Public ATLAS Luminosity Results for Run-2 of the LHC

Introduction

Here any results related to ATLAS luminosity measurements in Run-2 are given. For Run-3 results, please see Luminosity Results for Run-3 (2022). For Run-1 results, please see Luminosity Results for Run-1 (2010-2012).

Publications and Conference Results

Luminosity:

• Luminosity determination in pp collisions at √s=13 TeV using the ATLAS detector at the LHC: Documents the final Run-2 luminosity calibration for pp collisions at the given center-of-mass energy
• Luminosity determination using Z→ ℓ⁺ℓ^- events at √s=13 TeV with the ATLAS detector: Documents the run-2 Z-counting analysis
• Luminosity determination for low-pileup datasets at sqrt(s)=5 and 13 TeV using the ATLAS detector at the LHC: Documents the preliminary (July 2020) Run 2 low-pileup dataset luminosity results
• Luminosity determination in pp collisions at sqrt(s) = 13 TeV using the ATLAS detector at the LHC and associated ATLAS physics briefing and CERN Courier article: Documents the preliminary (Spring 2019) Run 2 luminosity results
• Luminosity determination in pp collisions at sqrt(s) = 8 TeV using the ATLAS detector at the LHC: Documents the final 2012 luminosity results
• Improved luminosity determination in pp collisions at sqrt(s) = 7 TeV using the ATLAS detector at the LHC: Documents the final 2010 and 2011 luminosity results
• Luminosity Determination in pp Collisions at sqrt(s) = 7 TeV using the ATLAS Detector in 2011
• Updated Luminosity Determination in pp Collisions at sqrt(s) = 7 TeV using the ATLAS Detector − ATLAS-CONF-2011-011
• Luminosity Determination in pp Collisions at sqrt(s)=7 TeV Using the ATLAS Detector at the LHC − CERN-PH-EP-2010-069

Satellite Bunches:

• Measurement of the Rate of Collisions from Satellite Bunches for the April-May 2010 LHC Luminosity Calibration − ATLAS-CONF-2010-102

Luminosity Plots from offline luminosity analysis

Luminosity Plots from the October 2021 beam test

• LUMI-2021-03 Initial results from the magnetic non-linearity scans at ATLAS in the October 2021 LHC pilot beam run.

Luminosity Plots for multiple Run-2 years

Plots supporting the preliminary Run-2 analyses:

• LUMI-2022-02 Calibration transfer and long-term stability of the ATLAS luminosity measurement
• LUMI-2022-01 Long-term stability of the track-counting luminometer in Run-2 (complemented by some figures from 2018 below).
• LUMI-2020-02 Z-counting analysis and stability plots with both electron and muon channels.
• LUMI-2021-02 Track multiplicity plots for the track-counting luminosity measurement at 13 TeV.

Luminosity Plots for the 2018 Analysis

Plots supporting the preliminary analysis of the 2018 pp running period:

• LUMI-2018-001 Long-term stability and LUCID mu-correction plots.
• LUMI-2018-002 Emittance scan analysis using LUCID bunch-by-bunch data.
• LUMI-2020-01 Track-counting luminosity studies from the 2018 √s=13 TeV dataset.

Luminosity Plots for the 2017 Analysis

Plots supporting the preliminary analysis of the 2017 pp running period:

• LUMI-2017-001 vdM scan, non-factorisation analysis, bunch train, mu-correction and long-term stability plots.
• LUMI-2017-002 Track-counting selection plots.
• LUMI-2017-003 Z-counting analysis and stability plots.
• LUMI-2017-004 Plots for the low-luminosity running period

Luminosity Plots for the 2016 Analysis

The results can be found here: Luminosity2016Analysis

Luminosity Plots for the 2015 Analysis

The results can be found here: Luminosity2015Analysis

Luminosity Plots for the dedicated ALFA datasets with high β* in Run-2

Plots supporting the preliminary analysis of the 2018 13 TeV ALFA running period:

• LUMI-2023-03 Luminosity figures for 2018 ALFA runs at √s=13 TeV

Online Luminosity Summary Plots over Multiple Years including full Run 2

These plots are based on the online luminosity estimate.

