c4e125a27ac2bbc717b13f52e4
6e0884aa73d4ac33fc6180942d3e66812bf00f9c
57799 F20101112_AADEPV drozdetski_a_Page_096.jpg
c3e684d4461cf6608198d0ae3fd36593
96fed7954cc4574ef1a4f18e4a6d627ff07b1ef7
59669 F20101112_AADEQL drozdetski_a_Page_116.jpg
7a2df38cb9f56ec05d37dede1680cc94
bc7a79755d1723e494d69d080af12a7b19cba8b1
78316 F20101112_AADEPW drozdetski_a_Page_097.jpg
dbd76a906b370142c4f8ccc84ef3c80c
17fd3fd0b697917999c946ff8af0fff32a1850d4
46762 F20101112_AADEQM drozdetski_a_Page_117.jpg
7d06213b2c6acafbd60bdf9176386257
7715420309c546c6c9338c5ef8d3b94231c2ebfd
83659 F20101112_AADEPX drozdetski_a_Page_098.jpg
4cd8afda31fed407495bd872a198fffa
6ec4d559d2ce7c7970d7f67f2341310dcee586af
82294 F20101112_AADERA drozdetski_a_Page_135.jpg
625fc2726de74530665dc6c2555c5527
5b53d70613e01b389daed655856a04b54e9d3c2c
53167 F20101112_AADEQN drozdetski_a_Page_118.jpg
37693add9e73f56321bde1826c9bab63
98f5869d8cf4c60e17eefe38ed63efa3f9c4e88a
86388 F20101112_AADEPY drozdetski_a_Page_099.jpg
071c6ce1c282e59ddd1ea1fd2f2bcf9f
0fb4fa8cbfe03641a10ad05683cd4754342b816b
35213 F20101112_AADERB drozdetski_a_Page_138.jpg
edebedecb1d124b13a2ff83e1143820f
780f3958ceb32bf0b857e7b7e2fd3294f39789ae
80253 F20101112_AADEQO drozdetski_a_Page_119.jpg
f1f50a15bd5b1d5be8cc9675d52ae2be
7cabfcd80843f0a26b850cc43a0b122f47907100
89181 F20101112_AADEPZ drozdetski_a_Page_100.jpg
68ae78e9dc235d8fda24812a6a09543e
8f0b771f1bdc43c599ebe01c9e8175dbc128e60c
20645 F20101112_AADERC drozdetski_a_Page_140.jpg
764ecd7ae321a03a131ba3d09f1285c8
ab2a251e9e7aa21379d57f7db4d4bcdb481a2dae
77635 F20101112_AADEQP drozdetski_a_Page_120.jpg
926d9d4f28c3baed6def031ccedd483a
dc8efd7763a7ce817f07ecdefc55acd7cdb6c398
28406 F20101112_AADERD drozdetski_a_Page_001.jp2
46b52f7fb291b9514bdba3285fed1c31
8cd661b44dc3c4136cc91670f1e0876f5702e18a
56530 F20101112_AADEQQ drozdetski_a_Page_121.jpg
f888e4c6c030ba55472d3119dc200b21
04fef32826f12fa3792a2da2b74351063ad84200
6536 F20101112_AADERE drozdetski_a_Page_002.jp2
e10be34207c681b30ea5af1c1894be3d
29f0dacd0278e7b41759d772e8257c31cc1e9956
48810 F20101112_AADEQR drozdetski_a_Page_123.jpg
d9df454df462071111913f24552be582
f3d5f7edc51d6db60bb6edb82ecd597273c82fd4
112078 F20101112_AADERF drozdetski_a_Page_004.jp2
9ea3d1e4f913408d367b603835ee7f0e
5386daaa740536a2aa559c0481a9e984ef8dd960
47863 F20101112_AADEQS drozdetski_a_Page_124.jpg
94d2251a73e509cef117ad8de47e3321
864e1336dc8b5faca56947edd5f26570c62df05c
1051979 F20101112_AADERG drozdetski_a_Page_006.jp2
73afe54571186148bee12862dcdbf95a
72340a6e7809a241ff8d7adaa5bc31f48f2c6de1
33245 F20101112_AADEQT drozdetski_a_Page_125.jpg
c2daf94bafa34332411305af51485b6e
2f6b9aed47d76b88c7087b75ba102490fcdec86f
1051986 F20101112_AADERH drozdetski_a_Page_007.jp2
9e3e3ec1d5209f638df059c8149fd757
a559b9565b05df1824df7949a05ed551b9ffc367
1051923 F20101112_AADERI drozdetski_a_Page_009.jp2
06e3f324567d7b22fff16fcb053c9ccb
55d89d3b7063f9533549b3b4e6cc40362df3760e
36482 F20101112_AADEQU drozdetski_a_Page_126.jpg
9b1c8366c2380d51d64a69014228519b
4a1c98f39a2334fc2d7b83b22760f4b647ce55ce
1051966 F20101112_AADERJ drozdetski_a_Page_013.jp2
062333612345e2106a9ee755de6265b9
41725e6bbe7e4362757df76db84b0220b93acb35
66679 F20101112_AADEQV drozdetski_a_Page_127.jpg
8435504d4d652cf6d251833ae6014ac4
5560c672273be02a05e69215970881261101e0f9
1051985 F20101112_AADERK drozdetski_a_Page_014.jp2
9d593899dd00b5e10f20f37d937fb84a
6383b0abee79283ac199e2dbbc7e3cfc687c8cda
258274 F20101112_AADERL drozdetski_a_Page_015.jp2
628b0b47c6b1c66c1276772fd550062a
24855174fb8c1fd0e653717849510b60af3bebdf
46685 F20101112_AADEQW drozdetski_a_Page_129.jpg
52b16ec14dc45089494c1d5d27851613
0ad65e1feaa53f02d56b51a7e5d7f9ac1160c091
F20101112_AADESA drozdetski_a_Page_041.jp2
838ba29c3fdefb70e386e669dd9d7f41
d4e14bbb039f51b00f8232fe70fb35788abe7d41
100375 F20101112_AADERM drozdetski_a_Page_016.jp2
26ef263028e1cdddad77b0d2227c2910
839bbc7a56db62e5e74bd0d58a34cb4932c2cf0e
79073 F20101112_AADEQX drozdetski_a_Page_131.jpg
18a9849b0f7d6619a9288b6fff2f6e99
884faa320571f4fb26d685f02d397fe45a8e1d20
F20101112_AADESB drozdetski_a_Page_042.jp2
222eb01ec620998180b1d88d82febc34
19a3cc63ebc1645b93ab2ba29961c276fb4e1031
23852 F20101112_AADERN drozdetski_a_Page_019.jp2
6a7cfacb0094e68e127cd3bf9e31b65a
00e6f9c7c19fa844ede1bd4d3f6ba017ad866508
84796 F20101112_AADEQY drozdetski_a_Page_132.jpg
3e222075095e2e871f52c3f7f8a929cf
3ba6413efab08452e014daf8b51ca4968759dc25
82113 F20101112_AADESC drozdetski_a_Page_044.jp2
7bc7bb47b62226c90d398c793b23af69
9ece46b67d39ff0f109209eab573675e7463b603
927389 F20101112_AADERO drozdetski_a_Page_020.jp2
4984ee6d867ded6567856e6150ad441f
19d74cc31e514e074d325034fa814a6e662767b0
80638 F20101112_AADEQZ drozdetski_a_Page_133.jpg
385d7059c9891f95b2aed6706ae686c1
42117a802941aa429d4c31903253324a1855545f
1051960 F20101112_AADESD drozdetski_a_Page_045.jp2
f48ca4f081d21a492d3b09e36d036be9
7294651ea86144c955bff122ebadf8df1a937cec
918974 F20101112_AADERP drozdetski_a_Page_021.jp2
8e884a63d4cce10662fba9603bbee707
c0e0af0e4cfb2edb6693648138a3dc3a0e5b727a
1051953 F20101112_AADESE drozdetski_a_Page_047.jp2
4de8aea2fc4b36e853829d8c041f258e
