Characterisation and mitigation of beam-induced backgrounds observed in the ATLAS detector during the 2011 proton-proton run

This paper presents a summary of beam-induced backgrounds observed in the ATLAS detector and discusses methods to tag and remove background contaminated events in data. Trigger-rate based monitoring of beam-related backgrounds is presented. The correlations of backgrounds with machine conditions, such as residual pressure in the beam-pipe, are discussed. Results from dedicated beam-background simulations are shown, and their qualitative agreement with data is evaluated. Data taken during the passage of unpaired, i.e. non-colliding, proton bunches is used to obtain background-enriched data samples. These are used to identify characteristic features of beam-induced backgrounds, which then are exploited to develop dedicated background tagging tools. These tools, based on observables in the Pixel detector, the muon spectrometer and the calorimeters, are described in detail and their efficiencies are evaluated. Finally an example of an application of these techniques to a monojet analysis is given, which demonstrates the importance of such event cleaning techniques for some new physics searches.

6 May 2013

Contact: DAPR coordinators internal

JINST 8 (2013) P07004

e-print arXiv:1303.0223 - pdf from arXiv

Inspire record

Figures

Figures

Figure 01

The general layout of the LHC. The dispersion suppressors (DSL and DSR) are sections between the straight section and the regular arc. In this paper they are considered to be part of the arc, for simplicity. LSS denotes the Long Straight Section -- roughly 500 m long parts of the ring without net bending. All insertions (experiments, cleaning, dump, RF) are located in the middle of these sections. Beams are injected through transfer lines TI2 and TI8.

png (162kB) eps (4MB) pdf (361kB)

Figure 02

Detailed layout of the ATLAS interaction region. The inner triplet consists of quadrupole magnets Q1, Q2 and Q3. The tertiary collimator (TCT) is not shown but is located between the neutral absorber (TAN) and the D2 magnet.

png (65kB) eps (1MB) pdf (31kB)

Figure 03

Schematic illustration of the LHC cleaning system. Primary and secondary collimators and absorbers in the cleaning insertions remove most of the halo. Some tertiary halo escapes and is intercepted close to the experiments by the TCT.

png (118kB) eps (141kB) pdf (55kB)

Figure 04

Inelastic beam-gas interaction rates per proton, calculated for beam-1 in LHC fill 2028. The machine elements of main interest are indicated. The pressure in the arc is assumed to be constant from 270 m onwards. The letters, a-e, identify the different sections, for which rates are given separately in other plots.

png (61kB) eps (11kB) pdf (6kB)

Figure 05

Simulated distribution of the z-coordinates of inelastic beam-gas events from which a muon with more than 100 GeV has reached the interface plane at 22.6 m. The two curves correspond to muons at radii below and above 1 m at the interface plane.

png (57kB) eps (12kB) pdf (7kB)

Figure 06

Simulated radial distribution at the interface plane of muons from inelastic beam-gas collisions and from the beam-2 TCT. The four solid curves correspond to muons originating from beam-gas events in different regions of the LSS and the adjacent LHC arc. The dashed curve shows the distribution of muons from the beam-2 TCT, normalised to 10⁵ {it p}/s lost on the TCT. The letters refer to the regions indicated in Fig. 4.

png (92kB) eps (18kB) pdf (8kB)

Figure 07

Simulated azimuthal distribution of beam-gas muons at the interface plane in four different radial ranges. The values have not been normalised to unit area, but represent the rate over the entire surface - which is different in each plot. The letters refer to the regions indicated in Fig. 4. The contribution from nearby regions drops quickly with radius. Histograms with negligible contribution have been suppressed.

png (89kB) eps (43kB) pdf (18kB)

Figure 08

Pressures, beam parameters and BCM background rates during the start of a typical LHC fill.

png (85kB) eps (15kB) pdf (7kB)

Figure 09

Correlation between P22 and BCM background rate. Each dot represents one LB.

png (166kB) eps (328kB) pdf (2MB)

