Compact Muon Solenoid
LHC, CERN
Visit us: CMS Public Website, CMS Physics ;
Contact us: CMS Publications Committee
| CMS-PAS-JME-18-001 | ||
| Pileup mitigation at CMS in 13 TeV data | ||
| CMS Collaboration | ||
| July 2019 | ||
| Abstract: With increasing instantaneous luminosity at the Large Hadron Collider come additional reconstruction challenges. At high luminosity, many collisions occur simultaneously within one proton-proton bunch crossing. Isolating a collision of interest from the additional "pileup'' collisions is needed so as to allow for effective physics performance. In the CMS Collaboration, a number of techniques have been developed capable of mitigating the impact of these pileup collisions. Such methods include charged-hadron subtraction, pileup jet identification, neutral isospin based "$\Delta\beta$'' correction, and most recently pileup per particle identification. This note represents a survey of the performance of these techniques covering jet and missing transverse momentum reconstruction as well as lepton isolation using 35.9 and 41.4 fb$^{-1}$ of data collected with the CMS experiment in 2016 and 2017 at a center-of-mass energy of 13 TeV with up to 50 simultaneous collisions per bunch crossing. The performance of each algorithm is discussed including projections for the future, higher luminosity, running of the LHC. Significant improvements in jet energy, mass, and angular resolution, the identification of pileup jets, missing transverse momentum resolution, and lepton isolation are found. | ||
|
Links:
CDS record (PDF) ;
CADI line (restricted) ;
These preliminary results are superseded in this paper, JINST 15 (2020) P09018. The superseded preliminary plots can be found here. |
||
| Figures | |
png pdf |
Figure 1:
Distribution of the mean number of inelastic interactions per crossing (pileup) in data for pp collisions in 2016 (orange), 2017 (light blue), and 2018 (navy blue). A minimum bias cross section of 69.2 mb is chosen. The mean number per year of inelastic interactions is written in the legend. |
png pdf |
Figure 2:
Data-to-simulation comparison for three different variables of the PUPPI algorithm. The markers show a subset of the data taken in 2016 while the solid lines are QCD multijet simulation. The lower panel of each plot shows the ratio of data to simulation. Only statistical errors are displayed. Each distribution is normalized to one. The upper plot shows the $\alpha $ distribution for charged particles associated to the LV (red triangles), charged particles associated to PU vertices (blue circles) and neutral particles (black crosses) in the central region of the detector (0 $ < |\eta | < $ 2.5). The lower left plot shows the signed $\chi ^{2}=(\alpha - \bar{\alpha}_{{\text {PU}}})|\alpha - \bar{\alpha}_{{\text {PU}}}|/ {\text {RMS}}_{{\text {PU}}}^{2}$ for charged particles associated to PU vertices. The lower right plot shows the PUPPI weight distribution for neutral particles. This distribution is normalized to unity only taking into account particles with weight greater than 0.01, i.e., those that are not rejected by the PUPPI algorithm. |
png pdf |
Figure 2-a:
Data-to-simulation comparison for three different variables of the PUPPI algorithm. The markers show a subset of the data taken in 2016 while the solid lines are QCD multijet simulation. The lower panel of each plot shows the ratio of data to simulation. Only statistical errors are displayed. Each distribution is normalized to one. The upper plot shows the $\alpha $ distribution for charged particles associated to the LV (red triangles), charged particles associated to PU vertices (blue circles) and neutral particles (black crosses) in the central region of the detector (0 $ < |\eta | < $ 2.5). The lower left plot shows the signed $\chi ^{2}=(\alpha - \bar{\alpha}_{{\text {PU}}})|\alpha - \bar{\alpha}_{{\text {PU}}}|/ {\text {RMS}}_{{\text {PU}}}^{2}$ for charged particles associated to PU vertices. The lower right plot shows the PUPPI weight distribution for neutral particles. This distribution is normalized to unity only taking into account particles with weight greater than 0.01, i.e., those that are not rejected by the PUPPI algorithm. |
png pdf |
Figure 2-b:
Data-to-simulation comparison for three different variables of the PUPPI algorithm. The markers show a subset of the data taken in 2016 while the solid lines are QCD multijet simulation. The lower panel of each plot shows the ratio of data to simulation. Only statistical errors are displayed. Each distribution is normalized to one. The upper plot shows the $\alpha $ distribution for charged particles associated to the LV (red triangles), charged particles associated to PU vertices (blue circles) and neutral particles (black crosses) in the central region of the detector (0 $ < |\eta | < $ 2.5). The lower left plot shows the signed $\chi ^{2}=(\alpha - \bar{\alpha}_{{\text {PU}}})|\alpha - \bar{\alpha}_{{\text {PU}}}|/ {\text {RMS}}_{{\text {PU}}}^{2}$ for charged particles associated to PU vertices. The lower right plot shows the PUPPI weight distribution for neutral particles. This distribution is normalized to unity only taking into account particles with weight greater than 0.01, i.e., those that are not rejected by the PUPPI algorithm. |