a9a6ad92ca3f286caafd1e2eec92cb4c2e54f25d
917632 F20101112_AADERQ drozdetski_a_Page_026.jp2
cff495063cf98dabbbde21e6d5c9f8b1
3063058fc20fdfe657aa377f8a255d09dc0e21d4
F20101112_AADESF drozdetski_a_Page_048.jp2
b608529885f37217c6c1e20ea982cbb8
b4b4b1ddc75c2f67c061a124af3e3feeedb8722d
942914 F20101112_AADERR drozdetski_a_Page_027.jp2
2c60ed3fc6396408d014af37b9272813
40bf0b217b0c2547fe1ec60c624808ca8cf61728
1051975 F20101112_AADESG drozdetski_a_Page_049.jp2
1bb0c2e41295f5a8229133974a0e8978
beb8c80e72e68448df034749fc70af3e34974e57
F20101112_AADERS drozdetski_a_Page_029.jp2
8bd3cd3223de82eebc4e439fff93afc5
c3efaa57813d0fd87d06bad2eb71330239c5249c
116723 F20101112_AADESH drozdetski_a_Page_050.jp2
abb606da60ff229d5df73a59c6ed843e
40ec7c8aea70fe447628eee3100657de20f7eae5
F20101112_AADERT drozdetski_a_Page_030.jp2
13847e0ea0b1379a2dc7a6b990414bb3
3b8d0ac17dbc531c37f8ef43a0270cf375571cd2
F20101112_AADESI drozdetski_a_Page_051.jp2
7a45e69f43c1aa4993d570da092fc4c7
a2bbdc7dfa8689d738bdb3fc22358b3edd9e47cd
1051976 F20101112_AADERU drozdetski_a_Page_032.jp2
c0ee030f56a1d50c339c8c15533220c7
d764e31500ca4dca30f2462d2c5f6563fb19ad56
1051939 F20101112_AADESJ drozdetski_a_Page_052.jp2
3740ad71e175aa2eb71355d166a325cb
19e9ada4be45974a5c36a0e1f22f8a063841365c
887341 F20101112_AADESK drozdetski_a_Page_053.jp2
67c1da1648bf82f0ca8362602b4e1de8
4b68ba5e911fa7d9a89cc552429d0470b59a5eab
859747 F20101112_AADERV drozdetski_a_Page_033.jp2
b34e7a0a1e71cef41bea32f95bb1e372
d2d105d47e92fa46cc76bb0f95f85f1ad837df7f
1051933 F20101112_AADESL drozdetski_a_Page_054.jp2
73360176a3d7bb190651778715f1a7d6
beb7bcbda3d02cbd6b8e7927a9eb6d43d023509d
F20101112_AADERW drozdetski_a_Page_036.jp2
9fbe2a7443589196cd3d886e8d283c8f
b23b833b6b661fa5f86b14d64f884acf6b919d1a
1050484 F20101112_AADESM drozdetski_a_Page_055.jp2
5d1d06303a0c5770655bfe7b82c988ed
f855d18bc8b7e8f5fc1405030f35ac56ea8a7e4b
1051963 F20101112_AADERX drozdetski_a_Page_037.jp2
2d0529d986cdcee42af25805b7f9b82a
0e2811d3ae66bdd2feebcca7c31672e09f713037
814932 F20101112_AADETA drozdetski_a_Page_077.jp2
bd3eebc07b3eb2ee29882cb3bb23223e
d92cba2ec2244e6e337177e8d0fe08d679e650ff
985159 F20101112_AADESN drozdetski_a_Page_056.jp2
13022c0e234cec00f065f4912afa2e60
3bc75434ffcd68f47e7d42f3d8a6867b6c536b65
1051972 F20101112_AADERY drozdetski_a_Page_039.jp2
218f5c18ef77fa3fc8b496501490c86c
47900cf6d3f316f788c9a98b37f18d93055583ae
914806 F20101112_AADETB drozdetski_a_Page_078.jp2
c81ef2395b320d8c1db5315e721811f8
96fa43bb7f6e55ae47b5b72940456f298c4f544f
984412 F20101112_AADESO drozdetski_a_Page_057.jp2
029dbbd1f31659f86bf470eff98a1e9a
2c4703286b1f3ea89733e67001111e54be17c856
1051948 F20101112_AADERZ drozdetski_a_Page_040.jp2
0e8b8ebe797da7df4b02013271ebdd45
de84d00ece6ad0dafe3f08e577d5cb60af90e4fd
1049598 F20101112_AADETC drozdetski_a_Page_080.jp2
31dd1f5a35a18429346d5d426daec4b1
92e2519424778d3a60496d6396ac41ed5e0ba21c
F20101112_AADESP drozdetski_a_Page_060.jp2
13da69be005438ba9289f7bd3575ee96
846e5208b6489a1d1dd5db5bf58c2c9d77eac8c0
F20101112_AADETD drozdetski_a_Page_081.jp2
2f8a57bc7ccdefc27e49519dba385399
7b6281dae04d620ddd86982158337e2092090703
F20101112_AADESQ drozdetski_a_Page_061.jp2
b0350bf4c2c77615cd75537df32bcdde
accf1e98ab6f901e6c182942af080d465a8ec915
1051765 F20101112_AADETE drozdetski_a_Page_082.jp2
f3893cdf0ce9832a45f07cb79e3655c6
458ed1e0455a62cdd1dd32b07a339f750c689d36
1051949 F20101112_AADESR drozdetski_a_Page_062.jp2
31128d02c61fb6af3227c8dc3ac7a539
be076edd9103bafe5fdc0d74dde3e1f973c9fbf4
1005683 F20101112_AADETF drozdetski_a_Page_083.jp2
64485b509ca11e9765d7fe1f3e65bf24
7ce7ca8df34fadbcce5edffcc5b7ef52197f8631
1051970 F20101112_AADESS drozdetski_a_Page_063.jp2
a87c18c571399aaf1fdc09bd9aa345bf
12fbdd0a0695fc45dff7836ba1850848fdeae7ef
F20101112_AADETG drozdetski_a_Page_084.jp2
13f481c7acbde949e9afc908fccc86f6
33a0d2e98c10097d57d4755aa2f5b2f97dcdc45f
1051977 F20101112_AADEST drozdetski_a_Page_064.jp2
efc86073a5cdaf6ae0e95109e89c4e9c
3865c626acb9c1b45fbd64fa8176f70c0573a043
1051961 F20101112_AADETH drozdetski_a_Page_086.jp2
d0169701377d603019b943e94f062f1b
d7d9460ac7035d0d682d3f2d1920dcb93c14f16d
93843 F20101112_AADESU drozdetski_a_Page_067.jp2
632b9909df4b430b94098569e903f01b
00deb30ae1077e08389fc0120ae7b0b21d039cf7
588485 F20101112_AADETI drozdetski_a_Page_087.jp2
6a2106b9c370eed780423ad4860b2d46
c233e7419f859fb9bc4fcbdae29629c7e4a23ff9
F20101112_AADESV drozdetski_a_Page_068.jp2
03f4fb56035f2c33d0aa9b308e842075
618229c483c670fa40360f31d3a02995661ac04f
1051967 F20101112_AADETJ drozdetski_a_Page_088.jp2
ef5d5b9dc68bef03b6df3a799ce020db
6651fe738c62e5add377446b0f517be00aec7aa2
996079 F20101112_AADETK drozdetski_a_Page_089.jp2
0c09955a6186db2e95e2c8b8846f33c1
0b36cbd26b2a65f265010353a86c63bdfd65e6e9
1051968 F20101112_AADESW drozdetski_a_Page_070.jp2
b62eb5d8ee9eb9fd1f2f6dfd64c83fdc
54e15de85a62db7f0ba362ada78b2f0b050ea101
F20101112_AADETL drozdetski_a_Page_091.jp2
7e049e3d03f6b66ec7c0d4d3610e865c
b133274f414091554d7acf803ee1b1b5a3c290fb
986151 F20101112_AADESX drozdetski_a_Page_073.jp2
89bfdb2b130597d88a77299e3910dcbb
8adb0888b265d00a1a3ff08bd6e66f5479689604
936607 F20101112_AADEUA drozdetski_a_Page_110.jp2