Figure 10

P58, P22 and background seen at the BCM during the ``solenoid test'' in LHC fill 1803. The dashed line is not a fit, but the actual P22 value, which just happens to agree numerically with the BCM rate on this scale.

png (73kB) eps (17kB) pdf (17kB)

Figure 11

BCM background rate normalised to 10¹¹ protons for the 2011 proton-proton running period starting from mid-May. The rate is shown together with the P22 average residual pressure.

png (76kB) eps (43kB) pdf (28kB)

Figure 12

BCM collision rate (top) and background rates for beam-1 (middle) and beam-2 (bottom) per BCID before and during collisions. The data are averaged for several LHC fills over roughly 15 minute periods: at full energy but before bringing beams into collision (symbols) and after declaring stable beams (histogram). Thus the error bars reflect both the fill-to-fill variation and differences of the intensity decay during the averaging time. The data are not normalised by intensity, but only fills with comparable luminosities at the start of the fill are used in the average.

png (117kB) eps (67kB) pdf (28kB)

Figure 13

L1_J10 trigger rate in different classes of unpaired bunch isolation as a function of the luminosity of colliding bunches after subtraction of the luminosity-dependent pedestal, determined from the empty bunches.

png (107kB) eps (28kB) pdf (33kB)

Figure 14

Correlation of L1_J10 and BCM collision trigger rates in different classes of unpaired bunch isolation.

png (66kB) eps (886kB) pdf (1MB)

Figure 15

A high-multiplicity BIB event in the Pixel detector, showing the typically long pixel clusters deposited in the barrel region. On the left is the layout of the Pixel detector barrel viewed along the beam-line and the right shows the event display in a zoomed region.

png (1MB) eps (5MB) pdf (122kB)

Figure 16a

Pixel cluster width (in η direction) versus pseudorapidity for collision data.

png (80kB) eps (14kB) pdf (10kB)

Figure 16b

Pixel cluster width (in η direction) versus pseudorapidity for background data.

png (73kB) eps (15kB) pdf (10kB)

Figure 16c

Pixel cluster width (in η direction) versus pseudorapidity for collision Monte Carlo simulation.

png (80kB) eps (14kB) pdf (10kB)

Figure 16d

Pixel cluster width (in η direction) versus pseudorapidity for background Monte Carlo simulation.

png (77kB) eps (15kB) pdf (10kB)

Figure 17a

Pixel cluster deposited charge versus cluster width (in η direction), for Pixel barrel clusters.

png (130kB) eps (155kB) pdf (28kB)

Figure 17b

Pixel cluster deposited charge versus cluster width (in η direction), for Pixel barrel clusters.

png (116kB) eps (113kB) pdf (22kB)

Figure 17c

Pixel cluster deposited charge versus cluster width (in η direction), for Pixel barrel clusters.

png (111kB) eps (100kB) pdf (20kB)

Figure 17d

Pixel cluster deposited charge versus cluster width (in η direction), for Pixel barrel clusters.

png (122kB) eps (117kB) pdf (22kB)

Figure 18a

Conditional probability distributions, P^c and P^b, for clusters in the innermost pixel barrel for colliding bunches.

png (34kB) eps (29kB) pdf (8kB)

Figure 18b

Conditional probability distributions, P^c and P^b, for clusters in the innermost pixel barrel for unpaired bunches respectively.

png (32kB) eps (29kB) pdf (7kB)

Figure 18c

Calculated BIB compatibility for the innermost pixel barrel layer.

png (39kB) eps (29kB) pdf (8kB)

Figure 19a

The average pixel cluster compatibility distributions for simulated collisions. The intensity scale represents the number of events normalised by the maximum bin.

png (51kB) eps (15kB) pdf (6kB)

Figure 19b

The average pixel cluster compatibility distributions for simulated beam-gas. The intensity scale represents the number of events normalised by the maximum bin.

png (75kB) eps (55kB) pdf (12kB)

Figure 19c

The average pixel cluster compatibility distributions for background data. The intensity scale represents the number of events normalised by the maximum bin.

