H??? FTK update

1
H??? FTK update

• Catalin, Tony
• FTK meeting, 9/7/2006

2
Recap H??? with FTK
LVL1 (new trigger)
LVL2 (FTK)
200 Hz
40 kHz

• Step1 look at LVL1 side
• Step2 look at LVL2 side
• Similar to Koheis study

3
Recap EM and tau at LVL1

• Definition of EM
• L1Em_EmClus gt threshold
• L1EmIsol lt 4 GeV
• L1Em_HdIsolL1Em_HdCore lt 2 GeV
• Definition of ?
• L1Em_TauClus gt threshold
• L1EmIsol lt 4 GeV
• L1Em_HdIsol lt 2 GeV

tau
electron
Had sumlt2 GeV
EM isolationlt4 GeV
4
LVL1 EM and tau dijet rates

• Eric and Alan pointed us to the latest dijet
  study by Pilar Casado (IFAE, Barcelona)
• https//twiki.cern.ch/twiki/bin/view/Atlas/LVL1Tau
  Studies
• LVL1 trigger rates are given for 1031 for
  various thresholds
• L1em_EmClusgt9GeV L1EmIsollt4GeV 1.6 kHz (we
  1.4kHz)
• L1em_TauClusgt9GeV L1EmIsollt4GeV 2.8 kHz (we
  2.3kHz)
• Different cross-sections could account for this
  13 and 18 diff.
• Only J1 and J2 samples contribute significantly
  to the rates
• EM trig J1,J2,J3,(J4J8) 81, 18, 1,
  lt0.06
• TAU trig J1,J2,J3,(J4J8) 83, 16, 1,
  lt0.05

5
LVL1 trigger study

• According to Alan Watson
• LVL1 trigger decision based on multiplicities
  alone (does not use any roi info at LVL1)
• Run over dijet and keep track of the following
• Isolated EM cluster count EM
• Isolated TAU cluster count (could be the same as
  above) ?EM-had
• Isolated TAU cluster different from EM count
  ?!EM
• (?EM-had, ?EM-had) pair count
• (?!EM, ?!EM) pair count
• (EM?2 ?EM-had) count
• (EM?1 ?!EM) count

6
LVL1 Trigger Rates at 21033 cm-2s-1

• ? 1EM (black) up to 300 kHz
• ?1?EM-had (red) up to 150 kHz
• LVL1 projected 40 kHz, designed max output 75 kHz

7
EM? LVL1 trigger rate at 21033

• EMgt1?EM-had rate up to 10 kHz
• Shown versus EM-threshold or ? threshold
• Stronger dependence on ? threshold than on EM
  threshold

8
2? LVL1 trigger rate at 21033

• Require ?2?EM-had objects
• Rate up to 20 kHz

9
2?!EM LVL1 trigger rate at 21033

• Not possible to trigger on, but interesting
• As much as 3kHz comes from 2 hadronic taus
• Could be interesting to study if and how much
  increases signal acceptance in the all-had channel

10
Step2 LVL2 rejection

• Thanks to Paola for pointing us to the CMS
  algorithm

Jet axis
leading PT track
RS
RI
RO
Recipe - find leading PT track in a RI0.1 cone
around the jet - require PT(lead. track) gt 3
GeV - find tracks within an RS0.07 cone around
lead. - count, as before, tracks in the
isolation reg RSltRltRO
11
Higgs 130 GeV
Curves correspond to different RO min leading
PT . RS0.07, RI0.1 for all.
RO0.3, PTSiggt3.0 GeV/c
factor of 1.4
optimum
RO0.5, PTSiggt10.0 GeV/c
Max tracks in,out
3,1 2,2 2,1 4,0 3,0 1,2
1,1 2,0 1,0
in tracks are here w/ Ptgt1.0
out tracks are here w/ Ptgt0.5
12
Sherpa Dijet
RO0.3, PTSiggt3.0 GeV/c
factor of 10
RO0.5, PTSiggt10.0 GeV/c
Curves correspond to different RO min leading
PT . RS0.07, RI0.1 for all.
Max tracks in,out
3,1 2,2 2,1 4,0 3,0 1,2
1,1 2,0 1,0
in tracks are here w/ Ptgt1.0
out tracks are here w/ Ptgt0.5
13
Conclusions and Outlook

• We believe we have the correct LVL1 trigger
  implementation, and we verified that we can
  reproduce some recent dijet trigger rates to
  within 10-20
• At LVL2 we can dial dijet rejections corresp to
  factors of 5-100.
• Questions are
• How much is it admissible to increase the LVL1 or
  LVL2 output rates?
• Which of the above limits us?
• For example, we can add EMtau trigger with
  thresholds 10-25 GeV which would increase the
  LVL1 output rate by lt10 kHz
• Remains to be investigated
• How much signal we gain
• If we would gain significantly in the hadronic
  ditau Higgs channel

14
Recap Trigger Study May 06
Lepton ET
LOW-LUM
HIGH-LUM
10,15,20
0
MU20 (20)
2MU6
1
EM25I (30)
2EM15I (20)
2
J200 (290)
3
3J90 (130)
4
4J65 (90)
5
J60xE60 (100100)
6
TAU25xE30 (6060)
MU10EM15I
Missing ET
Tau ET
15,20,25,30
10,15,20,25
15
Recap Low Luminosity RegimeSignal 130 GeV H???
metgt15
metgt30
Trigger efficiency
ET(?) threshold
ET(?) threshold
ET(?) threshold
ET(lep) gt 10 GeV
ET(lep) gt 15 GeV
ET(lep) gt 20 GeV
Reminder - only lepton, tau, and met thresholds
are varied - rest of the trigger table
stays the same Biggest effect lowering the
lepton ET threshold (up to 15 gain)
16
More Backup

• ?1?!EM rates vs EM or ? thresholds (at L21033)
• Stronger dependence on ? threshold