380f0026f61b3eb755af6be0c662798a
de51d62528494c1e22ae6e0c09d5d0efcac5471b
F20101112_AADETM drozdetski_a_Page_092.jp2
396f59d29c58a6a485fe25cea382b79d
fab4440bd7cbf591ea505385e8819cd1ccf2c7ef
660402 F20101112_AADESY drozdetski_a_Page_074.jp2
a18ae41dc0802b085bc73b385150bcea
eb9f149d6bd500b8da709cea2d403c689e4f761f
64303 F20101112_AADEUB drozdetski_a_Page_111.jp2
3c8e1032aa83a731e539a49662c2bd6c
f5d3b7c0bdac5df506e59f733c28358c9fb96dc9
F20101112_AADETN drozdetski_a_Page_093.jp2
38492be7ae614f561cbb4c09b17ac713
4e0da9cd0d8028947b3baad06da3aa60e9973892
829720 F20101112_AADESZ drozdetski_a_Page_076.jp2
b16c3f33a3892a7fcecfbb61c903822c
6e4ac6c9d49ea56579229d419ab14142f7d3ea59
846177 F20101112_AADEUC drozdetski_a_Page_112.jp2
0ab17f68ccec1df7a0c7edb19bce561e
de50448ff215459a421e062c0fa243a263bcd3de
F20101112_AADETO drozdetski_a_Page_094.jp2
54e736d7dd24e8972f4ba9d888ba31b4
5db8f62fb287a89aa601ee4c7560c4513eedd6e9
852572 F20101112_AADEUD drozdetski_a_Page_114.jp2
499ad2aa684ef759b91fe7372ec8f681
4fa1a116aace9d135a4c0c84f7440e2fdca674c4
1051886 F20101112_AADETP drozdetski_a_Page_095.jp2
e49b76791f405056d2ac20e8e9699df7
04b92ebfbf8b5ec96032a4fe4458d58f16738c56
914034 F20101112_AADEUE drozdetski_a_Page_115.jp2
359b2b5a7789d974d99688fa6ea0b486
2cc40129eb714bfef6da6afa109a468b905fed6d
1051954 F20101112_AADETQ drozdetski_a_Page_097.jp2
50b511d6b2789efc42d3174036904a6e
b7f090b9103a20db897dcc0a80ccb6492ca9b5ce
860385 F20101112_AADEUF drozdetski_a_Page_116.jp2
8d01cdbd040bdf325e3f5fbb35844ca5
1ab1ac1ed326f3d2bdb114555560187230d70bf9
F20101112_AADETR drozdetski_a_Page_098.jp2
085ccdf7c81b7c20c0917e9580535c70
7015327573dd7dd6e7eea8079f6348e4bddba678
40261 F20101112_AADFAA drozdetski_a_Page_055.pro
82ebce3e475b1d60345cfa4b1a532314
8628ec706ee3cd708e157ec5c02187c963d540ff
726534 F20101112_AADEUG drozdetski_a_Page_117.jp2
550a093ec469e714b987f7682ab6753e
0a745c4dcac9eb6f82fd77a8e374a900ea7d704f
F20101112_AADETS drozdetski_a_Page_099.jp2
0d96d96e499261231993798465612563
db03c1bcf5efd2e9912c3f6bb80bd185f4898344
47753 F20101112_AADFAB drozdetski_a_Page_056.pro
be2414d82837ba1b3eba5bc8c3670512
58bcbde89675bac9957b5145eb6bab8dcb9b8f93
800087 F20101112_AADEUH drozdetski_a_Page_118.jp2
0fc27dea0c12927bf37d8d7873ceb889
28ef449329ef62f6d6a4ae1192f6e752c41ae1d3
F20101112_AADETT drozdetski_a_Page_101.jp2
d17d181e232f5c74b4b3686c09cfbb60
b4108c9bc37a98f78f48057fd0cca8fef3ad3b1d
54364 F20101112_AADFAC drozdetski_a_Page_058.pro
d4e063d2ec1a49ec9229e2a875c3177d
c32d03b8bb7e0f0fa0d67a111eabd0020bda946d
757178 F20101112_AADEUI drozdetski_a_Page_121.jp2
4b873550b57f3baff55506cc2608cb7d
6076d7e19bfaf5d1fe32cecf9e1bd5766f6685a0
F20101112_AADETU drozdetski_a_Page_102.jp2
922baa500464ea212447933b2de65d5a
21ecb0c8f3ffe351503d5854cb106faab2fc5de7
112633 F20101112_AADEUJ drozdetski_a_Page_122.jp2
be23ac5a73c110205a5a36996147529e
e20348298bca6cc5ebf741f2da26250a410b5085
F20101112_AADETV drozdetski_a_Page_104.jp2
4464b2b43937aef7bfffd356deed028e
653e534b6d0d51224ad1e52daf15fa8db3802646
13235 F20101112_AADFAD drozdetski_a_Page_059.pro
3d78884c05c21884654fc30a7cf325e6
c52a06fcc52ab8a1fc9402d8be6b70e2aa31fa0b
74958 F20101112_AADEUK drozdetski_a_Page_123.jp2
630c8b5641b0584ea082aa263158ad3b
75aa5826f7915cf6fe3f58974d498e7cf4387e4a
912339 F20101112_AADETW drozdetski_a_Page_106.jp2
f9a7cc59561eb7555b1c49f964e0291b
3de0b1169aadd70cfd3e21c3ec61ba03eb6de41b
65840 F20101112_AADFAE drozdetski_a_Page_061.pro
7ebb1bd30185303617fb1db125e197db
66dd5c532f4a170947c089c13e96295b520e7820
624591 F20101112_AADEUL drozdetski_a_Page_124.jp2
dbd2ea16a6ebc4fcbd2237793d8f61f9
f6bfb12a89dd04bd736e74231f7088d6a3ce8369
55375 F20101112_AADFAF drozdetski_a_Page_064.pro
c23b2f72d7ccab813d2d353e13f4d527
aa0f1a6aa6a07fedc06a22f9cd9641710f127805
F20101112_AADEVA drozdetski_a_Page_008.tif
e1fccec4f9b37d0d6aeb6e1e4234cd9e
6f7f94209beadcb54f406553ddd8e7cbabb99f0c
403955 F20101112_AADEUM drozdetski_a_Page_125.jp2
1a8ac8017dff97618f4af0c6a63dcec3
c0e1512479372d122719a7f219e4b3c52b36c3c9
1051962 F20101112_AADETX drozdetski_a_Page_107.jp2
ced84b971f92bd20c146eb0fd0b1b51e
d4dcfde969f33a313834fe1bf6a8061fc286f5b0
56043 F20101112_AADFAG drozdetski_a_Page_065.pro
3fa9db9db3bb61a9e339efd027617a74
08ef2d34fead016a95b4018b1d13a0a49e0b356b
F20101112_AADEVB drozdetski_a_Page_009.tif
a363fbc94966afb71b6308b7052fa175
89a9a74635f746323614066c9e5fe7deed8caa53
51647 F20101112_AADEUN drozdetski_a_Page_126.jp2
9627771807558d51a1d2fdcb8088a186
86bc33d469b372d4834b59e856225aef88d7eb77
124314 F20101112_AADETY drozdetski_a_Page_108.jp2
8fa772535a9726fd63469b079338ee1b
248e5c88087e0b27e66a934dd477e000f3e76f2a
55305 F20101112_AADFAH drozdetski_a_Page_066.pro
7c84bfc95a615d1da7e0933a1d73eb25
fa84ef4865303654f1fccc7210b75cfd771f5717
F20101112_AADEVC drozdetski_a_Page_010.tif
90741eeb372826a2114c1c6271995153
cb6649dfc6597b925f17088e885617e90e162f4e
664173 F20101112_AADEUO drozdetski_a_Page_128.jp2
1773edd3a49b61d3113b566187a8b739
d3bae7ac9e14b51640dea11fe81b324dc68f91be
F20101112_AADETZ drozdetski_a_Page_109.jp2
e0feed26fd37fd220b3eb55f2e681b7e
d68b657887de2f05858bdc3390a8f2d73598fa78
