ECAL hit timing with Run1 and Run2 energy reconstruction in early 25ns data at 13 TeV

timing in ECAL barrel

The plot shows the timing distribution of the hits in the ECAL barrel with a reconstructed energy above 1 GeV. Different energy reconstruction methods are compared: in red the "weight method" which was the default method used in Run I, in blue the "multifit method" which is the default method used in Run II. It can be seen that the multifit is effective in suppressing the contribution to the reconstructed energy from energy deposited in early or late bunch crossings. The apparent lack of events in the peak at t=0 ns with the multifit method is due to a slightly different response of the two methods at low energy. The component of the distribution around -10 ns is due to anomalous signals ascribed to direct energy deposition by particles in the APDs. Unlike the hits in an electromagnetic shower, the anomalous signals generally affect single crystals in the calorimeter. This feature is employed in the reconstruction of higher level objects to reject them by a combination of a topological selection and a cut on the hit timing. For more details see CMS Note 2010/012, sec. 4.2. For this plot the timing cut is not applied and the contribution of the anomalous signal surviving the topological selection is visible.