A Search for Supersymmetry in CMS Photon +Jet + ME T … · A Search for Supersymmetry in CMS...
Transcript of A Search for Supersymmetry in CMS Photon +Jet + ME T … · A Search for Supersymmetry in CMS...
A Search for Supersymmetry in CMS
Photon +Jet + MET EventsRachel Yohay
University of VirginiaMarch 29, 2012
R. Yohay March 29, 2012
Outline
• Detecting photons with the CMS ECAL
• Distinct challenges of the ECAL endcaps
• SUSY photon + jet + MET search
• Motivation
• Event selection
• Background estimation
• Results
• Interpretation
2
R. Yohay March 29, 2012
!"#$"%&&'()'*+,-'./0(1"/2#)'
3()$0%'%0%.4"#)'%)%"$5'&./0%'6*789'&4/1(0(45'()'4:%'*+,-';/""%0'<#"'=>??''+@3'A/4/'
B%/&C"%A'C&()$'D'!%E'%F%)4&'
GF%"/00'%H%.4'#<'&()$0%'.:/))%0'()4%"I./0(1"/2#)'/)A'4"/)&8/"%).5'
.#""%.2#)&'#)'4:%'J!%%'()F/"(/)4'B/&&'()'4:%'*+,-';/""%0'
Energy resolution:(σ/E)2 = (a/√E)2 + (b/E)2 + c2
ECAL operating principles
• Constructed of ~75,000 compact, dense (8.3 g/cm3), relatively radiation hard scintillating lead tungstate (PbWO4) crystals
• Crystal dimensions: ~1 Molière radius (~22 mm) x ~1 Molière radius x ~25 X0 (~9 mm) ⇒ most of an
electromagnetic shower is contained within 1 crystal
• Short scintillation time (~80% of scintillation light is emitted in 25 ns, the LHC collision frequency) ⇒ can
easily resolve events in different LHC buckets
3
[1]
[2]
PbWO4 crystal
[3]
CMS–ECAL TDR 1 General O verview
11
– stabilize the temperature of the calorimeter to ! 0.1 °C.
A 3-D view of the barrel and endcap electromagnetic calorimeter is shown in Fig. 1.5.
Fig. 1.5: A 3-D view of the electromagnetic calorimeter.
1.6.1 The barrel calorimeter
The barrel part of the ECAL covers the pseudorapidity range |"| < 1.479 (see Fig. 1.6).
The front face of the crystals is at a radius of 1.29 m and each crystal has a square cross-section of
# 22 $ 22 mm2 and a length of 230 mm corresponding to 25.8 X0. The truncated pyramid-shaped
crystals are mounted in a geometry which is off-pointing with respect to the mean position of the
primary interaction vertex, with a 3° tilt in both % and in ". The crystal cross-section corresponds
to &" $ &% = 0.0175 $ 0.0175 (1°). The barrel granularity is 360-fold in % and (2 $ 85)-fold in ",
resulting in a total number of 61 200 crystals. The crystal volume in the barrel amounts to 8.14 m3
(67.4 t). Crystals for each half-barrel will be grouped in 18 supermodules each subtending 20° in
%. Each supermodule will comprise four modules with 500 crystals in the first module and
400 crystals in each of the remaining three modules. For simplicity of construction and assembly,
crystals have been grouped in arrays of 2 $ 5 crystals which are contained in a very thin wall
(200 µm) alveolar structure and form a submodule.
Goal: 0.05
R. Yohay March 29, 2012
Distinct challenges of the ECAL endcaps
• ECAL endcaps extend crystal coverage from η = 1.5 out to η = 3.0
• Larger acceptance for rare H→γγ, H→ZZ, and H→WW processes
• Calorimetry at high η (+ particle flow techniques) ⇒ better MET reconstruction
⇒ better sensitivity to SUSY processes
• EE faces a much harsher environment than EB
• Strong magnetic field
• Higher occupancy
• More radiation damage
• Significant design difference between EB and EE: choice of vacuum phototriode as photodetector
• Calibrating EE is a considerable challenge
4
Carbon fiber frameMetal supports for attaching to the dee
Crystal
Vacuum phototriode (VPT)
Disassembled EE supercrystal
[4]
R. Yohay March 29, 2012
Vacuum phototriodes
• Chosen for their radiation hardness and good performance in strong magnetic fields
• Cathode at 0 V, dynode at 600 V, and anode mesh between them at 800 V
5
BELL et al.: VACUUM PHOTOTRIODES FOR THE CMS ELECTROMAGNETIC CALORIMETER ENDCAP 2285
Fig. 2. Diagram of a vacuum phototriode.
The VPTs for the CMS ECAL endcap are manufactured byResearch Institute Electron, St. Petersburg, Russia, followinga period of development in collaboration with PNPI Gatchina,Russia. A total of 6200 devices have been delivered at the timeof writing, comprising a preproduction batch of 500, and 5700devices manufactured under production conditions. Further in-formation on the development of VPTs for CMS may be foundin [5] and [6].
The VPTs delivered so far have a mean quantum efficiencyof 22% at 430 nm, with a standard deviation of 3%, and amean gain of 10.2 at 0 T, with a standard deviation of 1.8. EachVPT/crystal pair will be individually calibrated, so that thesevariations do not compromise the resolution of the calorimeter.The specification requires that the loss in output after irradia-tion to 20 kGy must be less than 10%, while the response ina magnetic field of 4 T, with the VPT axis at 15 to the fielddirection, must be at least 75% of that achieved in zero field.
III. MEASUREMENTS OF VPT CHARACTERISTICS
The measured variation in the gain of a typical VPT as a func-tion of the applied voltage is shown in Fig. 3. Two curves areshown, giving the gain as a function of dynode voltage with theanode held at 1000 or 800 V, and the photocathode at ground.The operating point in CMS will be V, V;at this point, the gain is insensitive to variations in the appliedvoltage.
All VPTs are tested at the Rutherford Appleton Laboratory(RAL) to determine their response as a function of magneticfield up to 1.8 T, and as a function of angle to the applied field.The automated testing rig at RAL is able to make measure-ments on 24 VPTs simultaneously, to determine their responseto pulses of light. Each VPT is illuminated by a set of four blueLEDs of peak wavelength approximately 420 nm, with the lightdiffused by a frosted perspex plate to ensure even illuminationacross the photocathode surface.
A typical set of measurements is shown in Fig. 4. The varia-tion of the VPT output shows a characteristic periodic behavioras a function of angle in a 1.8-T field. When the response is mea-sured as a function of magnetic field, at a fixed angle of 15 , thegain is observed to be constant at fields above 1 T, but variableat lower fields. When the magnetic field is not parallel to theVPT axis, the electrons acquire a transverse drift velocity pro-portional to the cross-product of the electric and magnetic fields;
the maxima and minima seen in Fig. 4 arise because the prob-ability of electron capture on the anode grid varies periodicallywith the transverse drift velocity. [7].
The VPTs will be installed in CMS at angles between 8 and24 to the magnetic field, and the mean response over this rangeis taken as a figure of merit characterising the VPT.
Approximately 10% of the devices, selected at random,are also tested at Brunel University in a 4-T superconductingmagnet, at a fixed angle of 15 . These measurements are wellcorrelated with the measurements made at 1.8 T, indicating thatthe response is stable at high magnetic fields, and that the 1.8-Tmeasurement is a reliable indicator of the VPT performance inCMS.
The measured yield of the VPTs is found to meet the require-ments of the CMS ECAL. No systematic variations in the VPTperformance have been observed during the period of manufac-ture.
Detailed measurements of the photocathode response havebeen made at the University of Split, Croatia, again using blueLED light peaking at 420 nm. The quantum efficiency is foundto be uniform over the photocathode area; a typical example isgiven in Fig. 5, which shows the variation in quantum efficiencyin steps of 0.5 mm across a diameter of the faceplate of a VPT.
IV. RADIATION TOLERANCE
The high radiation dose expected at LHC implies that the ra-diation tolerance of the VPTs must be carefully monitored. Testshave shown that the performance of the VPTs is not degradedsignificantly by neutron irradiation at the levels expected in thecentral region of the endcap over ten years of LHC running.Gamma irradiation tests have shown that the VPT response isreduced by an amount consistent with the observed loss of face-plate transparency, so that the main area of concern for CMS isthe induced absorbance of the faceplate glass expected duringthe 10-year lifetime of the experiment.
The faceplates of the CMS VPTs are made of a radiation-hardUV-transmitting borosilicate glass, which is manufactured insmall batches. Before being used for VPT production, severalfaceplates from each batch are irradiated in the dark at a tem-perature of 19 C to a dose of 20 kGy over a 48-h period using a
Co source at Brunel University. This dose is close to that ex-pected in the central region of the endcap after ten years runningin CMS. After irradiation the induced absorbance of the sampleis determined as a function of wavelength; a measurement ona typical glass is shown in Fig. 6. The total transmission lossafter convolution with the lead tungstate scintillation spectrum,shown as an inset in Fig. 6, must be less than 10% for the batchto be accepted for use in VPT production.
V. SUMMARY AND CONCLUSION
Vacuum Phototriodes have been developed that meet the de-manding requirements of the CMS ECAL endcap. Approxi-mately 15 000 of these devices are required, of which 6200 havealready been delivered by the manufacturer (at mid-October2003). The performance of all of these VPTs has been measuredin a 1.8-T magnetic field; the measurements show that the per-formance of the VPTs reaches the level required to allow the
0 V 800 V
600 V
Schematic of a CMS VPT
Faceplate
Anode, dynode, and cathode HV wires
[5]
[6]
R. Yohay March 29, 2012
VPT testing at UVa
• During the spring of 2008, extensive VPT testing was carried out in the UVa 4 T magnet, commissioned during winter 2008-2009
• Certified VPTs were installed on the endcap crystals
• Issues to be understood:
• Response vs. angle with respect to the magnetic field direction: is it smooth and in rough agreement with theoretical calculation?
• How do VPTs with skewed anodes or crinkled anodes compare to nominal?
6
Angle (deg)-20 0 20
Res
pons
e (A
DC
cou
nts)
200
300
400
500
VPT 02351VPT 02438VPT 02543VPT 03539VPT 07318VPT 08108VPT 08128VPT 13164VPT 13286
VPTs: 02351 02438 02543 03539 07318 08108 08128 13164 13286
Response of 9 VPTs vs. angle of VPT with respect to the magnetic field direction
Apparatus for measuring VPT response vs. angle
[7]
2/9/09 2:15 PM/afs/cern.ch/user/y/yohay/private/69486_over_70128.svg
Page 1 of 1file:///Users/rachelyohay/Documents/CERN/DPG_Feb1209_4T-0T_ratios/69486_over_70128.svg
ix20 30 40 50
iy
50
55
60
65
70
Entries 200
0 0 0
0 167 0
0 0 0
0.7
0.75
0.8
0.85
0.9
0.95
Entries 200
0 0 0
0 167 0
0 0 0
run 69486 (3.8 T) / run 70128 (0 T), blue LED
B field effect in one section of EE
CMS Work in ProgressCMS Work in Progress
R. Yohay March 29, 2012
VPT stability
7
Response of VPTs to SPS spill simulation
CMS Work in ProgressCMS Work in Progress
VPT response vs. time for different background pulsing rates
• VPTs first used in the OPAL electromagnetic calorimeter endcaps [8]
• VPT gain varies with frequency/amplitude of incident light [9]
• Effect strongly suppressed in 4 T magnetic field
• High→low frequency: gain increases
• Low→high frequency: gain increases or decreases
• All VPT responses are different
• Provide a constant rate of stability LED pulses to the VPTs to suppress gain changes at LHC on/off transitions
R. Yohay March 29, 2012
Radiation damage
8
• Sustained ionizing radiation causes crystal radiation damage, reducing crystal transparency and ultimately the amount of light collected by the photodetectors
• Crystal transparency loss correlated with LHC integrated luminosity, and increases faster for higher instantaneous luminosity
• Continuously pulse crystals from known light source to track and correct for transparency loss from radiation damage
!"