43446 F20101112_AADFAI drozdetski_a_Page_067.pro
da03843f15cb42b0e7bc0aaed6f100c9
22e35721a3b7276c65abe381701435ea7f8fbb8a
F20101112_AADEVD drozdetski_a_Page_012.tif
1808aa11e54343728cf4b09fba340e8b
7559f6eba8f689f9edb1b9ed09b5ef3d5e4c9152
65616 F20101112_AADEUP drozdetski_a_Page_129.jp2
f4f9f556b3aaafe6fa5e20d93db608eb
6131e8263ab641067f36c49062bbd5896dc70c21
51889 F20101112_AADFAJ drozdetski_a_Page_068.pro
e3f7cede88a43e1fb75b391e3277f561
eb73e7cb44713c336d1a990165c098b89b1f8cb6
F20101112_AADEVE drozdetski_a_Page_014.tif
e1bfbf6df918b71f0086f8d405b908d3
22f6a399af03c11c4341a39db13da2f00c221e13
124272 F20101112_AADEUQ drozdetski_a_Page_131.jp2
0018560db820f9d61e441b7af1112fa6
6485f98088fe8245a9403cd24650e7a68f761cad
42994 F20101112_AADFAK drozdetski_a_Page_069.pro
b4ecc933ce6435fea19fb3bc06ad8046
99040c415eefa03fa0e396f20e6271539f7bafd6
F20101112_AADEVF drozdetski_a_Page_015.tif
0d6bd915e4f91e1def62c90bc5b12127
f3b910a8db9b5e739091adbbb69a4dfb1f2a43a5
109912 F20101112_AADEUR drozdetski_a_Page_134.jp2
a38f3e067f4e2f6c9ea6929f78ef30e3
95866103f96b068e56396cea5a972910ce1aa229
50358 F20101112_AADFBA drozdetski_a_Page_092.pro
34ab6559482d0acbe4a322333e61f73c
ede573be21b8f22c6f2586533783884476948b0a
38678 F20101112_AADFAL drozdetski_a_Page_070.pro
a1797577f418f952e004f1714482471c
96fbe154685c0e9ca1ba4d22b5038ba165fb3299
F20101112_AADEVG drozdetski_a_Page_016.tif
473d31b8e70ded00934a0c4aef015be7
1ba97ca701eb8c8553b891d8f899cb9d91657bbf
123950 F20101112_AADEUS drozdetski_a_Page_136.jp2
4705eb94dad41b626f40f0ab97b922c9
b63e5f8376df308179a756056d322bd9ad07c842
52727 F20101112_AADFBB drozdetski_a_Page_093.pro
f7356c8db91a23d380c9e82c4c39bed2
56c4fedfb79a0805836cbac93f6a289245b45ee7
63911 F20101112_AADFAM drozdetski_a_Page_071.pro
d040e0ba91aafbad95b4a49a076bf44e
4a7d2e7d255f6a82e31a7705f310162ab250d850
F20101112_AADEVH drozdetski_a_Page_017.tif
dcb2ccd7f328d38ac028e12dfd59abb3
26dce5a68e04b2afb6033ce14f46752851b5e2de
113023 F20101112_AADEUT drozdetski_a_Page_137.jp2
2455f6bbda283de724c1628fef0f7e48
4a5e590d1a542d9d6cea64a431a681061a942936
56672 F20101112_AADFBC drozdetski_a_Page_094.pro
702bdfc1e1bbc038aa80e47ffd3116ca
1fdfaac0a7f9f6dfef70d471bc62cae5cba2b2a0
37331 F20101112_AADFAN drozdetski_a_Page_072.pro
951d05c612c967c44be77ce28762fe76
de112008f3a1d207014fc2e536c6b0a57cbca288
F20101112_AADEVI drozdetski_a_Page_019.tif
b114a2e2e78076009d74fdb0f883b2fd
e79d60b5928e9cae95cb82c87320e47454acb7a6
25901 F20101112_AADEUU drozdetski_a_Page_140.jp2
8872c5b6613812d0f52070ed50dc81f9
67360745716c1de46fe71a2004a49e0d25e58d3e
67443 F20101112_AADFBD drozdetski_a_Page_095.pro
5451235b6bb1402496fa73fbb81b796d
ee8bb9ff857a18df2457634967c73f4e915735b5
26233 F20101112_AADFAO drozdetski_a_Page_073.pro
943007806e5575eac6bb67c8366c01d9
c5f4a76bb26f02c5da7ea315f8f845dff8b1042f
F20101112_AADEVJ drozdetski_a_Page_024.tif
c6fcb0cb7bb25eb2f5855cb8945792cc
7ba3c631d67f58258d73218e1eb698ccddd5e5cf
F20101112_AADEUV drozdetski_a_Page_002.tif
589de76704ba28289329fd4b73b6420d
c5710813a5a26ee9693245745b4d159d6756116b
24475 F20101112_AADFAP drozdetski_a_Page_074.pro
3e6fdedfc6392ba143d4ceb65dad19da
89b602640fc112d313142f86d34c2288329a27be
F20101112_AADEVK drozdetski_a_Page_027.tif
908b01aff8bdf77fd73e51dbb7b1d928
6e39a97f859ca1bf6e21cc683c5886de14c53b3c
F20101112_AADEUW drozdetski_a_Page_004.tif
d0f75d182abb0b8839b6b06e550b74ca
6730162a3b5e006989a8321b809c9c0620f7ef01
37941 F20101112_AADFBE drozdetski_a_Page_096.pro
a0b00a3f3fbecdb54516ba6c4e4e8851
8d250325e7ffdc0a97845a8f762bae26f1a51624
22135 F20101112_AADFAQ drozdetski_a_Page_075.pro
eee6c58c26cec04bc3c14047a2a07042
9d58b4e3d9758ee1fb13b4978da80c2030d25cc2
F20101112_AADEVL drozdetski_a_Page_031.tif
08f4d21ad77a524cfa61e47054bf371c
0e8dd8dd1152ea53c840a7b77b46f3caf388a819
F20101112_AADEUX drozdetski_a_Page_005.tif
e21854c616e77e1d267398fde6842ee7
327585d0c01fff73d63340ce2dc99a240a0eb509
50438 F20101112_AADFBF drozdetski_a_Page_097.pro
7a69abc73cf0cb76092f66e3907321ad
f5e840776b46a3c55ed787018d79603dea13d5ea
37336 F20101112_AADFAR drozdetski_a_Page_076.pro
eab42446c6bbc16271644bc47668596f
189915d698e691a0acbad534433536159968a187
F20101112_AADEVM drozdetski_a_Page_032.tif
585c52bfddc329338db96d2a0232b3ee
63820ee358363a3a8c6577dad3f454c70bc162af
55262 F20101112_AADFBG drozdetski_a_Page_098.pro
5f54ff1b80cbee66b56539fe2f089573
fc8170c4d77bff9ba4b9961a12e0ace73022bbdc
F20101112_AADEWA drozdetski_a_Page_054.tif
efc762a879fdb4cca93a0c27a9d8bd58
fe6ed7c9f600fd2815f4fc87cf73eaab98df63ab
28255 F20101112_AADFAS drozdetski_a_Page_077.pro
574cf58bd293d0d35ce2ad17f2e934e9
d3dcf738295027a8d41aa24201c81dd528d790da
F20101112_AADEVN drozdetski_a_Page_033.tif
ed45acb98af39d888619c2148adf0493
f12a94665b8dc2fe7e61825837cc119102840c50
F20101112_AADEUY drozdetski_a_Page_006.tif
dd0f9cef7d64fbdacb8ac647136ba114
34c0f51cb702d55cb2f1a64dc69f064ba1243454
59309 F20101112_AADFBH drozdetski_a_Page_100.pro
1be6a0e2829ea23796f6f8348ee13009
c8121883d40d3a863609cf74e0ddb2645d0d78d2
F20101112_AADEWB drozdetski_a_Page_056.tif
8628c480c82b21dd0a419f6318bf5f1c
b2a9eb728c6213b3f27373ffdc58e74605ef0df0
