Mitigation of anomalous missing transverse energy measurements in data collected by CMS at sqrt(s) = 13 TeV during the LHC Run 2 (DP-2020/018)

Link to DPS note on CDS

Introduction

Proton-proton collision events with a signature of high missing transverse momentum are generally interpreted as containing high-momentum neutrinos, or possibly other invisible particles as expected in a variety of scenarios beyond the standard model, e.g. dark matter or Supersymmetry. Experimentally, detector noise or other issues can also lead to the measurement of anomalously high missing transverse momentum, thus mimicking an interesting physics signature. Various algorithms (filters) have been developed in CMS to identify and suppress such spurious events. In this note the performance of these filters is presented in events that contain at least one jet with high transverse momentum ($p_{T}$). The events were recorded by the CMS experiment during the LHC Run2.

Glossary

•PF $  p_{T}^{miss} $ : defined as the magnitude of the missing transverse momentum ($  \vec{p}_{T}^{miss} $) which is the projection onto the plane perpendicular to the beam axis of the negative vector sum of the transverse momentum ($ p_{T}$) of all the candidates reconstructed by particle-flow (PF) algorithm in an event [1].

•PUPPI $  p_{T}^{miss} $ : advanced reconstruction algorithm involving pileup mitigation based on the “pileup per particle identification” (PUPPI) technique [1][2].

•Anomalous high-$p_{T}^{miss}$ filters: a set of algorithms developed using information from various subdetectors such as electromagnetic calorimeter (ECAL), hadronic calorimeter (HCAL) and muon systems of CMS [3] to reject the events with spurious signals and machine-induced backgrounds. In addition, dedicated algorithms targeting mismeasured or misrecontructed events have been employed [1].

•Jets: the jets reconstructed with the anti-$ k_{T}$ algorithm with a distance parameter of 0.4.

•CMS.JetID: criteria used for the identification of jets based on the jet energy fractions and the number of jet constituents [1].

•Data sample: events passing a set of triggers requiring the presence of high $p_{T}$ jets and/or large hadronic activity in an event.

Event display overview

Below are the details about physics objects shown in the following event displays

•High quality particle flow (PF) tracks with $p_{T} > $ 2.0 GeV are shown as green lines.

•Yellow lines represent the reconstructed jets with $p_{T} > $ 30 GeV.

•Red and blue bars/towers represent the energy deposited by PF candidates in ECAL and HCAL respectively.

•The muons reconstructed using either global muon or tracker muon reconstruction algorithms with $p_{T} > $ 5 GeV are shown by red lines.

•The reconstructed vertices with number of degrees of freedom greater than 4 are shown by yellow marker.

•The violet color arrow shows the magnitude and direction of PF $  \vec{p}_{T}^{miss} $.

Figures

An event rejected by the dedicated HCAL noise filter based on the pulse shape information and the hit multiplicity in the hybrid photodiode (HPD) and the readout box. This event is an example of HPD discharge noise producing high energy hits in several HPD channels at phi=0.66 leading to unphysical missing transverse momentum.

An event rejected by the dedicated HCAL noise filter based on the pulse shape information and the hit multiplicity in the hybrid photodiode (HPD) and the readout box. This event is an example of HPD discharge noise producing high energy hits in several HPD channels at phi=0.66 leading to unphysical missing transverse momentum.

During Run 2, four supercrystals (groups of 5x5 crystals) in the ECAL endcap were found to occasionally produce anomalous high energy signals leading to fake reconstructed missing transverse momentum. This event has several bad quality hits with energy greater than 1 TeV in an affected ECAL supercrystal, and is rejected by the dedicated ECAL noise filter.

During Run 2, four supercrystals (groups of 5x5 crystals) in the ECAL endcap were found to occasionally produce anomalous high energy signals leading to fake reconstructed missing transverse momentum. This event has several bad quality hits with energy greater than 1 TeV in an affected ECAL supercrystal, and is rejected by the dedicated ECAL noise filter.

An event rejected by the filter identifying muons mistakenly reconstructed with high $p_{T}$. The filter decision is based on the presence of a high $p_{T}$ muon with a poor quality track and a large momentum measurement uncertainty.

An event rejected by the filter identifying muons mistakenly reconstructed with high $p_{T}$. The filter decision is based on the presence of a high $p_{T}$ muon with a poor quality track and a large momentum measurement uncertainty.

