Search for New Physics in Multijets and Missing Momentum Final State in Proton-Proton Collisions at sqrt{s} = 8 TeV (SUS-13-012)

Further information

This analysis is documented in SUS-13-012.
The Table and Figure numbers in this twiki correspond to the table and figure numbers in the paper. Additional approved figures which are not in the paper are attached below.

Abstract

A search for new physics is performed in multijet events with large missing transverse momentum produced in proton-proton collisions at sqrt{s} = 8 TeV using a data sample corresponding to an integrated luminosity of 19.5 fb^{-1} collected with the CMS detector at the LHC. The data sample is divided into three jet multiplicity categories (3–5, 6–7, and 8 jets), and studied further in bins of two variables: the scalar sum of jet transverse momenta and the missing transverse momentum. The observed numbers of events in various categories are consistent with backgrounds expected from standard model processes. Exclusion limits are presented for several simplified supersymmetric models of squark or gluino pair production.

Approved tables and plots ( click on plot to get larger version )

Figure	Caption
	Figure 1a : The simulated ratio R_Z/g as a function of (a) MHT, (b) HT, (c) NJets, where the values for three MHT bins are shown with linear fits, and (d) the double ratio of R:Z(mu+mu-)/g, using events from data to those from simulation; the linear fit and its uncertainty band are overlaid.
	Figure 1b : The simulated ratio R_Z/g as a function of (a) MHT, (b) HT, (c) NJets, where the values for three MHT bins are shown with linear fits, and (d) the double ratio of R:Z(mu+mu-)/g, using events from data to those from simulation; the linear fit and its uncertainty band are overlaid.
	Figure 1c : The simulated ratio R_Z/g as a function of (a) MHT, (b) HT, (c) NJets, where the values for three MHT bins are shown with linear fits, and (d) the double ratio of R:Z(mu+mu-)/g, using events from data to those from simulation; the linear fit and its uncertainty band are overlaid.
	Figure 1d : The simulated ratio R_Z/g as a function of (a) MHT, (b) HT, (c) NJets, where the values for three MHT bins are shown with linear fits, and (d) the double ratio of R:Z(mu+mu-)/g, using events from data to those from simulation; the linear fit and its uncertainty band are overlaid.
	Figure 2a : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the lost lepton background evaluation method to simulated tt and W+jets events (solid points) in comparison to the genuine tt and W+jets background from simulation (shaded curves). Only statistical uncertainties are shown.
	Figure 2b : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the lost lepton background evaluation method to simulated tt and W+jets events (solid points) in comparison to the genuine tt and W+jets background from simulation (shaded curves). Only statistical uncertainties are shown.
	Figure 2c : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the lost lepton background evaluation method to simulated tt and W+jets events (solid points) in comparison to the genuine tt and W+jets background from simulation (shaded curves). Only statistical uncertainties are shown.
	Figure 3a : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the th background evaluation method to simulated tt and W+jets events (solid points) in comparison to the genuine tt and W+jets background from simulation (shaded curve). Only statistical uncertainties are shown.
	Figure 3b : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the th background evaluation method to simulated tt and W+jets events (solid points) in comparison to the genuine tt and W+jets background from simulation (shaded curve). Only statistical uncertainties are shown.
	Figure 3c : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the th background evaluation method to simulated tt and W+jets events (solid points) in comparison to the genuine tt and W+jets background from simulation (shaded curve). Only statistical uncertainties are shown.
	Figure 4a : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the “rebalance-and-smear” method to simulated QCD multijet events (solid points) in comparison with the genuine QCD multijet background from simulation (shaded curve). The distributions are shown for events that satisfy the baseline selection, except that the MHT selection is not applied, and in addition HT > 1000GeV is required for the events used in the H/T distribution. The statistical uncertainties are indicated by the hatched band for the expectation and by error bars for the prediction.
	Figure 4b : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the “rebalance-and-smear” method to simulated QCD multijet events (solid points) in comparison with the genuine QCD multijet background from simulation (shaded curve). The distributions are shown for events that satisfy the baseline selection, except that the MHT selection is not applied, and in addition HT > 1000GeV is required for the events used in the H/T distribution. The statistical uncertainties are indicated by the hatched band for the expectation and by error bars for the prediction.