48604 F20101112_AADFAT drozdetski_a_Page_081.pro
2a5f314fcc4357885679010dea2dc535
2464142d56bfc313f569013d9b0d2c0eebf6ec91
F20101112_AADEVO drozdetski_a_Page_034.tif
b4a4b77056a7f2e70cab681e24550fb0
0527a24c82bd8f875315557afb913623ed855da8
F20101112_AADEUZ drozdetski_a_Page_007.tif
befbb02fb06659334c6158aa9c2dd51a
db191a814c6c49e6074905d1b19ca81fe8e43327
48465 F20101112_AADFBI drozdetski_a_Page_101.pro
11b1d28d2af6d5ba62520ee183d1a05d
a84fbb3ce65ac2e93f9afcd77a58dc6004cd37eb
F20101112_AADEWC drozdetski_a_Page_059.tif
1395c2d362e7d3e57a7af8c60414f1a4
15d15443940636a6a3b511ef142a529201ef510c
44193 F20101112_AADFAU drozdetski_a_Page_084.pro
c67f718bbfa75ff9818e8e6057cab9bc
1472a88d47969b42060d2978578c93ddfe133cd9
F20101112_AADEVP drozdetski_a_Page_035.tif
1a829fc0b204a62f9b288032d036e26a
b45956c47d3cec364898a86248ae87853a5354a5
57197 F20101112_AADFBJ drozdetski_a_Page_104.pro
957161616d21f635ef087cbdffb399fa
a94ece07c1b5a021159532dc998b823ffae21e38
F20101112_AADEWD drozdetski_a_Page_063.tif
4a8d634394dcdedeb52abfbbbdece493
377c017099d8ed4e62a510a6520b849ceb005f04
36109 F20101112_AADFAV drozdetski_a_Page_086.pro
94fc63724d9718655e8f46dd2f1e7b66
a3d7d69655b95a3041f1ef57c03067b079e28207
F20101112_AADEVQ drozdetski_a_Page_036.tif
dd0b371284f9a8906ea26825398eec84
07a2dd918a984dbc2ae94ff12fcbadea439cbf0c
56169 F20101112_AADFBK drozdetski_a_Page_105.pro
57b285b348034b293386921e0b9b1d36
6e2f0179452691140c1e8332a0352a9767d223e7
F20101112_AADEWE drozdetski_a_Page_064.tif
508cb9dd809c27edead5b8e5962e43aa
b0d6db57c1e3eb768fb5e61a1094df3c38b8f926
52904 F20101112_AADFAW drozdetski_a_Page_088.pro
8275887925cde8ddba68547d12932e3d
f7d4d11330de538908d887b9469f1ab42569689b
F20101112_AADEVR drozdetski_a_Page_038.tif
6cba0e0af226d99e06c2b21b86417d2b
1b79872a437ce0025cf74a3baaf0d94c6f3e7423
51843 F20101112_AADFCA drozdetski_a_Page_122.pro
3249cdbe0008f154cfddc8a4848b8d57
e5995e4cb4c9d7b51fa88947cd745a5a9e78e1b2
53378 F20101112_AADFBL drozdetski_a_Page_107.pro
b6261d8b54ffe6f9f2731a5decebc4cd
a8bdc317bc5cc6dd202d6e1951f72af10ea09fd4
F20101112_AADEWF drozdetski_a_Page_066.tif
cf3278c411e2fa2cfb77563783e1484e
4f100a2c0c9de1d691af018536a5557c86149151
44139 F20101112_AADFAX drozdetski_a_Page_089.pro
e6c18035aee3413d7ca94f68862f7db7
f4d1e3c62c3fcf3db7e888a326003430c2270f65
F20101112_AADEVS drozdetski_a_Page_039.tif
d6cf1883a4afcee8c05c21d4c58ac6b3
d232aaee610337dbe6be1cddc01f0770b3bc54c8
33744 F20101112_AADFCB drozdetski_a_Page_123.pro
ebf8da8e947978d2ad0c2a350dcbd5d9
0d20b112cc1b472bd6a23dafed7f8186e990a82a
64486 F20101112_AADFBM drozdetski_a_Page_108.pro
ae4954237e7fedb42d9f562d5cea1d5e
6d8663a1694b54c0aa1a317ba79d4b9f0fee039a
F20101112_AADEWG drozdetski_a_Page_067.tif
4d436042086399cb701dad77a62125fa
b2502144d8557991230f020b5fe6a310832b359d
F20101112_AADFAY drozdetski_a_Page_090.pro
f5b6e5fc0c42a0b2f660de10c31eac49
055c34b6ef2d773e112e2afdc607393c5d557b61
F20101112_AADEVT drozdetski_a_Page_042.tif
219159c562e7f50acc94a2290cb6181f
16f3dd1f40436abd4d5f047b83dba260e0cc7136
27654 F20101112_AADFCC drozdetski_a_Page_124.pro
26a9005970c612127455fdce64e6ec79
36c06924746e1b9d7a13a489580d034f9702dd2b
56471 F20101112_AADFBN drozdetski_a_Page_109.pro
42af67099e838b540122d44567790abd
123506bf58cfe0a852eecfff49b39401e3149bbe
F20101112_AADEWH drozdetski_a_Page_069.tif
f1e22a9af44f936c5b723c6a89b7926c
44c6cc8d0b603017f01fd181583f2dea25448541
59147 F20101112_AADFAZ drozdetski_a_Page_091.pro
e0858721271c03b117d14ff232e87c96
2839cd69f1a192472969f0e5d1b070cff7d625e6
14407 F20101112_AADFCD drozdetski_a_Page_128.pro
b9b9c5a42744cbb0368b25e6e17d26f9
6ae00941873f32d5c756657c57175df8dd497106
29623 F20101112_AADFBO drozdetski_a_Page_110.pro
c636aafe077ddf16fcba206fbc8f0967
c9ca672e179cfa7170b53bdef84e43ea54116ee1
F20101112_AADEWI drozdetski_a_Page_070.tif
b083191e2fc1fb506f89d76b9e526426
fa845aa95874099b2cf67ffac87606becf65c37a
F20101112_AADEVU drozdetski_a_Page_043.tif
31df6da246e3a6f9523d55bad465f69f
676d76a179d11fa15325eb8129baad96d2896252
31257 F20101112_AADFCE drozdetski_a_Page_129.pro
44f6165bab2b39ce9a816cb1a8b4ccc0
5dec41e34ca296055da4df0560f5a0b401602852
19714 F20101112_AADFBP drozdetski_a_Page_111.pro
d83ed51e6e446ec23d248df66f2b0dca
e052a513d35573afbc2179be7dbe51db0fe3eaa1
F20101112_AADEWJ drozdetski_a_Page_071.tif
c21d715a6d7e6c59f1b56730a4481f1d
1c020ee70307ed9661003133c50beeaa11ec1f45
F20101112_AADEVV drozdetski_a_Page_045.tif
4028a71539157ceefe4ac420dce8793e
9e39e4eb7c1b6e4561290ea355f878b389ddd20c
27801 F20101112_AADFBQ drozdetski_a_Page_112.pro
93196a11396a6a555bca6e47c0417601
5af9f2951b4f7447d07f2f3741f93cbc56728663
F20101112_AADEWK drozdetski_a_Page_072.tif
ca2f5f7db4e11e5c60820f2c9b2019de
f332f3d4663c952f5ae3f653cea5d298d9a63c80
F20101112_AADEVW drozdetski_a_Page_046.tif
88c4c43f320cb8429689ec9150712ff5
38bdb2392bf1ebb916297b33aec0dfb0720ac426
46891 F20101112_AADFCF drozdetski_a_Page_130.pro
f95774d7c29de2b7e63b7f6c368122ae
622f28ce6486e121a65c081ad523aa8442a5d1a8
21575 F20101112_AADFBR drozdetski_a_Page_113.pro
2205c79fe991839c4ad79a24323524d4
d04b1a2486569d87ef3f6e32a1c0840df9eae14f
F20101112_AADEWL drozdetski_a_Page_073.tif