!"#$%&'()*+(%*),&-*#+(%.
norm
aliz
ed A
PD/P
N[10]
R. Yohay March 29, 2012
LED stability and calibration system
• Dual wavelength LED stability and monitoring system designed at UVa
• Blue (450 nm) LED: near the peak of crystal scintillation and VPT photocathode efficiency, so ideal for transmitting the maximum amount of light to VPTs for stability pulsing
• Orange (617 nm): transparent to crystals but still efficient for VPT photocathode, so ideal for disentangling crystal damage from VPT gain changes
• PN diodes for normalization
9
Suppressing Classic VPT Effect with LED Pulser
January 22, 2008 A.Bornheim - ECAL Endcap Stability 4
Indication that the extend to which the VPT effect is suppressed is a function of the rate we pulse the LED pulser. See also Rachels talk.
classic VPT effect approximately halved for ~1
kHz stability pulse vs. ~100 Hz stability pulse 2008 JINST 3 S08004
Distribution system for VPT stabilisation light pulses
To ~ 250 crystals
PbWO4 Crystal
LED light source
(from laser monitoring distribution system)
VPT
'Level 1' diffusing sphere
Figure 4.8: Distribution system for VPT stabilisation light pulses.
An all-silica optical fibre (CF01493-43 from OFS (Furukawa)) is inserted into a hole drilledinto the lens of each LED and collects light by proximity focusing. The seven fibres from a givenlight source are combined into a single bundle that transports light to a diffusing sphere which hasa dual role, acting also as part of the distribution network of the laser monitoring system. Lightfrom each diffusing sphere is distributed to up to 220 individual detector channels through the setof optical fibres that also carry the laser monitoring pulses. Light is injected via the rear face of acrystal, which carries the VPT, and reaches the VPT via reflection from the front of the crystal. Thesystem is synchronized to pulse during a fraction of the 3 µs abort gaps that occur during every 89µs cycle of the LHC circulating beams.
4.4 On-detector electronics
The ECAL read-out has to acquire the small signals of the photo-detectors with high speed andprecision. Every bunch crossing digital sums representing the energy deposit in a trigger tower aregenerated and sent to the trigger system. The digitized data are stored during the Level-1 triggerlatency of ≈ 3 µs.
The on-detector electronics has been designed to read a complete trigger tower ( 5×5 crystalsin η ×φ ) or a super-crystal for EB and EE respectively. It consists of five Very Front End (VFE)boards, one Front End (FE) board, two (EB) or six (EE) Gigabit Optical Hybrids (GOH), one LowVoltage Regulator card (LVR) and a motherboard.
The motherboard is located in front of the cooling bars. It connects to 25 photo-detectorsand to the temperature sensors using Kapton flexible printed circuit boards and coaxial cables forEB and EE respectively. In the case of the EB the motherboard distributes and filters the APDbias voltage. Two motherboards are connected to one CAEN HV supply located at a distance ofabout 120m with remote sensing. In the case of the EE the operating voltages for the VPTs aredistributed and filtered by a separate HV filter card, hosting as well the decoupling capacitor forthe anode signals. Five of these cards serving five VPTs each are installed into each super-crystal.One LVR and five VFE cards plug into the motherboard.
Each LVR card [81] uses 11 radiation-hard low voltage regulators (LHC4913) developed byST-microelectronics and the RD49 project at CERN. The regulators have built in over-temperature
– 100 –
4 blue
3 orange
{{
PN diode
PN diode
LED system on EE dee patch panel
[11]
[12]
CMS Work in Progress
R. Yohay March 29, 2012
Hardware setup
• LED amplitudes set via I2C commands communicated to the hardware via Ethernet-to-serial and serial-to-I2C bridges
• Trigger pulse originates in counting room and is regenerated on the LED circuit board on the detector
10
reshaping
delay option
veto option
trigger counting
internal pulser
dee A
LVPECLdee B
LVPECL
f( )
dee A
LVDS
dee B
LVDS
to endcap
!"#$%&'$(#
)*+$,$-*"."$/0$!"#1*(!$232(/0-$,$#04/-0556-
Description
!"#$%&'#$%()*%+,-%.)/0/%123"+4
%",#235+621
%#$"1%"1%+%#3748%7,"9231+4%
:/;<)=*()<)>:<?@!"32%",#235+62A
%5'3%#$2%B"-21#%C'11"&
42%3+,
D2%'5%6'EC7#",D%
$+3-B+32A%B
"#$%#$2%1"E
C421#%+,
-%E'1#%C'3#+&42%1'
5#B+32%5'
3E+#F
!"#"$#%&'%!
"#%()%!"#%'*(+,%-
.%()%!"#%/-)&0-.1%!0%&'%(
%$%&#'(%-)%02,%!"#34$!%56'%-)*71%
*7"4-",D%+,%:/;%&+12-
%+CC4"6+#"'
,%,22-1%,'#$",D%E'32%#$+,%7-6.%:/;%6$"CA%29238#$",D%
2412%"1%",647-2-F%
G$232%"1%+,
%",#23,+4%H
I%+,-%0F0I%C'B23%17
CC48%#'%37,%#$2%",#235+62
%+,-%8'73%#+3D
2#%
$+3-B+32F%)
"EC42%181#2E
1%B"44%37
,%53'E%#$2%()*%&71%C'B23F
(,4"J2%1'E2%6'EC2#",D%C3'-76#1A%"#%6+,
%171#+",
%$"D$%&+7-%3+#21A%+,
-%$+1%E
74#"C42%:/;%
*71121
;'EC+32-
%#'%637
-23%1'
47#"',1A%"#%-
'21,K#%3248
%',%&"#@&+1$",D%>;%1'5#B+32A%"44%-
25",2-%
:<L%6$+3+6#23"1"61A%'
3%C3'6211'
3%+,-%L)%-2C2,-2,#%#"E
",DF
>'B23@L
,@.212#%+6#"'
,1%6+,
%&2%12#%",
%MM>3'E%1'%#$+#%18
1#2E1%+32
%C3'C2348
%","#"+4"12-
%
B$2,%#$2%>;%"1%,
'#%C3212,
#F
!+#6$
-'D%+6#"'
,1%6+,
%&2%C3'D3+EE2-%#'%+7#'E+#"6+448
%321C',-%#'%>;%5+"47
32F
N+49+,"6%"1'4+#"'
,%6+,
%&2%",1#+442-
%B$232%#$2%:/;%&71121%,
22-%#'%&2%"1'4+#2-
%53'E%#$2%
6'EC7#23%5'
3%,'"12A%D
3'7,-%4''CA%'3%1+52#8
%6',1"-23+#"'
,1F
O,%'C#"',+4%0
3-%",#235+62
%"1%+9+"4+&
42%5'3%P"&32%LC#"6A%*
472#''#$A%.)Q//%+,-%1"E
"4+3%
",#235+621
7
R%/SS/@T%*3'+-6+1#%M
U7"CE2,#%V#-%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%$$$$$$$$$$$$$$$$$%"/
6/C6F'-#%3H?%%C+D2%/%'5%?W
C$%XYQ%/?%Y/0@QS/%%%%B
BBF"/66$"CF6'E%%%%%1+421Z
"/66$"CF6'E
*71%[
?
*71%[
/
*71%[
0
HI%<%0F0I
!"#$%&'$(#
)*+$,$-*"."$/0$!"#1*(!$232(/0-$,$#04/-0556-
Description
!"#$%&'#$%()*%+,-%.)/0/%123"+4
%",#235+621
%#$"1%"1%+%#3748%7,"9231+4%
:/;<)=*()<)>:<?@!"32%",#235+62A
%5'3%#$2%B"-21#%C'11"&
42%3+,
D2%'5%6'EC7#",D%
$+3-B+32A%B
"#$%#$2%1"E
C421#%+,
-%E'1#%C'3#+&42%1'
5#B+32%5'
3E+#F
!"#"$#%&'%!
"#%()%!"#%'*(+,%-
.%()%!"#%/-)&0-.1%!0%&'%(
%$%&#'(%-)%02,%!"#34$!%56'%-)*71%
*7"4-",D%+,%:/;%&+12-
%+CC4"6+#"'
,%,22-1%,'#$",D%E'32%#$+,%7-6.%:/;%6$"CA%29238#$",D%
2412%"1%",647-2-F%
G$232%"1%+,
%",#23,+4%H
I%+,-%0F0I%C'B23%17
CC48%#'%37,%#$2%",#235+62
%+,-%8'73%#+3D
2#%
$+3-B+32F%)
"EC42%181#2E
1%B"44%37
,%53'E%#$2%()*%&71%C'B23F
(,4"J2%1'E2%6'EC2#",D%C3'-76#1A%"#%6+,
%171#+",
%$"D$%&+7-%3+#21A%+,
-%$+1%E
74#"C42%:/;%
*71121
;'EC+32-
%#'%637
-23%1'
47#"',1A%"#%-
'21,K#%3248
%',%&"#@&+1$",D%>;%1'5#B+32A%"44%-
25",2-%
:<L%6$+3+6#23"1"61A%'
3%C3'6211'
3%+,-%L)%-2C2,-2,#%#"E
",DF
>'B23@L
,@.212#%+6#"'
,1%6+,
%&2%12#%",
%MM>3'E%1'%#$+#%18
1#2E1%+32
%C3'C2348
%","#"+4"12-
%
B$2,%#$2%>;%"1%,
'#%C3212,
#F
!+#6$
-'D%+6#"'
,1%6+,
%&2%C3'D3+EE2-%#'%+7#'E+#"6+448
%321C',-%#'%>;%5+"47
32F
N+49+,"6%"1'4+#"'
,%6+,
%&2%",1#+442-
%B$232%#$2%:/;%&71121%,
22-%#'%&2%"1'4+#2-
%53'E%#$2%
6'EC7#23%5'
3%,'"12A%D
3'7,-%4''CA%'3%1+52#8
%6',1"-23+#"'
,1F
O,%'C#"',+4%0
3-%",#235+62
%"1%+9+"4+&
42%5'3%P"&32%LC#"6A%*
472#''#$A%.)Q//%+,-%1"E
"4+3%
",#235+621
7
R%/SS/@T%*3'+-6+1#%M
U7"CE2,#%V#-%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%$$$$$$$$$$$$$$$$$%"/
6/C6F'-#%3H?%%C+D2%/%'5%?W
C$%XYQ%/?%Y/0@QS/%%%%B
BBF"/66$"CF6'E%%%%%1+421Z
"/66$"CF6'E
*71%[
?
*71%[
/
*71%[
0
HI%<%0F0I
!"#$%&'$(#
)*+$,$-*"."$/0$!"#1*(!$232(/0-$,$#04/-0556-
Description
!"#$%&'#$%()*%+,-%.)/0/%123"+4
%",#235+621
%#$"1%"1%+%#3748%7,"9231+4%
:/;<)=*()<)>:<?@!"32%",#235+62A
%5'3%#$2%B"-21#%C'11"&
42%3+,
D2%'5%6'EC7#",D%
$+3-B+32A%B
"#$%#$2%1"E
C421#%+,
-%E'1#%C'3#+&42%1'
5#B+32%5'
3E+#F
!"#"$#%&'%!
"#%()%!"#%'*(+,%-
.%()%!"#%/-)&0-.1%!0%&'%(
%$%&#'(%-)%02,%!"#34$!%56'%-)*71%
*7"4-",D%+,%:/;%&+12-
%+CC4"6+#"'
,%,22-1%,'#$",D%E'32%#$+,%7-6.%:/;%6$"CA%29238#$",D%
2412%"1%",647-2-F%
G$232%"1%+,
%",#23,+4%H
I%+,-%0F0I%C'B23%17
CC48%#'%37,%#$2%",#235+62
%+,-%8'73%#+3D
2#%
$+3-B+32F%)
"EC42%181#2E
1%B"44%37
,%53'E%#$2%()*%&71%C'B23F
(,4"J2%1'E2%6'EC2#",D%C3'-76#1A%"#%6+,
%171#+",
%$"D$%&+7-%3+#21A%+,
-%$+1%E
74#"C42%:/;%
*71121
;'EC+32-
%#'%637
-23%1'
47#"',1A%"#%-
'21,K#%3248
%',%&"#@&+1$",D%>;%1'5#B+32A%"44%-
25",2-%
:<L%6$+3+6#23"1"61A%'
3%C3'6211'
3%+,-%L)%-2C2,-2,#%#"E
",DF
>'B23@L
,@.212#%+6#"'
,1%6+,
%&2%12#%",
%MM>3'E%1'%#$+#%18
1#2E1%+32
%C3'C2348
%","#"+4"12-
%
B$2,%#$2%>;%"1%,
'#%C3212,
#F
!+#6$
-'D%+6#"'
,1%6+,
%&2%C3'D3+EE2-%#'%+7#'E+#"6+448
%321C',-%#'%>;%5+"47
32F
N+49+,"6%"1'4+#"'
,%6+,
%&2%",1#+442-
%B$232%#$2%:/;%&71121%,
22-%#'%&2%"1'4+#2-
%53'E%#$2%
6'EC7#23%5'
3%,'"12A%D
3'7,-%4''CA%'3%1+52#8
%6',1"-23+#"'
,1F
O,%'C#"',+4%0
3-%",#235+62
%"1%+9+"4+&
42%5'3%P"&32%LC#"6A%*
472#''#$A%.)Q//%+,-%1"E
"4+3%
",#235+621
7
R%/SS/@T%*3'+-6+1#%M
U7"CE2,#%V#-%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%$$$$$$$$$$$$$$$$$%"/
6/C6F'-#%3H?%%C+D2%/%'5%?W
C$%XYQ%/?%Y/0@QS/%%%%B
BBF"/66$"CF6'E%%%%%1+421Z
"/66$"CF6'E
*71%[
?