png pdf |
Figure 2-c:
Data-to-simulation comparison for three different variables of the PUPPI algorithm. The markers show a subset of the data taken in 2016 while the solid lines are QCD multijet simulation. The lower panel of each plot shows the ratio of data to simulation. Only statistical errors are displayed. Each distribution is normalized to one. The upper plot shows the $\alpha $ distribution for charged particles associated to the LV (red triangles), charged particles associated to PU vertices (blue circles) and neutral particles (black crosses) in the central region of the detector (0 $ < |\eta | < $ 2.5). The lower left plot shows the signed $\chi ^{2}=(\alpha - \bar{\alpha}_{{\text {PU}}})|\alpha - \bar{\alpha}_{{\text {PU}}}|/ {\text {RMS}}_{{\text {PU}}}^{2}$ for charged particles associated to PU vertices. The lower right plot shows the PUPPI weight distribution for neutral particles. This distribution is normalized to unity only taking into account particles with weight greater than 0.01, i.e., those that are not rejected by the PUPPI algorithm. |
png pdf |
Figure 3:
Jet energy resolution as a function of the generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 3-a:
Jet energy resolution as a function of the generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 3-b:
Jet energy resolution as a function of the generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 3-c:
Jet energy resolution as a function of the generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 4:
Jet energy resolution as a function of the number of pileup interactions for jets with CHS (solid red line) and with PUPPI (dashed blue line) applied in QCD multijet simulation for different jet ${p_{\mathrm {T}}}$ values (different markers). The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 4-a:
Jet energy resolution as a function of the number of pileup interactions for jets with CHS (solid red line) and with PUPPI (dashed blue line) applied in QCD multijet simulation for different jet ${p_{\mathrm {T}}}$ values (different markers). The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 4-b:
Jet energy resolution as a function of the number of pileup interactions for jets with CHS (solid red line) and with PUPPI (dashed blue line) applied in QCD multijet simulation for different jet ${p_{\mathrm {T}}}$ values (different markers). The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 4-c:
Jet energy resolution as a function of the number of pileup interactions for jets with CHS (solid red line) and with PUPPI (dashed blue line) applied in QCD multijet simulation for different jet ${p_{\mathrm {T}}}$ values (different markers). The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 5:
Jet $\eta $ resolution as a function of generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 5-a:
Jet $\eta $ resolution as a function of generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 5-b:
Jet $\eta $ resolution as a function of generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 5-c:
Jet $\eta $ resolution as a function of generator-level jet ${p_{\mathrm {T}}}$ for raw PF jets (orange circles), PF jets with CHS applied (red triangles), and PF jets with PUPPI applied (blue squares) in QCD multijet simulation. The number of pileup interactions is chosen to be between 20 and 30. The resolution is shown for AK4 jets with $|\eta | < $ 0.5 (upper left) and 3.2 $ < |\eta | < $ 4.7 (upper right) as well as for AK8 jets with $|\eta | < $ 0.5 (lower). |
png pdf |
Figure 6:
The charged and neutral multiplicities for CHS and PUPPI in a dijet selection (real jets) and minimum bias selection (noise jets) in data. The multiplicities are shown for AK4 jets using CHS reconstructed real jets (red dashed), CHS reconstructed noise jets (black long dashed), PUPPI reconstructed real jets (blue solid), and PUPPI reconstructed noise jets (orange varied dashes). The upper plots show the charged (upper left) and neutral particle multiplicities (upper right) for jets with $|\eta | < $ 0.5. The lower left plot shows the neutral particle multiplicity for jets with 3 $ < |\eta | < $ 5. The lower right plot shows the neutral particle multiplicity of AK4 jets with $|\eta | < $ 0.5 in a dijet selection in data using CHS (red dashed and black long dashed) and PUPPI (blue solid and orange varied dashed) for PU of 20-30 and 50-60, respectively. |
png pdf |
Figure 6-a:
The charged and neutral multiplicities for CHS and PUPPI in a dijet selection (real jets) and minimum bias selection (noise jets) in data. The multiplicities are shown for AK4 jets using CHS reconstructed real jets (red dashed), CHS reconstructed noise jets (black long dashed), PUPPI reconstructed real jets (blue solid), and PUPPI reconstructed noise jets (orange varied dashes). The upper plots show the charged (upper left) and neutral particle multiplicities (upper right) for jets with $|\eta | < $ 0.5. The lower left plot shows the neutral particle multiplicity for jets with 3 $ < |\eta | < $ 5. The lower right plot shows the neutral particle multiplicity of AK4 jets with $|\eta | < $ 0.5 in a dijet selection in data using CHS (red dashed and black long dashed) and PUPPI (blue solid and orange varied dashed) for PU of 20-30 and 50-60, respectively. |