png (101kB) eps (116kB) pdf (21kB)

Figure 20a

The pixel BIB tagging algorithm applied to a Monte Carlo sample of beam-gas events.

png (103kB) eps (15kB) pdf (7kB)

Figure 20b

The pixel BIB tagging algorithm applied to a Monte Carlo sample of minimum-bias collisions with pile-up of 21 events per bunch crossing.

png (82kB) eps (11kB) pdf (6kB)

Figure 20c

The pixel BIB tagging algorithm applied to 2011 background data.

png (95kB) eps (15kB) pdf (7kB)

Figure 20d

The tagging efficiency as a function of the pixel cluster multiplicity, evaluated from the beam-gas simulation in Fig 20a.

png (36kB) eps (10kB) pdf (7kB)

Figure 21

Azimuthal distribution of background tagged Pixel clusters, normalised by the cluster distribution in collision events.

png (73kB) eps (24kB) pdf (14kB)

Figure 22a

Leading jet φ (a) and time (b) in unpaired bunches and collision data.

png (55kB) eps (11kB) pdf (7kB)

Figure 22b

Leading jet φ (a) and time (b) in unpaired bunches and collision data.

png (52kB) eps (11kB) pdf (8kB)

Figure 23a

Polar position and direction of the BIB objects compared to the collision objects.

png (33kB) eps (125kB) pdf (21kB)

Figure 23b

Reconstructed time of the BIB objects compared to the collision objects.

png (27kB) eps (136kB) pdf (26kB)

Figure 24a

Difference between the reconstructed polar position θ_{rm{scriptsize pos}} and the reconstructed polar direction θ_{rm{scriptsize dir}} for the muon segments in the CSC. Data from cleaned unpaired bunches (points) are compared to collisions (filled histogram).

png (66kB) eps (14kB) pdf (11kB)

Figure 24b

Difference between the reconstructed polar position θ_{rm{scriptsize pos}} and the reconstructed polar direction θ_{rm{scriptsize dir}} for the muon segments in the inner MDT endcap. Data from cleaned unpaired bunches (points) are compared to collisions (filled histogram).

png (66kB) eps (15kB) pdf (11kB)

Figure 25

Position of the muon segments in the CSC and the inner MDT endcaps with a direction nearly parallel to the beam-pipe in the cleaned unpaired bunches. The arrow indicates the direction towards the centre of the LHC ring (positive x-axis). Units correspond to the number of entries per bin.

png (84kB) eps (140kB) pdf (28kB)

Figure 26a

Reconstructed time of the CSC muon segments with a direction nearly parallel to the beam-pipe. Data from cleaned unpaired bunches (points) are compared to collisions (filled histogram).

png (64kB) eps (13kB) pdf (10kB)

Figure 26b

Reconstructed time of the inner MDT endcap segments with a direction nearly parallel to the beam-pipe. Data from cleaned unpaired bunches (points) are compared to collisions (filled histogram).

png (57kB) eps (13kB) pdf (10kB)

Figure 27a

Cluster time plotted as a function of its z position in the LAr for the cleaned unpaired bunches. Only the clusters matching a BIB muon segment are shown. The two bands, covering the radial extent of the detectors, show the expected time for the BIB clusters in 1.5 rm{m} <r<2 rm{m} for LAr and 2 rm{m} <r<4.25 rm{m} for TileCal going in the A→C or C→A direction. Units correspond to the number of entries per bin.

png (78kB) eps (24kB) pdf (8kB)

Figure 27b

Cluster time plotted as a function of its z position in the TileCal for the cleaned unpaired bunches. Only the clusters matching a BIB muon segment are shown. The two bands, covering the radial extent of the detectors, show the expected time for the BIB clusters in 1.5 rm{m} <r<2 rm{m} for LAr and 2 rm{m} <r<4.25 rm{m} for TileCal going in the A→C or C→A direction. Units correspond to the number of entries per bin.

png (78kB) eps (25kB) pdf (9kB)

Figure 28

Time of the leading jet as a function of its η in the cleaned unpaired bunches. Only the events identified by the one-sided method are shown.