d06085733e3d22e984c2efa44f9fa636
ec4fe7706c936a9fbe0bf96b744f0977016c050a
F20101112_AADEVX drozdetski_a_Page_047.tif
3e0f64fda722fb9d67b7f6a011e0531f
602f8b58b539f16b28860fca82487c9dd0786e59
58107 F20101112_AADFCG drozdetski_a_Page_133.pro
809cfacf44ec0e7c29453ab8861d8d51
507ff832861e62f5aa34eb04b7c72ee5a8487af6
F20101112_AADEXA drozdetski_a_Page_093.tif
26704ee9fd3331d5ef8eb69883409814
31ac12e3dbb15f41df06f101b3252c185e163ab0
26415 F20101112_AADFBS drozdetski_a_Page_114.pro
2d7e15b18ebb4dc06a603f32a5258cd7
994866117575ccbb96357d3651c3dd8e40bd1861
F20101112_AADEWM drozdetski_a_Page_074.tif
15c1667fc098c20f0cb9403f0417a5c3
26bfcba28ea41cf7b789d489d36a3021b1b2ac3b
F20101112_AADEVY drozdetski_a_Page_049.tif
1742a0014c94136e8cea2bee96109b6f
0b2a4a7538265a2e9a9e20dc365b0befe58870f3
51228 F20101112_AADFCH drozdetski_a_Page_134.pro
4befa1890576af9c9f406452aa0cfc8f
1bcbc65b3e29ffda14f9658a411dbf7a323787b1
F20101112_AADEXB drozdetski_a_Page_094.tif
96a213082e5164cb43ecd313c6054507
eb220a27e9eaa6f2071fe2b7094ce64467153fa6
29272 F20101112_AADFBT drozdetski_a_Page_115.pro
1bbf5d9b1af4f454068a2d925591e68c
5c7a6f96380b90144ec8aad80f7d3798ac642634
F20101112_AADEWN drozdetski_a_Page_075.tif
1db1cec1781c8ddd2140aa87f4a2244e
0900e73da85254979f86a12eeba2f79561c8bef5
20997 F20101112_AADFCI drozdetski_a_Page_138.pro
9eb753ac3d378aeb3b8b696cd21c67e1
992662e9775351391f2891c3f0645e657520e788
F20101112_AADEXC drozdetski_a_Page_095.tif
cc9e04c50dd8122a695856211525cad3
bc62597a2759a9323d6a0ecbe963d7d559fc038e
33597 F20101112_AADFBU drozdetski_a_Page_116.pro
b6a41cb4a208a29c7fdc21bb2d370e78
4c471b5af20c3bef4415f53f449d86a76be7c20c
F20101112_AADEWO drozdetski_a_Page_076.tif
163f25514fab1dd433c76af1ed3cc2b0
2bbb3a85a62fd43e3d9f867edce4c7766a5243d5
F20101112_AADEVZ drozdetski_a_Page_052.tif
21e353979d501f4f7af105aad294c54d
b0e7d6b988925b82044cd891a5e5080185b58681
51368 F20101112_AADFCJ drozdetski_a_Page_139.pro
47ca08982def4ef5955096ac434c989b
8bd1aaa71ec5b5fbba32159679925e4b2d790f0e
F20101112_AADEXD drozdetski_a_Page_096.tif
c84d14db685ae7ce0a81c0b17e03fab6
62be9d2f9a8c3c67999c668e50c714a4d04aab03
32534 F20101112_AADFBV drozdetski_a_Page_117.pro
c868f2f1b098ed05b16e76e28735065d
8294e92e26e1bc24b522652dbaa08d91f97c8366
F20101112_AADEWP drozdetski_a_Page_077.tif
ffc63399e7a9011ec3ffd29693156cab
46d0e40132575da340d3d3f1202b47cadfa2c86d
10450 F20101112_AADFCK drozdetski_a_Page_140.pro
6109204f117be0d42f049896d4718739
cf248757b967ad76b156dbc634c35f38c9e20056
F20101112_AADEXE drozdetski_a_Page_097.tif
5a3573c08283de65b5e4eef433c7ae2c
6cc5179d6f6fe77802638a3fe85cfd8356d93d4a
40932 F20101112_AADFBW drozdetski_a_Page_118.pro
7bfcca36891c150db4b5efc17e53a2c6
922d7778719b7d4ac1b32ee27981b56331600e04
F20101112_AADEWQ drozdetski_a_Page_078.tif
b6b42ba55f19ecd9d435f951d843cb42
9376147edc9673bc5700b64de4985d720d473d6b
104 F20101112_AADFCL drozdetski_a_Page_002.txt
de8fa03218f9583ba8415c5043aa1396
84548a2f890bfeed0d284331dfdc621a76c0a5f9
F20101112_AADEXF drozdetski_a_Page_098.tif
05a5e5c55f7b410053baed1718506c1c
eaf0c4a40520e82d2b64c0c37d958d46d3412c46
57184 F20101112_AADFBX drozdetski_a_Page_119.pro
817b13a7b3b52b836902518d20b9356c
833366189e26b651a76086862d77288cc402c806
F20101112_AADEWR drozdetski_a_Page_079.tif
6078ab54179556399b69fe169faa6f39
c7ee4748251ccac699dda5f5867abbdb1299eb3d
1686 F20101112_AADFDA drozdetski_a_Page_026.txt
e8dc0e3dc8237f2d211183ce4dd8e58e
3e152f731c70e87d8788c00f2e2a2c45ade5665e
2379 F20101112_AADFCM drozdetski_a_Page_004.txt
8118a79073a71562a511892415f72ce0
621bc1f0923c5ab9d9bfaff7420e3b2004423efa
F20101112_AADEXG drozdetski_a_Page_099.tif
dd8e74a82c5b0a65219239e8bfc43318
469ce6ef38198a4f0d14635262c507ce4448234a
50944 F20101112_AADFBY drozdetski_a_Page_120.pro
90de64955593c335fac89815eb0b178f
02caaf8eefca21df025d31c616e44407a230888a
F20101112_AADEWS drozdetski_a_Page_080.tif
3cd37227ef5713abb3a9194ba1909094
372f76fc468d1586c6bb194a83fa4d420b1362c2
F20101112_AADFDB drozdetski_a_Page_027.txt
24fdfc9dcb7381ec95d629420e47411a
369fa056ee5bcfbf2786da4d3dbcb0b88884d020
1048 F20101112_AADFCN drozdetski_a_Page_005.txt
7817ff668291c9707272dbec0d5ed112
8410b10892f797440d0d9fcbe232e42d209bf179
F20101112_AADEXH drozdetski_a_Page_101.tif
72972cfb0ceb735fcf3eae26912d056a
fe85f5be1760a696297ce866306f877115562909
24426 F20101112_AADFBZ drozdetski_a_Page_121.pro
4abda0959df954f696ab28609ea53cd0
0cad1639dcd0d4e3d37f98c16b9ee1ea0f66a26e
F20101112_AADEWT drozdetski_a_Page_081.tif
be6ae858f048dcca9e1ebdbdec0cdc53
fcb0b0a40c2a5002dbff78ee27c7e9e0d6e5b0aa
2085 F20101112_AADFDC drozdetski_a_Page_028.txt
bc24741ed37aeaf642c38b0b8ea7c50d
69df8325b3024fd1522fa5e29dcfc0e12cd18ba0
F20101112_AADFCO drozdetski_a_Page_006.txt
49ba39fbf6f0879b0c7a6680f9eeb53f
feab05f53f1d83b05655ba3d862e25a454c7b548
F20101112_AADEXI drozdetski_a_Page_102.tif
8fa83c4bde059319df077517fe5ff49d
2d4fb472b2312bcb2b50f816da70019d93d76f62
F20101112_AADEWU drozdetski_a_Page_083.tif
0231d414dd678e1bd2db83b38058b87d
9bb06911236689ad5276718ad663806dcdfe6ef1
2732 F20101112_AADFDD drozdetski_a_Page_029.txt
360db27c5aa03f48af7a915e4ab58a22
0b1e5181633a4d09dcadc62374942be47db302ed