	Figure 4c : Predicted (a) HT, (b) MHT, and (c) NJets distributions found from applying the “rebalance-and-smear” method to simulated QCD multijet events (solid points) in comparison with the genuine QCD multijet background from simulation (shaded curve). The distributions are shown for events that satisfy the baseline selection, except that the MHT selection is not applied, and in addition HT > 1000GeV is required for the events used in the H/T distribution. The statistical uncertainties are indicated by the hatched band for the expectation and by error bars for the prediction.
	Table 1: Predicted event yields for the different background components in the search regions defined by HT,MHT and NJets. The uncertainties of the different background sources are added in quadrature to obtain the total uncertainties.
	Figure 5: Summary of the observed number of events in each of the 36 search regions in comparison to the corresponding background prediction. The hatched region shows the total uncertainty of the background prediction.
	Figure 6a: Observed MHT distributions compared to the predicted backgrounds for search regions with HT > 500 GeV and jet multiplicity intervals of (a) 3–5, (b) 6–7, and (c) 8. The background distributions are stacked. The last bin contains the overflow. The hatched region indicates the uncertainties of the background predictions. The ratio of data to the background is shown in the lower plots. The MHT distributions expected from events with gluino and squark pair production, with either m(squark) = 700 GeV and m(LSP) = 125 GeV or m(gluino) = 1.1 TeV and m(LSP) = 100 GeV, are overlaid.
	Figure 6b: Observed MHT distributions compared to the predicted backgrounds for search regions with HT > 500 GeV and jet multiplicity intervals of (a) 3–5, (b) 6–7, and (c) 8. The background distributions are stacked. The last bin contains the overflow. The hatched region indicates the uncertainties of the background predictions. The ratio of data to the background is shown in the lower plots. The MHT distributions expected from events with gluino and squark pair production, with either m(squark) = 700 GeV and m(LSP) = 125 GeV or m(gluino) = 1.1 TeV and m(LSP) = 100 GeV, are overlaid.
	Figure 6c: Observed MHT distributions compared to the predicted backgrounds for search regions with HT > 500 GeV and jet multiplicity intervals of (a) 3–5, (b) 6–7, and (c) 8. The background distributions are stacked. The last bin contains the overflow. The hatched region indicates the uncertainties of the background predictions. The ratio of data to the background is shown in the lower plots. The MHT distributions expected from events with gluino and squark pair production, with either m(squark) = 700 GeV and m(LSP) = 125 GeV or m(gluino) = 1.1 TeV and m(LSP) = 100 GeV, are overlaid.
	Figure 7a: The observed and expected 95% CL upper limits on the (a) squark-squark and (b-d) gluino-gluino production cross sections in either the m(squark)-m(LSP) or the m(gluino)-m(LSP) plane obtained with the simplified models. For the squark-squark production the upper set of curves corresponds to the scenario when the first two generations of squarks are degenerate and light, while the lower set corresponds to only one light accessible squark. Figure content as a root file. The acceptance times efficiency maps for each search region are available electronically in this file.
	Figure 7b: The observed and expected 95% CL upper limits on the (a) squark-squark and (b-d) gluino-gluino production cross sections in either the m(squark)-m(LSP) or the m(gluino)-m(LSP) plane obtained with the simplified models. For the squark-squark production the upper set of curves corresponds to the scenario when the first two generations of squarks are degenerate and light, while the lower set corresponds to only one light accessible squark. Figure content as a root file. The acceptance times efficiency maps for each search region are available electronically in this file.
	Figure 7c: The observed and expected 95% CL upper limits on the (a) squark-squark and (b-d) gluino-gluino production cross sections in either the m(squark)-m(LSP) or the m(gluino)-m(LSP) plane obtained with the simplified models. For the squark-squark production the upper set of curves corresponds to the scenario when the first two generations of squarks are degenerate and light, while the lower set corresponds to only one light accessible squark. Figure content as a root file. The acceptance times efficiency maps for each search region are available electronically in this file.
	Figure 7d: The observed and expected 95% CL upper limits on the (a) squark-squark and (b-d) gluino-gluino production cross sections in either the m(squark)-m(LSP) or the m(gluino)-m(LSP) plane obtained with the simplified models. For the squark-squark production the upper set of curves corresponds to the scenario when the first two generations of squarks are degenerate and light, while the lower set corresponds to only one light accessible squark. Figure content as a root file. The acceptance times efficiency maps for each search region are available electronically in this file.