Delivered Luminosity versus time for 2011-2018 (p-p data only) Cumulative luminosity versus day delivered to ATLAS during stable beams and for high energy p-p collisions.	pdf file
Total Integrated Luminosity at Run 2 (13 TeV pp data only) Cumulative luminosity versus time delivered to ATLAS (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in LHC Run 2. A version showing just the delivered luminosity (green) is also available as pdf and png for making overlay animations.	pdf file
Total Integrated Luminosity and Data Quality in 2015-2018 Cumulative luminosity versus time delivered to ATLAS (green), recorded by ATLAS (yellow), and certified to be good quality data (blue) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2015-2018. The complete pp data sample in 2018 is shown. The delivered luminosity accounts for the luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so‐ called "warm start": when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. The data quality assessment shown corresponds to the All Good efficiency shown in the 2015-2018 Full Dataset DQ tables here. The All Good Data Quality criteria require all reconstructed physics objects to be of good data quality. A version showing just the delivered luminosity (green) is available as pdf and png. A version showing the delivered luminosity (green) and recorded luminosity (yellow) is available as pdf and png.	pdf file
Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2015 -- 2018 pp collision data at 13 TeV centre-of-mass energy. All data recorded by ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency.	pdf file eps file

Integrated and Instantaneous Luminosities for Individual Years

2018 pp data taking

These plots show the estimated integrated luminosity with an initial calibration for the 2018 data.

Total Integrated Luminosity in 2018 Cumulative luminosity versus time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2018. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. <The luminosity shown represents an initial 13 TeV luminosity calibration for the 2018 data based on van-der-Meer beam-separation scans in 2017.	pdf file
Total Integrated Luminosity and Data Quality in 2018 Cumulative luminosity versus time delivered to ATLAS (green), recorded by ATLAS (yellow), and certified to be good quality data (blue) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2018. The complete pp data sample in 2018 is shown. The delivered luminosity accounts for the luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so‐ called "warm start": when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. The data quality assessment shown corresponds to the All Good efficiency shown in the 2018 Full Dataset DQ tables here. The All Good Data Quality criteria require all reconstructed physics objects to be of good data quality. A version showing just the delivered luminosity (green) is available as pdf and png. A version showing the delivered luminosity (green) and recorded luminosity (yellow) is available as pdf and png.	pdf file
Integrated Luminosity by Day in 2018 Integrated luminosity per day time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2018. The data shown is identical to the data shown above, except not cumulative over the year.	pdf file
Peak Luminosity per Fill in 2018 The peak instantaneous luminosity delivered to ATLAS during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each LHC fill as a function of time in 2018. The luminosity is determined using counting rates measured by the luminosity detectors, and is based on an initial estimate from van-der-Meer beam-separation scans during 2017.	pdf file
Peak Interactions per Crossing in 2018 The maximum number of inelastic collisions per beam crossing (µ) during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each fill in 2018. The preliminary luminosity measurement is used to determine the number of interactions per beam crossing as µ = Lbunch x σinel / fr where Lbunch is the per-bunch instantaneous luminosity, σinel is the inelastic cross-section at 13 TeV, which is taken to be 80 mb, and fr is the LHC revolution frequency of 11.245 kHz. The number of interactions shown is averaged over all colliding bunch pairs, and only the peak value per fill during stable beams is shown.	pdf file
Total Stable Beam Time by Day in 2018 Total Stable Beams Time - shown is the amount of time in stable beams delivered by the LHC each calendar day integrated over 2018. Also shown is the stable beams time multiplied by the L1 livefraction to show the amount of stable beams time recorded by ATLAS.	pdf file
Stable Beam Time in 2018 Stable Beams Time per day - shown is the amount of time in stable beams delivered by the LHC for each calendar day. Also shown is the stable beams time multiplied by the L1 livefraction to show the amount of stable beams time recorded by ATLAS.	pdf file

2018 PbPb data taking

Total Integrated Luminosity in 2018 Cumulative luminosity versus time delivered to (blue) and recorded by ATLAS (cyan) during stable beams for PbPb collisions at 5 TeV centre-of-mass energy in 2018. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors using a preliminary calibration.	pdf file
Integrated Luminosity by Day in 2018 Integrated luminosity per day time delivered to (blue) and recorded by ATLAS (cyan) during stable beams for PbPb collisions at 5 TeV centre-of-mass energy in 2018. The data shown is identical to the data shown above, except not cumulative over the running period.	pdf file
Peak Luminosity per Fill in 2018 The peak instantaneous luminosity delivered to ATLAS during stable beams for PbPb collisions at 5 TeV centre-of-mass energy is shown for each LHC fill as a function of time in 2018. The luminosity is determined using counting rates measured by the luminosity detectors, and is based on an initial calibration estimated from the previous 2015 PbPb data.	pdf file

2017 pp data taking

These plots show the estimated integrated luminosity with a preliminary calibration for the 2017 data.

