=Paper= {{Paper |id=Vol-2507/362-366-paper-66 |storemode=property |title=ATLAS Muon Trigger Performance |pdfUrl=https://ceur-ws.org/Vol-2507/362-366-paper-66.pdf |volume=Vol-2507 |authors=Antonio Policicchio }} ==ATLAS Muon Trigger Performance== https://ceur-ws.org/Vol-2507/362-366-paper-66.pdf

Proceedings of the 27th International Symposium Nuclear Electronics and Computing (NEC’2019)
Budva, Becici, Montenegro, September 30 – October 4, 2019

ATLAS MUON TRIGGER PERFORMANCE
A. Policicchio1 on behalf of the ATLAS Collaboration
1
Sapienza Università di Roma and INFN Roma1, Roma, Italy
E-mail: antonio.policicchio@roma1.infn.it

Events containing muons in the final state are an important signature for many analyses being carried
out at the Large Hadron Collider (LHC), including both standard model measurements and searches
for new physics. To be able to study such events, it is required to have an efficient and well-
understood muon trigger. The ATLAS muon trigger consists of a hardware based system, as well as a
software based reconstruction. Due to the high luminosity in Run 2, several improvements have been
implemented to keep the trigger rate low, while still maintaining a high efficiency. Some examples of
recent improvements include requiring coincidence of hits in the muon spectrometer and the
calorimeter and optimised muon isolation. An overview of the muon trigger, the recent improvements,
and the performance in Run 2 data are presented. The improvements planned for Run 3 are also
discussed.

Keywords: ATLAS, LHC, Muon, Trigger, TDAQ

Antonio Policicchio

Copyright © 2019 for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

362
Proceedings of the 27th International Symposium Nuclear Electronics and Computing (NEC’2019)
Budva, Becici, Montenegro, September 30 – October 4, 2019

1. Introduction

Muons are a signature of many of the physics processes studied and searched for at the ATLAS
experiment [1] at the Large Hadron Collider (LHC). The muon trigger system is thus an essential
component of the experiment and it is designed to efficiently select events of interest for physics
analyses. It consists of two components: the hardware-based Level-1 (L1) and the software-based
high-level trigger (HLT). Events accepted at the fast L1 step are passed on to the HLT, which runs
algorithms close to offline reconstruction to make a final trigger decision. The ATLAS muon trigger
system has undergone continuous improvements in order to cope with the instantaneous luminosity
and the large number of interactions per bunch crossing provided by the LHC during the Run 2. This
allowed the muon trigger system to provide high quality muons over a large spectrum of transverse
momentum (pT) with high efficiency. In Run 3, the instantaneous luminosity of the LHC will be
increased up to 3.0 x 1034 cm-2s-1, 50% more with respect to the Run 2 highest luminosity. The L1
trigger rate exceeds the detector readout limitations when extrapolated to the luminosity expected
during Run 3. Various upgrades of the muon trigger system will be deployed, in order to reduce the
trigger rate while keeping the same performance as in Run 2. These proceedings present an overview
of the muon trigger system and of the recent upgrades, as well as its performance measured in 2018
proton-proton collision data. The improvements planned for Run 3 are also discussed.

2. The ATLAS muon trigger

Muon triggers are based on the information provided by the muon spectrometer (MS) and the
inner detector (ID). The MS consists of three large air-core superconducting toroidal magnets (two
endcaps and one barrel) providing a field of approximately 0.5 T on average. The deflection of the
muon trajectories in the bending plane of the magnetic field is measured via hits in three layers of
monitored drift tube (MDT) precision chambers covering the region in pseudorapidity |η| < 2.7. In the
innermost endcap wheels of the MS, cathode strip chambers (CSC) are used instead of MDTs in the
region 2.0 < |η| < 2.7. Three layers of resistive plate chambers (RPC) in the barrel (|η| < 1.05) and 3-4
layers of thin gap chambers (TGC) in the endcaps (1.05 < |η| < 2.4) allow for fast read-out to make the
initial trigger decision at L1. The ID provides track measurements based on hits in the pixel,
semiconductor tracker and transition radiation tracker, arranged in successive layers (with the pixel
detector closest to the interaction point) surrounded by a solenoid providing a 2 T magnetic field. It
covers a fiducial region up to |η| = 2.5. Higher resolution and precise tracking information provided by
MDT and CSC chambers and ID are used to refine the trigger decision at the HLT. The locations of
the detectors are shown in Figure 1 (left). The inner endcap disks made of the CSC and MDT
chambers are also referred to as the Small Wheels (SW), the external disks as the Big Wheels (BW).