A monojet candidate event passing all the recommended filters. The missing transverse momentum is balanced by a single jet. No other jet with $p_{T} > $ 50 GeV is found in the event.

A monojet candidate event passing all the recommended filters. The missing transverse momentum is balanced by a single jet. No other jet with $p_{T} > $ 50 GeV is found in the event.

Rho-phi view of a monojet candidate event passing all the recommended filters. The missing transverse momentum is balanced by a single jet. No other jet with $p_{T} > $ 50 GeV is found in the event.

Rho-phi view of a monojet candidate event passing all the recommended filters. The missing transverse momentum is balanced by a single jet. No other jet with $p_{T} > $ 50 GeV is found in the event.

PF $ p_{T}^{miss} $ distribution in events with at least one jet with $ p_{T} > $ 200 GeV for the 2018 collision data before (black dots) and after (blue line) the various $ p_{T}^{miss}$ filters are applied. The red markers correspond to events where the highest $ p_{T}$ jet additionally satisfies the jetID criteria. The last bin includes the overflow bin. The bottom pads present the fraction of events not rejected by the anomalous high-$p_{T}^{miss}$ filters (blue) and that for events where the highest $ p_{T} $ jet additionally satisfies the jetID criteria (red). The fraction of events passing the filters is close to unity in the low $p_{T}^{miss}$ region and a greater fraction of events are rejected in the tails of the distribution where the effects of the detector or reconstruction issues that these filters are designed to tackle are expected to be more pronounced.

PUPPI $ p_{T}^{miss} $ distribution in events with at least one jet with $ p_{T} > $ 200 GeV for the 2018 collision data before (black dots) and after (blue line) the various $ p_{T}^{miss}$ filters are applied. The red markers correspond to events where the highest $ p_{T}$ jet additionally satisfies the jetID criteria. The last bin includes the overflow bin. The bottom pads present the fraction of events not rejected by the anomalous high-$p_{T}^{miss}$ filters (blue) and that for events where the highest $ p_{T} $ jet additionally satisfies the jetID criteria (red). The fraction of events passing the filters is close to unity in the low $p_{T}^{miss}$ region and a greater fraction of events are rejected in the tails of the distribution where the effects of the detector or reconstruction issues that these filters are designed to tackle are expected to be more pronounced.

PF $ p_{T}^{miss} $ distribution in events with at least one jet with $ p_{T} > $ 200 GeV for the 2017 collision data before (black dots) and after (blue line) the various $ p_{T}^{miss}$ filters are applied. The red markers correspond to events where the highest $ p_{T}$ jet additionally satisfies the jetID criteria. The last bin includes the overflow bin. The bottom pads present the fraction of events not rejected by the anomalous high-$p_{T}^{miss}$ filters (blue) and that for events where the highest $ p_{T} $ jet additionally satisfies the jetID criteria (red). The fraction of events passing the filters is close to unity in the low $p_{T}^{miss}$ region and a greater fraction of events are rejected in the tails of the distribution where the effects of the detector or reconstruction issues that these filters are designed to tackle are expected to be more pronounced.

PUPPI $ p_{T}^{miss} $ distribution in events with at least one jet with $ p_{T} > $ 200 GeV for the 2017 collision data before (black dots) and after (blue line) the various $ p_{T}^{miss}$ filters are applied. The red markers correspond to events where the highest $ p_{T}$ jet additionally satisfies the jetID criteria. The last bin includes the overflow bin. The bottom pads present the fraction of events not rejected by the anomalous high-$p_{T}^{miss}$ filters (blue) and that for events where the highest $ p_{T} $ jet additionally satisfies the jetID criteria (red). The fraction of events passing the filters is close to unity in the low $p_{T}^{miss}$ region and a greater fraction of events are rejected in the tails of the distribution where the effects of the detector or reconstruction issues that these filters are designed to tackle are expected to be more pronounced.

PF $ p_{T}^{miss} $ distribution in events with at least one jet with $ p_{T} > $ 200 GeV for the 2016 collision data before (black dots) and after (blue line) the various $ p_{T}^{miss}$ filters are applied. The red markers correspond to events where the highest $ p_{T}$ jet additionally satisfies the jetID criteria. The last bin includes the overflow bin. The bottom pads present the fraction of events not rejected by the anomalous high-$p_{T}^{miss}$ filters (blue) and that for events where the highest $ p_{T} $ jet additionally satisfies the jetID criteria (red). The fraction of events passing the filters is close to unity in the low $p_{T}^{miss}$ region and a greater fraction of events are rejected in the tails of the distribution where the effects of the detector or reconstruction issues that these filters are designed to tackle are expected to be more pronounced.