Total Integrated Luminosity in 2017 Cumulative luminosity versus time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2017. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors. The luminosity shown represents a preliminary 13 TeV luminosity calibration for the 2017 data based on van-der-Meer beam-separation scans in 2016.	pdf file
Integrated Luminosity by Day in 2017 Integrated luminosity per day time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2017. The data shown is identical to the data shown above, except not cumulative over the year.	pdf file
Peak Luminosity per Fill in 2017 The peak instantaneous luminosity delivered to ATLAS during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each LHC fill as a function of time in 2017. The luminosity is determined using counting rates measured by the luminosity detectors, and is based on a preliminary analysis of van-der-Meer beam-separation scans during 2016.	pdf file
Peak Interactions per Crossing in 2017 The maximum number of inelastic collisions per beam crossing (µ) during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each fill in 2017. The preliminary luminosity measurement is used to determine the number of interactions per beam crossing as µ = Lbunch x σinel / fr where Lbunch is the per-bunch instantaneous luminosity, σinel is the inelastic cross-section at 13 TeV, which is taken to be 80 mb, and fr is the LHC revolution frequency of 11.245 kHz. The number of interactions shown is averaged over all colliding bunch pairs, and only the peak value per fill during stable beams is shown.	pdf file
Total Stable Beam Time by Day in 2017 Total Stable Beams Time - shown is the amount of time in stable beams delivered by the LHC each calendar day integrated over 2017. Also shown is the stable beams time multiplied by the L1 livefraction to show the amount of stable beams time recorded by ATLAS.	pdf file
Stable Beam Time in 2017 Stable Beams Time per day - shown is the amount of time in stable beams delivered by the LHC for each calendar day. Also shown is the stable beams time multiplied by the L1 livefraction to show the amount of stable beams time recorded by ATLAS.	pdf file

2016 pp data taking

These plots show the estimated integrated luminosity with the preliminary calibration update from February 2017.

Total Integrated Luminosity in 2016 Cumulative luminosity versus time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2016. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors. These detectors have been calibrated with the use of the van-der-Meer beam-separation method, where the two beams are scanned against each other in the horizontal and vertical planes to measure their overlap function. The luminosity shown represents the preliminary 13 TeV luminosity calibration based on van-der-Meer beam-separation scans in 2016.	pdf file
Integrated Luminosity by Day in 2016 Integrated luminosity per day time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2016. The data shown is identical to the data shown above, except not cumulative over the year.	pdf file
Peak Luminosity per Fill in 2016 The peak instantaneous luminosity delivered to ATLAS during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each LHC fill as a function of time in 2016. The luminosity is determined using counting rates measured by the luminosity detectors, and is based on a preliminary analysis of van-der-Meer beam-separation scans during 2016.	pdf file
Peak Interactions per Crossing in 2016 The maximum number of inelastic collisions per beam crossing (µ) during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each fill in 2016. The preliminary luminosity measurement is used to determine the number of interactions per beam crossing as µ = Lbunch x σinel / fr where Lbunch is the per-bunch instantaneous luminosity, σinel is the inelastic cross-section at 13 TeV, which is taken to be 80 mb, and fr is the LHC revolution frequency of 11.245 kHz. The number of interactions shown is averaged over all colliding bunch pairs, and only the peak value per fill during stable beams is shown.	pdf file
Total Stable Beam Time by Day in 2016 Total Stable Beams Time - shown is the amount of time in stable beams delivered by the LHC each calendar day integrated over 2016. Also shown is the stable beams time multiplied by the L1 livefraction to show the amount of stable beams time recorded by ATLAS.	pdf file
Stable Beam Time in 2016 Stable Beams Time per day - shown is the amount of time in stable beams delivered by the LHC for each calendar day. Also shown is the stable beams time multiplied by the L1 livefraction to show the amount of stable beams time recorded by ATLAS.	pdf file