*71%[
/
*71%[
0
HI%<%0F0I
from EMTC
20.000.800
0.80020.00
undervoltage/overcurrent trip
settable I/V limits
I/V monitoring (digital and
analog)
channel enable/disable
Ethernet
DB9 serial
DB9 serial x3
dee A I2C to DACS
dee B I2C TO DACs
trigger control I2C
power monitoring control I2C
dee A 20 Vdee B 20 V
20 V x8 (x4 dee A, x4 dee B)
LED PC
power supply
serial device server
I2C controller
x 3
power
monitoring
box
trigger
control
module
Figure 2.4: LED control and monitoring system. Photographs shown in the
diagram can be found in refs. [6]-[9].
output programmable power supply is used per endcap (one channel per dee).
This model supports remote control via a standard RS-232 DB9 connection
on its rear. Of the over 40 remote commands recognized by the supply, 15
have been implemented in the custom PL330TP control software running on
the LED PC:
1. Enable/disable channel
2. Set current limit for channel
3. Set voltage for channel
4. Set voltage increment for channel
5. Decrement voltage on channel by voltage increment (as set with the
previous command)
6. Read current limit for channel
7. Read voltage set point for channel
8. Read current for channel
9. Read voltage for channel
8
R. Yohay March 29, 2012
Control and monitoring software
• Use existing LaserSupervisor XDAQ executive as interface to ECAL DAQ
• Execute LED on/off commands sent from the LaserSupervisor
• Monitoring of electronics every 10 minutes
11
EE+ power monitor box
EE- power monitor box
Run Control and Monitoring System
CMS control room
ECALSupervisor
ECAL DAQ PC
OtherSupervisor
other ECAL DAQ PC
LaserSupervisor
laser PC
configuration
state
configuration state
USC
USCconfiguration state
LEDController
laser barracksconfiguration database
configuration parameters
LEDHardwareController
LED PC
LEDCommandController
USC S2F18
laser hardware
LEDHardwareController
commandsstatus requests
status
status
commands
EE- control box EE+ control box
EE- power supply
EE+ power supply
EE- EE+
commands
I2C2PC status register & counter values (from control boxes)I/V limits & channel power status (from power monitor boxes)I & V (from power supply)
EE- HV rack EE+ HV rackI, V I, V
power power
set LED amplitudes set LED amplitudespower power
Monday, April 20, 2009
R. Yohay March 29, 2012
Control and monitoring software
12
EE+ power monitor box
EE- power monitor box
Run Control and Monitoring System
CMS control room
ECALSupervisor
ECAL DAQ PC
OtherSupervisor
other ECAL DAQ PC
LaserSupervisor
laser PC
configuration
state
configuration state
USC
USCconfiguration state
LEDController
laser barracksconfiguration database
configuration parameters
LEDHardwareController
LED PC
LEDCommandController
USC S2F18
laser hardware
LEDHardwareController
commandsstatus requests
status
status
commands
EE- control box EE+ control box
EE- power supply
EE+ power supply
EE- EE+
commands
I2C2PC status register & counter values (from control boxes)I/V limits & channel power status (from power monitor boxes)I & V (from power supply)
EE- HV rack EE+ HV rackI, V I, V
power power
set LED amplitudes set LED amplitudespower power
Monday, April 20, 2009
Wrote this software
Wrote monitoring programs to assess the health of this hardware
R. Yohay March 29, 2012
LED and VPT performance
• Stable, reliable LED system important for ECAL calibration
• VPT effect currently dwarfed by transparency loss, but system in place to mitigate gain changes in order to achieve best performance
13
2/9/09 1:57 PM/afs/cern.ch/user/y/yohay/private/3.8T_over_3.8T_different_times.svg
Page 1 of 1file:///Users/rachelyohay/Documents/CERN/DPG_Feb1209_4T-0T_ratios/3.8T_over_3.8T_different_times.svg
Entries 200
Underflow 3
Overflow 0
/ ndf2! 11.97 / 14
Constant 4.40!42.46
Mean 0.0!1
Sigma 0.00012!0.00171
Ratio of VPT/PN values for two 3.8 T runs
0.98 0.99 1 1.010
10
20
30
40 Entries 200
Underflow 3
Overflow 0
/ ndf2! 11.97 / 14
Constant 4.40!42.46
Mean 0.0!1
Sigma 0.00012!0.00171
run 69841 (3.8 T) / run 69486 (3.8 T), blue LED
18
A.L
edovskoy
-U
niv
ers
ity
ofV
irgin
ia-
25-M
ay-2
011
T [s]
1303 1303.5 1304
610!
vpt / pnA
B
0.98
0.99
EE+, DEE2, DS=9
Laser has smaller “HF” fluctuations
Disagree in shape and level
Blue: Laser
Dark Cyan: Blue LED
18
an example ofmeasured timedistribution for one LEDchannel
measured time jitter for each of 76 LED soak channelsis below 100 ps
[ns] MAXAVERAGE T10 10.5 11
EVEN
TS
0
10
20
30
40 0.002± = 0.076 !
A.L
edov
skoy
-U
nive
rsity
ofV
irgin
ia-
22-A
pr-2
010
10
10
CMS Work in ProgressCMS Work in Progress
CMS Work in Progress
VPT/PN vs. time for blue laser (green) and blue LED (blue) Apr. 17 - May 2 2011
Ratio of PN-normalized VPT amplitudes for two different runs (channels experienced HV and B field cycling between runs)
Distribution of average LED timing for a single diffusing sphere
R. Yohay March 29, 2012
SUSY with photons• Why search for supersymmetry?
• Provides a way to control loop corrections to the Higgs mass
• Provides a stable, feebly interacting particle ⇒ dark matter candidate
• SUSY particles are heavier than their SM counterparts, so SUSY is a broken symmetry
• In gauge-mediated SUSY breaking (GMSB), ordinary gauge interactions link the SUSY-breaking and visible sectors
• ~eV-keV gravitino is the lightest SUSY particle ⇒ escapes CMS undetected,
leading to large MET
• Neutralino is the next-to-lightest SUSY particle (NLSP) ⇒ neutralino
usually decays to photon + gravitino
14
spin changed by half integer
H
f
(a)
S
H
(b)
Figure 1.1: One-loop quantum corrections to the Higgs squared mass parameter m2H , due to (a) a Dirac
fermion f , and (b) a scalar S.
The Standard Model requires a non-vanishing vacuum expectation value (VEV) for H at the minimum
of the potential. This will occur if ! > 0 and m2H < 0, resulting in !H" =
!#m2
H/2!. Since we
know experimentally that !H" is approximately 174 GeV, from measurements of the properties of theweak interactions, it must be that m2
H is very roughly of order #(100 GeV)2. The problem is that m2H
receives enormous quantum corrections from the virtual e!ects of every particle that couples, directlyor indirectly, to the Higgs field.
For example, in Figure 1.1a we have a correction to m2H from a loop containing a Dirac fermion
f with mass mf . If the Higgs field couples to f with a term in the Lagrangian #!fHff , then theFeynman diagram in Figure 1.1a yields a correction
"m2H = # |!f |2
8"2#2
UV + . . . . (1.2)
Here #UV is an ultraviolet momentum cuto! used to regulate the loop integral; it should be interpretedas at least the energy scale at which new physics enters to alter the high-energy behavior of the theory.The ellipses represent terms proportional to m2
f , which grow at most logarithmically with #UV (andactually di!er for the real and imaginary parts of H). Each of the leptons and quarks of the StandardModel can play the role of f ; for quarks, eq. (1.2) should be multiplied by 3 to account for color. Thelargest correction comes when f is the top quark with !f $ 1. The problem is that if #UV is of orderMP, say, then this quantum correction to m2
H is some 30 orders of magnitude larger than the requiredvalue of m2
H % #(100 GeV)2. This is only directly a problem for corrections to the Higgs scalar bosonsquared mass, because quantum corrections to fermion and gauge boson masses do not have the directquadratic sensitivity to #UV found in eq. (1.2). However, the quarks and leptons and the electroweakgauge bosons Z0, W± of the Standard Model all obtain masses from !H", so that the entire massspectrum of the Standard Model is directly or indirectly sensitive to the cuto! #UV.
One could imagine that the solution is to simply pick a #UV that is not too large. But then onestill must concoct some new physics at the scale #UV that not only alters the propagators in the loop,but actually cuts o! the loop integral. This is not easy to do in a theory whose Lagrangian does notcontain more than two derivatives, and higher-derivative theories generally su!er from a failure of eitherunitarity or causality [2]. In string theories, loop integrals are nevertheless cut o! at high Euclideanmomentum p by factors e!p2/!2
UV . However, then #UV is a string scale that is usually† thought to benot very far below MP. Furthermore, there are contributions similar to eq. (1.2) from the virtual e!ectsof any arbitrarily heavy particles that might exist, and these involve the masses of the heavy particles,not just the cuto!.
For example, suppose there exists a heavy complex scalar particle S with mass mS that couples tothe Higgs with a Lagrangian term #!S |H|2|S|2. Then the Feynman diagram in Figure 1.1b gives acorrection
"m2H =
!S
16"2
"#2
UV # 2m2S ln(#UV/mS) + . . .
#. (1.3)
†Some recent attacks on the hierarchy problem, not reviewed here, are based on the proposition that the ultimatecuto! scale is actually close to the electroweak scale, rather than the apparent Planck scale.
3
mH2 ∝
SM large SUSY large
- = finite
Example of Higgs mass regularization via a SUSY loop that cancels its SM counterpart loop.
[14]
[13]
R. Yohay March 29, 2012
GMSB final states
15 4
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
γ+jets+MET
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01�G
1
g�gqq�qγ�χ01�G
1
Bino NLSP: neutralino→γ+gravitino or neutralino→Z(→jets)+gravitino
g
g
g
g
j j
jj
!
"
l!
G
G
q
q
W 0
W! W!
l+γ+jets+MET
Wino NLSP: neutralino→γ+gravitino and chargino→W(→lν)+gravitino
g�gqq�qγ�χ01
1
g�g
q
q�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
2γ+jets+MET
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01�G
1
g�gqq�qγ�χ01�G
1
Bino NLSP: neutralino→γ+gravitino
g�g
q
q�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
Z0
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g
g
g
g
j j
jj
!
"
l!
G
G
q
q
W 0
W! W!
g�gqq�qγ�χ01
1
g�g
q
q�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
1
g�gqq�qγ�χ01
1
g�g
q
q
q�
�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
Z0
1
Wino NLSP: neutralino→γ+gravitino and chargino→W(→jets)+gravitino
γ+jets+MET
R. Yohay March 29, 2012
GMSB final states
16 4
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
γ+jets+MET
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01�G
1
g�gqq�qγ�χ01�G
1
Bino NLSP: neutralino→γ+gravitino or neutralino→Z(→jets)+gravitino
g
g
g
g
j j
jj
!
"
l!
G
G
q
q
W 0
W! W!
l+γ+jets+MET
Wino NLSP: neutralino→γ+gravitino and chargino→W(→lν)+gravitino
g�gqq�qγ�χ01
1
g�g
q
q�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
2γ+jets+MET
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01�G
1
g�gqq�qγ�χ01�G
1
Bino NLSP: neutralino→γ+gravitino
g�g
q
q�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
Z0
1
g�gqq�qγ�χ01
1
g�gqq�qγ�χ01
1
g
g
g
g
j j
jj
!