Additional approved plots not in PAS ( click on plot to get larger version )

Figure	Description
	Figure A : Comparison of the HT distribution from simulated events and data after the baseline selection.
	Figure B : Comparison of the MHT distribution from simulated events and data after the baseline selection.
	Figure C : Comparison of the jet multiplicity distribution (pT > 50 GeV, abs(eta) < 2.5) from simulated events and data after the baseline selection.
	Figure D : Comparison of the distribution of muon pT for simulated events and data after the baseline selection with inverted lepton veto (muon control sample). For simulated events the pT of generated prompt leptons is shown.
	Figure E : Comparison of the HT distribution for simulated events and data after the baseline selection with inverted lepton veto (muon control sample).
	Figure F : Comparison of the MHT distribution for simulated events and data after the baseline selection with inverted lepton veto (muon control sample).
	Figure G : Comparison of the jet multiplicity distribution (pT > 50 GeV, abs(eta) < 2.5) for simulated events and data after the baseline selection with inverted lepton veto (muon control sample).
	Figure H : Comparison of the transverse mass distribution for simulated events and data after the baseline selection with inverted lepton veto (muon control sample). The blue line shows the distribution for a signal model (gluino induced stop quark production with four top quarks in the final state, m(gluino) = 1000 GeV & m(LSP) = 400 GeV).
	Figure I : Ratio of the data driven prediction on simulated events and the expectation from simulation for the background from lost leptons in each search bin. The green horizontal lines enclose the assigned uncertainty band to cover any potential bias.
	Figure J : Ratio of the data driven prediction on simulated events and the expectation from simulation for the background with hadronically decaying taus in each search bin. The red line shows the average non-closure in each jet multiplicity region; the prediction is corrected for this bias. The green horizontal lines enclose the assigned uncertainty band.
	Figure K : Transverse momentum response (jet pT / tau pT) templates of hadronically decaying taus in four different bins of transverse momentum.
	Figure L : Comparison and relative difference of the data driven R+S prediction on simulated QCD multijet events and the expectation from simulation for 6>=NJets>=7 and HT > 1000 GeV. The hatched area displays the statistical uncertainty on the distribution from simulation, while the error bars show the estimated statistical uncertainty on the prediction.
	Figure M : Comparison and relative difference of the data driven R+S prediction on simulated QCD multijet events and the expectation from simulation for NJets>=8 and HT > 1000 GeV. The hatched area displays the statistical uncertainty on the distribution from simulation, while the error bars show the estimated statistical uncertainty on the prediction.
	Figure N : Jet transverse momentum response template for one particular pT and eta bin. The contribution from heavy flavor jets is overlaid, showing the non-gaussian tail at low response values.
	Figure O : Comparison of the MHT distributions from simulated events and data with 3 to 5 jets (pT > 50 GeV, abs(eta) < 2.5) for different HT bins.
	Figure P : Comparison of the MHT distributions from simulated events and data with 6 to 7 jets (pT > 50 GeV, abs(eta) < 2.5) for different HT bins.
	Figure Q : Comparison of the MHT distributions from simulated events and data with 8 or more jets (pT > 50 GeV, abs(eta) < 2.5) for different HT bins.

Interpretation of the results within the pMSSM framework

Introduction

We show results of a phenomenological MSSM interpretation of the 8 TeV HT+MHT analysis SUS-13-012.

We follow the approach of the phenomenological MSSM interpretation of 7 TeV CMS results, documented in the approved PAS SUS-12-030: About 7300 points in pMSSM parameter space are sampled from an evidence-based prior probability density, based on theoretical predictions and measurements of flavour observables, Higgs mass, top mass, bottom mass and anomalous magnetic moment of the muon. For each pMSSM point theta we calculate the likelihood L(SUS-13-012|theta).

Results are presented as distributions of pMSSM parameters, masses and other observables, in two ways:

1. A fully Bayesian approach: the prior distribution is compared to the posterior distribution including the SUS-13-012 data. The prior distribution is simply the distribution of the 7300 pMSSM points. The posterior distribution is obtained by weighting each of the 7300 pMSSM points with L(SUS-13-012|theta). Posterior and prior distribution are normalized to one.
2. excluded vs not excluded: we compare the distribution of the 7300 pMSSM points to the distribution of those pMSSM points that are excluded by SUS-13-012 and the distribution of those points not excluded by SUS-13-012. A pMSSM point is considered excluded if Z = sign(ln(B_10))sqrt(2*|ln B_10|) < -1.64, with B_10 = L(SUS-13-012|theta)/L(SUS-13-012|H_0), where L(SUS-13-012|H_0) is the likelihood for the background only hypothesis H_0. Z is a signed analog of the frequentist "n-sigma".