651 F20101112_AADFCP drozdetski_a_Page_008.txt
f2d7f149b08d92a925b7dfd1cfa57711
02a7740cde1d93878ab7be6604d314fc79de2049
F20101112_AADEXJ drozdetski_a_Page_103.tif
db59f4c5da110caa810b692bea3fa388
17c9ccb755783616c326ba40589f7ac525d67e85
F20101112_AADEWV drozdetski_a_Page_086.tif
58693b0ff1c28ec82c5e6e14f794ef21
2ddae04af304ee60278c651ff3967050f4d7ddf1
838 F20101112_AADFDE drozdetski_a_Page_031.txt
c045e4d0ed9ca3ef4506fa9595094227
254db3e01af41f44b737c9a05421a8393732a351
1226 F20101112_AADFCQ drozdetski_a_Page_009.txt
496e5b2f150a78fdf04fca47e822018c
763d2fc3b31a0570f0f793946e8deed8d581452a
F20101112_AADEXK drozdetski_a_Page_104.tif
b9a101473ad8ba1cfe1d757a27824f72
329ca641cb9ba0e5349b94d54cfb94fb67ebe71a
F20101112_AADEWW drozdetski_a_Page_088.tif
30bf63710e64c8e195ba5dcdd1114f0d
8b67f54e5ba34234f8be734087403cdc06b2009c
2216 F20101112_AADFDF drozdetski_a_Page_035.txt
0dcf59317651076b4870c081132a7d6e
0693e6017e772ffec197ca99db86f7f1bafe5f7f
F20101112_AADEXL drozdetski_a_Page_105.tif
74b6e400ebfdfb1f711a8f7797f258c2
9912d07921415c9229693a768c51d52ed9a08ae8
F20101112_AADEWX drozdetski_a_Page_090.tif
2a43bfef67ac2e60af06cfc5f0693ece
7f56776097d4c3b63eb37c879802cf72c596554c
2773 F20101112_AADFCR drozdetski_a_Page_012.txt
095d950753b4950c3953b9bf024d7d7c
e1b47bc9add09df52da1d58bac7135bb2e40b2eb
F20101112_AADEXM drozdetski_a_Page_106.tif
fd23448a5d08b21517cecaf58e4dbee1
9fe57a6a71278d0bc3039a1850875f5f121c4340
F20101112_AADEWY drozdetski_a_Page_091.tif
bd794dc5606d6c517351d0f3ef0890b6
60bf1dbad72808ee4de8f1999396610977a1dcd0
2319 F20101112_AADFDG drozdetski_a_Page_036.txt
19d31aca99c45a8d1e8de90917f2dfaa
73610f637fd3fc4a16fb9c388c7b3e089668d09e
F20101112_AADEYA drozdetski_a_Page_123.tif
5b19f4c62f5a59e5d7ba21685c91fb97
f639d89e2ec775f4449fb1f4aab749e0ddb03c98
2763 F20101112_AADFCS drozdetski_a_Page_013.txt
fb079a2d959f00cf67f9745b62a993b8
e8067b6cfab1a0fb8d51f9e57c43bcd32892ede5
F20101112_AADEXN drozdetski_a_Page_108.tif
bb828b6acd262f47efeab3d3b3d72c59
5b3e035e8974973af0e48163ea8b6136f7856229
F20101112_AADEWZ drozdetski_a_Page_092.tif
f91f0ac767572409529341270fe83c1a
e5fe5a1e3f3ef5cc8c272b79f5978965aef79ea0
1351 F20101112_AADFDH drozdetski_a_Page_037.txt
29bf35bffe34e416bde84393a5de0749
e96b54278dd9967819f6be34c8a3c2455decb8b5
F20101112_AADEYB drozdetski_a_Page_124.tif
16e4348929b54cef6abe1742b7df44ef
d0fc509b403dc17d8086289e16dae885e9a5578d
2970 F20101112_AADFCT drozdetski_a_Page_014.txt
c82556c46f8a4bcc2018cdeb2728843d
ba2c7bf32c0ed40e8dbda8bb297055d5a15f205c
F20101112_AADEXO drozdetski_a_Page_110.tif
ed136eba783f0b1dd430ff3cda2d5cf6
8877420bd970f3b97e988588c670cb5fb6a69d62
2274 F20101112_AADFDI drozdetski_a_Page_038.txt
05344693f587ce3e2b7ee36d999ed9ca
57845ab2312914823e6267a6225001d9d8b14804
F20101112_AADEYC drozdetski_a_Page_125.tif
3e35acdb94ce80b847898f178d8b85bf
bddee389287b2fc74aaf65aa7776837cab674d80
186 F20101112_AADFCU drozdetski_a_Page_017.txt
1cb6f0a013a9bb0c4c7c75c17ecd2a97
a6a5643ac79a488b376b9e032048616841359a6a
F20101112_AADEXP drozdetski_a_Page_111.tif
37fa254f849cd25deaefa508f18d2fa2
b9205a6c889ea8107903e71cf73a9bf75771cd88
2358 F20101112_AADFDJ drozdetski_a_Page_040.txt
45f0322f951e62129f3960efe75afcc2
b1a9bd69f21f25bea9f114aecc83b3d8b54537ab
F20101112_AADEYD drozdetski_a_Page_126.tif
5252af0f7a1d53fbd08ac529a40a3970
871b26ec8bcb1893600ba958d8c66495e292b13d
394 F20101112_AADFCV drozdetski_a_Page_019.txt
5ce0653e89ac2f40763f80f7ef65142e
031d4485efb266bffd309fd2d24b6fd93c73d094
F20101112_AADEXQ drozdetski_a_Page_112.tif
18fab82ed6f8686f06fb83fb5331824e
f14b423fe7f6486e6999dded7fbd4c5597f93529
2114 F20101112_AADFDK drozdetski_a_Page_042.txt
d076529e4dcd8c256dbf7b5c00d0c8c7
0585cd95b00c8b07de76d50bd345e0a951df28f6
F20101112_AADEYE drozdetski_a_Page_129.tif
b770eab0e92e42d69dbc8914c24068e1
0fe2525ed1ee19fdb4c61e9bf64516eac65cd3d6
1761 F20101112_AADFCW drozdetski_a_Page_021.txt
fdcb8d621780c0dfbd3d797d045d051b
3ce934e756ee86ccbf4791ccd05f5de145448b05
F20101112_AADEXR drozdetski_a_Page_113.tif
d59fd31f740028ce1bb43947a6fe5948
fb96b0fcccd4e363690d067b2daf59b959f1d050
2199 F20101112_AADFEA drozdetski_a_Page_064.txt
98617161c25e700d209332a43beb2d83
03b162aa88b8dde5d4bd252fbcc637dcfd24eca5
2079 F20101112_AADFDL drozdetski_a_Page_043.txt
bc63bf7dd4581ff2cca84dcd602d8318
97cb1bf432b40bfb205bbaf3fb548b6c09ce8230
F20101112_AADEYF drozdetski_a_Page_130.tif
93823405526e3c388a3b37b63171f8fc
d184cf4c3c1794a0b85fc2bc0037034ea3112893
F20101112_AADFCX drozdetski_a_Page_023.txt
b113c7ea5b8774027046291890481606
b9c0e2b9b35006d98e92659e243e753dc46134bd
F20101112_AADEXS drozdetski_a_Page_114.tif
114a17bf45e6b7f09b5ec48efe59a7ab
4f45b9eddda19c9d6dc63e0e8aba3a70fabf6882
F20101112_AADFEB drozdetski_a_Page_065.txt
cffd058608234f7398016ae2e5488662
93eee79c85ae563474fbbb86affc557e9a4b036d
1510 F20101112_AADFDM drozdetski_a_Page_044.txt
daec500fff80d2ca21db02f4be52b6aa
29af68a3d68ed494a30f891f5544d9e9fbbd1aa6
F20101112_AADEYG drozdetski_a_Page_131.tif
8b6b4502fb68f203c66d75ad57b14ee3
0f51ad5c945022373b23a8febaf745247ccd6679
1222 F20101112_AADFCY drozdetski_a_Page_024.txt
b5bb0f5f790249f0f2cf0afa1c2030e2