"
l!
G
G
q
q
W 0
W! W!
g�gqq�qγ�χ01
1
g�g
q
q�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
1
g�gqq�qγ�χ01
1
g�g
q
q
q�
�q
γ�χ01m�χ01
= 50 GeV
m�χ01= 150 GeV
m�χ01= 500 GeV
�G
ν
Z0
1
Wino NLSP: neutralino→γ+gravitino and chargino→W(→jets)+gravitino
γ+jets+MET
This talk
4.7 fb-1
R. Yohay March 29, 2012
Photon selection
17
• Single L1-seeded diphoton triggers with 36 and 22 GeV thresholds
• Combined detector isolation cuts in ΔR = 0.3 cone to improve acceptance in jet-rich events
• Shower shape cuts further reduce jet fakes and anomalous energy deposits
• Pileup subtraction from isolation cone using Fastjet [15]
• Minimum ΔR between the photons to avoid isolation cone overlap
• ~±3 ns timing cut removes cosmics and beam halo
• Pixel veto rejects electrons
65
Table 4.1: Selection criteria for γγ, eγ, ee, and ff events.
VariableCut
γγ eγ ee ff
HLT match IsoVL IsoVL IsoVL IsoVL || R9Id
ET> 40/
> 25 GeV> 40/
> 25 GeV> 40/
> 25 GeV> 40/
> 25 GeVSC |η| < 1.4442 < 1.4442 < 1.4442 < 1.4442H/E < 0.05 < 0.05 < 0.05 < 0.05R9 < 1 < 1 < 1 < 1
Pixel seed No/No Yes/No Yes/Yes No/No
Icomb, σiηiη< 6 GeV &&
< 0.011< 6 GeV &&
< 0.011< 6 GeV &&
< 0.011
< 20 GeV &&(≥ 6 GeV ||≥ 0.011)
JSON Yes Yes Yes YesNo. good PVs ≥ 1 ≥ 1 ≥ 1 ≥ 1
∆REM > 0.6 > 0.6 > 0.6 > 0.6∆φEM ≥ 0.05 ≥ 0.05 ≥ 0.05 ≥ 0.05
Table 4.2: Definition of HB/HE/HF hadronic jets. Add a footnote describing thePF electron and PF muon definitions, with references.
Variable Cut
Algorithm L1FastL2L3Residual corrected PF (cf. Sec. 3.1.3)pT > 30 GeV|η| < 5.0
Neutral hadronicenergy fraction
< 0.99
Neutral electromagneticenergy fraction
< 0.99
Number of constituents > 1Charged hadronic energy > 0.0 GeV if |η| < 2.4
Number of charged hadrons > 0 if |η| < 2.4Charged electromagnetic
energy fraction< 0.99 if |η| < 2.4
∆R to nearest electron, muon,or one of the two
primary EM objects> 0.5
R. Yohay March 29, 2012
Photon ID efficiency
• Photon ID efficiencies taken from MC and corrected by (data electron efficiency)/(MC electron efficiency)
• Use Z→ee events to measure the electron efficiencies
• Photon ID cuts designed to behave similarly for electrons and photons
• Signal MC acceptance × efficiency multiplied by 1 factor of εdata/εMC per photon
• Pixel match veto efficiency estimated from MC: (96.4 ± 0.5)% (stat. ⊕ syst. due to tracker material budget variation)
• Data/MC efficiency scale factor: 0.99 ± 0.04, with errors due to:
• Z signal and background shape variation
• Signal fit over/underestimation
• Pileup effects
• MC electron/photon difference
18
R. Yohay March 29, 2012
Photon ID scale factor dependence
19
PVN2 4 6 8 10 12 14 16 18 20
MC
!/da
ta!
0.85
0.90
0.95
1.00
1.05
1.10
1.15 / ndf 2" 16.56 / 19
p0 0.002351± 0.9925
/ ndf 2" 16.56 / 19
p0 0.002351± 0.9925
(GeV)TE20 40 60 80 100 120 140 160 180
MC
!/da
ta!
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4 / ndf 2" 53.93 / 12
p0 0.001737± 0.9977
/ ndf 2" 53.93 / 12
p0 0.001737± 0.9977
!-1.0 -0.5 0.0 0.5 1.0
MC
"/da
ta"
0.92
0.94
0.96
0.98
1.00
1.02
1.04
1.06
1.08 / ndf 2# 17.54 / 11
p0 0.002067± 0.996 / ndf 2# 17.54 / 11
p0 0.002067± 0.996
jetN0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
MC
!/da
ta!
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1.10
1.15 / ndf 2" 2.126 / 4
p0 0.002282± 0.9937
/ ndf 2" 2.126 / 4
p0 0.002282± 0.9937
PVN0 5 10 15 20
Effic
ienc
y
0.80
0.85
0.90
0.95
1.00
1.05
Data (runs 165088-180252)
)t+jet, QCD, W+jet, t!ee+jet, "MC (Drell-Yan
ET η Nj
NPV
CMS Work in Progress CMS Work in Progress CMS Work in Progress
CMS Work in Progress CMS Work in Progress
R. Yohay March 29, 2012
Jet selection
• ≥1 jet not overlapping any electron, photon, or fake (loosely isolated) photon
20
44
Table 3.2: Definition of HB/HE hadronic jets. Add a footnote describing the PFelectron and PF muon definitions, with references.
Variable Cut
Algorithm L1FastL2L3Residual corrected PF
pT > 30 GeV
|η| < 2.6Neutral hadronic
energy fraction< 0.99
Neutral electromagnetic
energy fraction< 0.99
Number of constituents > 1
Charged hadronic energy > 0.0 GeV if |η| < 2.4Number of charged hadrons > 0 if |η| < 2.4Charged electromagnetic
energy fraction< 0.99 if |η| < 2.4
∆R to nearest electron, muon,
or one of the two
primary EM objects
> 0.5
6.5 Effect of multiple interactions 17
!"#
Figure 13: Calibrated E/ x,y resolution versus calibrated PF ∑ ET for Calo E/T, TC E/T, and PF E/T
in data and in simulation.
For ∑ ET, we use the PF ∑ ET as measured by the particle-flow algorithm for all types of E/T, as it
gives the best estimate of the true ∑ ET, and hence is an accurate evaluation of the event activity.
We use PF ∑ ET for all algorithms to ensure their measure is the same. We calibrate PF ∑ ET to
the particle-level ∑ ET, on average, using the predicted average mean value as a function of the
particle-level ∑ ET from a simulation of events from the PYTHIA 8 event generator [22].
Figure 13 shows the calibrated E/ x,y Gaussian core resolution versus the calibrated PF ∑ ET for
different E/T reconstruction algorithms in events containing at least two jets with pT > 25 GeV.
Both TC E/T and PF E/T show improvements in the E/T resolution compared to the Calo E/T, and
the PF E/T yields the smallest E/T resolution.
Figure 14 shows the PF E/T distributions for different intervals of Calo ∑ ET and for jet mul-
tiplicities varying from two to four, normalized to the same area. The jets are required to be
above a pT-threshold of 20 GeV. The good agreement of the normalized shapes in Fig. 14 in-
dicates that PF E/T-performance in events without genuine E/T is driven by the total amount of
calorimetric activity (parametrized by Calo ∑ ET) and no residual non-linear contribution from
jets to PF E/T is visible. Similar behaviour is also observed for Calo E/T and TC E/T.
6.5 Effect of multiple interactions
Pile-up, namely multiple proton collisions within the same bunch crossing, occurs because of
high LHC bunch currents and can play an important role in �E/T performance.
Because there is no true �E/T in minimum bias events and because the average value for a com-
ponent of �E/T in these events is zero (e.g., the x or y component), pile-up should have only a
small effect on the scale of the component of the measured �E/T projected along the true �E/T di-
rection. Pile-up, however, will have a considerable effect on the resolution of the parallel and
perpendicular components.
We investigate the effect of pile-up using multijet samples, γ, and Z data.
Figure 3.10: σ of a Gaussian fit to the x- and y-components of calibrated �ET vs. the
calibrated PF ET sum in a sample of events containing at least two jets with pT > 25
GeV. σ is calibrated such that the �ET scale is equal for all three algorithms. PF�
ET
is corrected, on average, to the particle level using a Pythia v8 simulation [48]. The
blue markers (data) and line (MC) refer to PF jets. Reprinted from Fig. 13 of ref.
[46].
R. Yohay March 29, 2012
Backgrounds
• Dominant: QCD with fake MET
• Diphoton
• γ + jet: 1 jet misidentified as a photon
• Multijet: at least 2 jets misidentified as photons
• Subdominant: electroweak processes with real MET
• W(→eν)γ: electron misidentified as a photon
• W(→eν) + jet: electron and jet misidentified as photons
21
R. Yohay March 29, 2012
Estimating the QCD background
• EM superior to hadronic energy resolution ⇒ fake MET due
entirely to jet mismeasurement
• Measure QCD background from data—control sample with well-measured EM objects to model the QCD fake MET spectrum
• Reweight events in control sample based on di-EM pT (kinematics) and Nj (hadronic activity)
• Normalize the predicted QCD fake MET spectrum to a signal-depleted region with MET < 20 GeV
22
Jets (poorly measured kinematics, source of fake MET)
EM objects (well measured kinematics, no fake MET)
di-EM pT (well-measured handle on the kinematics of the jet system)
Most energetic EM object2nd most energetic EM object
y
x
z (beam direction)
R. Yohay March 29, 2012
Reweighting
23
MET
Step 1: Find a control sample similar to the γγ search sample EM 1
EM 2
jets
Di-EM pT
Step 2: pT of di-EM system different between control and γγ samples ⇒
different MET ⇒ assign weight to
control event based on di-EM pT
Di-EM pT
(γγ
di-E
M p
T) /
(c
ontr
ol d
i-EM
pT)
Weights per di-EM pT bin
EM 1EM 2
jets
di-EM pT
γγ samplecontrol sample
MET
Step 4: Weight each event in control MET distribution with weights from steps 2-3.
Step 3: Repeat step 2 for events with 0 jets, 1 jet, and ≥2 jets.
R. Yohay March 29, 2012
QCD control samples• Z dielectron (ee)
• 81 GeV ≤ mee < 101 GeV
• Photon with inverted pixel seed veto ⇒ similar energy resolution as photons
• Di-EM pT reweighting significant because the kinematics of Z and QCD diphoton production are different
• Subtract tt contribution to ee sample using invariant mass sidebands
• Electromagnetic dijets (ff)
• Photon with inverted isolation or shower shape, below a maximum allowed isolation
• Tends to have a little bit of HCAL energy, so use PF ET instead of ECAL ET
• Similar kinematics to diphoton sample, so reweighting has small effect
24
Tight Isolation / shower shape Loose
Pixe
l see
d
No
Yes
ff control region
ee control region
γγ search region
R. Yohay March 29, 2012
Di-EM pT spectra
25
(GeV)T
Di-EM p0 20 40 60 80 100 120 140 160 180 200
1
10
210
310
, 0 jets-1CMS 4.7 fb!!
ee
ff
(GeV)T
Di-EM p0 20 40 60 80 100 120 140 160 180 200
1
10
210
, 1 jet-1CMS 4.7 fb!!
ee
ff
(GeV)T
Di-EM p0 20 40 60 80 100 120 140 160 180 200
1
10
2102 jets!, -1CMS 4.7 fb
""
ee
ff
• ee (red) and ff (blue) spectra area-normalized to γγ (black) spectrum
• wij = (Ncontrol/Nγγ)(Nγγij/Ncontrolij)
• i runs over di-EM pT bins
• j runs over Nj bins
0 jets 1 jet ≥2 jets
CMS Work in Progress CMS Work in Progress CMS Work in Progress
R. Yohay March 29, 2012
Estimating the electroweak background
• W(→eν)γ and W(→eν) + jet can fake γγ if the electron pixel seed is missed
• Estimate the electron→photon mis-ID rate fe→γ by fitting for the Z contribution in the ee and eγ samples
• fe→γ = 0.015 ± 0.002(stat.) ± 0.005(syst.)
