Inclusive search for new physics at CMS with the jets and missing momentum signature

Abstract

A search for new physics is presented based on an event signature of at least three jets accompanied by large missing transverse momentum using a data sample of 36/pb of proton-proton collisions at √s=7TeV acquired by the CMS detector at the LHC in 2010. There are three main background components. One is irreducible, from Z+jets events, with the Z decaying to neutrinos. A second comes from W+jets and top events where W decay yields a lepton that is not identified or a tau that decays hadronically. The third arises from QCD multi-jets where the missing momentum originates from jet mis-measurements, heavy-flavour decays, or instrumental effects. All these backgrounds are estimated from the data. No excess of events over backgrounds is observed. Exclusion limits are presented for the Constrained Minimal Supersymmetric extension of the Standard Model (CMSSM). Cross section limits are also presented using simplified models with generic particles decaying to one or two jets and an undetected particle.

Plots from SUS-10-005 PAPER ( click on plot to get .pdf )

Figure	Abbreviated Caption
	CL upper limits on the gluino pair-production cross sections for the high-MHT selection, as a function of the gluino mass and the LSP mass. The contours where the reference cross section and three times this cross section can be excluded are shown.
	CL upper limits on the squark pair-production cross sections for the high-MHT selection, as a function of the squark mass and the LSP mass. The contours where the reference cross section and three times this cross section can be excluded are shown.
	Contributions to the total resolution distribution of simulated jets, categorised into populations as stated in the legend.
	HT distribution data-MC comparison.The data HT distribution is saved in this file
	MHT distribution data-MC comparison. The data MHT distribution is saved in the this file
	Resolution distribution of simulated jets for various values of expected momentum loss due to masked ECAL channels, as stated in the legend.
	The expected and observed 95% CL upper limits in the CMSSM gluino-squark mass planes for LO and NLO cross sections. The \pm 1 standard deviation (sigma) band corresponds to the expected limit. The contours are the combination of the HT and MHT selections such that the contours are the enveloppe with respect to the best sensitivity. CMSSM param are tb10, mu>0, A0=0.
	The expected and observed 95% CL upper limits in the CMSSM m0-m1/2 mass planes for LO and NLO cross sections. The \pm 1 standard deviation (sigma) band corresponds to the expected limit. The contours are the combination of the HT and MHT selections such that the contours are the enveloppe with respect to the best sensitivity. CMSSM param are tb10, mu>0, A0=0.
	The HT distribution from the R&S method applied to simulation events, compared to the actual MC distribution (MC truth).
	The MHT distribution from the R&S method applied to simulation events, compared to the actual MC distribution (MC truth).
	Total high-MHT selection efficiency for gluino production as a function of the gluino mass and the LSP mass.
	Total high-MHT selection efficiency for squark production as a function of the squark mass and the LSP mass.
	Total signal efficiency for the HT selection as a function of m0 and m1/2. Other param are tb10, mu>0, A0=0.
	Total signal efficiency for the MHT selection as a function of m0 and m1/2. Other param are tb10, mu>0, A0=0.

Plots from SUS-10-005 PAS (not appearing in the paper) ( click on plot to get .pdf )

Figure	Abbreviated Caption
	Closure Test of LostLepton Method in MHT
	Closure Test of LostLepton Method in HT
	Spectra of R+S prediction compared to fully simulated MC distributions, for events passing cuts as indicated in the legend.
	Ratio of R+S predicted over fully simulated MC distributions, for events passing cuts as indicated in the legend. The grey shaded area indicate where all denominators vanish.