577af04d0ac235a4a5ecd8a472775b7e6c1bf162
F20101112_AADEXT drozdetski_a_Page_116.tif
ceb2c2859434af6737b0050eaa7ba544
4c6615fb029c4b4c3e67287df88360b1c9bc6585
1748 F20101112_AADFEC drozdetski_a_Page_067.txt
4f14c9f5692f283b11796944bd228774
cc3b33d7e76cee641a9858d2711652354d620e71
1485 F20101112_AADFDN drozdetski_a_Page_046.txt
78f40e4ee717309be4fb45e90cf6389b
85fa8f966a7970fe7526dfdf88ccc8e9a8618004
F20101112_AADEYH drozdetski_a_Page_132.tif
c51f16bbda8cdd271b02b706597486ac
9f9eee615f3b32bdecfb5f21fd72d92b36112cc3
1810 F20101112_AADFCZ drozdetski_a_Page_025.txt
b9b7b14251d7c027616b16bd45182fd8
e04b31795908840d95311343570e264916045f45
F20101112_AADEXU drozdetski_a_Page_117.tif
2f9ac9650fd7061de4119b1ef1283296
c88fb59ff0e8fe2783a58e4edb77d3942c0c0b0f
F20101112_AADFED drozdetski_a_Page_068.txt
008dda5f5c8e0a3f0cc3eb178f7684c0
2cf297eff8e86f932f0b19f9b9e62bfa58e7ceca
389 F20101112_AADFDO drozdetski_a_Page_047.txt
779dd599fdbec45fb87043c4f922644a
1984a2191fc8ed2f6f25a47219d360b9fa1543cc
F20101112_AADEYI drozdetski_a_Page_133.tif
e188196e3a7a83b296db980f32e78645
7c9d127c0a8095d1aabde042408b781eb26909a3
F20101112_AADEXV drozdetski_a_Page_118.tif
d94056f4ad55485566c5b70b538c7509
2e538eaf822825fe1b773f2e6aac62bfebc9ca2c
1990 F20101112_AADFEE drozdetski_a_Page_069.txt
1442827a0508a268ca4a9c29fe467c09
a05ae58336e9d80fe85e30b63ad5c73b598983b7
2401 F20101112_AADFDP drozdetski_a_Page_048.txt
3a45f4cfca6ea90ed557c1a1084673e7
e5bab69ab6b91af5ec84376d66c8211405dce52d
F20101112_AADEYJ drozdetski_a_Page_135.tif
3ca1a8d3f6b8ec58f9611ddfc776322f
bb22760eacb6ea2ceb3a542cd1d7bf4910a6836f
F20101112_AADEXW drozdetski_a_Page_119.tif
ae645043bf8c0c9b2f482deb0e2737ce
f48c817c1002a7e9ef4f1481e73f3401b1d4a91f
1561 F20101112_AADFEF drozdetski_a_Page_070.txt
d43ae2f68642776b508d168a636d2d18
fcc3bb0a2e45e22b87ca8979aacc1986bbc283c9
1852 F20101112_AADFDQ drozdetski_a_Page_049.txt
1ef318e65b89a8c23e660c3016411462
3ad7e01c726e4deb1db81015a5d8e334035c943f
F20101112_AADEYK drozdetski_a_Page_136.tif
b9305f359f8f82c4d89fa1380a351e4e
9b5dba8a921b7ec862214bd43b2467c23e2d3fb9
F20101112_AADEXX drozdetski_a_Page_120.tif
a4a7b245690347522e6336b0f4762121
4900bd9e69038ba91d9744496f4c516f84ce8c5b
2600 F20101112_AADFEG drozdetski_a_Page_071.txt
120001db481a9a958fbab095ae1c1941
f3a50112d0a241d2aa2880874b0f71857182df71
2229 F20101112_AADFDR drozdetski_a_Page_050.txt
538a6323ba3e8bc1cfecbf3c20071341
4da1ba69dece212ae3328dd54c0d0ec302778a44
F20101112_AADEYL drozdetski_a_Page_137.tif
d59ceefeb7563680263186dde50e1b61
899ea0dae42866bb2c17db8b59b6643c06feef5f
F20101112_AADEXY drozdetski_a_Page_121.tif
641e54141f44fc3944f1f9a959ce26c3
ce09ae334c42801797a0e947c8692bd17095f023
40146 F20101112_AADEZA drozdetski_a_Page_021.pro
1a8f837a5570ef5abe004facbe289f1d
5c9431468db917b9cb78c2f17e5904e947344cba
904 F20101112_AADFDS drozdetski_a_Page_053.txt
181096642703c39ad8308460fd3ab7dd
ffe8b48dbc8870496f6176529f20a6e39ce5b143
F20101112_AADEYM drozdetski_a_Page_139.tif
3225774107addca0292a5cec4c9d2ded
c6a10571804973e226e8f4bb3e91c392ecd1a200
F20101112_AADEXZ drozdetski_a_Page_122.tif
41c364a0f7c0e7f6560071583a29c13d
7dc9747c02e2bd8e79ec81a1692e80ae2c22c05e
1635 F20101112_AADFEH drozdetski_a_Page_072.txt
a626b953b27c6130334362d5888393b8
b9dff77cd105aa079117fa166fc3a9063c326c24
43088 F20101112_AADEZB drozdetski_a_Page_022.pro
70445d280b59d2aa27b51e78effabf23
e3826b14df97edfea03de7c7adfaaa6f2519c849
1854 F20101112_AADFDT drozdetski_a_Page_055.txt
9f96229d6814c90e05133efe2625c90e
81882ee6123cd73e6cd57141ab5d9270e2b41932
F20101112_AADEYN drozdetski_a_Page_140.tif
7df7f2565934df24373b702bb2462bc0
f4c2a84bd418a2b5502b2716aa5a68d6d24d2af4
1385 F20101112_AADFEI drozdetski_a_Page_073.txt
9190e895b8f4c90735289908f5445191
94b7de36c81e752698f91262a73a5586c95f5e68
48106 F20101112_AADEZC drozdetski_a_Page_023.pro
a8696e7c7ee5be8a8144142052fc2067
98d3094eec48bdac6e68808b085d9077066f82e5
2633 F20101112_AADFDU drozdetski_a_Page_056.txt
f05a583923781569100c9f57b2337a3e
5fb969eef89c85ad1cb35111be8d81257356f757
1254 F20101112_AADEYO drozdetski_a_Page_002.pro
c0dbb3d22be553d3ede873bf4c59f51e
f6b4b39fe54142d08e5a534fbed00314d9c1b762
1166 F20101112_AADFEJ drozdetski_a_Page_074.txt
4ba310cbce71d2b85d97c6bc66db0202
e0f5315a3d2d50b7d361278ccb44b4a771348dce
30048 F20101112_AADEZD drozdetski_a_Page_024.pro
ac7fcdd19d2e76ce5398d99d152bbad1
055db5585becd79bab6019405380daf3e3927936
2160 F20101112_AADFDV drozdetski_a_Page_058.txt
e0198550f0674810f18b34cede6b4d9f
b547ad2974cfc935c313394912b0d675fa7e74e3
3217 F20101112_AADEYP drozdetski_a_Page_003.pro
9f12c892be4f13c24d2f91e9fa2dd887
616780306a77e4df1be1d8d7ec988b2873e18a67
1133 F20101112_AADFEK drozdetski_a_Page_075.txt
9497aab7b81d5568f004852d465481e9
b96dcfc2794986d6997c9d8aff01a868e4eedc3f
40960 F20101112_AADEZE drozdetski_a_Page_027.pro
acb258ecfb8837617169507fc0017d47
4c249dfe31b06add13e9fb143ef3043abf8fc79e
2028 F20101112_AADFDW drozdetski_a_Page_060.txt
7ff56a4a95ae02d2f6fa6b1f15a2eb4e
703540bec9e7255e05c9d9c26d9af31fc911afd2
54584 F20101112_AADEYQ drozdetski_a_Page_004.pro
aa9381543835381731c1801674fcf0fc
b2e829fbb0704c12095d345e314f97a5de45b98e