png (41kB) eps (20kB) pdf (7kB)

Figure 29a

BIB rate in filled bunches of 2011 proton-proton runs. The rates in filled (a) and unpaired (b) bunches cannot be compared quantitatively because of different trigger requirements. One entry in the plot corresponds to one LHC fill. Only the statistical uncertainties are shown. Technical stops are indicated in the plot.

png (29kB) eps (13kB) pdf (14kB)

Figure 29b

BIB rate in unpaired isolated and unpaired non-isolated bunches of 2011 proton-proton runs. The rates in filled (a) and unpaired (b) bunches cannot be compared quantitatively because of different trigger requirements. One entry in the plot corresponds to one LHC fill. Only the statistical uncertainties are shown. Technical stops are indicated in the plot.

png (56kB) eps (20kB) pdf (20kB)

Figure 29c

Ratio of the BIB rate in beam-1 and beam-2 in the unpaired bunches of 2011 proton-proton runs. The rates in filled (a) and unpaired (b) bunches cannot be compared quantitatively because of different trigger requirements. One entry in the plot corresponds to one LHC fill. Only the statistical uncertainties are shown. Technical stops are indicated in the plot.

png (29kB) eps (18kB) pdf (13kB)

Figure 30a

Distributions of jet kinematic and discriminating variables for the sample enriched in fake jets before and after applying the jet selection criteria. Distributions for the sample enriched in collision jets, labelled as ``good jets sample'' in the figures, are also superimposed where applicable. Distributions for jets from collisions are re-weighted in a way to reproduce the two-dimensional jet p_T^jet versus jet η distribution obtained from the sample enriched in fake jets after Tight selection cuts.

png (59kB) eps (20kB) pdf (6kB)

Figure 30b

Figure 30c

Figure 30d

Figure 30e

Figure 30f

Figure 30g

Figure 30h

Figure 31a

Azimuthal distribution (left) and the charged particle fraction (right) of the leading jet in the monojet analysis signal region before and after the cleaning cuts. The monojet analysis signal region events are emphasised by the red line. The residual level of BIB as estimated by the two-sided method is also shown.

png (16kB) eps (12kB) pdf (5kB)

Figure 31b

Figure 32a

Transverse momentum (left) and the charged particle fraction (right) of the leading jet in the monojet analysis signal region. The non-collision background is the BIB evaluated by the two-sided method.

png (63kB) eps (19kB) pdf (7kB)

Figure 32b

Figure 33

Example of an event in the monojet analysis signal region with a BIB muon entering from the right and causing a fake jet. In the longitudinal projection (bottom left), CSC chambers with hits (highlighted in red) are seen on both sides. LAr calorimeter cells (yellow) in-between contain large energy (green towers) that forms a fake jet. A muon track (red line) parallel to the z-axis is reconstructed on side C. The transverse projection (top left) shows E_T^miss (dashed line) opposite to the fake jet. The reconstructed tracks (blue) in the inner tracking detector do not point towards the fake jet. A detailed view (middle right) shows that the calorimeter cells and the muon track are aligned in φ. Focusing on the LAr energy depositions in the longitudinal projection (bottom right) reveals that the fake jet consists of a cluster elongated in the z direction.

png (4MB) eps (8MB) pdf (115kB)

Figure 34a

Selection criteria for the TileCal muon filter. The blue rectangles correspond to the η--t regions used by the TileCal muon filter to select the events. The numbers correspond to the minimum number of selected clusters required in each region.

png (18kB) eps (6kB) pdf (4kB)

Figure 34b

Figure 34c

Figure 35

Example of an event selected by the TileCal muon filter in unpaired bunches. The clusters are shown in red marks and belongs to the same φ slice.

png (22kB) eps (8kB) pdf (4kB)

Figure 36

Ratio of the standard deviation of r and z position of the cells contained within a cluster in unpaired bunches (solid) and simulated collision events (dashed).

png (36kB) eps (10kB) pdf (5kB)