2016 pPb and Pbp data taking

Total Integrated Luminosity in 2016 Cumulative luminosity versus time delivered to (blue) and recorded by ATLAS (cyan) during stable beams for pPb and Pbp collisions at 5 TeV and 8 TeV centre-of-mass energy in 2016. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors using a preliminary calibration.

pdf file

Integrated Luminosity by Day in 2016 Integrated luminosity per day time delivered to (blue) and recorded by ATLAS (cyan) during stable beams for pPb and Pbp collisions at 5 TeV and 8 TeV centre-of-mass energy in 2016. The data shown is identical to the data shown above, except not cumulative over the running period.

pdf file

2015 pp data taking

These plots show the estimated integrated luminosity with the most recent offline calibration. (December 2015)

Total Integrated Luminosity and Data Quality in 2015 Cumulative luminosity versus time delivered to ATLAS (green), recorded by ATLAS (yellow), and certified to be good quality data (blue) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2015. The delivered luminosity accounts for the luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so‐ called "warm start": when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. The data quality assessment shown corresponds to the All Good efficiency shown in the 2015 EOYE Dataset DQ table here. The All Good Data Quality criteria require all reconstructed physics objects to be of good data quality. The IBL was turned off for two runs, corresponding to 0.2 fb-1 and analyses that don’t rely on the IBL can use 3.4 fb-1 of data with a corresponding DQ efficiency of 93.1%. The luminosity shown represents the preliminary 13 TeV luminosity calibration which was updated in December 2015. The table corresponds to the DQ tag DetStatus-v73-pro19-08 with lumi tag OflLumi-13TeV-003.	pdf file For animations: delivered (png) delivered (pdf) recorded (png) recorded (pdf)
Total Integrated Luminosity in 2015 Cumulative luminosity versus time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2015. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors. These detectors have been calibrated with the use of the van-der-Meer beam-separation method, where the two beams are scanned against each other in the horizontal and vertical planes to measure their overlap function. The luminosity shown represents the preliminary 13 TeV luminosity calibration which was updated in December 2015.	pdf file
Integrated Luminosity by Day in 2015 Integrated luminosity per day time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 13 TeV centre-of-mass energy in 2015. The data shown is identical to the data shown above, except not cumulative over the year.	pdf file
Peak Luminosity per Fill in 2015 The peak instantaneous luminosity delivered to ATLAS during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each LHC fill as a function of time in 2015. The luminosity is determined using counting rates measured by the luminosity detectors, and is based on a preliminary 13 TeV calibration determined using van-der-Meer beam-separation scans.	pdf file
Peak Interactions per Crossing in 2015 The maximum number of inelastic collisions per beam crossing (µ) during stable beams for pp collisions at 13 TeV centre-of-mass energy is shown for each fill in 2015. The preliminary luminosity measurement is used to determine the number of interactions per beam crossing as µ = Lbunch x σinel / fr where Lbunch is the per-bunch instantaneous luminosity, σinel is the inelastic cross-section at 13 TeV, which is taken to be 80 mb, and fr is the LHC revolution frequency of 11.245 kHz. The number of interactions shown is averaged over all colliding bunch pairs, and only the peak value per fill during stable beams is shown.	pdf file

2015 PbPb data taking

These plots show the estimated integrated luminosity of the PbPb collisions at 5 TeV per nucleon. The luminosity shown is a preliminary estimation based on van-der-Meer beam-separation scans.