png pdf |
Figure 6-b:
The charged and neutral multiplicities for CHS and PUPPI in a dijet selection (real jets) and minimum bias selection (noise jets) in data. The multiplicities are shown for AK4 jets using CHS reconstructed real jets (red dashed), CHS reconstructed noise jets (black long dashed), PUPPI reconstructed real jets (blue solid), and PUPPI reconstructed noise jets (orange varied dashes). The upper plots show the charged (upper left) and neutral particle multiplicities (upper right) for jets with $|\eta | < $ 0.5. The lower left plot shows the neutral particle multiplicity for jets with 3 $ < |\eta | < $ 5. The lower right plot shows the neutral particle multiplicity of AK4 jets with $|\eta | < $ 0.5 in a dijet selection in data using CHS (red dashed and black long dashed) and PUPPI (blue solid and orange varied dashed) for PU of 20-30 and 50-60, respectively. |
png pdf |
Figure 6-c:
The charged and neutral multiplicities for CHS and PUPPI in a dijet selection (real jets) and minimum bias selection (noise jets) in data. The multiplicities are shown for AK4 jets using CHS reconstructed real jets (red dashed), CHS reconstructed noise jets (black long dashed), PUPPI reconstructed real jets (blue solid), and PUPPI reconstructed noise jets (orange varied dashes). The upper plots show the charged (upper left) and neutral particle multiplicities (upper right) for jets with $|\eta | < $ 0.5. The lower left plot shows the neutral particle multiplicity for jets with 3 $ < |\eta | < $ 5. The lower right plot shows the neutral particle multiplicity of AK4 jets with $|\eta | < $ 0.5 in a dijet selection in data using CHS (red dashed and black long dashed) and PUPPI (blue solid and orange varied dashed) for PU of 20-30 and 50-60, respectively. |
png pdf |
Figure 6-d:
The charged and neutral multiplicities for CHS and PUPPI in a dijet selection (real jets) and minimum bias selection (noise jets) in data. The multiplicities are shown for AK4 jets using CHS reconstructed real jets (red dashed), CHS reconstructed noise jets (black long dashed), PUPPI reconstructed real jets (blue solid), and PUPPI reconstructed noise jets (orange varied dashes). The upper plots show the charged (upper left) and neutral particle multiplicities (upper right) for jets with $|\eta | < $ 0.5. The lower left plot shows the neutral particle multiplicity for jets with 3 $ < |\eta | < $ 5. The lower right plot shows the neutral particle multiplicity of AK4 jets with $|\eta | < $ 0.5 in a dijet selection in data using CHS (red dashed and black long dashed) and PUPPI (blue solid and orange varied dashed) for PU of 20-30 and 50-60, respectively. |
png pdf |
Figure 7:
Data-to-simulation comparison for two input variables to the PU jet ID calculation: $\beta $ (upper left) and charged particle multiplicity (upper right). The lower plots are data-to-simulation comparisons of the PU jet ID boosted decision tree (BDT) output for AK4 jets with 30 $ < {p_{\mathrm {T}}} < $ 50 GeV for the detector region within the tracker volume (left) and 3 $ < |\eta | < $ 5 (right). Black markers represent the data while the colored areas are DY+jets simulation events. The simulation sample is split into jets originating from quarks (red), gluons (purple), PU (green), and jets that could not be assigned (gray). The distributions are normalized to unity. In the lower plots, the shape of a sample showered with {herwig} is superimposed and included in the total uncertainty band in the ratio panel (light gray). Also included in the ratio panel is the PU rate uncertainty (dark gray). |
png pdf |
Figure 7-a:
Data-to-simulation comparison for two input variables to the PU jet ID calculation: $\beta $ (upper left) and charged particle multiplicity (upper right). The lower plots are data-to-simulation comparisons of the PU jet ID boosted decision tree (BDT) output for AK4 jets with 30 $ < {p_{\mathrm {T}}} < $ 50 GeV for the detector region within the tracker volume (left) and 3 $ < |\eta | < $ 5 (right). Black markers represent the data while the colored areas are DY+jets simulation events. The simulation sample is split into jets originating from quarks (red), gluons (purple), PU (green), and jets that could not be assigned (gray). The distributions are normalized to unity. In the lower plots, the shape of a sample showered with {herwig} is superimposed and included in the total uncertainty band in the ratio panel (light gray). Also included in the ratio panel is the PU rate uncertainty (dark gray). |
png pdf |
Figure 7-b:
Data-to-simulation comparison for two input variables to the PU jet ID calculation: $\beta $ (upper left) and charged particle multiplicity (upper right). The lower plots are data-to-simulation comparisons of the PU jet ID boosted decision tree (BDT) output for AK4 jets with 30 $ < {p_{\mathrm {T}}} < $ 50 GeV for the detector region within the tracker volume (left) and 3 $ < |\eta | < $ 5 (right). Black markers represent the data while the colored areas are DY+jets simulation events. The simulation sample is split into jets originating from quarks (red), gluons (purple), PU (green), and jets that could not be assigned (gray). The distributions are normalized to unity. In the lower plots, the shape of a sample showered with {herwig} is superimposed and included in the total uncertainty band in the ratio panel (light gray). Also included in the ratio panel is the PU rate uncertainty (dark gray). |