• Systematic error due to small pT dependence of the mis-ID rate
• Scale eγ sample by fe→γ/(1 - fe→γ)
26
Tight Isolation / shower shape Loose
Pixe
l see
d
No
Yes
fake
electron
photon
eγ control region: 1 e + 1 γ
R. Yohay March 29, 2012
Results
27
[GeV]TE0 20 40 60 80 100 120 140 160 180
Num
ber o
f Eve
nts
/ GeV
-210
-110
1
10
210
310
410 Candidate Sample!!
Total Background UncertaintyQCDElectroweak
(1040/1520/375)!!GGM (1600/1280/375)!!GGM
-1 Ldt = 4.7 fb" = 7 TeV, sCMS Preliminary
At least 1 Jet Requirement
• ff sample used for the primary QCD background estimate
• Difference between ee and ff prediction taken as a systematic error
• Use MET > 50 GeV as search region
R. Yohay March 29, 2012
Upper limit calculation• CLS 95%
• Limits calculated in multiple MET bins and then combined
• 50-60 GeV, 60-80 GeV, 80-100 GeV, 100-140 GeV, 140-180 GeV, and ≥180 GeV
• Uncertainties on signal acceptance × efficiency, background, and integrated luminosity modeled with log-normal distributions
28
Source Systematic uncertainty (%)
Integrated luminosity 4.5
Background estimate ~48
Statistics ~32
Reweighting ~2.3
Normalization ~0.23
Electron→photon mis-ID rate 31
Difference between ff and ee ~35
Acceptance × efficiency 7-72
Photon ID efficiency correction 4
PDF error on cross section 4-66
PDF error on acceptance 0.1-9
Renormalization scale 4-28
R. Yohay March 29, 2012
Modeling the GMSB signal
29
)2 (GeV/cq~m500 1000 1500 2000
)2 (G
eV/c
g~m
500
1000
1500
2000
95%
CL
Upp
er L
imit
(pb)
-210
CMS Preliminary = 7 TeVs, -1dt = 4.7 fbL !
)2 (GeV/c0!m200 400 600 800 1000
)2 (G
eV/c
g~m
500
1000
1500
2000
95%
CL
Upp
er L
imit
(pb)
-210
CMS Preliminary = 7 TeVs, -1dt = 4.7 fbL "
NLSPg~
M2 decoupled, mneutralino = 375 GeV,
mgluino vs. msquark
Light squarks decoupled, mgluino vs.
mneutralino
R. Yohay March 29, 2012
Model exclusions
30
M2 decoupled, mneutralino = 375 GeV,
mgluino vs. msquark
Light squarks decoupled, mgluino vs.
mneutralino
)2 (GeV/cq~m500 1000 1500 2000
)2 (G
eV/c
g~m
500
1000
1500
2000
0!"GGM bino-like )2 = 375 (GeV/c0
!"mAt least 1 jet requirementNLO Limits
Observed (theory)#1±
Expected (theory)#1± (experimental)#1±
CMS Preliminary = 7 TeVs, -1dt = 4.7 fbL $
Excluded
)2 (GeV/c0!"m
500 1000 1500
)2 (G
eV/c
g~m
500
1000
1500
2000
)2 = 2500 (GeV/cq~mAt least 1 jet requirementNLO Limits
Observed (theory)#1±
Expected (theory)#1± (experimental)#1±
CMS Preliminary = 7 TeVs, -1dt = 4.7 fbL $
Excluded NLSPg~
R. Yohay March 29, 2012
Conclusions
31
• Top performing ECAL is an important instrument in the quest for new physics at CMS
• Steadily moving toward design calibration precision
• Utilized in Higgs, SUSY, and Exotica searches
• Searches in the diphoton final state are powerful tools for observing SUSY
• Clean trigger objects
• Dominant background estimated from data
• 4.7 fb-1 search has excluded gluinos and light squarks in the GMSB scenario below ~1 TeV
• Best exclusions to date on gauge mediation
• As energy and luminosity increase, different variants on the diphoton search can be explored to give the best coverage of possible SUSY scenarios
• Looking forward to 2012!
R. Yohay March 29, 2012
Backup
32
R. Yohay March 29, 2012
ECAL performance
33
!"#$"%&&'()'*+,-'./0(1"/2#)'
3()$0%'%0%.4"#)'%)%"$5'&./0%'6*789'&4/1(0(45'()'4:%'*+,-';/""%0'<#"'=>??''+@3'A/4/'
B%/&C"%A'C&()$'D'!%E'%F%)4&'
GF%"/00'%H%.4'#<'&()$0%'.:/))%0'()4%"I./0(1"/2#)'/)A'4"/)&8/"%).5'
.#""%.2#)&'#)'4:%'J!%%'()F/"(/)4'B/&&'()'4:%'*+,-';/""%0'
!"#$"%&&'()'*+,-'./0(1"/2#)'
3()$0%'%0%.4"#)'%)%"$5'&./0%'6*789'&4/1(0(45'()'4:%'*+,-';/""%0'<#"'=>??''+@3'A/4/'
B%/&C"%A'C&()$'D'!%E'%F%)4&'
GF%"/00'%H%.4'#<'&()$0%'.:/))%0'()4%"I./0(1"/2#)'/)A'4"/)&8/"%).5'
.#""%.2#)&'#)'4:%'J!%%'()F/"(/)4'B/&&'()'4:%'*+,-';/""%0'
[3]
[3]
Effect of intercalibration and transparency corrections on the Z→ee mass peak position and width
W relative E/p vs. time
!
!"#$"%&'(%")*'&+,-,'.
!"#$%&'('%)%*$+%&'('
[10] RMS deviation of supercrystal temperature measurements over 2010 running
R. Yohay March 29, 2012
Comparison of crystal properties
34
1 General Overview CMS–ECAL TDR
4
the operating conditions at the LHC. As a consequence of these efforts, the initial drawback of lowlight yield has been overcome by progress in crystal growth and through the development of large-area silicon avalanche photodiodes. In Table 1.1 the properties of PbWO
4
are compared with thoseof other crystals used in electromagnetic calorimeters.
1.3.1 Developments since the Technical Proposal
Since the submission of the CMS Technical Proposal in December 1994 [1.4] substantialprogress has been made on the following aspects of the electromagnetic calorimeter project:
– crystal parameters such as scintillation speed (decay time constant), mechanical toleranceand light yield have been significantly improved;
– it has been shown that irradiation does not affect the scintillation mechanism and that theintrinsic energy resolution is not degraded. Nevertheless, radiation damage affects thecrystal transparency through the formation of colour centres, causing a loss in the amountof light collected. It has been shown that the calorimeter performance can be maintainedby following the light loss using a laser-based monitoring system. Extensive studies onradiation damage have led to a better understanding of its underlying causes and to theproduction of acceptably radiation-hard crystals in a reproducible way;
– avalanche photodiodes (APD) are used to collect the scintillation light in the barrel regionsince they are able to provide gain in the presence of the high transverse magnetic field.Their performance has been improved in a successful collaboration with two industrialfirms. This R&D programme has achieved a better quantum efficiency, a reducedsensitivity to the passage of charged particles, and a higher radiation tolerance.
In the endcaps the photodetectors are required to survive a much higher integratedradiation dose (50 kGy) and neutron fluence (7
!
10
14
n/cm
2
). At these fluences theinduced leakage current for APDs would lead to an unacceptable energy equivalent of
Table 1.1:
Comparison of properties of various crystals
NaI(Tl) BGO CSI BaF
2
CeF
3
PbWO
4
Density [g/cm
3
] 3.67 7.13 4.51 4.88 6.16 8.28
Radiation length [cm] 2.59 1.12 1.85 2.06 1.68 0.89
Interaction length [cm] 41.4 21.8 37.0 29.9 26.2 22.4
Molière radius [cm] 4.80 2.33 3.50 3.39 2.63 2.19
Light decay time [ns] 230 60300
16 0.9630
825
5 (39%)15 (60%)100 (1%)
Refractive index 1.85 2.15 1.80 1.49 1.62 2.30
Maximum of emission [nm] 410 480 315 210310
300340
440
Temperature coefficient [%/
°
C] ~0 –1.6 –0.6 –2/0 0.14 –2
Relative light output 100 18 20 20/4 8 1.3
[2]
R. Yohay March 29, 2012
Expected radiation dose
35
CMS–ECAL TDR 1 General Overview
9
The dose
The integrated dose at the shower maximum in the crystals and around the crystals atvarious values of the pseudorapidity is given in Fig. 1.4. The integrated dose at shower maximumin the barrel crystals is about 4 kGy whereas it rises to about 90 kGy and 200 kGy at |!| = 2.6 and|!| = 3 respectively.
Fig. 1.4: The dose and neutron fluence in and around the crystals as a function ofpseudorapidity. Numbers in bold italics are doses, in kGy, at shower maximum and atthe rear of the crystals. The other numbers are fluences immediately behind the crystals,in the space for endcap electronics surrounded by moderators and in the silicon of thepreshower in units of 1013 cm–2. All values correspond to an integrated luminosity of5 " 105 pb–1 appropriate for the first ten years of LHC operation.
The neutron fluence
Hadron cascades in the crystals lead to a large neutron albedo emerging from the ECAL.The fluence of neutrons with energies above 100 keV is shown in Fig. 1.4 for various values ofpseudorapidity. Knowledge of the neutron fluence behind the barrel part of the ECAL is importantfor estimating the increase in APD leakage currents due to radiation damage.
Because of the large neutron fluence in the endcap, VPTs have been chosen for thisregion. The front-end electronics are separated from the VPTs and located in a space surroundedby polyethylene moderators. The neutron fluence and absorbed dose in this region, at a radius aslow as 60 cm, are below 7 " 1013 n cm–2 and 13 kGy respectively.
With an optimized design of moderators, the neutron fluence at the location of Si detectorsof the endcap and barrel preshowers can be kept below 15 " 1013 n cm–2 and 2 " 1013 n cm–2
respectively.
EE
Tracker
HB
EB
Patch Panel space
! = 1.479
! = 2
! = 2.6
! = 3
4 2.5
0.20.35
15
90200
42
1.2
2
5
1.75
0 m
1.29
0 m
3.045 m3.900 m
15
6.5
2
SE
EB
33
13
0.50.75
2.5
7
1.6
702050
[2]
R. Yohay March 29, 2012
Previous VPT rate effect
36
Fig. 5. Relative variation of the VPT gain averaged over 20
spills as a function of time for different average VPT anode
currents.
[16]
R. Yohay March 29, 2012
LED running modes
• Calibration sequence
• Few hundred LED pulses read out (readout rate ~100 Hz) for each EE monitoring region
• Continuous monitoring of the VPTs and crystals, to complement the laser monitoring
• Local run
• Short sequence of a few hundred LED pulses triggered by ECAL-generated trigger and read out
• Useful for debugging the system and checking the health of the ECAL
• Soaking
• Fire the blue LED stability pulses all the time in the abort gaps to dampen VPT gain changes
• Frequency up to ~11.4 kHz (use only 100 Hz right now)
37
R. Yohay March 29, 2012
Flavor blindness of GMSB
38
• SUSY must be a broken symmetry — if not, each SM particle and its superpartner would have the same mass, and the superpartners would have been discovered already
• SUSY-breaking terms in the minimal SUSY Lagrangian generically allow for lepton flavor violating decays:
• The relation cee = cµµ = cττ = m2 (and similar for the other fermion families) arises naturally in GMSB, because the sparticle masses (i.e. the diagonal terms) only depend on their gauge couplings, which are identical for the three families
�eR
�µR
�τR
cee ceµ ceτ
cµe cµµ cµτ
cτe cτµ cττ
��eR �µR �τR
�
1
if not proportional to the unit matrix, will lead to, for example, μ → eγ
R. Yohay March 29, 2012
GMSB searches with photons
39Figure 3: The E/T significance, showing the separation power for mismeasured QCD and electroweak sources ofE/T (left), and the limits in the mχ0
1and χ0
1 lifetime plane (right).
all selection, the dominant background originates from electroweak production of a Z0with two
photons, where the Z0decays to neutrinos and results in significant E/T . The total predicted
background in this search is 1.4± 0.4 events, and zero are observed. Limits are set at 95% C.L.
in the mχ01and χ0
1 lifetime plane, shown in Figure 3. For a short-lived neutralino with lifetime
less than a nanosecond, the limit is 150 GeV/c2.