Total Integrated Luminosity in 2015 Cumulative luminosity versus time delivered to (dark blue) and recorded by ATLAS (light blue) during stable beams for PbPb collisions at 5 TeV centre-of-mass energy per nucleon in 2015. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors. These detectors have been calibrated with the use of the van-der-Meer beam-separation method, where the two beams are scanned against each other in the horizontal and vertical planes to measure their overlap function.	pdf file
Integrated Luminosity by Day in 2015 Integrated luminosity per day time delivered to (green) and recorded by ATLAS (yellow) during stable beams for PbPb collisions at 5 TeV centre-of-mass energy per nucleon in 2015. The data shown is identical to the data shown above, except not cumulative over the year.	pdf file
Peak Luminosity per Fill in 2015 The peak instantaneous luminosity delivered to ATLAS during stable beams for PbPb collisions at 5 TeV centre-of-mass energy per nucleon is shown for each LHC fill as a function of time in 2015. The luminosity is determined using counting rates measured by the luminosity detectors, and is based on a preliminary calibration determined using van-der-Meer beam-separation scans.	pdf file
Peak Interactions per Crossing in 2015 The maximum number of inelastic collisions per beam crossing (µ) during stable beams for PbPb collisions at 5 TeV centre-of-mass energy per nucleon is shown for each fill in 2015. The preliminary luminosity measurement is used to determine the number of interactions per beam crossing as µ = Lbunch x σinel / fr where Lbunch is the per-bunch instantaneous luminosity, σinel is the inelastic PbPb cross-section at 5 TeV, which is taken to be 7.7 barns, and fr is the LHC revolution frequency of 11.245 kHz. The number of interactions shown is averaged over all colliding bunch pairs, and only the peak value per fill during stable beams is shown.	pdf file

2015 5 TeV pp data taking

These plots show the estimated integrated luminosity of the 5 TeV reference data with the online calibration.

Total Integrated Luminosity in 2015 Cumulative luminosity versus time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 5 TeV centre-of-mass energy in 2015. The delivered luminosity accounts for luminosity delivered from the start of stable beams until the LHC requests ATLAS to put the detector in a safe standby mode to allow for a beam dump or beam studies. The recorded luminosity reflects the DAQ inefficiency, as well as the inefficiency of the so-called ‘warm start’: when the stable beam flag is raised, the tracking detectors undergo a ramp of the high-voltage and, for the pixel system, turning on the preamplifiers. Shown is the luminosity as determined from counting rates measured by the luminosity detectors. An online calibration based on Monte Carlo estimates has been used to set the luminosity scale.	pdf file
Integrated Luminosity by Day in 2015 Integrated luminosity per day time delivered to (green) and recorded by ATLAS (yellow) during stable beams for pp collisions at 5 TeV centre-of-mass energy in 2015. The data shown is identical to the data shown above, except not cumulative over the year.	pdf file
Peak Luminosity per Fill in 2015 The peak instantaneous luminosity delivered to ATLAS during stable beams for pp collisions at 5 TeV centre-of-mass energy is shown for each LHC fill as a function of time in 2015. An online calibration based on Monte Carlo estimates has been used to set the luminosity scale.	pdf file

Annual plots on the Number of Pileup Interactions and Data Taking Efficiencies

The following figures are based on the online luminosity determination

2018 pp Collisions

Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2018 pp collision data at 13 TeV centre-of-mass energy. All data recorded by ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the preliminary 13 TeV luminosity calibration for 2018, released in February 2019, that is based on van-der-Meer beam-separation scans. Data collected by ATLAS for the entire 2018 run through the end of October are shown.

pdf file eps file

2017 pp Collisions

Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2017 pp collision data at 13 TeV centre-of-mass energy. All data recorded by ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the preliminary 13 TeV luminosity calibration released in February 2018, based on van-der-Meer beam-separation scans in 2017.

pdf file eps file

Number of Interactions per Crossing Same as above, but additionally showing the 13 TeV data from 2015 -- 2017.