png pdf |
Figure 7-c:
Data-to-simulation comparison for two input variables to the PU jet ID calculation: $\beta $ (upper left) and charged particle multiplicity (upper right). The lower plots are data-to-simulation comparisons of the PU jet ID boosted decision tree (BDT) output for AK4 jets with 30 $ < {p_{\mathrm {T}}} < $ 50 GeV for the detector region within the tracker volume (left) and 3 $ < |\eta | < $ 5 (right). Black markers represent the data while the colored areas are DY+jets simulation events. The simulation sample is split into jets originating from quarks (red), gluons (purple), PU (green), and jets that could not be assigned (gray). The distributions are normalized to unity. In the lower plots, the shape of a sample showered with {herwig} is superimposed and included in the total uncertainty band in the ratio panel (light gray). Also included in the ratio panel is the PU rate uncertainty (dark gray). |
png pdf |
Figure 7-d:
Data-to-simulation comparison for two input variables to the PU jet ID calculation: $\beta $ (upper left) and charged particle multiplicity (upper right). The lower plots are data-to-simulation comparisons of the PU jet ID boosted decision tree (BDT) output for AK4 jets with 30 $ < {p_{\mathrm {T}}} < $ 50 GeV for the detector region within the tracker volume (left) and 3 $ < |\eta | < $ 5 (right). Black markers represent the data while the colored areas are DY+jets simulation events. The simulation sample is split into jets originating from quarks (red), gluons (purple), PU (green), and jets that could not be assigned (gray). The distributions are normalized to unity. In the lower plots, the shape of a sample showered with {herwig} is superimposed and included in the total uncertainty band in the ratio panel (light gray). Also included in the ratio panel is the PU rate uncertainty (dark gray). |
png pdf |
Figure 8:
Leading vertex jet efficiency (upper) and purity (lower) in DY+jets simulation as a function of the number of pileup interactions for PUPPI (blue closed squares), CHS (red closed triangles), CHS+tight PU jet ID (magenta open squares), CHS+medium PU jet ID (orange crosses), and CHS+loose PU jet ID (black triangles). Plots are shown with AK4 jets having a $ {p_{\mathrm {T}}} > $ 20 GeV and (left) $|\eta | < $ 2.5 and (right) $|\eta | > 3$. The LV jet efficiency is defined as the number of matched reconstruction-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of generator-level jets with $ {p_{\mathrm {T}}} > $ 30 GeV that originate from the main interaction. For the lower plots, the purity is defined as the number of matched generator-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of reconstructed jets that have $ {p_{\mathrm {T}}} > $ 30 GeV. |
png pdf |
Figure 8-a:
Leading vertex jet efficiency (upper) and purity (lower) in DY+jets simulation as a function of the number of pileup interactions for PUPPI (blue closed squares), CHS (red closed triangles), CHS+tight PU jet ID (magenta open squares), CHS+medium PU jet ID (orange crosses), and CHS+loose PU jet ID (black triangles). Plots are shown with AK4 jets having a $ {p_{\mathrm {T}}} > $ 20 GeV and (left) $|\eta | < $ 2.5 and (right) $|\eta | > 3$. The LV jet efficiency is defined as the number of matched reconstruction-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of generator-level jets with $ {p_{\mathrm {T}}} > $ 30 GeV that originate from the main interaction. For the lower plots, the purity is defined as the number of matched generator-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of reconstructed jets that have $ {p_{\mathrm {T}}} > $ 30 GeV. |
png pdf |
Figure 8-b:
Leading vertex jet efficiency (upper) and purity (lower) in DY+jets simulation as a function of the number of pileup interactions for PUPPI (blue closed squares), CHS (red closed triangles), CHS+tight PU jet ID (magenta open squares), CHS+medium PU jet ID (orange crosses), and CHS+loose PU jet ID (black triangles). Plots are shown with AK4 jets having a $ {p_{\mathrm {T}}} > $ 20 GeV and (left) $|\eta | < $ 2.5 and (right) $|\eta | > 3$. The LV jet efficiency is defined as the number of matched reconstruction-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of generator-level jets with $ {p_{\mathrm {T}}} > $ 30 GeV that originate from the main interaction. For the lower plots, the purity is defined as the number of matched generator-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of reconstructed jets that have $ {p_{\mathrm {T}}} > $ 30 GeV. |
png pdf |
Figure 8-c:
Leading vertex jet efficiency (upper) and purity (lower) in DY+jets simulation as a function of the number of pileup interactions for PUPPI (blue closed squares), CHS (red closed triangles), CHS+tight PU jet ID (magenta open squares), CHS+medium PU jet ID (orange crosses), and CHS+loose PU jet ID (black triangles). Plots are shown with AK4 jets having a $ {p_{\mathrm {T}}} > $ 20 GeV and (left) $|\eta | < $ 2.5 and (right) $|\eta | > 3$. The LV jet efficiency is defined as the number of matched reconstruction-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of generator-level jets with $ {p_{\mathrm {T}}} > $ 30 GeV that originate from the main interaction. For the lower plots, the purity is defined as the number of matched generator-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of reconstructed jets that have $ {p_{\mathrm {T}}} > $ 30 GeV. |
