Higgs_Public

Search for the SM Higgs Boson in the MET+b-jets Final State
with 0.973/fb Run II data at CDF

Abstract

We present a search for the Standard Model Higgs boson produced in association with a Z or a W boson in pp collisions at sqrt(s)= 1.96 TeV. We focus our search on the H->bb, and Z->vv or W->lv (where the charged lepton escapes detection) decay modes. The selected events shall have exactly two jets in the final state, one or two b-tagged jet(s), and large missing transverse momentum due to the undetected leptons. The exclusive single and double tag events are treated separately throughout the analysis. We processed 0.973/fb data collected in CDF Run II. The observed number of events is consistent with the Standard Model background prediction within the errors. 95% C.L. upper limits are set on the SM Higgs production cross-section in the ZH, WH, and combined channels at various Higgs masses. The observed cross-section limits divided by the Standard Model cross-section range from 17.8 to 22.8 between 110 GeV and 130 GeV of Higgs masses. The observed limits are also consistent with the Monte Carlo expectations.

Authors: V. Veszpremi, O. Gonzalez, D. Bortoletto, A. Garfinkel, Song Ming Wang
Blessed: July 20, 2006
Godparents: none
Data: Run II, 0.973/fb
Documentation: CDF8362
Internal Web Page
Public: Note CDF8442 (or PS)

Introduction

According to the latest results from LEP2, the lower bound on the mass of the Higgs is 114.4 GeV at a 95 % C.L. [Phys.Lett. B565 (2003) 61-75] The previous results of the CDF collaboration from Run I has been published in [FERMILAB-PUB-05-042-E]. In the CDF Run I analysis, the luminosity of the data was not enough to observe any trace of the Higgs beyond the region excluded by LEP. With the projected total luminosity of Run II, we expect to have a sensitivity for observing or excluding the Higgs near and over the current lower mass limit [FERMILAB-PUB-03/320-E]. In this analysis, updated in 2006, we target the lower mass region using 0.973/fb of Run II data.

In the final state, exactly two b-jets, the decay produts of the Higss, are expected with transverse energy over 20 GeV (tagging threshold) and no identified leptons. Since the two leptons escape the detector without being observed, a large amount of transverse momentum appears missing from the events. The diagram of the ZH process is shown below. When the Higgs is created along with a W instead of a Z, one of the two leptons is charged. We select such WH events when the charge lepton is not identified by a loose electron or isolated track id. The selection criteria for these leptons are loose enough to assure mutual exclusiveness with the dedicated WH search.

Control Region and Extended Signal Region definitions:

For all the regions:

1. Quality cuts:

    - Consider runs from the data in which all the calorimeters and the tracking system were functional

    - A set of quality cuts are applied to partially filter events with large fake missing transverse energy

2. Basic Selection cuts:

- Select events with two jets (one is central):
    - Leading jet Et (corrected) > 35 GeV
    - Second most energetic jet Et (corrected) >20 GeV

- No other jets with Et > 20 GeV (Njets=2)

- Missing Et > 55 GeV (Missing Et corrected for jets)

- At least 1 b-tagged jet is required / Jet Et(corrected) > 20 GeV /

The sample is split into two by the number of tags in order to take advantage of the better S/B ratio in the double tag sample. The limits are calculated separately in the two mutually exclusive sets of single and double tag events and combined with each other at the end. Although, the final results are driven by the double tag sample, the single tag limits also improve the final limits despite the weaker signal separation.

Note: The requirements on the Missing Et and the Leading Jet Et are driven by the trigger we used to collect the data.

Simulating the heavy flavor background:

Heavy flavor background (and signal) events passing the quality and basic selection cuts are simulated in Pythia. The simulated background processes are: QCD dijet (bb, cc), W+h.f., Z+h.f., Top (including single top), and Diboson production. All these processes are normalized to luminosity generally using the theoretical NLO cross-sections with the exception of the QCD h.f. production. QCD events are normalized in such a way that the total amount of estimated background is equal to the number of data events passing the quality and basic selection cuts.

Estimating light flavor contribution:

Only the heavy flavor events (containing tagged b-, c- or in some cases tau-jets) are simulated, the mistagged light flavor background is estimated from the data using the so called mistag rate matrix. This matrix is used to estimate the rate at which a generic data event is considered to contain one or two heavy flavor jet(s) based on the Et, pseudo-rapidity, and azimuth of the jet(s), and the total Et and total number of tracks in the event.

3. Definition of Extended Signal Region and the Control Regions:

Control Region 1 (QCD):
    - All events with an identified lepton (by a set of loose electron or isolated track id-cuts) are rejected.
    - Azimuthal angular separation between 2nd leading jet and missing Et is less than 23 degrees (0.4 rad)
    - This control region is dominated by heavy flavor QCD dijet events. These events are simulated by b/c-filtered QCD Pythia Monte Carlo. It also contains mistagged light flavor jets, which are derived from the data using the mistag matrix. This control region has been defined so that the normalization (and in a limited way the shape) of the QCD simulation can be confirmed. (See plots below.)

