The LHCb ECAL and Preshower calibration with isolated electrons

Stenyakin O.V., Yushchenko O.P. The LHCb ECAL and Preshower calibration with isolated electrons: IHEP Preprint 2012-25. – Protvino, 2012. – p. 18, figs. 10, tables 3, refs.: 8. The results of the calibration of the LHCb electromagnetic calorimeter ECAL and Preshower detector on the real data recorded in 2010, 2011 and 2012 are presented. The calibration procedure and track selection requirements are described. The implemented calibration method is very fast and allows to perform the ECAL and Preshower calibration simultaneously. Аннотация Стенякин О.В., Ющенко О.П. Калибровка электромагнитного калориметра ECAL и предливневого детектора Preshower эксперимента LHCb с помощью изолированных электронов: Препринт ИФВЭ 2012-25. – Протвино, 2012. – 18 с., 10 рис., 3 табл., библиогр.: 8. В работе представлены результаты калибровки электромагнитного калориметра ECAL и предливневого детектора Preshower эксперимента LHCb с помощью изолированных электронов, полученных из реальных данных в 2010, 2011 и 2012 гг. Приведено описание процедуры калибровки и необходимого для её выполнения метода отбора треков. Данный метод калибровки отличается высокой скоростью и позволяет проводить калибровку калориметра ECAL и детектора Preshower одновременно. c © State Research Center of Russia Institute for High Energy Physics, 2012 Introduction A precise calibration of the electromagnetic calorimeter is a key task in any high energy physics experiment. This calibration should allow a correct reconstruction of electromagnetic shower energies. It is well known that the calibration can be performed with an electron beam of fixed energy. The methods based on comparison of reconstructed π mass with true one are also used [1]. Another interesting possibility consists in a usage of electrons/positrons produced in real events. The main idea consists in the identification of electrons without usage of the ECAL (by RICH, for example), measurement of their momenta by tracking system and comparison of the energy deposition in the ECAL with measured momentum. In this case we do not care about sources of electrons. In particular, in the case of the LHCb experiment they can be produced in Band D-mesons semileptonic inclusive decays as well as due to photons conversion in material of VELO detector and other upstream detector elements [2]. In this note we present selection procedures, the method of calibration coefficient calculations and final correction factors to be used in the ECAL and Preshower calibration with the data samples collected by the LHCb detector in 2010, 2011 and 2012. The calorimeter system of the LHCb experiment is described in Sec. 1. The problem definition to determine the calibration coefficients is described in Sec. 2. The details of the electron selections and data samples used for the calibration are described in Sec. 3. The calibration procedure and the correction factors to the energy deposition of the calorimeters are described in Sec. 4. In Sec. 5 results of the calibration and time stability of the electromagnetic calorimeter and Preshower detector are presented. 1 1. The LHCb calorimeter system The calorimeter system [3, 4, 5, 6, 7] is composed of the Scintillator Pad Detector and Preshower Detector (SPD/PS), an electromagnetic (”shashlik” type) calorimeter (ECAL) and a hadronic (Fe and scintillator tiles) calorimeter (HCAL). The thickness of the ECAL is chosen to be 25 radiation lengths to provide complete shower absorption in the energy range of the detector. Due to the strong variation of the secondary particles flow over the calorimeter surface, three sections with different transverse cell size are used in the SPD/PS and ECAL. Each of the SPD, PS and ECAL detectors has 6016 channels. The Fig. 1 shows the details of the calorimeters structure. Figure 1. Lateral segmentation of the SPD/PS and ECAL (left) and the HCAL (right). One quarter of the detector front face is shown. In the left figure the cells dimensions are given for ECAL. For SPD/PS they should be reduce by ≈ 1.5%. 2. Problem definition In original formulation the problem consists in the determination of the calibration coefficients for the calculation of energy deposition in the electromagnetic calorimeter and Preshower detector based on data from recorded ADC values. In general, this can be done starting from the RAW (not reconstructed) data where original ADC counts are presented. This implies also electrons identification, pattern recognition in the tracking system and momentum reconstruction and can be done in separate reconstruction stream. Nevertheless, the task can be significantly simplified if one will relay on the DSTs with complete ID and tracking procedures performed in the standard reconstruction stream. The DST (Data Summary ”Tape”) is the format used to store the reconstructed LHCb data. The DST data contain cells energy deposition for the ECAL and PS obtained with the dedicated calibration procedure which uses different corrections and calibration methods. Using data presented in DST which are essentially energy depositions we will provide additional multiplicative corrections obtained with method described in this note. 2 The Preshower detector (in which an electromagnetic shower starts) is located in front of the ECAL. Therefore the total energy deposition of electron/photon should be considered as a linear combination of energy depositions in the ECAL and Preshower. 3. Data samples and selections As mentioned above all the analysis is based on the DSTs where ID and track information as well as calibrated energy deposition are given. The tracks ID information on the DST provides a possibility to perform RICH-based ID as well as combined identification based on RICH and calorimeters and muon system information. To make a selection of the electron we use the variable bestParticleID which is based on RICH information only. It is chosen on the basis of a combined DLL value (Delta Log Likelihood) which is calculated with respect to charged π-meson (DLL of π is assumed to be 0). We could make better selection of the electrons by making a stronger cut on DLL value. But this method does not lead to a noticeable improvement and significantly reduces the final statistics. So, finally, we rely on the standard RICH-based electron identification. In order to ensure the purity of the data sample only isolated tracks are used. That means that there is no other track in some neighbourhood at the ECAL entrance. We choose this one as a circle with 30 cm radius to avoid overlap of showers in the ECAL. The selected size of the circle is sufficient to cover a shower in 3× 3 cluster in each zone of the ECAL. For the same reason we use tracks with large momentum value. For 2010 data there are electrons with momentum pe > 10 GeV/c and for 2011 and 2012 — with pe > 3 GeV/c. As a result, the data sample is sufficiently clean without significant hadronic contribution (see Fig. 2 as an example). So we want to calibrate the Preshower as well we select electromagnetic showers which start in the PS by the requirement Etot(PS) > 20 MeV for PS 3 × 3 cluster associated with the track. On the other hand, to reduce a background from hadrons we should limit the energy deposition in the HCAL cluster. It is set to Etot(HCAL) 10 GeV/c (2010) or pe > 3 GeV/c (2011, 2012); • Only isolated electrons are used: no other charged tracks within the circle with R = 30 cm at ECAL entrance; • Total Preshower energy deposition associated with the track Etot(PS) > 20 MeV; • Total HCAL energy deposition associated with the track Etot(HCAL) < 1 GeV.