2.1. Level-1 muon trigger
The L1 trigger identifies muons using coincidence requirements on the hits provided by the
RPCs or TGCs. To estimate the pT of the muons, the deviation from the hit pattern of an infinite
momentum assumption is used to define six L1 thresholds in the pT range 4 - 20 GeV. The L1 trigger
identifies Regions of Interest (RoIs) in the detector, and the information is passed to the HLT. In the
barrel region, for the three high pT thresholds, hits in all three layers are required, while for the
remaining low pT thresholds (up to 10 GeV) the hit requirement is loosened to two coincidence hits.

363
Proceedings of the 27th International Symposium Nuclear Electronics and Computing (NEC’2019)
Budva, Becici, Montenegro, September 30 – October 4, 2019

Figure 1. Left: schematic picture showing one quarter of a cross section of the ATLAS detector [2]. Right:
Efficiency of L1 10 GeV muon trigger including (green) or excluding (yellow) the trigger chambers in the feet
region of the detector. The efficiency is computed with respect to offline muon candidates and it is plotted as a
function of the η for a specific sector -2.16 < φ < -1.77 rad of the feet region [3]

Figure 2. Left: The η distributions of the L1 20 GeV muon candidates (grey) and the rate reduction effect by the
EI/FI coincidence (yellow) [4]. Right: Comparison between η of the L1 20 GeV muon candidates passing the
EI/FI coincidence (white) and the subset which passes the hadronic tile calorimeter coincidence (blue) [3]

In the endcaps the coincidence in three layers is required for all thresholds. To increase the
coverage, additional RPC chambers have been installed in 2015 in the regions of the feet that support
the ATLAS detector. Figure 1 (right) shows the L1 muon trigger efficiency for the L1 muon trigger
with 10 GeV pT threshold in the feet region as a function of the pseudorapidity. The increase of
efficiency due to the additional RPC chambers is shown by the green area. In the endcaps, to reduce
fake muon contributions from particles not originating from the interaction point, an additional
coincidence of the SW TGCs in the forward inner (FI) and endcap inner (EI) chambers has been
introduced in 2016. As shown in Figure 2 (left) for the 20 GeV L1 muon trigger, fake muons are
significantly suppressed (~20%). Further rate reduction has been achieved enabling a TGC
coincidence with the extended barrel region of the hadronic tile calorimeter (TileCal). The resulting
rate reduction (~6%) is shown in Figure 2 (right) for the 20 GeV L1 muon trigger.

2.2. High level muon trigger
The muon reconstruction at the HLT is split into fast and precise reconstruction steps. Starting
from the L1 RoI, in the fast reconstruction step, the MDT hits are used to refine the muon candidate
provided by the L1. First, muon candidates are built by performing a track fit based on the precision
measurement of drift times and positions in the MDT chambers (MS-only track). The transverse
momentum is assigned using look-up tables. The MS-only track is then extrapolated to the interaction
point and combined with tracks reconstructed in the ID, forming combined muon candidates. In 2016

364
Proceedings of the 27th International Symposium Nuclear Electronics and Computing (NEC’2019)
Budva, Becici, Montenegro, September 30 – October 4, 2019

Figure 3. Absolute L1 (black), and absolute (red) and relative (blue) HLT muon trigger efficiencies measured in
Z→μμ events as a function of the offline muon pT in the barrel (left) and endcap (right) regions [4]. At L1, a
muon trigger with a pT threshold of 20 GeV is shown, the events at the HLT step pass either a trigger with a 26
GeV pT requirement and an isolation cut, or a trigger requiring pT > 50 GeV

an additional measurement outside the magnetic field has been introduced to extrapolate to hits in the
CSC layer and improve the pT resolution in the highest pseudorapidity endcap region. If the fast
reconstruction step is passed, the muon candidate enters the precision step. Track reconstruction
exploiting the information from all MS detectors is performed. MS-only muon candidates are built
first and are subsequently combined with precision ID tracks to form combined muons. In addition to
the extrapolation from the MS to the ID (outside-in), there exists another algorithm that extrapolates
ID tracks to the MS in case the former fails (inside-out). This recovers about 1-5% of muons
depending on the muon pT. To recover efficiency losses from the L1 trigger, another algorithm
searches for muon candidates in the full detector. As this is very CPU expensive, it is used in di-muon
triggers where one muon passed the baseline sequence and the second muon is searched for in the full
detector. MS-only muon triggers are also available. In this case the muon reconstruction is only run in
the MS. MS-only triggers are vital e.g. in searches for long-lived particles (displaced decays of long-
lived particles may result in muons with no connecting tracks in the ID). To cope with the rate
limitations, isolation criteria may be applied on the muon candidates built at the HLT. Isolation
requires that there is little activity in the detector around the muon candidate by cutting on the relative
pT sum of ID tracks in a cone around the muon candidate with respect to the pT of the muon candidate.