	Di-jet asymmetry for adjusted Monte Carlo and data. The area starting from the red line defines the window which is kept fixed during the extrapolation down to prelT,3 = 0. For comparison, the green shaded area show the fraction of events in the asymmetry window under the hypothesis of a purely Gaussian shape
	Extrapolation of non fraction of events with high diJet asymmetry down to prelT,3 = 0. For comparison, green markers show the fraction of events in the asymmetry window under the hypothesis of a purely Gaussian shape
	Not in PAS: Tail scaling factors as function of pT for ΙetaΙ<1.1
	Not in PAS: Tail scaling factors as function of pT for 1.1<ΙetaΙ<1.7
	Distribution of DeltaPhiMin in slices of HT for simulated QCD events. Gaussian fits with a fixed mean=0 are applied
	Sigma of the Gaussian fits of MHT slices from simulation and compared to data. The deviation at high MHT from the exponential dependence for the data is due to contamination from SM background with real MHT.
	Closure test of the factorization method for the baseline selection of the ratio r for a MADGRAPH QCD sample
	Factorization method: Background prediction for different simulations. Bins 1–2 show PYTHIA and MADGRAPH QCD simulation, bins 3–7 show pseudo-simulated QCD samples (4–5: scaling the jet response tail, 6: additional jet resolution smearing of 20%, 7: increased low jet energy response tail).

Predicted event yields from the different background estimation methods for the baseline selection and for the high-MHT and high-HT search selections. In the limit calculations the background estimates used are the R+S method for QCD, Z→νν from photons, and the W/ttbar lost-lepton and hadronic tau predictions.

	Signal Acceptance for the HT selection in the m0 - m1/2 plane for CMSSM for tan(beta)=10.
	Signal Acceptance for the MHT selection in the m0 - m1/2 plane for CMSSM for tan(beta)=10.
	Exclusion contour in the m0 - m1/2 plane for CMSSM for tan(beta)=10. The shown exclusion contour is the best envelope of the obtained contours using the MHT and HT search selection. For low m0 ~< 500 GeV the MHT selection is more sensitive than the HT selection. The background estimation is obtained by data-driven methods and therefore subject to possible signal-contamination. The possible signal contamination at a given point in the parameter space, which varies with the CMSSM cross-section and acceptance, is subtracted from the background event yield. The lines of the observed and expected limits can cross each other, because (i) of the two different selections combined in this plot, and because (ii) the signal contamination and therefore the corrected backgound estimation varies slightly with m0 and m1/2.
	Exclusion contour in the m0 - m1/2 plane for CMSSM for tan(beta)=10. Two independent statistical methods (CLs and Bayes) were used to calculated the two NLO limits compared in this plot. See above for a description of the limit contours.
	Exclusion contour in the gluino mass - squark mass plane for CMSSM for tan(beta)=10. See above for a description of the limit contours.
	Exclusion contour in the m0 - m1/2 plane for CMSSM for tan(beta)=3. See above for a description of the limit contours.
	Exclusion contour in the m0 - m1/2 plane for CMSSM for tan(beta)=50. See above for a description of the limit contours.
	Exclusion contour in the m0 - m1/2 plane for CMSSM for the three different values of tan(beta)=3, 10, 50. The NLO expected and observed limits are shown. This plot shows, that there is basically no dependence of the limits on the value of tan(beta). See above for a description of the limit contours.

topology	Acceptance	Relative Experimental Uncertainty	Relative Theoretical uncertainty	95% exclusion limit cross section x BR	Abbreviated Caption
		jet energy scale and resolution, MCstatistics	initial and final state radiation, PDF	this take into account the Experimental and Theoretical Uncertainties and an additional absolute uncertainty of 2% to taken into account for trigger and cleaning efficiency and 4% for the luminosity; the reference cross section is drawn in center of the bin excluded
					Signal acceptance, experimental and theoretical uncertainties, and exclusion limit cross section for high MHT search region and gluino pair production.
					Signal acceptance, experimental and theoretical uncertainties, and exclusion limit cross section for high HT search region and gluino pair production.
					Signal acceptance, experimental and theoretical uncertainties, and exclusion limit cross section for high MHT search region and squark pair production.
					Signal acceptance, experimental and theoretical uncertainties, and exclusion limit cross section for high HT search region and squark pair production.

Event 70626194, in luminosity section 49 of run 148953. r-φ view of the highest-/HT event passing the search selections.