Control Region 2 (EWK):
    - At least 1 isolated high-pt lepton is reqired.
    - Azimuthal angular separation between 2nd leading jet and missing Et is greater than 23 degrees (0.4 rad)
    - The single tag part of this control region contains all the physical backgrounds (EWK, Top, QCD) in approximately equal amount, however, the uncertainty of the lepton id in the QCD events represent a large systematic error. The error assigned to the lepton id is 2% wrt events passing the basic selection, this translates to a ~50% error in the number of QCD events faking an isolated lepton.
       The double tag set contains less QCD h.f. It is dominated by top, W+h.f., and mistagged light flavor; therefore, it provides a better control on the top and electroweak background components.

Extended Signal Region (SIG):
    - Events with isolated leptons are vetoed
    - Azimuthal angular separation between 2nd leading jet and missing Et is greater than 23 degrees (0.4 rad)
    - After a succesful MC simulation in the two other control regions, a cut optimization is performed in this region. The optimization is done on the MC and the Mistag samples and aims to maximize the signal to square-root background ratio.

Optimization cuts:
    - Azimuthal angle between leading jet and missing transverse energy, Phi(1st jet, MET) > 0.8
    - Leading jet Et > 60 GeV
    - Missing Et > 75 GeV

Final limits:
The table below shows the expected limits in the ZH and WH, and the expected and observed limits in the combined channels at various Higgs masses.

Higgs mass (GeV)	ZH Exp.	WH Exp.	Combined Exp.	Combined Obs.
110	25.5	34.8	14.9	17.8
115	28.4	34.0	15.4	16.0
120	31.7	35.1	16.8	15.6
125	40.5	39.2	20.0	21.4
130	50.0	41.5	22.6	22.8

The limit calculation was done on the invariant dijet mass in the single and double tag samples. (See plots below.) We used Tom Junk's limit calculator. The code and more details how to use it can be found here.

Blessed Plots:

	Leading order diagram of ZH production. BMP
	Production cross section of the Standard Model Higgs boson BMP
	Branching ratios of the Higgs boson decay BMP
	Dijet invariant mass in the single tag signal region after applying the optimized selection cuts GIF EPS PNG
	Dijet invariant mass in the double tag signal region after applying the optimized selection cuts GIF EPS PNG
	95% C.L. exclusion limits in the ZH, WH, and combined (MET+b-jets) channels wrt Standard Model cross-section GIF EPS PNG
	Single Tagged Candidate Event Jet Et1 = 84.7 GeV Jet Et2 = 71.9 GeV - TAGGED Missing Et = 98.0 GeV Dijet mass = 128.6 GeV GIF EPS
	same event... GIF EPS
	Double Tagged Candidate Event Jet Et1 = 100.3 GeV - TAGGED Jet Et2 = 54.7 GeV - TAGGED Missing Et = 144.8 GeV Dijet mass = 82.1 GeV GIF EPS
	same event... GIF EPS

Further plots:

Control Region 1 (QCD), Single Tag
- dominated by QCD heavy flavor production

Leading jet Et in log scale	eps gif png eps gif png
Second leading jet Et in log scale	eps gif png eps gif png
Dijet mass in log scale	eps gif png eps gif png
Missing transverse energy log scale	eps gif png eps gif png
Azimuthal angle between jets log scale	eps gif png eps gif png
Azimuthal angle between leading jet and missing transverse energy log scale	eps gif png eps gif png

Control Region 1 (QCD), Double Tag
- also dominated by QCD

Dijet mass in log scale	eps gif png eps gif png
Missing transverse energy in log scale	eps gif png eps gif png

Control Region 2 (EWK), Single Tag
- mix of all background sources

Dijet mass	eps gif png
Missing transverse energy	eps gif png

Control Region 2 (EWK), Double Tag
- dominated by Top and W+ h.f.

Leading jet Et	eps gif png
Second leading jet Et	eps gif png
Dijet mass	eps gif png
Missing transverse energy	eps gif png
Azimuthal angle between jets	eps gif png
Azimuthal angle between leading jet and missing transverse energy	eps gif png
(Missing Ht)/Ht	eps gif png

Signal Region (SIG, before optimization), Single Tag

Leading jet Et in log scale	eps gif png eps gif png
Second leading jet Et in log scale	eps gif png eps gif png
Dijet mass	eps gif png
Missing transverse energy in log scale	eps gif png eps gif png
Azimuthal angle between jets	eps gif png
Azimuthal angle between leading jet and missing transverse energy	eps gif png
Missing Ht (vectorial sum of the jet Et-s) in log scale	eps gif png eps gif png
Ht (scalar sum of the jet Et-s) in log scale	eps gif png eps gif png
(Missing Ht)/Ht	eps gif png

Signal Region (SIG, before optimization), Double Tag

Leading jet Et	eps gif png
Second leading jet Et	eps gif png
Dijet mass	eps gif png
Missing transverse energy	eps gif png
Azimuthal angle between jets	eps gif png
Azimuthal angle between leading jet and missing transverse energy	eps gif png
Missing Ht (vectorial sum of the jet Et-s)	eps gif png
Ht (scalar sum of the jet Et-s)	eps gif png
(Missing Ht)/Ht	eps gif png

Page was last updated August 8th, 2006 by Viktor Veszpremi
Contact, Legal Notices, and Fermilab Policy on Computing