pdf file eps file

2016 pp Collisions

Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2016 pp collision data at 13 TeV centre-of-mass energy. All data delivered to ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the preliminary 13 TeV luminosity calibration released in February 2017, based on van-der-Meer beam-separation scans in 2016.	pdf file eps file
Number of Interactions per Crossing Same as above, but showing the combined 13 TeV data from 2015 and 2016.	pdf file eps file
Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2016 pp collision data recorded from 16 April - 16 September at 13 TeV centre-of-mass energy. All data delivered to ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the preliminary 13 TeV luminosity calibration based on van-der-Meer beam-separation scans in 2016. This plot was first shown at the September 2016 LHCC meeting.	pdf file eps file
Number of Interactions per Crossing Same as above, but showing the combined 13 TeV data from 2015 and 2016.	pdf file eps file
Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2016 pp collision data recorded from 16 April - 25 July at 13 TeV centre-of-mass energy. All data delivered to ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the preliminary 13 TeV luminosity calibration based on van-der-Meer beam-separation scans in 2016. This plot was first shown at ICHEP 2016.	pdf file eps file
Number of Interactions per Crossing Same as above, but showing the combined 13 TeV data from 2015 and 2016.	pdf file eps file
Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2016 pp collision data recorded from 16 April - 13 June at 13 TeV centre-of-mass energy. All data delivered to ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the initial 2016 13 TeV luminosity calibration which is based on the 2015 calibration. This plot was first shown at LHCP 2016.	pdf file eps file
Number of Interactions per Crossing Same as above, but showing the combined 13 TeV data from 2015 and 2016.	pdf file eps file
Data Taking Efficiency per Day ATLAS data taking efficiency in 2016. The denominator is the luminosity delivered between the declaration of stable beams and the LHC request to turn the sensitive detectors off to allow a beam dump or beam studies. The numerator is the luminosity recorded by ATLAS. Each bin represents one day. The empty bins are due to days in which no stable beams were delivered by the LHC. The inefficiency accounts for the time needed to turn on the high-voltage of the Pixel, SCT, and some of the muon detectors at the start of an LHC fill and any inefficiencies due to deadtime or due to individual problems with a given sub detector that prevents the ATLAS data taking to proceed.	pdf file
Data Taking Efficiency per Week Same as above but per week.	pdf file

2015 pp Collisions

Number of Interactions per Crossing Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2015 pp collision data recorded from 3 June - 3 November at 13 TeV centre-of-mass energy. All data delivered to ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency. The luminosity shown represents the preliminary 13 TeV luminosity calibration which was updated in December 2015.	pdf file eps file
Number of Interactions per Crossing Same as above, but with data delivered in fills with 50ns and 25ns bunch spacing shown separately.	pdf file eps file
Data Taking Efficiency per Day ATLAS data taking efficiency in 2015. The denominator is the luminosity delivered between the declaration of stable beams and the LHC request to turn the sensitive detectors off to allow a beam dump or beam studies. The numerator is the luminosity recorded by ATLAS. Each bin represents one day. The empty bins are due to days in which no stable beams were delivered by the LHC. The inefficiency accounts for the time needed to turn on the high-voltage of the Pixel, SCT, and some of the muon detectors at the start of an LHC fill and any inefficiencies due to deadtime or due to individual problems with a given sub detector that prevents the ATLAS data taking to proceed.	pdf file
Data Taking Efficiency per Week Same as above but per week.	pdf file

2015 LHCP dataset

Number of Interactions per Crossing - LHCP Dataset Shown is the luminosity-weighted distribution of the mean number of interactions per crossing for the 2015 pp collision data recorded from 3 June - 25 August at 13 TeV centre-of-mass energy (LHCP dataset). All data delivered to ATLAS during stable beams is shown, and the integrated luminosity and the mean mu value are given in the figure. The mean number of interactions per crossing corresponds to the mean of the poisson distribution of the number of interactions per crossing calculated for each bunch. It is calculated from the instantaneous per bunch luminosity as μ=Lbunch x σinel / fr where Lbunch is the per bunch instantaneous luminosity, σinel is the inelastic cross section which we take to be 80 mb for 13 TeV collisions, and fr is the LHC revolution frequency.

pdf file eps file

Number of Interactions per Crossing - LHCP Dataset Same as above, but with data delivered in fills with 50ns and 25ns bunch spacing shown separately.

