CMS-DP-2019/040

CMS ECAL Trigger

●The plots show the amplitude bias and resolution for different sets of amplitude weights versus signal ET by strip

○The fractional bias is defined as: ( online/offline - 1)

○The fractional spread is defined as: 𝜎(online/offline - 1) / (online/offline)

○Error on the points: error on the mean of the bias and error on the spread of the bias

●Dataset: Full-Readout ZeroBias data from November 2017 with PU=40 with the 8b4e filling scheme (8 colliding bunch trains separated by 4 empty bunches) for TPs with abs(η) > 2.3

●There is a strong ET dependence to the amplitude bias and resolution. It is more striking for low energy (< 2 GeV) TPs (by strip)

●There are significant improvements observed when updating from the current weights, especially at the lowest ET values.

●PU optimised weights perform the best in term of resolution for all energies.

●Fractional TP Amplitude bias versus position in LHC bunch train

○for 2017 data and simulation, with 8b4e filling scheme (8 colliding bunch trains separated by 4 empty bunches) for TPs with abs(η) > 2.3

■Data: Full-Readout ZeroBias data from November 2017 with PU=40

■Standalone MC: Simulated signals with same TP ET spectrum and BX distribution as data, with PU=40

■Mean TP energy by strip : 1 GeV

●The plots show the amplitude bias for different sets of amplitude weights:

○The fractional bias is defined as:

■data: ( online/offline - 1)

■Standalone MC: ( reconstructed/true -1 )

○There is a strong bunch position dependence to the amplitude bias, seen in both data and simulation with the same trend

○There are significant improvements observed when updating from the current weights