Motivation
The purpose of these studies is to optimise the L1 trigger for light Higgs -> tau tau and Z -> tau tau.Background trigger rates
Background rates are momentarily calculated using old ntuples produced with ORCA (ORCA 8.13.3, L = 2x1033 cm-2s-1)(Info on rates and samples can be found on Monicas page: http://monicava.web.cern.ch/monicava/hlt_rates.htm)
Two set of samples used to calculate background rates, minimum bias and QCD samples:
Minimum bias sample
| Cross Section (mb) | Events (PU / no PU) |
| 79.3 | 7.2 M / 50 M |
QCD sample details
| PtHat Bin (GeV) | Cross Sections (mb) | Events |
| 15-20 | 1.455E+00 | 49491 |
| 20-30 | 6.320E-01 | 49244 |
| 30-50 | 1.629E-01 | 49742 |
| 50-80 | 2.156E-02 | 99486 |
| 80-120 | 3.079E-03 | 96238 |
| 120-170 | 4.943E-04 | 99736 |
| 170-230 | 1.008E-04 | 99226 |
| 230-300 | 2.447E-05 | 99481 |
| 300-380 | 6.237E-06 | 98739 |
| 380-470 | 1.782E-06 | 46491 |
| 470-600 | 6.832E-07 | 47496 |
| 600-800 | 2.042E-07 | 48986 |
| 800-1000 | 3.512E-08 | 45741 |
| TOTAL | 930099 |
Rates from Trigger group. Mainly minbias samples used.
(PTDR2:
CMS Note AN 2006/085):
L1 trigger table
| Trigger | L1 Threshold (GeV) | L1 Rate (kHz) | Cumulative L1 Rate (kHz) |
| Single Egamma | >22 | 3.9 ± 0.3 | 3.9 ± 0.3 |
| Double Egamma | >11 | 1.0 ± 0.1 | 4.6 ± 0.3 |
| Single Muon | >=14 | 2.5 ± 0.2 | 7.1 ± 0.3 |
| Double Muon | >=3 | 4.1 ± 0.3 | 11 ± 0.4 |
| Inclusive Tau | >100 | 2.2 ± 0.2 | 12.9 ± 0.5 |
| Double Tau | >60 | 3.0 ± 0.2 | 14.9 ± 0.5 |
| 1, 2, 3, 4 jets | >150, >100, >70, >50 | 2.2 ± 0.2 | 15.8 ± 0.5 |
| HT | >275 | 2.0 ± 0.2 | 16.2 ± 0.5 |
| Missing ET | >60 | 0.4 ± 0.2 | 16.3 ± 0.5 |
| HT + Missing ET | >200, >40 | 1.1 ± 0.1 | 16.6 ± 0.5 |
| jet + Missing ET | >100, >40 | 1.1 ± 0.1 | 16.7 ± 0.5 |
| Tau + Missing ET | >60, >40 | 2.7 ± 0.2 | 18.8 ± 0.5 |
| Muon + Missing ET | >5, >30 | 0.3 ± 0.1 | 19.0 ± 0.6 |
| Egamma + Missing ET | >15, >30 | 0.5 ± 0.1 | 19.1 ± 0.6 |
| Muon + jet | >=7, >100 | 0.2 ± 0.1 | 19.1 ± 0.6 |
| Egamma + jet | >15, >100 | 0.6 ± 0.1 | 19.2 ± 0.6 |
| Muon + Tau | >=7, >40 | 1.2 ± 0.1 | 19.8 ± 0.6 |
| Egamma + Tau | >15, >60 | 2.6 ± 0.2 | 20.5 ± 0.6 |
| Egamma + Muon | >15, >=7 | 0.2 ± 0.1 | 20.5 ± 0.6 |
| Prescaled Triggers | |||
| TOTAL L1 RATE | 22.3 ± 0.6 | ||
Reproducing PTDR2 L1 rates:
The first exercise was to reproduce all PTDR2 rates using the Minbias
PU sample.This can be done with high accuracy, as can be shown in the lower plots, which
show the integrated rate es a function of the Et threshold:
Inclusive single & double jet rate:
Single & double tau rate:
Single & double isol. Egamma rate:
Single muon & MET rate:
Geant bug
It has been found that the minbias samples are affected by a Geant bug, which result in an fault production
of low Pt neutral particles. This error causes an overestimating of the egamma rates. A workaround has been
applied by the trigger group to account for this fault rate. Egamma candidates with an reconstructed energy
larger than generated energies around a cone of R=0.7 were rejected.
But during this study I also found strong indications that the Geant bug also affects the tau rates in a certain
phase space. The lower plots show a correlation between MET and the highest tau Et.
After applying a similar workaround on taujet candidates the same correlation plots look like this:
The integrated rate for the MET + taujet cross trigger as a function of the taujet Et at a fixed MET of 40 GeV
before and after the Geant bug correction are shown below:
The single & double tau rates don't change too much after that bugfix:
Signal trigger acceptances
Fast Level1 Calorimeter Trigger Simulation
For an estimate of the signal acceptances I have implemented an Fast L1
Calorimeter Simulation package(details can be found here: https://twiki.cern.ch/twiki/bin/view/CMS/FastL1CaloSim).
Validation of the FastL1CaloSim package
The package is validated by comparing the candidates spectra of the ORCA L1 output with the onesfrom the FastL1CaloSim output.
Below the comparison for the QCD sample with Pt_hat between 50 and 80 GeV are shown.
The ORCA version was 8_13_3, the CMSSW sample was generated with version 1_0_6.
All histograms are normalised to the absolute number of events used.
Highest & 2nd highest Et Taujet:
Highest & 2nd highest Et Jet:
Highest Et isolated Egamma & MET:
Similar results are obtained for other Pt_hat bins.
New tau triggers: expected background rates & signal acceptances
The L1 signal acceptances are obtained by running the FastL1CaloSim
code over a Z-> tau tau sample generated with CMSSW_1_0_1.
The acceptances are defined the following way:
Tau+Muon:
Nominator: Number of events with at least one L1 Muon (Et>7 GeV) and at least one L1 Tau (Et> threshold) and
one generated TauJet (vis. Et>15 GeV) and one generated Muon (Et>7 GeV)
Denominator: Number of events with one generated TauJet (vis. Et>15 GeV) and one generated Muon (Et>7 GeV)
Tau+isoEgamma:
Nominator: Number of events with at least one L1 isoEgamma (Et>10 GeV) and at least one L1 Tau (Et> threshold) and
one generated TauJet (vis. Et>15 GeV) and one generated Electron (Et>10 GeV)
Denominator: Number of events with one generated TauJet (vis. Et>15 GeV) and one generated Electron (Et>10 GeV)
Tau+MET:
Nominator: Number of events with L1 MET >40 GeV and at least one L1 Tau (Et> threshold) and
at least one generated TauJet (vis. Et>15 GeV)
Denominator: Number of events with at least one generated TauJet (vis. Et>15 GeV)
Muon + Tau (corr.):
isol. Egamma + Tau (corr.) - collinear and non-collinear definition:
MET + Tau (corr.):
Muon + Tau (uncorr.):
isol. Egamma + Tau (uncorr.) - collinear and non-collinear definition:
MET + Tau (uncorr.):
more coming soon ...