pdf file eps file

Pixel Cluster Counting Luminosity Plots

Pixel Cluster Counting in 2016 and 2017 data

Figure 1 Longitudinal cluster-size distributions in the most forward IBL 3D module on the negative-z side, obtained from simulated single-interaction minimum-bias events. Only clusters originating from primary particles are used. The broken or on-module-edge clusters are shorter than the unbroken off-module-edge clusters. On-module-edge clusters are identified on the basis of their position, length and width. If two clusters have a one-pixel gap along the z direction and their hits are in the same or adjacent rows along the azimuthal direction, they are identified as broken clusters from the same long cluster.	pdf file, eps file
Figure 3 Number of clusters in the 3D modules on the positive-z (red) and the negative-z (blue) side of the IBL as a function of the longitudinal position of the interaction, obtained from simulated single-interaction minimum-bias events. The distribution of all clusters is fitted with a second-order polynomial.	pdf file, eps file
Figure 4 Two-component fit to the cluster length distribution of whole clusters in the most forward IBL 3D module on the negative-z side. Clusters on the module edges and broken clusters have been removed. The data correspond to about one minute of data-taking and are extracted from randomly triggered events. The fit components are a Gaussian to describe the clusters from primary particles and a template derived from simulation to describe the clusters generated by particles from secondary interactions. Clusters shorter than three pixels are excluded from the fit to minimize the systematic uncertainty from background sources that are not simulated, such as noisy pixels.	pdf file, eps file
Figure 5 Azimuthal dependence of the number of primary clusters for the modules in the most forward IBL µ ring on the negative-z side. The data correspond to about one minute of data-taking and are extracted from randomly triggered events. A plain average of all modules at the same µ can be biased when a module has a transient problem. In order to exclude modules identified as having a transient problem, the azimuthal distribution of the number of clusters is fitted to a sinusoidal function, in order to account for the beams not being perfectly centered in the IBL. The area under the fitted curve is then taken as the correct cluster count. In this example, the module at azimuthal index 8 has been excluded.	pdf file, eps file
Figure 6 Dependence of the number of clusters on the luminous length. The black points show the number of clusters obtained in the simulated minimum-bias samples for different values of the luminous length σz but the same longitudinal position of the luminous centroid (µz = -2mm). A correction is applied to take into account the σz dependence; the green points show the number of clusters after correction. The correction is designed to return the ideal number of clusters that would have been observed if all interactions happened at the IBL center (µz = σz = 0) and therefore the corrected number of clusters is expected no longer to depend on µz or σz.	pdf file, eps file
Figure 7 Dependence of the number of clusters on the longitudinal position of the luminous centroid. The black points show the number of clusters obtained in the simulated minimum-bias samples for different longitudinal positions (µz) of the luminous centroid, but with the same luminous length (σz = 35mm). A correction is applied to take into account the µz dependence; the green points show the number of clusters after the correction. The correction is designed to return the ideal number of clusters that would have been observed if all interactions happened at the IBL center (µz = σz = 0) and therefore the corrected number of clusters is expected no longer to depend on µz or σz.	pdf file, eps file
Figure 8 Time stability of the PCC/LUCID luminosity ratio over the course of LHC fill 6024, after application of the luminous-region-dependent corrections illustrated in the previous two figures. LUCID is the baseline ATLAS luminometer. The rate of primary pixel clusters, integrated over the duration of the fill, is normalized to the LUCID-based integrated luminosity over the same period. Only statistical errors are shown.	pdf file, eps file