5 Search for Hidden Valley Dark Photons
Recently, a subset of HV models with SUSY have emerged which may simultaneously explain
anomalies of the observed spectra of high-energy cosmic ray positrons and electrons, and provide
a solution to the origin of Dark Matter. They propose a scenario where dark matter can
annihilate to pairs of so-called dark photons, which then would potentially themselves decay to
pairs of e+e− or µ+µ−through their mixing with the SM photon. At the Tevatron, gaugino
production could lead to cascade decays of charginos and neutralinos which could include decays
to both SM photons and these dark photons, whose favored mass range is on the order of 1 GeV.
The signature, then, will be a SM photon, two opposite charge leptons spatially close to each
other in the detector, and E/T from the escaping LSP. DØ has searched for this signature with
4.1fb−1
of Run II data4.
Events are selected by requiring a single photon and two leptons. The proximity of the
leptons to each other requires a careful treatment of the isolation requirement placed on the
leptons. The main background is QCD with photon conversions producing pairs of leptons.
After selection, the dilepton invariant mass is used as a signal discriminant, where the signal
would appear as an excess at the dark photon mass. No significant excess is observed, and
exclusion limits are set on the lightest chargino mass as a function of the dark photon mass,
shown in Figure 4.
6 Search for Hidden Valley Long-lived Particles
Another class of HV models predicts a new, confining gauge group, weakly coupled to the SM.
Stable configurations of HV hadrons, “v-hadrons” in this model, can be long-lived. In one
benchmark model, the SM Higgs boson mixes with the HV Higgs boson. In this scenario, the
SM Higgs could decay, through the mixing with the HV Higgs, to pairs of v-hadrons. The
v-hadrons then preferentially couple to the heavier SM quarks due to a helicity suppression
[20]
LEP (1989-2000)•Minimal GMSB model•Neutralino pair production•mneutralino > 97 GeV for short-lived neutralino
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
50 60 70 80 90 100
m(e~) = 2.0m(!
~o1)
m(e~) = 1.1m(!
~o1)
!~o
1 mass (GeV/c2)
"LI
MIT
at #
s = 2
08 G
eV (p
b)
ExpectedObservedObserved
ALEPH DELPHI L3 OPAL130$#s$209 GeV
preliminary
50
60
70
80
90
100
110
120
100 200 300 400 500m(e
~ ) (GeV/c2)
m(!~ o 1)
(GeV
/c2 )
Excluded at 95% C.L.
e+e- % !~o
1!~o
1 % G~
G~&&
130$#s$209 GeV
ALEPH DELPHI L3 OPAL
Fig. 1. Left: cross section lower limit versus neutralino mass from the combined two-photonsearches. The production cross sections are calculated in two scenarios. Right: excluded area inthe selectron-neutralino mass plane assuming a purely bino neutralino. The dashed overlayedarea is the region consitent with the GMSB interpretation of the ee!!+ !ET CDF event [9], nowcompletely excluded.
results from the ALEPH, DELPHI and OPAL collaborations exists with data from 192to 208GeV and yields no excess with respect to the SM expectation [4]. The individualresults can be found in Refs. [5,6,7,8]. Figure 1 shows the interpretation of these resultsassuming a purely bino neutralino and mass degeneracy between eL and eR. Neutralinopair-production cross sections above 0.02 pb are excluded at 95% C.L. Since the crosssection depends on the mass of the exchanged selectron, the lower limit on the neutralinomass can be conservatively estimated to be 97 GeV/c2 if me = 2 ·m!0
1, improving for larger
!01 " e mass di!erences.
If the gravitino mass lies in the range between a few eV/c2 and a few hundred eV/c2, theneutralino NLSP will decay somewhere inside the detector. The probability for only one ofthe pair-produced neutralinos to decay before the electromagnetic calorimeter is greaterthan for both of them [2], thus the expected final topology consists in a single non-pointingphoton. ALEPH [5] and DELPHI [6] have searched for single photons with a minimumimpact parameter of 40 cm and have found 2/1.0 and 5/3.0 candidates/background, re-spectively in their
#s = 189 " 209 and
#s = 202 " 209 data sets. The e"ciency for
this search reaches a maximum of around 10% for decay lengths of 8 m. This yields anapproximate upper limit of 0.4 pb in the production cross section.
3 Two leptons and !E signatures
We discuss next the pair-production of sleptons in the slepton NLSP scenario. In generalthe "1 will be the lighter slepton, but a degenerate case is possible in which the three "1,eR and µR act as co-NLSP’s. Thus searches for pairs of sleptons decaying into their SMcounterpart and a gravitino were developed by all four collaborations taking into accountthe possible lifetime of the slepton.
3
[17]7
[TeV]!80 90 100 110 120 130 140
[fb]
"
1
10
210
[GeV]01
#m
120 140 160 180
[GeV]±
1#m
220 240 260 280 300 320 340 360
NLO cross sectionobserved limitexpected limit
" 1 ±expected limit " 2 ±expected limit
-1 6.3 fb$DSPS8 GMSB SUSY (Prospino 2.1)
[GeV]-1cR
380 400 420 440 460 480 500
[fb]
"
1
10
210
[GeV]D
q*m460 480 500 520 540 560 580
[GeV]g*m480 500 520 540 560 580 600 620
LO cross sectionobserved limitexpected limit
" 1 ±expected limit " 2 ±expected limit
-1 6.3 fb$D=20cR!%=1 UED (PYTHIA 6.421), &
FIG. 2: The predicted cross section for the benchmark GMSB and UED models, and 95% C.L. expected and observed exclusion limits, as a
function of Λ and R−1c , respectively. For the GMSB model, corresponding masses are shown for the lightest chargino, χ±
1, and neutralino, χ0
1.
For the UED model, corresponding masses are shown for the KK quark, q∗D, and KK gluon, g∗. The γ∗ mass is approximately equal to R−1c .
092002 (1999).
[14] D. Acosta et al. (CDF Collaboration), Phys. Rev. D 71, 031104
(2005).
[15] T. Aaltonen et al. (CDF Collaboration), Phys. Rev. Lett. 104,
011801 (2010).
[16] V.M. Abazov et al. (D0 Collaboration), Phys. Rev. Lett. 94,
041801 (2005).
[17] V.M. Abazov et al. (D0 Collaboration), Phys. Lett. B 659, 856
(2008).
[18] V.M. Abazov et al. (D0 Collaboration), Phys. Rev. Lett. 102,
231801 (2009); V.M. Abazov et al. (D0 Collaboration), Phys.
Lett. B 690, 108 (2010).
[19] V.M. Abazov et al. (D0 Collaboration), Nucl. Instrum. Methods
Phys. Res. A 565, 463 (2006).
[20] T. Andeen et al., FERMILAB-TM-2365 (2007).
[21] G.C. Blazey et al., arXiv:hep-ex/0005012.
[22] T. Sjostrand et al., Comput. Phys. Commun. 135, 238 (2001).
[23] R. Brun and F. Carminati, CERN Program Library Long
Writeup W5013, 1993 (unpublished).
[24] U. Baur et al. Phys. Rev. D 48, 5140 (1993).
[25] J. Cortes et al. Nucl. Phys. B 278, 26 (1986).
[26] J. Alwall et al., J. High Energy Phys. 09, 028 (2007).
[27] B.C. Allanach et al., Eur. Phys. J. C 25, 113 (2002).
[28] A. Djouadi et al., Acta Phys. Polon. B 38, 635 (2007).
[29] M. ElKacimi et al., Comput. Phys. Commun. 181, 122 (2010).
[30] J. Pumplin et al., J. High Energy Phys. 07, 012 (2002); D.
Stump et al., J. High Energy Phys. 10, 046 (2003).
[31] T. Junk, Nucl. Instrum. Methods A 434, 435 (1999); W. Fisher,
FERMILAB-TM-2386-E (2006).
Tevatron Run II (2001-2010)•Minimal GMSB model (SPS8) [18]•Chargino and neutralino pair production•mneutralino > ~170 GeV (Λ > 124 TeV) for short-lived neutralino
[19]
LHC7 (2009-2011)•Simplified model parametrized in terms of tan β and squark, gluino, and wino/bino/higgsino masses•No assumptions on the number of messengers, the messenger mass, or the SUSY breaking scale•Squark and gluino production•Short-lived neutralino
R. Yohay March 29, 2012
CMS longitudinal cross section
40
[2]
interaction pointbeam line
EB edge|η| = 1.479
HE edge|η| = 2.6
ME edge|η| = 2.1
HF edge|η| = 5.0
R. Yohay March 29, 2012
Trigger definitions
41
•IsoVL
• IECAL < 0.012ET + 6 GeV
• IHCAL < 0.005ET + 4 GeV
• Itrack < 0.002ET + 4 GeV
•R9Id
•R9 > 0.85
R. Yohay March 29, 2012
Photon isolation criteria
42
• IECAL - 0.0792ρ + IHCAL - 0.0252ρ + Itrack < 6 GeV
• ρ = average energy density per unit area in the calorimeters as measured by Fastjet
HCAL energy in R < 0.15 cone around photon candidateECAL energy of photon candidate < 0.05
CMS–ECAL TDR 1 General Overview
5
electronics noise and hence vacuum phototriodes (VPTs) have been chosen for thisregion. Test results using VPTs demonstrate that they can fulfil the endcap requirements;
– considerable progress has been made in developing the readout electronics. The analogpart consists of a multi-slope preamplifier and a gain-ranging ADC. The analogcomponents have been produced in radiation-hard technology;
– a prototype crystal matrix (7
!
7 crystals) read out with APDs has been tested in a high-energy electron beam at CERN and has achieved an excellent energy resolution of 0.5%at 120 GeV;
– the Proto97 matrix with near-final mechanics for crystal support and preamplifier-crystalinterface, as well as a full light-to-light readout including fibre-optic communication hasbeen successfully tested during September and November 1997;
– a preshower detector consisting of two lead/silicon detector layers will be placed in frontof the endcap calorimeter. A test of a small prototype including the complete electronicchain operating at 40 MHz has shown that the measured position and energy resolutionmeet the design requirements;
– detailed performance studies, carried out using GEANT simulations of the ECALincluding the effects of electronics and pileup noise as well as the material in front of thecalorimeter, have shown that the design figures for resolution and efficiency can beachieved.
In addition to this progress achieved since the submission of the Technical Proposal,overall optimization of the calorimeter project has been vigorously pursued. This optimization hasalso taken into account the desire to ensure full geometrical coverage, the requirements of thesurrounding detectors, as well as matching the cost to the available financial resources.
A schematic view of the calorimetry and tracking system is shown in Fig. 1.2.
Fig. 1.2:
Schematic view of one quadrant of the calorimetry and tracking system.
EE
COIL
! = 2.6! = 3 Tracker
HB
EB HE
! = 0.9
Silicon Strips + Pixels
! = 1.479
MSCGs
1.71
1 m
3.900 m
1.75
0 m
3.045 m
SE
1.29
0 m
PATCHPANEL
CRYSTALS
0.185m
SB
3.170 m
HCAL
Highest energy (photon seed) crystal
2
neous luminosity, rate-reduction factors were applied to the triggers at 20 GeV. Consequently,
photons with EγT< 26 GeV are taken from a restricted data-set having an integrated luminosity
of 2.1 ± 0.2 pb−1
. No photon isolation criteria are applied at the trigger level.
The event selection requires at least one reconstructed primary interaction vertex consistent
with a pp collision [17]. The time of the ECAL signals is required to be compatible with that
of collision products [18]. Topological selection criteria are used to suppress direct interactions
in the ECAL APDs [19]. The residual contamination has an effect smaller than 0.2% on the
measured cross section over the entire EγT
range considered. Contamination from non-collision
backgrounds is estimated to be negligible [16].
Photon candidates are built from ECAL energy clusters fully contained in the barrel section.
The photon candidate pseudorapidity is corrected for the position of the primary interaction
vertex. The absolute photon energy scale is determined with electrons from reconstructed Z-
boson decays with an uncertainty estimated to be less than 1%. Consistent results are obtained
with low-energy photons from π0decays. The linearity of the response of detector and elec-
tronics has been measured with laser light and test beams, to a precision better than 1% in the
energy range probed in this Letter [13]. Showers initiated by charged hadrons are rejected by
requiring EHCAL/Eγ < 0.05, where EHCAL
is the sum of energy in the HCAL towers within
R < 0.15, with R2 = (η − ηγ)2 + (φ − φγ)2. Electrons are rejected by requiring the absence of
hits in the first two layers of the pixel detector that would be consistent with an electron track
matching the observed location and energy of the photon candidate (pixel veto requirement).