png pdf |
Figure 8-d:
Leading vertex jet efficiency (upper) and purity (lower) in DY+jets simulation as a function of the number of pileup interactions for PUPPI (blue closed squares), CHS (red closed triangles), CHS+tight PU jet ID (magenta open squares), CHS+medium PU jet ID (orange crosses), and CHS+loose PU jet ID (black triangles). Plots are shown with AK4 jets having a $ {p_{\mathrm {T}}} > $ 20 GeV and (left) $|\eta | < $ 2.5 and (right) $|\eta | > 3$. The LV jet efficiency is defined as the number of matched reconstruction-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of generator-level jets with $ {p_{\mathrm {T}}} > $ 30 GeV that originate from the main interaction. For the lower plots, the purity is defined as the number of matched generator-level jets with $ {p_{\mathrm {T}}} > $ 20 GeV divided by the number of reconstructed jets that have $ {p_{\mathrm {T}}} > $ 30 GeV. |
png pdf |
Figure 9:
Rate of jets in the PU-enriched region divided by the rate of jets in the LV-enriched region as a function of the number of vertices for CHS jets (red triangles), CHS jets with medium PU jet ID applied (orange crosses) and PUPPI jets (blue squares) in DY+jets simulation (full markers) and data (open markers). The plots show the ratio for events in the central region (left) and in the forward region (right). The lower panels show the data-to-simulation ratio along with a gray band corresponding to the one-sided uncertainty that is the difference between simulated DY+jets events showered with the {pythia} parton shower to those showered with the {herwig} parton shower. |
png pdf |
Figure 9-a:
Rate of jets in the PU-enriched region divided by the rate of jets in the LV-enriched region as a function of the number of vertices for CHS jets (red triangles), CHS jets with medium PU jet ID applied (orange crosses) and PUPPI jets (blue squares) in DY+jets simulation (full markers) and data (open markers). The plots show the ratio for events in the central region (left) and in the forward region (right). The lower panels show the data-to-simulation ratio along with a gray band corresponding to the one-sided uncertainty that is the difference between simulated DY+jets events showered with the {pythia} parton shower to those showered with the {herwig} parton shower. |
png pdf |
Figure 9-b:
Rate of jets in the PU-enriched region divided by the rate of jets in the LV-enriched region as a function of the number of vertices for CHS jets (red triangles), CHS jets with medium PU jet ID applied (orange crosses) and PUPPI jets (blue squares) in DY+jets simulation (full markers) and data (open markers). The plots show the ratio for events in the central region (left) and in the forward region (right). The lower panels show the data-to-simulation ratio along with a gray band corresponding to the one-sided uncertainty that is the difference between simulated DY+jets events showered with the {pythia} parton shower to those showered with the {herwig} parton shower. |
png pdf |
Figure 10:
Median soft drop jet mass (upper left), median $\tau _{21}$ (upper right), and soft drop jet mass resolution (lower) for AK8 jets from boosted W bosons with 400 $ < {p_{\mathrm {T}}} < $ 600 GeV for CHS (red triangles) and PUPPI (blue squares) of a bulk graviton signal sample as a function of the number of vertices. |
png pdf |
Figure 10-a:
Median soft drop jet mass (upper left), median $\tau _{21}$ (upper right), and soft drop jet mass resolution (lower) for AK8 jets from boosted W bosons with 400 $ < {p_{\mathrm {T}}} < $ 600 GeV for CHS (red triangles) and PUPPI (blue squares) of a bulk graviton signal sample as a function of the number of vertices. |
png pdf |
Figure 10-b:
Median soft drop jet mass (upper left), median $\tau _{21}$ (upper right), and soft drop jet mass resolution (lower) for AK8 jets from boosted W bosons with 400 $ < {p_{\mathrm {T}}} < $ 600 GeV for CHS (red triangles) and PUPPI (blue squares) of a bulk graviton signal sample as a function of the number of vertices. |
png pdf |
Figure 10-c:
Median soft drop jet mass (upper left), median $\tau _{21}$ (upper right), and soft drop jet mass resolution (lower) for AK8 jets from boosted W bosons with 400 $ < {p_{\mathrm {T}}} < $ 600 GeV for CHS (red triangles) and PUPPI (blue squares) of a bulk graviton signal sample as a function of the number of vertices. |
png pdf |
Figure 11:
W boson identification performance using a selection on $\tau _{21}$ for CHS (red triangles and dark red crosses) and PUPPI (blue squares and circles) AK8 jets as function of the number of vertices for loose and tight selection cuts, respectively. Shown on the left is the W boson identification efficiency evaluated in simulation for a bulk graviton decaying into a WW boson pair and on the right the misidentification rate evaluated with QCD multijet simulation. |
png pdf |
Figure 11-a:
W boson identification performance using a selection on $\tau _{21}$ for CHS (red triangles and dark red crosses) and PUPPI (blue squares and circles) AK8 jets as function of the number of vertices for loose and tight selection cuts, respectively. Shown on the left is the W boson identification efficiency evaluated in simulation for a bulk graviton decaying into a WW boson pair and on the right the misidentification rate evaluated with QCD multijet simulation. |
png pdf |
Figure 11-b:
W boson identification performance using a selection on $\tau _{21}$ for CHS (red triangles and dark red crosses) and PUPPI (blue squares and circles) AK8 jets as function of the number of vertices for loose and tight selection cuts, respectively. Shown on the left is the W boson identification efficiency evaluated in simulation for a bulk graviton decaying into a WW boson pair and on the right the misidentification rate evaluated with QCD multijet simulation. |
png pdf |
Figure 12:
Top quark identification efficiency (left) and misidentification rate (right) as function of the number of vertices for CHS (open symbols) and PUPPI (closed symbols) using different combinations of substructure variables: soft drop mass cut between 105 and 210 GeV (blue rectangles), $\tau _{32} < $ 0.54 cut (orange circles), and both together (red triangles). |
png pdf |
Figure 12-a:
Top quark identification efficiency (left) and misidentification rate (right) as function of the number of vertices for CHS (open symbols) and PUPPI (closed symbols) using different combinations of substructure variables: soft drop mass cut between 105 and 210 GeV (blue rectangles), $\tau _{32} < $ 0.54 cut (orange circles), and both together (red triangles). |
png pdf |
Figure 12-b:
Top quark identification efficiency (left) and misidentification rate (right) as function of the number of vertices for CHS (open symbols) and PUPPI (closed symbols) using different combinations of substructure variables: soft drop mass cut between 105 and 210 GeV (blue rectangles), $\tau _{32} < $ 0.54 cut (orange circles), and both together (red triangles). |
png pdf |
Figure 13:
The hadronic recoil response ($-< u_{\parallel} > /< q_{T} > $) of the Z boson computed for PUPPI and PF ${{p_{\mathrm {T}}} ^\text {miss}}$ as a function of $q_{T}$ in data ${{\mathrm{Z} \rightarrow ee}}$ events. The lower panel shows the data-to-simulation ratio. A gray shaded band is added in the lower panel showing the systematic uncertainties resulting from JES and JER variations, and variations in the unclustered energy added in quadrature. |
png pdf |
Figure 14:
The hadronic recoil components ${u_{||}}$ (left) and ${u_{\perp}}$ (right) for PUPPI and PF ${{p_{\mathrm {T}}} ^\text {miss}}$ resolution as a function of the number of vertices in data in ${{\mathrm{Z} \rightarrow ee}}$ events. The lower panel shows the data-to-simulation ratio. A gray shaded band is added in the lower panel showing the systematic uncertainties resulting from JES and JER variations, and variations in the unclustered energy added in quadrature. |
png pdf |
Figure 14-a:
The hadronic recoil components ${u_{||}}$ (left) and ${u_{\perp}}$ (right) for PUPPI and PF ${{p_{\mathrm {T}}} ^\text {miss}}$ resolution as a function of the number of vertices in data in ${{\mathrm{Z} \rightarrow ee}}$ events. The lower panel shows the data-to-simulation ratio. A gray shaded band is added in the lower panel showing the systematic uncertainties resulting from JES and JER variations, and variations in the unclustered energy added in quadrature. |
png pdf |
Figure 14-b:
The hadronic recoil components ${u_{||}}$ (left) and ${u_{\perp}}$ (right) for PUPPI and PF ${{p_{\mathrm {T}}} ^\text {miss}}$ resolution as a function of the number of vertices in data in ${{\mathrm{Z} \rightarrow ee}}$ events. The lower panel shows the data-to-simulation ratio. A gray shaded band is added in the lower panel showing the systematic uncertainties resulting from JES and JER variations, and variations in the unclustered energy added in quadrature. |
png pdf |
Figure 15:
The identification efficiency for prompt muons in DY+jets simulated events (left) and the misidentification rate for nonprompt muons in QCD multijet simulated events (right) for the different definitions of the isolation: $\Delta \beta $-corrected isolation (black circles), PUPPI-with-lepton (blue triangles), PUPPI-no-lepton (red crosses), PUPPI-combined (green squares) as a function of the number of vertices. The threshold of each isolation is set to yield 12% misidentification rate for reconstructed muons in QCD multijet simulation. |
png pdf |
Figure 15-a:
The identification efficiency for prompt muons in DY+jets simulated events (left) and the misidentification rate for nonprompt muons in QCD multijet simulated events (right) for the different definitions of the isolation: $\Delta \beta $-corrected isolation (black circles), PUPPI-with-lepton (blue triangles), PUPPI-no-lepton (red crosses), PUPPI-combined (green squares) as a function of the number of vertices. The threshold of each isolation is set to yield 12% misidentification rate for reconstructed muons in QCD multijet simulation. |
png pdf |
Figure 15-b:
