L1 Trigger / ECAL Performance Plots For Run2

Link to the DPS note: CMS-DP-2016-007

Abstract

The ECAL trigger performance has been investigated using the data at 13 TeV from 2015 data taking. The effect of the new ECAL response correction strategy has been studied preprocessing events from the end of the 8 TeV data taking to separate the effect of the new approch from the effect of the new calibration that has been introduced in the L1 trigger for the 2015 data taking. The global performance of the L1 Ecal trigger is thus illustrated by the turn-on curves and resolution plots for 13 TeV run of 2015 and they are compared to those from the 8 TeV run of 2012.

Results

Figure	Caption
plot_LC_2012.pdf	electron triggering efficiency (EG threshold: 20 GeV ET ) as a function of the offline reconstructed electron ET from 2012 data at √s=8TeV, using 2012 (red for barrel, blue for endcap) and 2015 (black for barrel, green for endcap) ECAL response corrections at the ECAL TP level. An unbinned likelihood fit was used. Log scale on x-axis
plot_LC_2012_noLog.pdf	L1 electron triggering efficiency (EG threshold: 20 GeV ET ) as a function of the offline reconstructed electron ET from 2012 data at √s=8TeV, using 2012 (red for barrel, blue for endcap) and 2015 (black for barrel, green for endcap) ECAL response corrections at the ECAL TP level. An unbinned likelihood fit was used. Linear scale on x-axis
plot_EB_12-15.pdf	Electron trigger efficiency at L1 (EG threshold: 20 GeV ET ) as a function of the offline reconstructed electron transverse energy (ET) for electrons in Barrel ECAL from 2012 (black dots) and 2015 data (red dots). \ In addition to the new ECAL response correction strategy, the new ECAL calibrations have been applied at L1 trigger level during 2015 data taking. The higher pT granularity of these new calibrations explains the shift of the turn-on curve.\ An unbinned likelihood fit was used.
plot_EB_12-15_noLog.pdf	Electron trigger efficiency at L1 (EG threshold: 20 GeV ET ) as a function of the offline reconstructed electron ET for electrons in Barrel ECAL from 2012 (black dots) and 2015 data (red dots). In addition to the new ECAL response correction strategy, the new ECAL calibrations have been applied at L1 trigger level during 2015 data taking. The higher pT granularity of these new calibrations explains the shift of the turn-on curve. An unbinned likelihood fit was used.\ linear scale on x-axis
plot_EE_12-15.pdf	Electron trigger efficiency at L1 (EG threshold: 20 GeV ET ) as a function of the offline reconstructed electron ET for electrons in Endcap ECAL from 2012 (black dots) and 2015 data (red dots). In addition to the new ECAL response correction strategy, the new ECAL calibrations have been applied at L1 trigger level during 2015 data taking. The higher pT granularity of these new calibrations explains the shift of the turn-on curve. An unbinned likelihood fit was used.
plot_EE_12-15_noLog.pdf	Electron trigger efficiency at L1 (EG threshold: 20 GeV ET ) as a function of the offline reconstructed electron ET for electrons in Endcap ECAL from 2012 (black dots) and 2015 data (red dots). In addition to the new ECAL response correction strategy, the new ECAL calibrations have been applied at L1 trigger level during 2015 data taking. The higher pT granularity of these new calibrations explains the shift of the turn-on curve. An unbinned likelihood fit was used.\ linear scale on x-axis
plot_iEG_2015.pdf	L1 electron triggering efficiency in Barrel ECAL as a function of the offline reconstructed electron ET . The efficiency is shown for the 15, 20, 30, 40 GeV EG trigger thresholds.An unbinned likelihood fit was used.
plot_iEG_2015_noLog.pdf	L1 electron triggering efficiency in Barrel ECAL as a function of the offline reconstructed electron E . The efficiency is shown for the 15, 20, 30, 40 GeV EG trigger thresholds. An unbinned likelihood fit was used. \ linear scale on x-axis
plot_EE_iEG_2015.pdf	L1 electron triggering efficiency in Endcap ECAL as a function of the offline reconstructed electron ET . The efficiency is shown for the 15, 20, 30, 40 GeV EG trigger thresholds. An unbinned likelihood fit was used
plot_EE_iEG_2015_noLog.pdf	L1 electron triggering efficiency in Endcap ECAL as a function of the offline reconstructed electron E. The efficiency is shown for the 15, 20, 30, 40 GeV EG trigger thresholds. An unbinned likelihood fit was used. \ linear scale on x-axis
frame.pdf	L1 EG resolution for all electrons located in the Barrel for the last 7.3 fb-1 of 2012 (black) and 2015 (red) data. For both distributions, a Cruijff fit is superimposed. Two main contributions affect the resolution: the ecal response corrections applied to 2015 data and not applied to 2012 data and the higher pT granularity in the calibration applied to L1. \ σR: 0.040 (2012), 0.029 (2015) \ σL: 0.029 (2012), 0.035 (2015)
frameEE.pdf	L1 EG resolution for all electron pT located in the Barrel for the last 7.3 fb-1 2012 (black) and 2015 (red) data. For both distributions, a Cruijffb fit is superimposed. Two main contributions affect the resolution: the change in strategy for the ecal response corrections and the higher pT granularity in the calibration applied to L1. \ σR: 0.109 (2012), 0.090 (2015) \ σL: 0.092 (2012), 0.093 (2015)
Etareso_2015_2012.pdf	L1 EG resolution for all electron pT as a function of the pseudorapidity η for 2015 data. For each η bin, a Cruijff function was used to model the data distribution. The points correspond to the peak position in each η bin. The vertical bars on each point represent the σL (lower error bars) and σR (upper error bars) of each fitted distribution.
totalSigma.pdf	L1 EG resolution for all electron pT as a function of the pseudorapidity η for 2015 and the last 7.3fb-1 of 2012 data. For each η bin, a Cruijff function was used to model the data distribution. The points correspond to the width ( σR +σL )/2 of each fitted distribution.

Note:

CruijffPDF: a bifurcated Gaussian with asymmetric tails: f(x)=exp((x−m)^2/(2σ^2_{L,R} +α_L,R(x−m)^2)) from doi:10.1103/PhysRevD.82.051101