2.3. Efficiency measurement in 2018 pp collision data

Muon trigger efficiencies are measured exploiting Z→μμ (J/ψ for low pT triggers) events in a
tag-and-probe method [5]. Figure 3 shows the L1 and HLT muon trigger efficiencies as a function of
the offline reconstructed muon pT using data collected in 2018 for barrel and endcaps. At L1, a muon
trigger with a pT threshold of 20 GeV is shown, the events at the HLT step pass either a trigger with a
26 GeV pT requirement and an isolation cut, or a trigger requiring pT > 50 GeV. The L1 trigger
efficiency is limited by the coverage of trigger chambers, while the HLT with respect to L1 is almost
100% efficient. A sharp turn-on curve of the trigger efficiency at the HLT is visible above the trigger
pT threshold. The efficiencies in the endcaps are higher than in the barrel, mostly caused by detector
geometric coverage (~80% in barrel).

3. Outlook for LHC Run 3

In the ATLAS upgrade foreseen for the LHC Run 3, the SW will be replaced by the New Small
Wheels (NSW). The NSW will consist of small-strip TGC and Micro-Mesh Gaseous Structure
chambers that will be used for both triggering and precision tracking [6]. In the barrel-endcap
transition region (1.05 < |η| < 1.3), the MDT chambers will be replaced by integrated stations of new

365
Proceedings of the 27th International Symposium Nuclear Electronics and Computing (NEC’2019)
Budva, Becici, Montenegro, September 30 – October 4, 2019

generation RPC and small-diameter MDT chambers to enhance the trigger coverage
(BIS7/8 project [7]). The excellent online tracking capabilities of the new detectors will be used to
improve the online muon identification in order to identify and reject fake muons and thus reduce the
L1 trigger rate. A schematic diagram of the improved trigger algorithm for muon identification in the
endcap region is shown in Figure 4 (left). In the current scheme, tracks A, B and C are identified as L1
muons by the BW station. In the new trigger scheme, tracks are identified as muons only if they have
coincidence hits in both the NSW and BW stations. In addition, the candidate muon track segment
reconstructed in the NSW must point to the interaction point. Fake muons (tracks B and C) typically
do not have these features and would be rejected by the algorithm. For a L1 20 GeV muon trigger a
rate reduction by a factor ~2 is estimated, allowing to keep the same Run 2 trigger thresholds with the
same performance. Figure 4 (right) shows the pseudorapidity distribution of the L1 20 GeV muon
candidates in 2017 data, the distributions when enabling hadronic tile calorimeter coincidence, and
when enabling the RPC BIS7/8 and NSW coincidences.

Figure 4. Left: Schematic diagram of the proposed Run 3 trigger algorithm used for fake muon discrimination
viewed in one quarter of a cross section of the ATLAS detector [6]. Right: The η distribution of the L1 20 GeV
muon candidates (black line); the L1 20 GeV muon candidates rejected when enabling the hadronic tile
calorimeter (white), the RPC BIS7/8 (pale blue) and NSW (yellow) coincidences; the expected distribution of L1
20 GeV muon candidates in Run 3 after enabling all hadronic tile calorimeter, RPC BIS7/8, and NSW
coincidences (blue); the distribution of all offline muon candidates (red) and of the offline muon candidates with
pT > 20 GeV [3]

References
[1] ATLAS Collaboration, The ATLAS Experiment at the CERN Large Hadron Collider, JINST 3
S08003.
[2] ATLAS Collaboration, ATLAS muon spectrometer: Technical Design Report, CERN-LHCC-97-
022, https://cds.cern.ch/record/331068.
[3] ATLAS Collaboration, L1 Muon Trigger Public Results,
https://twiki.cern.ch/twiki/bin/view/AtlasPublic/L1MuonTriggerPublicResults.
[4] ATLAS Collaboration, Public muon trigger results,
https://twiki.cern.ch/twiki/bin/view/AtlasPublic/MuonTriggerPublicResults.
[5] ATLAS Collaboration, Muon reconstruction performance of the ATLAS detector in proton- proton
collision data at √s = 13 TeV, Eur. Phys. J. C (2016) 76:292.
[6] ATLAS Collaboration, New Small Wheel Technical Design Report, CERN-LHCC-2013-006,
https://cds.cern.ch/record/1552862.
[7] ATLAS Collaboration, Technical Design Report for the Phase-II Upgrade of the ATLAS Muon
Spectrometer, CERN-LHCC-2017-017, https://cds.cern.ch/record/2285580.

366