The identification efficiency for prompt muons in DY+jets simulated events (left) and the misidentification rate for nonprompt muons in QCD multijet simulated events (right) for the different definitions of the isolation: $\Delta \beta $-corrected isolation (black circles), PUPPI-with-lepton (blue triangles), PUPPI-no-lepton (red crosses), PUPPI-combined (green squares) as a function of the number of vertices. The threshold of each isolation is set to yield 12% misidentification rate for reconstructed muons in QCD multijet simulation. |
png pdf |
Figure 16:
The identification efficiency for prompt muons in DY+jets simulated events as a function of the misidentification rate for nonprompt muons in QCD multijet simulated events for the different definitions of the isolation: $\Delta \beta $-corrected isolation (black solid line), PUPPI-with-lepton (blue dashed line), PUPPI-no-lepton (red mixed dashed), PUPPI-combined (green long mixed dashed). The average number of pileup interactions is 27. |
png pdf |
Figure 17:
Mean relative isolation for PUPPI-with-lepton (left) and PUPPI-no-lepton (right) in data compared to DY+jets simulation. The relative isolation is split into separate charged hadron (CH, green squares), neutral hadron (NH, blue circles), photon (PH, red crosses) components, and combined (black triangles). Data and simulation are shown as full and open symbols, respectively. The lower panels show the data-to-simulation ratio of each component. |
png pdf |
Figure 17-a:
Mean relative isolation for PUPPI-with-lepton (left) and PUPPI-no-lepton (right) in data compared to DY+jets simulation. The relative isolation is split into separate charged hadron (CH, green squares), neutral hadron (NH, blue circles), photon (PH, red crosses) components, and combined (black triangles). Data and simulation are shown as full and open symbols, respectively. The lower panels show the data-to-simulation ratio of each component. |
png pdf |
Figure 17-b:
Mean relative isolation for PUPPI-with-lepton (left) and PUPPI-no-lepton (right) in data compared to DY+jets simulation. The relative isolation is split into separate charged hadron (CH, green squares), neutral hadron (NH, blue circles), photon (PH, red crosses) components, and combined (black triangles). Data and simulation are shown as full and open symbols, respectively. The lower panels show the data-to-simulation ratio of each component. |
png pdf |
Figure 18:
The identification efficiency for prompt muon isolation selection in $\mathrm{Z} \rightarrow \mu \mu $ data compared to DY+jets simulation as a function of number of vertices for PUPPI-combined (blue circles) and $\Delta \beta $-corrected isolation (red squares). Data and simulation are shown as full and open symbols, respectively. The lower panel shows the data-to-simulation ratio. |
png pdf |
Figure 19:
The misidentification rate defined as the number of events with a third isolated muon divided by the total number of events with a third muon in $\mathrm{Z} \rightarrow \mu \mu $ data for PUPPI-combined (blue closed circles) and $\Delta \beta $-corrected isolation (red open circles). The lower panel shows the ratio of PUPPI-combined and $\Delta \beta $-corrected isolation. |
| Tables | |
png pdf |
Table 1:
The tunable parameters of PUPPI optimized for application in 2016 data analysis. |
png pdf |
Table 2:
List of variables used in the PU jet ID for CHS jets. |
png pdf |
Table 3:
Data-to-simulation scale factors for the jet mass scale, jet mass resolution, and the $\tau _{21}$ selection efficiency for CHS and PUPPI. |
| Summary |
| The impact of pileup (PU) mitigation techniques on object reconstruction performance in the CMS experiment has been presented. The main techniques under study are charged-hadron subtraction (CHS) and pileup per particle identification (PUPPI), which both exploit particle information. The performance of these techniques is evaluated in the context of the reconstruction of jets and missing transverse momentum (${p_{\mathrm{T}}^{\text{miss}}}$), the calculation of jet substructure observables for boosted object tagging, and lepton isolation. CHS and PUPPI are further compared to other algorithmic approaches that act on jet, ${p_{\mathrm{T}}^{\text{miss}}}$, and lepton objects. While CHS rejects charged particles associated to PU vertices, PUPPI applies a more stringent selection to charged particles and rescales the four-momentum of neutral particles according to their probability to originate from the leading vertex. Both techniques reduce the dependence on PU interactions across all objects. A stronger reduction is achieved with PUPPI, in particular when considering events with more than 30 PU interactions. The best performance for jet mass and substructure observables, ${p_{\mathrm{T}}^{\text{miss}}}$ resolution, and rejection of misidentified leptons is also observed when using PUPPI. With respect to jet-momentum resolution and PU jet rejection, the preferred algorithm depends on the physics process under study: PUPPI provides a better jet-momentum resolution for jets with ${p_{\mathrm{T}}} < $ 100 GeV while CHS does so for ${p_{\mathrm{T}}} > $ 100 GeV. The highest rejection rate for jets originating purely from PU can be obtained when using a dedicated PU jet ID on top of CHS. However, when a looser working point for the PU jet ID is chosen such that it has an efficiency for jets coming from the leading vertex similar to PUPPI, PUPPI provides a similar rejection power. The PU suppression techniques studied in this note are proven to maintain reasonable object performance up to 50 PU interactions. Their use will be crucial for future running of the LHC, where yet higher PU conditions are expected. |