Pixel Cluster Counting in 2015 data

Typical distribution of the pixel-cluster size in the direction parallel to the beam axis, in the secondto-last forward inner layer (IBL) module Typical distribution of the pixel-cluster size in the direction parallel to the beam axis, in the secondto-last forward inner layer (IBL) module. The IBL is a cylinder that consists of 14 azimuthal “staves”. Each stave contains 20 modules arranged along the beam axis and covers the pseudorapidity range |eta| < 3. The 12413 clusters that populate this plot originate from 10433 randomly-triggered events during ATLAS physics data-taking in 2015. The distribution is fit to the sum of a Gaussian signal (mean from fit is 9-pixels cluster length), plus an exponentially-falling background of shorter clusters. The Gaussian area is proportional to the number of charged particles originating from the luminous region and therefore proportional to the luminosity integrated during the approximately one-minute data-taking period. The exponential area contains contributions from secondary interactions and afterglow that will deviate from proportionality to luminosity, and is therefore treated as a background.	pdf file eps file
Comparison of longitudinal cluster-size distributions for the four most forward modules Comparison of longitudinal cluster-size distributions for the four most forward modules in an arbitrarily chosen inner layer (IBL) stave. The IBL is a cylinder that consists of 14 azimuthal “staves”, each of which contains 20 modules arranged along the beam axis and covers the pseudorapidity range |eta| < 3. The modules farther from the nominal collision point produce, on the average, longer clusters, while the shape of the falling background is approximately the same in all modules. This confirms that the length of signal clusters is sensitive to the particle incidence angle, as expected from simple geometrical arguments.	pdf file eps file
Afterglow in the IBL During LHC operation, the IBL detector material is activated by pp collision debris. The resulting low energy decay products, collectively known as afterglow, induce delayed signals in the detector and produce single or small pixel clusters, thereby contributing to the background component of the cluster-length distributions rather than to the Gaussian-shaped long-cluster signal. The figure displays the bunch-slot dependence of the signal and background components, which correspond to, respectively, the area under the Gaussian curve and that under the falling exponential. These signal and background levels are extracted from fits to the cluster-length distributions from a forward IBL module, recorded in 8 consecutive 25 ns-long bunch slots. Only the first slot (numbered 2674) contained a colliding-bunch pair (and therefore a luminosity signal); the remaining 7 were nominally empty. The measured signal (Gaussian) component is largest in the first bunch slot, and drops by 4 orders of magnitude within 25 ns. The measured background (exponential) component, in contrast, decays much more slowly, with a time dependence suggestive of a mixture of radioactive decays. We have not attempted to identify those decays in this study.	pdf file eps file
Azimuthal dependence of the long-cluster (Gaussian) signal for the forward-most modules of the IBL. Azimuthal dependence of the long-cluster (Gaussian) signal for the forward-most modules of the IBL. The IBL is a cylinder that consists of 14 azimuthal staves, each of which contains 20 modules arranged along the beam axis and covers the pseudorapidity range |eta|<3. The sinusoidal dependence arises because in the plane perpendicular to the average beam direction, the luminous region is not perfectly centered on the IBL axis. In order to avoid being affected by occasional underperforming or noisy modules, the measured azimuthal dependence of the long-cluster signal is fit to the function A*cos((2*pi/14)*x + B)+C. Outliers are excluded from the fit on the basis of internal consistency. In this example, the 12th module was inefficient during the run considered; it was excluded form the fit based on its residual exceeding 4 sigma. The area under the fitted curve (rather than the sum of the 14 individual measurements) is used as a measure of the total number of signal clusters.	pdf file eps file
Time evolution, during an ATLAS run, of the longitudinal position of the luminous region Time evolution, during an ATLAS run, of the longitudinal position of the luminous region either as determined from the average position of reconstructed pp-collision vertices (light blue circles), or as inferred from the forward-backward asymmetry of the pixel-cluster counts in the IBL (black squares). This asymmetry is computed from the long-cluster signals in the 4 outermost IBL modules on either side of the interaction point. It is sensitive to the longitudinal position of the luminous region because the acceptance of each module depends on its distance from the average collision point. The pixel-cluster-based luminosity measurement can be corrected for this geometric effect by using the beamspot position derived from the measured asymmetry of the pixel-cluster signal.	pdf file eps file
Time evolution, during an ATLAS run, of the pixel-cluster luminosity signal Time evolution, during an ATLAS run, of the pixel-cluster luminosity signal normalized to the luminosity as measured by the ATLAS baseline luminometer (top), and of the RMS luminous length inferred from the longitudinal distribution of reconstructed pp-collision vertices (bottom). The luminosity-normalized pixel-cluster signal exhibits a similar time dependence to that of the luminous length, because the acceptance of each module depends on its distance from the corresponding interaction vertex. The pixel-cluster-based luminosity measurement must therefore be corrected for this geometric effect, using the luminous length derived from the longitudinal distribution of pp-collision vertices.	pdf file eps file
Comparison of the pixel-cluster counting (PCC) algorithm and the LUCID_EvtORA_Bi, TILE, and track-counting algorithms Fractional difference in run-integrated luminosity between the pixel-cluster counting (PCC) algorithm and the LUCID_EvtORA_Bi, TILE, and track-counting algorithms, for the subset of ATLAS runs with 25 ns bunch spacing for which the PCC data are available. The absolute scale of the PCC-based luminosity is cross-calibrated to the luminosity from the comparison algorithm averaged over the runs taken from 11 to 13 September 2015.	pdf file eps file
Fractional deviation of the bunch-averaged pile-up parameter average mu Fractional deviation of the bunch-averaged pile-up parameter average mu, obtained using different algorithms, from the pixel-cluster counting (PCC) value, as a function of average mu during a physics run on September 25, 2015. The data are normalized such that all algorithms yield the same integrated luminosity in the run considered.	pdf file eps file

---

Major updates:
-- EricTorrence - 2015-07-20

Responsible: EricTorrence
Subject: public