Fully Bayesian approach

Figure	Caption
	Marginalized 1D probability distributions for ~g mass. The filled blue histogram shows the prior density. The line histograms show posterior densities after including the HT + MHT. The solid curve shows the posterior density obtained from likelihoods calculated using the central values of estimated signal counts s, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Marginalized 1D probability distributions for ~uR and ~cR mass. The filled blue histogram shows the prior density. The line histograms show posterior densities after including the HT + MHT analysis. The solid curve shows the posterior density obtained from likelihoods calculated using the central values of estimated signal counts s, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Marginalized 1D probability distributions for the mass of the lightest colored sparticle. The filled blue histogram shows the prior density. The line histograms show posterior densities after including the HT + MHT analysis. The solid curve shows the posterior density obtained from likelihoods calculated using the central values of estimated signal counts s, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Marginalized 1D probability distributions for sparticle production cross section. The filled blue histogram shows the prior density. The line histograms show posterior densities after including the HT + MHT analysis. The solid curve shows the posterior density obtained from likelihoods calculated using the central values of estimated signal counts s, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Marginalized prior probability distribution for ~g mass versus ~χ10 mass. The grey and black contours enclose the 68% and 95% Bayesian credible regions respectively.
	Marginalized posterior probability distribution for ~g mass versus ~χ10 mass after including the HT + MHT analysis. The grey and black contours enclose the 68% and 95% Bayesian credible regions respectively.
	Marginalized prior probability distribution for ~uR mass versus ~χ10 mass. The grey and black contours enclose the 68% and 95% Bayesian credible regions respectively.
	Marginalized posterior probability distribution for ~uR mass versus ~χ10 mass after including the HT + MHT analysis. The grey and black contours enclose the 68% and 95% Bayesian credible regions respectively.

Excluded vs not-excluded

Figure	Caption
	Distributions of ~g mass. The filled blue histogram shows the distribution of pMSSM points samped from the prior density. The red (black) line histograms shows the distribution of pMSSM points not excluded (excluded) by the HT + MHT analysis. Solid curves show the posterior densities obtained from likelihoods calculated using the central values of estimated signal counts $s$, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Distributions of ~uR and ~cR mass. The filled blue histogram shows the distribution of pMSSM points samped from the prior density. The red (black) line histograms shows the distribution of pMSSM points not excluded (excluded) by the HT + MHT analysis. Solid curves show the posterior densities obtained from likelihoods calculated using the central values of estimated signal counts $s$, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Distributions of the mass of the lightest colored sparticle. The filled blue histogram shows the distribution of pMSSM points samped from the prior density. The red (black) line histograms shows the distribution of pMSSM points not excluded (excluded) by the HT + MHT analysis. Solid curves show the posterior densities obtained from likelihoods calculated using the central values of estimated signal counts $s$, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Distributions of the sparticle cross section. The filled blue histogram shows the posterior densities after preCMS measurements. The filled blue histogram shows the distribution of pMSSM points samped from the prior density. The red (black) line histograms shows the distribution of pMSSM points not excluded (excluded) by the HT + MHT analysis. Solid curves show the posterior densities obtained from likelihoods calculated using the central values of estimated signal counts $s$, whereas the dashed and dotted lines show the posterior densities obtained from likelihoods calculated using s-0.5s and s+0.5s respectively.
	Distribution of ~g mass versus ~χ10 mass for the sampled pMSSM points excluded by the HT + MHT analysis. The grey and black contours enclose the 68% and 95% of the excluded points.
	Distribution of ~g mass versus ~χ10 mass for the sampled pMSSM points non excluded by the HT + MHT analysis. The grey and black contours enclose 68% and 95% of the non-excluded points.
	Distribution of ~uR mass versus ~χ10 mass for the sampled pMSSM points excluded by the HT + MHT analysis. The grey and black contours enclose the 68% and 95% of the excluded points.
	Distribution of ~uR mass versus ~χ10 mass for the sampled pMSSM points not excluded by the HT + MHT analysis. The grey and black contours enclose the 68% and 95% of the non-excluded points.