| References | ||||
| 1 | CMS Collaboration | CMS Luminosity Measurements for the 2016 Data Taking Period | CMS-PAS-LUM-17-001 | CMS-PAS-LUM-17-001 |
| 2 | CMS Collaboration | Particle-flow reconstruction and global event description with the CMS detector | JINST 12 (2017) P10003 | CMS-PRF-14-001 1706.04965 |
| 3 | M. Cacciari, G. P. Salam, and G. Soyez | The catchment area of jets | JHEP 04 (2008) 005 | 0802.1188 |
| 4 | M. Cacciari and G. P. Salam | Pileup subtraction using jet areas | PLB 659 (2008) 119 | 0707.1378 |
| 5 | CMS Collaboration | Jet energy scale and resolution in the CMS experiment in pp collisions at 8 TeV | JINST 12 (2017) P02014 | CMS-JME-13-004 1607.03663 |
| 6 | CMS Collaboration | Jet algorithms performance in 13 TeV data | CMS-PAS-JME-16-003 | CMS-PAS-JME-16-003 |
| 7 | D. Bertolini, P. Harris, M. Low, and N. Tran | Pileup Per Particle Identification | JHEP 10 (2014) 059 | 1407.6013 |
| 8 | CMS Collaboration | Description and performance of track and primary-vertex reconstruction with the CMS tracker | JINST 9 (2014) P10009 | CMS-TRK-11-001 1405.6569 |
| 9 | CMS Collaboration | The CMS experiment at the CERN LHC | JINST 3 (2008) S08004 | CMS-00-001 |
| 10 | CMS Collaboration | The CMS trigger system | JINST 12 (2017) P01020 | CMS-TRG-12-001 1609.02366 |
| 11 | CMS Collaboration | 2017 tracking performance plots | CDS | |
| 12 | J. Alwall et al. | The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations | JHEP 07 (2014) 079 | 1405.0301 |
| 13 | T. Sjostrand, S. Mrenna, and P. Z. Skands | PYTHIA 6.4 Physics and Manual | JHEP 05 (2006) 026 | hep-ph/0603175 |
| 14 | T. Sjostrand, S. Mrenna, and P. Z. Skands | A Brief Introduction to PYTHIA 8.1 | CPC 178 (2008) 852 | 0710.3820 |
| 15 | B. Andersson, G. Gustafson, G. Ingelman, and T. Sjostrand | Parton Fragmentation and String Dynamics | PR 97 (1983) 31 | |
| 16 | T. Sjostrand | The Merging of Jets | PLB 142 (1984) 420 | |
| 17 | P. Nason | A New method for combining NLO QCD with shower Monte Carlo algorithms | JHEP 11 (2004) 040 | hep-ph/0409146 |
| 18 | S. Frixione, P. Nason, and C. Oleari | Matching NLO QCD computations with Parton Shower simulations: the POWHEG method | JHEP 11 (2007) 070 | 0709.2092 |
| 19 | S. Alioli, P. Nason, C. Oleari, and E. Re | A general framework for implementing NLO calculations in shower Monte Carlo programs: the POWHEG BOX | JHEP 06 (2010) 043 | 1002.2581 |
| 20 | K. Agashe, H. Davoudiasl, G. Perez, and A. Soni | Warped Gravitons at the LHC and Beyond | Phys.Rev. D 76 (2007) 036006 | hep-ph/0701186 |
| 21 | A. L. Fitzpatrick, J. Kaplan, L. Randall, and L.-T. Wang | Searching for the Kaluza-Klein Graviton in Bulk RS Models | JHEP 09 (2007) 013 | hep-ph/0701150 |
| 22 | O. Antipin, D. Atwood, and A. Soni | Search for RS gravitons via W(L)W(L) decays | Phys.Lett. B 666 (2008) 155 | 0711.3175 |
| 23 | NNPDF Collaboration | Parton distributions for the LHC Run II | JHEP 04 (2015) 040 | 1410.8849 |
| 24 | P. Skands, S. Carrazza, and J. Rojo | Tuning PYTHIA 8.1: the Monash 2013 Tune | EPJC 74 (2014) 3024 | 1404.5630 |
| 25 | CMS Collaboration | Event generator tunes obtained from underlying event and multiparton scattering measurements | EPJC 76 (Dec, 2015) 155. 68 p | CMS-GEN-14-001 |
| 26 | GEANT4 Collaboration | GEANT4---a simulation toolkit | NIMA 506 (2003) 250 | |
| 27 | M. Cacciari, G. P. Salam, and G. Soyez | The anti-$ k_t $ jet clustering algorithm | JHEP 04 (2008) 063 | 0802.1189 |
| 28 | M. Cacciari, G. P. Salam, and G. Soyez | FastJet user manual | EPJC 72 (2012) 1896 | 1111.6097 |
| 29 | CMS Collaboration | Identification techniques for highly boosted W bosons that decay into hadrons | JHEP 12 (2014) 017 | CMS-JME-13-006 1410.4227 |
| 30 | CMS Collaboration | Machine learning-based identification of highly Lorentz-boosted hadronically decaying particles at the CMS experiment | ||
| 31 | A. J. Larkoski, S. Marzani, G. Soyez, and J. Thaler | Soft Drop | JHEP 05 (2014) 146 | 1402.2657 |
| 32 | M. Dasgupta, A. Fregoso, S. Marzani, and G. P. Salam | Towards an understanding of jet substructure | JHEP 09 (2013) 029 | 1307.0007 |
| 33 | Y. L. Dokshitzer, G. D. Leder, S. Moretti, and B. R. Webber | Better jet clustering algorithms | JHEP 08 (1997) 001 | hep-ph/9707323 |
| 34 | J. Thaler and K. Van Tilburg | Maximizing Boosted Top Identification by Minimizing N-subjettiness | JHEP 02 (2012) 093 | 1108.2701 |
| 35 | CMS Collaboration | Performance of missing transverse momentum reconstruction in proton-proton collisions at $ \sqrt{s} = $ 13 TeV using the CMS detector | JINST 14 (2019), no. arXiv:1903.06078. CMS-JME-17-001-003. 07, P07004. | |
| 36 | CMS Collaboration | Performance of the CMS muon detector and muon reconstruction with proton-proton collisions at $ \sqrt{s}= $ 13 TeV | JINST 13 (2018) P06015 | CMS-MUO-16-001 1804.04528 |
| 37 | CMS Collaboration | Performance of CMS muon reconstruction in $ pp $ collision events at $ \sqrt{s}= $ 7 TeV | JINST 7 (2012) P10002 | CMS-MUO-10-004 1206.4071 |
|
|
Compact Muon Solenoid LHC, CERN |
|
|
|
|
|
|