The photon candidates must satisfy three isolation requirements that reject photons produced
in hadron decays: (1) IsoTRK < 2 GeV/c, where IsoTRK is the sum of the pT of tracks compatible
with the primary event vertex in an annulus 0.04 < R < 0.40, excluding a rectangular strip
of ∆η × ∆φ = 0.015 × 0.400 to remove the photon’s own energy if it converts into an e+e−pair; (2) IsoECAL < 4.2 GeV, where IsoECAL is the transverse energy deposited in the ECAL
in an annulus 0.06 < R < 0.40, excluding a rectangular strip of ∆η × ∆φ = 0.04 × 0.40; and
(3) IsoHCAL < 2.2 GeV, where IsoHCAL is the transverse energy deposited in the HCAL in an
annulus 0.15 < R < 0.40. The requirements were designed with two other objectives in mind.
First, the use of relatively loose photon identification and isolation selection criteria reduces the
dependence of the results on the details of the simulation of the detector noise, the underlying
event, and event pile-up. Second, the shape of the isolation regions is designed to allow the
use of electrons to determine the efficiency of the isolation requirements in data. The isolation
requirements also reduce the uncertainty on the signal due to the knowledge of the photon
fragmentation functions. In total, 4 × 105
photon candidates fulfill the selection criteria.
While the isolation requirements remove the bulk of the neutral-meson background, a substan-
tial contribution remains, mainly caused by fluctuations in the fragmentation of partons, where
neutral mesons carry most of the energy and are isolated. A modified second moment of the
electromagnetic energy cluster about its mean η position, σηη , is used to measure the isolated
prompt photon yield. It is calculated as
σ2
ηη =25
∑i=1
wi(ηi − η)2/
25
∑i=1
wi,
where wi = max(0, 4.7+ ln(Ei/E)), Ei is the energy of the ithcrystal in a group of 5× 5 centred
on the one with the highest energy, and ηi = ηi × δη, where ηi is the η index of the ith crystal
and δη = 0.0174; E is the total energy of the group and η the average η weighted by wi in the
same group [20]. Since σηη expresses the extent in η of the cluster, it discriminates between
clusters belonging to isolated prompt photons, for which the σηη distribution is very narrow
2
neous luminosity, rate-reduction factors were applied to the triggers at 20 GeV. Consequently,
photons with EγT< 26 GeV are taken from a restricted data-set having an integrated luminosity
of 2.1 ± 0.2 pb−1
. No photon isolation criteria are applied at the trigger level.
The event selection requires at least one reconstructed primary interaction vertex consistent
with a pp collision [17]. The time of the ECAL signals is required to be compatible with that
of collision products [18]. Topological selection criteria are used to suppress direct interactions
in the ECAL APDs [19]. The residual contamination has an effect smaller than 0.2% on the
measured cross section over the entire EγT
range considered. Contamination from non-collision
backgrounds is estimated to be negligible [16].
Photon candidates are built from ECAL energy clusters fully contained in the barrel section.
The photon candidate pseudorapidity is corrected for the position of the primary interaction
vertex. The absolute photon energy scale is determined with electrons from reconstructed Z-
boson decays with an uncertainty estimated to be less than 1%. Consistent results are obtained
with low-energy photons from π0decays. The linearity of the response of detector and elec-
tronics has been measured with laser light and test beams, to a precision better than 1% in the
energy range probed in this Letter [13]. Showers initiated by charged hadrons are rejected by
requiring EHCAL/Eγ < 0.05, where EHCAL
is the sum of energy in the HCAL towers within
R < 0.15, with R2 = (η − ηγ)2 + (φ − φγ)2. Electrons are rejected by requiring the absence of
hits in the first two layers of the pixel detector that would be consistent with an electron track
matching the observed location and energy of the photon candidate (pixel veto requirement).
The photon candidates must satisfy three isolation requirements that reject photons produced
in hadron decays: (1) IsoTRK < 2 GeV/c, where IsoTRK is the sum of the pT of tracks compatible
with the primary event vertex in an annulus 0.04 < R < 0.40, excluding a rectangular strip
of ∆η × ∆φ = 0.015 × 0.400 to remove the photon’s own energy if it converts into an e+e−pair; (2) IsoECAL < 4.2 GeV, where IsoECAL is the transverse energy deposited in the ECAL
in an annulus 0.06 < R < 0.40, excluding a rectangular strip of ∆η × ∆φ = 0.04 × 0.40; and
(3) IsoHCAL < 2.2 GeV, where IsoHCAL is the transverse energy deposited in the HCAL in an
annulus 0.15 < R < 0.40. The requirements were designed with two other objectives in mind.
First, the use of relatively loose photon identification and isolation selection criteria reduces the
dependence of the results on the details of the simulation of the detector noise, the underlying
event, and event pile-up. Second, the shape of the isolation regions is designed to allow the
use of electrons to determine the efficiency of the isolation requirements in data. The isolation
requirements also reduce the uncertainty on the signal due to the knowledge of the photon
fragmentation functions. In total, 4 × 105
photon candidates fulfill the selection criteria.
While the isolation requirements remove the bulk of the neutral-meson background, a substan-
tial contribution remains, mainly caused by fluctuations in the fragmentation of partons, where
neutral mesons carry most of the energy and are isolated. A modified second moment of the
electromagnetic energy cluster about its mean η position, σηη , is used to measure the isolated
prompt photon yield. It is calculated as
σ2
ηη =25
∑i=1
wi(ηi − η)2/
25
∑i=1
wi,
where wi = max(0, 4.7+ ln(Ei/E)), Ei is the energy of the ithcrystal in a group of 5× 5 centred
on the one with the highest energy, and ηi = ηi × δη, where ηi is the η index of the ith crystal
and δη = 0.0174; E is the total energy of the group and η the average η weighted by wi in the
same group [20]. Since σηη expresses the extent in η of the cluster, it discriminates between
clusters belonging to isolated prompt photons, for which the σηη distribution is very narrow
< 0.011
[2]
[21]
η
φ
not to scale
Tracker
0.30.15
included in isolation sumexcluded from isolation sum
0.3
0.06
0.087
ECAL HCAL
0.3
0.04
0.03
R. Yohay March 29, 2012
ECAL noise cleaning
43
• Form 3 × 3 matrix of crystals around the photon seed crystal
• Find the 2 highest energy crystals within the matrix
• If the sum of the energies of the 2 highest energy crystals divided by the sum of the energies of all 9 crystals within the matrix exceeds 0.95, reject the photon as ECAL noise
Highest energy crystal
2nd highest energy crystal
E + E E3×3
> 0.95 ⇒ reject
R. Yohay March 29, 2012
Photon ID variables
44
2.5 Photon identification and isolation 7
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442$|
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
900
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5$1.566 < |
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
900
Figure 6: N − 1 ECAL isolation distribution for data and MC, shown for barrel (right) and
endcap (left). The Monte Carlo results are normalized separately for each plot to the number
of entries in the data histogram.
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442$|
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
1600
1800
2000
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5$1.566 < |
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
Figure 7: N − 1 HCAL isolation distribution for data and MC, shown for barrel (right) and
endcap (left). The Monte Carlo results are normalized separately for each plot to the number
of entries in the data histogram.
2.5 Photon identification and isolation 7
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442$|
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
900
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5$1.566 < |
R = 0.4 (GeV)! Isolation T E"ECAL -2 0 2 4 6 8 10 12 14
can
dida
tes/
0.5G
eV#
0
100
200
300
400
500
600
700
800
900
Figure 6: N − 1 ECAL isolation distribution for data and MC, shown for barrel (right) and
endcap (left). The Monte Carlo results are normalized separately for each plot to the number
of entries in the data histogram.
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442$|
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
1600
1800
2000
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5$1.566 < |
R = 0.4 (GeV)! Isolation T E"HCAL 0 2 4 6 8 10
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
Figure 7: N − 1 HCAL isolation distribution for data and MC, shown for barrel (right) and
endcap (left). The Monte Carlo results are normalized separately for each plot to the number
of entries in the data histogram.
8 2 Supercluster and Photon reconstruction, corrections and observables
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442$|
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5$1.566 < |
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
Figure 8: N − 1 Track isolation distribution for data and MC, shown for barrel (right) andendcap (left). The Monte Carlo results are normalized separately for each plot to the numberof entries in the data histogram.
! i!i"Photon 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
can
dida
tes/
0.00
8#
0
200
400
600
800
1000
1200
1400 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442!|
! i!i"Photon 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
can
dida
tes/
0.00
8#
0
200
400
600
800
1000
1200
1400
! i!i"Photon 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
can
dida
tes/
0.02
#
0
100
200
300
400
500
600
700
800
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5!1.566 < |
! i!i"Photon 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
can
dida
tes/
0.02
#
0
100
200
300
400
500
600
700
800
Figure 9: The σiηiη shower shape variable for barrel and endcap photon candidates. The N − 1distributions are shown before cutting on the variables for photon identification. The MonteCarlo results are normalized separately for each plot to the number of entries in the data his-togram.
8 2 Supercluster and Photon reconstruction, corrections and observables
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442$|
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
1200
1400
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5$1.566 < |
R = 0.4 (GeV/c)! Isolation T
p"Track 0 5 10 15 20
can
dida
tes/
0.5G
eV#
0
200
400
600
800
1000
Figure 8: N − 1 Track isolation distribution for data and MC, shown for barrel (right) andendcap (left). The Monte Carlo results are normalized separately for each plot to the numberof entries in the data histogram.
! i!i"Photon 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
can
dida
tes/
0.00
8#
0
200
400
600
800
1000
1200
1400 Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 1.4442!|
! i!i"Photon 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
can
dida
tes/
0.00
8#
0
200
400
600
800
1000
1200
1400
! i!i"Photon 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
can
dida
tes/
0.02
#
0
100
200
300
400
500
600
700
800
Data partonic#MC ISR/FSR#MC
MC other
CMS Preliminary 2010 = 7 TeVs
-1L = 74 nb| < 2.5!1.566 < |
! i!i"Photon 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
can
dida
tes/
0.02
#
0
100
200
300
400
500
600
700
800
Figure 9: The σiηiη shower shape variable for barrel and endcap photon candidates. The N − 1distributions are shown before cutting on the variables for photon identification. The MonteCarlo results are normalized separately for each plot to the number of entries in the data his-togram.
14 5 Summary
Figure 16: Data. Super cluster position vs. time of photon seed. Prompt and candidate events arrive
at the EB at t = 0. Halo events primarily arrive out-of-time with respect to prompt events. The black
curves show the expected arrival time for halo particles due to the path length difference to reach ECAL.
contribution to the “candidate” sample below 5.9 events at 95% CL.
5 SummaryUsing the first 7 TeV LHC data delivered to CMS, we have carried out a series of studies on
photon candidates. The basic reconstruction quantities for photons have been compared to the
Monte Carlo simulation with good agreement. Using a selection which is intended to enrich
the sample in signal photons and suppress the background from QCD, we have illustrated that
the photon reconstruction and identification performance is similar to that expected from sim-
ulation. First results on the ECAL-seeded conversion finding also show good agreement with
simulation. Finally, we have shown that the number of fakes due to non-collision backgrounds
in the sample is under control.
The commissioning of the photon object at CMS is still in progress. Future results with a large
sample of high pT candidates and control samples of high-PT electrons will improve on the
results shown here and lessen the dependence on simulation.
[22]
R. Yohay March 29, 2012
Sample definitions
45
65
Table 4.1: Selection criteria for γγ, eγ, ee, and ff events.
VariableCut
γγ eγ ee ff
HLT match IsoVL IsoVL IsoVL IsoVL || R9Id
ET> 40/
> 25 GeV> 40/
> 25 GeV> 40/
> 25 GeV> 40/
> 25 GeVSC |η| < 1.4442 < 1.4442 < 1.4442 < 1.4442H/E < 0.05 < 0.05 < 0.05 < 0.05R9 < 1 < 1 < 1 < 1
Pixel seed No/No Yes/No Yes/Yes No/No
Icomb, σiηiη< 6 GeV &&
< 0.011< 6 GeV &&
< 0.011< 6 GeV &&
< 0.011
< 20 GeV &&(≥ 6 GeV ||≥ 0.011)
JSON Yes Yes Yes YesNo. good PVs ≥ 1 ≥ 1 ≥ 1 ≥ 1
∆REM > 0.6 > 0.6 > 0.6 > 0.6∆φEM ≥ 0.05 ≥ 0.05 ≥ 0.05 ≥ 0.05
Table 4.2: Definition of HB/HE/HF hadronic jets. Add a footnote describing thePF electron and PF muon definitions, with references.
Variable Cut
Algorithm L1FastL2L3Residual corrected PF (cf. Sec. 3.1.3)pT > 30 GeV|η| < 5.0
Neutral hadronicenergy fraction
< 0.99
Neutral electromagneticenergy fraction
< 0.99
Number of constituents > 1Charged hadronic energy > 0.0 GeV if |η| < 2.4
Number of charged hadrons > 0 if |η| < 2.4Charged electromagnetic
energy fraction< 0.99 if |η| < 2.4
∆R to nearest electron, muon,or one of the two
primary EM objects> 0.5
R. Yohay March 29, 2012
Effect of reweighting
46
(GeV)TME0 50 100 150200 250 300 350 400 450 500
10
210
310
410, no jet requirement-1CMS 4.7 fb
ee full reweighting
ee no reweighting
(GeV)TME0 50 100 150200 250 300 350 400 450 500
1
10
210
310
410, no jet requirement-1CMS 4.7 fb
ff full reweighting
ff no reweighting
(GeV)TME0 10 20 30 40 50 60 70
ffNeeN
0.75
0.8
0.85
0.9
0.95
1
1.05
1.1
, no jet requirement-1CMS 4.7 fbFull reweightingNo reweighting
, no jet requirement-1CMS 4.7 fbFull reweightingNo reweighting
(GeV)TME0 50 100 150 200 250 300 350 400 450 500
ffNeeN
0.5
1
1.5
2
2.5
3
3.5
, no jet requirement-1CMS 4.7 fbFull reweightingNo reweighting
CMS Work in Progress CMS Work in Progress
CMS Work in ProgressCMS Work in Progress
R. Yohay March 29, 2012
PF vs. ECAL ET (1)
47
1
10
210
310
410
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tj)/p
Te -
pTj
(p
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
1
10
210
310
410
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tj)/p
Te -
pTj
(p
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
1
10
210
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tj)/p
Tf -
pTj
(p-1
-0.8-0.6-0.4-0.2
00.20.40.60.8
1
1
10
210
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tj)/p
Tf -
pTj
(p
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
1
10
210
310
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tj)/p !T
- p
Tj(p
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
1
10
210
310
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Tj)/p !T
- p
Tj(p
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
Leading e
Trailing e
Leading f
Trailing f
Leading γ
Trailing γ
CMS Work in Progress CMS Work in Progress CMS Work in Progress
CMS Work in Progress CMS Work in Progress CMS Work in Progress
R. Yohay March 29, 2012
PF vs. ECAL ET (2)
48
EMF0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
>Tj
)/pTE
M -
pTj
<(p
0
0.1
0.2
0.3
0.4
0.5
, no jet requirement-1CMS 4.7 fbTrailing fLeading f
!Trailing !Leading
Trailing eLeading e
• Profile histograms of previous slide
CMS Work in Progress
R. Yohay March 29, 2012
Di-EM pT weights
49
(GeV)T
Di-EM p0 100 200 300 400 500 600
Wei
ght
0.5
1
1.5
2
2.5
3
3.5
4
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
5
10
15
20
25
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
1
2
3
4
5
6
7
(GeV)T
Di-EM p0 100 200 300 400 500 600
Wei
ght
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
2.4
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
0.5
1
1.5
2
2.5
3
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
2
4
6
8
10
CMS Work in Progress CMS Work in Progress CMS Work in Progress
CMS Work in Progress CMS Work in Progress CMS Work in Progress
ee 0 jets ee 1 jet ee ≥2 jets
ff 0 jets ff 1 jet ff ≥2 jets
R. Yohay March 29, 2012
Removing ttbar from the ee sample
50
• ee sample contains a non-negligible high-MET background of ttbar events
• 71 GeV ≤ mee < 81 GeV and 101 GeV ≤ mee < 111 GeV sidebands used to estimate the non-Z background in the 81 GeV ≤ mee < 101 GeV ee sample
• Reweight the low and high sidebands independently, using weights derived from events in those sidebands
• Subtract the low and high sideband MET distributions from the Z signal MET distribution
• Proceed with normalization of the sideband-subtracted ee sample
(GeV)TME0 50 100 150200 250 300 350 400 450 500
10
210
310
410, no jet requirement-1CMS 4.7 fb
< 101 GeVee m!81 GeV
< 81 GeVee m!71 GeV
< 111 GeVee m!101 GeV
R. Yohay March 29, 2012
ee sideband weights
51
(GeV)T
Di-EM p0 100 200 300 400 500 600
Wei
ght
0
2
4
6
8
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
1
2
3
4
5
6
7
8
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
0.5
1
1.5
2
2.5
3
3.5
(GeV)T
Di-EM p0 100 200 300 400 500 600
Wei
ght
0
1
2
3
4
5
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
0.5
1
1.5
2
2.5
(GeV)T
Di-EM p0 100 200 300 400 500 600 700
Wei
ght
0
0.5
1
1.5
2
2.5
3
CMS Work in Progress CMS Work in Progress CMS Work in Progress
CMS Work in Progress CMS Work in Progress CMS Work in Progress
R. Yohay March 29, 2012
fe→γ calculation
52
The number of events in the di-electron sample is given by
Nee = f2e→eNZ→ee
where fe→e is the efficiency to correctly identify an electron via pixel match andNZ→ee is the true number of Z→ee events. The number of events in the eγ sampledue to misidentification of 1 Z electron as a photon is given by
NZeγ = 2fe→e(1− fe→e)NZ→ee
Solving for fe→e,
fe→e =1
1+ 12
NZeγ
Nee
The number of events in the eγ sample due to correctly identifying a W electronis given by
NWeγ = fe→eNW
where NW is the number of true W→eν events. The number of γγ events from Welectron misidentification is given by
NEWγγ = (1− fe→e)NW
where we have neglected the contribution from Z electron misidentification sinceit is small (i.e., fe→γ is small and the Z contribution involves f2
e→γ , since bothelectrons have to be misidentified). Since
fe→e = 1− fe→γ
solving for NEWγγ
NEWγγ = fe→γ
1−fe→γNe→γ
2
The number of events in the di-electron sample is given by
Nee = f2e→eNZ→ee
where fe→e is the efficiency to correctly identify an electron via pixel match andNZ→ee is the true number of Z→ee events. The number of events in the eγ sampledue to misidentification of 1 Z electron as a photon is given by
NZeγ = 2fe→e(1− fe→e)NZ→ee
Solving for fe→e,
fe→e =1
1+ 12
NZeγ
Nee
The number of events in the eγ sample due to correctly identifying a W electronis given by
NWeγ = fe→eNW
where NW is the number of true W→eν events. The number of γγ events from Welectron misidentification is given by
NEWγγ = (1− fe→e)NW
where we have neglected the contribution from Z electron misidentification sinceit is small (i.e., fe→γ is small and the Z contribution involves f2
e→γ , since bothelectrons have to be misidentified). Since
fe→e = 1− fe→γ
solving for NEWγγ
NEWγγ = fe→γ
1−fe→γNe→γ
2
R. Yohay March 29, 2012
GMSB MC
53
• Signal spectrum generation via SuSpect v2.41 [23]
• Signal decays via SDECAY v1.3 [24]
• Event generation, parton showering, hadronization, and decay via Pythia 6 [25]
• CMS detector simulation (GEANT [26]) and reconstruction
• Gravitino LSP
• NLO production cross sections and renormalization and factorization scale uncertainties calculated with Prospino
• PDF uncertainties calculated using PDF4LHC [27] recommendations
• 2 different signal scenarios
• Bino NLSP (1): M1 = 375 GeV, M2 = 3.5 TeV, tan β = 2, squark and gluino masses in [400 GeV, 2000 GeV], sleptons and all gauginos except the lightest neutralino have mass 3.5 TeV, heavy right-handed squarks (GGM sum rules)
• Bino NLSP (2): M1 = 375 GeV, light squarks ~2.5 TeV, M3 in [300 GeV, 2000 GeV], M2 in [200 GeV, 1500 GeV]
R. Yohay March 29, 2012
References
1. http://cdsweb.cern.ch/record/929101/files/oreach-2001-001_01.jpg
2. CMS: The Electromagnetic Calorimeter Technical Design Report, ed. F. Pauss, CERN/LHCC 97-33 (1997).
3. CMS Collaboration, CMS ECAL Performance on 2011 Data, CERN-CMS-DP-2012-002 (2012).
4. http://cdsweb.cern.ch/record/936376/files/oreach-2004-002_04.jpg
5. http://cdsweb.cern.ch/record/1299085/files/VPT_RIE_front.jpg
6. K.W. Bell et al., Vacuum Phototriodes for the CMS Electromagnetic Calorimeter Endcap, IEEE Trans. Nucl. Sci. 51 (2004) 5.
7. Courtesy D. Phillips.
8. M. Akrawy et al., Development Studies for the OPAL End Cap Electromagnetic Calorimeter Using Vacuum Photo Triode Instrumented Leadglass, NIM A290 (1990) 76.
9. G. A. Akopdzhanov, A. V. Inyakin, and P. S. Shuvalov, Photomultiplier Short-Term Instability, NIM 161 (1979) 247.
10. CMS Collaboration, ECAL 2010 Performance Results, CERN-CMS-DP-2011-008 (2011).
11. CMS Collaboration, The CMS Experiment at the CERN LHC, JINST 3 (2008) S08004.
12. Courtesy A. Ledovskoy.
54
R. Yohay March 29, 2012
References
13. S. P. Martin, A Supersymmetry Primer, arXiv:hep-ph/9709356 (1997).
14. http://pdg.web.cern.ch/pdg/cpep/images/sypersymmetry.gif
15. M. Cacciari, FastJet: a code for fast k_t clustering, and more, LPTHE-P06-04 (2006).
16. Yu. Blinnikov et al., Radiation hardness, excess noise factor and short-term gain instability of vacuum phototriodes for the operation in pseudorapidity range 1.5 < η < 3.0 at CMS ECAL, NIM A504 (2003) 228.
17. A. García-Bellido, Searches for gauge mediated SUSY breaking at LEP, in Proceedings of the 10th International Conference on Supersymmetry and Unification of Fundamental Interactions, Hamburg, 2002 (unpublished).
18. B.C. Allanach et al., The Snowmass Points and Slopes: benchmarks for SUSY searches, Eur. Phys. J. C25 (2002) 113.
19. V. Abazov et al., Search for Diphoton Events with Large Missing Transverse Energy in 6.3 fb-1 of ppbar Collisions at √s = 1.96 TeV, Phys. Rev. Lett. 105 (2010) 221802.
20. B. Brau, Tevatron Searches for New Physics with Photons and Jets, arXiv:1007.2244v2 [hep-ex] (2011).
21. CMS Collaboration, Measurement of the Isolated Prompt Photon Production Cross Section in pp Collisions at sqrt(s) = 7 TeV, CERN-PH-EP/2010-053 (2010).
55
R. Yohay March 29, 2012
References
22. CMS Collaboration, Photon Reconstruction and Identification at √s = 7 TeV, CMS-PAS-EGM-10-005 (2010).
23. A. Djouadi, J.-L. Kneur, and G. Moultaka, SuSpect: A Fortran code for the supersymmetric and Higgs particle spectrum in the MSSM, Comput. Phys. Commun. 176 (2007) 426–455.
24. M. Muhlleitner, SDECAY: A Fortran code for SUSY particle decays in the MSSM, Acta Phys. Polon. B35 (2004) 2753–2766.
25. T. Sjöstrand, S. Mrenna, and P. Z. Skands, Pythia 6.4 Physics and Manual, JHEP 0605 (2006) 026.
26. J. Allison et al., Geant4 developments and applications, IEEE Trans. Nucl. Sci. 53, 270 (2006).
27. M. Botje et al., The PDF4LHC Working Group Interim Recommendations, arXiv:1101.0538v1 [hep-ph] (2011).
56