PUPPI $ p_{T}^{miss} $ distribution in events with at least one jet with $ p_{T} > $ 200 GeV for the 2016 collision data before (black dots) and after (blue line) the various $ p_{T}^{miss}$ filters are applied. The red markers correspond to events where the highest $ p_{T}$ jet additionally satisfies the jetID criteria. The last bin includes the overflow bin. The bottom pads present the fraction of events not rejected by the anomalous high-$p_{T}^{miss}$ filters (blue) and that for events where the highest $ p_{T} $ jet additionally satisfies the jetID criteria (red). The fraction of events passing the filters is close to unity in the low $p_{T}^{miss}$ region and a greater fraction of events are rejected in the tails of the distribution where the effects of the detector or reconstruction issues that these filters are designed to tackle are expected to be more pronounced.

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng Ecal_v1.png r1 manage 443.9 K 2020-04-01 - 13:13 AmandeepKaurKalsi  
PNGpng Ecal_v2.png r1 manage 446.0 K 2020-04-01 - 13:13 AmandeepKaurKalsi  
PNGpng Hcal_v1.png r1 manage 464.0 K 2020-04-01 - 13:13 AmandeepKaurKalsi  
PNGpng Hcal_v2.png r1 manage 470.8 K 2020-04-01 - 13:13 AmandeepKaurKalsi  
PNGpng Monojet_v1_1.png r1 manage 157.1 K 2020-04-01 - 13:38 AmandeepKaurKalsi  
PNGpng Monojet_v1_2.png r1 manage 281.7 K 2020-04-01 - 13:38 AmandeepKaurKalsi  
PNGpng Monojet_v2_1.png r1 manage 252.1 K 2020-04-01 - 13:38 AmandeepKaurKalsi  
PNGpng Monojet_v2_2.png r1 manage 273.4 K 2020-04-01 - 13:38 AmandeepKaurKalsi  
PNGpng Muon_v1.png r1 manage 333.8 K 2020-04-01 - 13:13 AmandeepKaurKalsi  
PNGpng Muon_v2.png r1 manage 322.4 K 2020-04-01 - 13:13 AmandeepKaurKalsi  
PDFpdf PF_MET_2016.pdf r1 manage 42.1 K 2020-04-01 - 13:36 AmandeepKaurKalsi  
PNGpng PF_MET_2016.png r2 r1 manage 86.2 K 2020-04-01 - 15:46 AmandeepKaurKalsi  
PDFpdf PF_MET_2017.pdf r1 manage 42.1 K 2020-04-01 - 13:36 AmandeepKaurKalsi  
PNGpng PF_MET_2017.png r1 manage 105.2 K 2020-04-01 - 14:42 AmandeepKaurKalsi  
PDFpdf PF_MET_2018.pdf r1 manage 42.0 K 2020-04-01 - 13:36 AmandeepKaurKalsi  
PNGpng PF_MET_2018.png r1 manage 105.4 K 2020-04-01 - 14:42 AmandeepKaurKalsi  
PDFpdf Puppi_MET_2016.pdf r1 manage 42.0 K 2020-04-01 - 13:36 AmandeepKaurKalsi  
PNGpng Puppi_MET_2016.png r2 r1 manage 86.9 K 2020-04-01 - 15:46 AmandeepKaurKalsi  
PDFpdf Puppi_MET_2017.pdf r1 manage 42.1 K 2020-04-01 - 13:36 AmandeepKaurKalsi  
PNGpng Puppi_MET_2017.png r1 manage 105.3 K 2020-04-01 - 14:42 AmandeepKaurKalsi  
PDFpdf Puppi_MET_2018.pdf r1 manage 42.0 K 2020-04-01 - 13:36 AmandeepKaurKalsi  
PNGpng Puppi_MET_2018.png r1 manage 105.5 K 2020-04-01 - 14:42 AmandeepKaurKalsi  
Edit | Attach | Watch | Print version | History: r8 < r7 < r6 < r5 < r4 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r8 - 2020-04-24 - AndreasHinzmann
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    CMSPublic All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright &© 2008-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback