1: \part[Vector boson scattering at the LHC]{Resonant vector boson scattering at high mass at the LHC}
2:
3: {\it G. Azuelos, P-A. Delsart and J. Id\'arraga}
4:
5:
6: \begin{abstract}
7: We examine, with full detector simuation, the reconstruction of $WZ$
8: resonances in the Chiral Lagrangian Model and the Higgsless model.
9: \end{abstract}
10:
11: \section{INTRODUCTION}
12: %\section{Introduction}
13:
14: In the absence of a light Higgs boson, either from the Standard
15: Model (SM), from supersymmetry (MSSM), or from a Little Higgs model,
16: electroweak symmetry breaking (EWSB) must find its origin in some
17: strongly interacting sector. Since the Goldstone bosons (GB)
18: breaking the symmetry become the longitudinal components of the
19: gauge bosons, the study of longitudinal vector boson ($V_L$)
20: scattering in the TeV region could reveal valuable information,
21: hopefully in the form of new resonances which should then be
22: discovered at the LHC. Previous ATLAS studies with full simulation
23: can be found in~\cite{Myagkov99,Azuelos99}. This note summarizes the
24: main conclusions from an analysis of WZ resonances with a more
25: realistic, full detector simulation~\cite{ATLAS_note}, performed in
26: the framework of the so-called {\it Data-Challenge 2} exercise of
27: the ATLAS collaboration.
28:
29: Dynamical EWSB is realized in many models, among which (i) models of
30: technicolor, where a new QCD-like gauge interaction is introduced,
31: along with chiral symmetry breaking producing the required GB's;
32: extended, multiscale, top-color assisted models of technicolor are
33: required to give mass to the fermions, including the top quark, while
34: avoiding FCNC effects (for a review,
35: see~\cite{Hill:2002ap,Lane:2002wv}); (ii) Little Higgs
36: models~\cite{Arkani-Hamed:2001ed,Arkani-Hamed:2001nc}, where a light
37: Higgs is present as a pseudo-GB resulting from the breaking of some
38: specific higher symmetries, (iii) higgsless
39: models~\cite{Csaki:2003zu}, where EWSB results from boundary
40: conditions at branes located in a warped fifth space dimension, and
41: (iv) string interactions. More generically, a Chiral Lagrangian (ChL)
42: model~\cite{Appelquist:1980vg,Longhitano:1980tm,Kilian:2003pc} of EWSB
43: provides a low energy effective description of electroweak
44: interactions. It is built as a covariant momentum (derivative)
45: expansion of GB fields, respecting the chiral symmetry $SU(2)_L \times
46: SU(2)_R$.
47:
48: Here, we consider a 1.15 TeV resonance resulting from a chosen set of
49: ChL parameters and a 700 GeV resonance from the Higgsless model.
50:
51: \section{Signals and backgrounds}
52: %\section{SIGNALS AND BACKGROUNDS}
53:
54: The chiral Lagrangian, in its expansion to fourth order, consists of
55: one term of dimension 2 completely determined by the symmetry
56: requirement and other interaction terms, of dimension four, with
57: arbitrary coefficients, serving as parameters of the model (see the
58: explicit form of the Lagrangian, for example,
59: in~\cite{Dobado:1999xb,Kilian:2003pc}). Among the dimension-four
60: terms, five of them describe vector boson scattering, but only two of
61: them, with coefficients $a_4$ and $a_5$, are important if one assumes
62: that custodial symmetry is conserved. The parameters $a_4$ and $a_5$,
63: together with the unitarization assumption, determine therefore the
64: phenomenology of high energy longitudinal vector boson scattering, and
65: can lead to the presence or absence of resonances, as predicted by
66: specific models. The partial waves can be calculated and, since the
67: effective Lagrangian is not renormalizable, a unitarization procedure
68: must be assumed. Here, we adopt the Inverse Amplitude Method (IAM), as
69: described in~\cite{Dobado:1999xb}. It gives an excellent description
70: of pion scattering at low energies~\cite{Pelaez:1999ey}. To generate
71: Monte Carlo samples of this signal, we have modified process 73 of
72: PYTHIA ($W_L Z_L$ scattering), replacing the partial waves by those
73: for vector boson scattering of the ChL model, as given
74: in~\cite{Haywood:1999qg} and choosing the Pade unitarization,
75: equivalent to the IAM. The parameters chosen were $a_4=0.00875$ and
76: $a_5=-0.00125$, corresponding to point $P2$ of
77: ref~\cite{Haywood:1999qg}, with a vector resonance of mass $\sim$1150
78: GeV and width $\Gamma = 85$ GeV.
79:
80: The signal of the Higgsless model was generated with PYTHIA, using the
81: QCD-like model of process 73, taking as reference a resonance mass of
82: 700 GeV, as in~\cite{Birkedal:2004au,Birkedal:2005yg}. SM vector
83: boson scattering background was added, choosing $m_h= 100$ GeV in
84: order to have a negligible contribution from diagrams with Higgs
85: exchange. The normalization of the resonance was obtained by
86: calculating the cross section in the resonance region in a model where
87: s-channel exchange of an additional $W_1$ Kaluza-Klein state of the
88: $W$, of mass 700 GeV, was introduced and where the Higgs diagram was
89: removed, as described in~\cite{Birkedal:2004au}.
90:
91: Three cases were studied:
92: (i) $qqWZ \to qq j j \ell \ell$ ,
93: (ii) $qqWZ \to qq\ell\nu j j$ and
94: (iii) $qqWZ \to qq\ell\nu \ell \ell$.
95: We discuss here only cases (1) and (3), as the $t\bar t$ and $W+j$
96: backgrounds are very important in the second case. In the ChL and
97: Higgsless models, for the cases studied, the cross sections for $qqWZ$
98: production are respectively 91.2 fb and 180 fb.
99:
100: The signals are characterized by the presence of the $W$ and $Z$ in
101: the final state, but also of two high energy jets in the forward and
102: backward directions originating from the primary quarks from which the
103: gauge bosons have been radiated. The backgrounds, therefore, are
104: processes with vector bosons and at least two jets. We have
105: considered the following because of their high cross-section.
106: \begin{itemize}
107: \item The main irreducible background is from SM $qqWZ$ processes
108: originating from gluon (QCD) or $Z/\gamma$ (QED) exchange diagrams
109: between quarks, with the $W$ and $Z$ radiated from the quarks. The
110: gauge bosons are mostly transverse ($V_T$), in this case, and
111: emitted less centrally than in the case of $V_LV_L$ scattering. This
112: background was generated with MADGraph~\cite{Maltoni:2002qb} with
113: some loose cuts: the two jets must have $p_T > 15$ GeV,
114: pseudorapidity $|\eta|<5$, with separation $|\Delta_r(qq)|>3$, where
115: $\Delta_r = \sqrt{(\Delta\phi)^2+(\Delta\eta)}$, and the invariant
116: mass $m_{qq}>250$ GeV. This leads to a cross section of 4.0 pb for
117: $qqW^+Z$ and 1.5 pb for $qqW^-Z$.
118: \item $t\bar t$ events with $P_T^{top} > 500$ GeV, generated with
119: (MC@NLO). Although the transverse momentum threshold is very high,
120: the cross section is high ($\sigma = 4.13$ pb). The number of events
121: used with full simulation, (18000), was therefore insufficient to
122: assess with good statistical accuracy the importance of this
123: background.
124: \item $W$+4jets events. This background sample was produced with
125: {\sc Alpgen}~\cite{Mangano:2002ea}. The cross-section is $\sigma = 1200$ pb and thus orders of
126: magnitude above the signal.
127: \end{itemize}
128:
129: All the events were fully simulated with Geant4 by the ATLAS
130: collaboration, using ATLAS {\it Rome} initial detector layout (and
131: using Athena version 9.0.4). They were digitized with electronic noise
132: but no-pileup. The events were reconstructed with the
133: default settings (Athena 10.0.4) but some extra jet collections were
134: added (see following section).
135:
136: \section{ANALYSIS}
137: %\section{Analysis}
138:
139: Details of the analysis can be found in~\cite{ATLAS_note}. Here, we
140: summarize the results, pointing out the main lessons from this full
141: detector simulation study. The analysis was performed within the
142: {\it Athena} framework at the level of the Analysis Object Data
143: (AOD). At this level, the potential physics objects (jets, muons,
144: electrons,... ) are reconstructed and one can access their kinematic
145: information as well as some identification or quality criteria.
146:
147: We chose a set of cuts according to the signal caracteristics
148: described in the introduction of the previous section :
149: \begin{enumerate}
150: \item {\bf identification and quality criteria}. We impose an
151: identification criterion (likelihood or a combined \emph{IsEM}
152: variable) for the electrons. The leptons ($e$'s or $\mu$'s) must
153: have high transverse momentum, 40 GeV and should be isolated, meaning that
154: there should be less than $\sim$6 GeV of
155: cumulated track energy in a 0.7 cone around their track.
156: Once candidate electrons or muons are chosen,
157: jets, reconstructed by various algorithms, are accepted if they do
158: not overlap with an electron and if they have a transverse momentum
159: $p_T > 15$ GeV.
160: \item {\bf Forward jets}. We require the presence of 2 \emph{forward
161: jets} with opposite directions. We consider a jet as a candidate
162: forward jet if its transverse momentum is greater than 15 GeV,
163: energy greater than 200 GeV and if it satisfies one of the
164: following conditions:
165: \begin{itemize}
166: \item It is the jet $j_1$ with highest pseudorapidity ($\eta$), but is
167: not also the jet with highest $P_T$.
168: \item $|\eta_j| > 2.5 $
169: \item The difference $|\eta|$ of this jet and $j_1$ is:
170: $\Delta \eta_{jj_1}>4$
171: \end{itemize}
172:
173: This complex selection was chosen in order to define central jets
174: relative to forward jets (rather than with absolute $\eta$
175: cuts). Other algorithms for tagging the forward jets were
176: considered, but did not result in an overall improvement.
177: \item {\bf central jets} Central jets expected from vector bosons are
178: required to have $p_T> 40$ GeV. They should lie, in $\eta$, between the
179: two forward jets.
180: \item {\bf Vector boson mass}. We impose that the mass of the
181: reconstructed W and/or Z be in a $\pm$15 GeV window around the
182: Standard Model value.
183: \item {\bf Central Jet Veto}. We reject all events with any excess
184: central jet (with $p_T > 40$ GeV).
185: \item { \bf $\mathbf{\Delta \phi}$ between vector bosons} Due to the
186: high mass of the resonance, it is produced almost at rest and the
187: vector bosons are essentially back-to-back. We impose, therefore,
188: that they be well separated azimuthally:
189: $\Delta\phi_{WZ}>1.0$.
190: \item {\bf Resonance mass}. To evaluate the efficiencies of the
191: selection criteria, we imposed a window cut of $\pm$150 GeV (100 GeV
192: for Higgsless case of a resonance at 700 GeV) around the
193: reconstructed mass. The significance of the signal is then estimated
194: from the number of signal ($S$) and background ($B$) events as
195: $S/\sqrt{B}$
196: \end{enumerate}
197: One important characterisitic of the signal, for cases 1 and 2, is
198: that the two central jets from the energetic vector boson decays are
199: highly boosted. They have a small opening angle and are often
200: reconstructed as a single jet. To study this effect and account for
201: it, we added to standard reconstruction various sets of jets with
202: different radius size. In general, if only one jet was found, with
203: mass close to the vector boson mass, we required that it be composed
204: of two subjets, when the cone radius was reduced to 0.2. Details can
205: be found in~\cite{ATLAS_note}.
206:
207: Preliminary results are shown in Fig~\ref{restopo4} for the two
208: signals considered. In both cases, a strong signal is seen, although
209: the shape of the background must be well understood, especially for
210: the Higgsless resonance.
211:
212: \begin{figure}[!ht]
213: \begin{center}
214: \includegraphics[width=0.48\textwidth,angle=0]{\nonSUSYBSMfigure{WZ_resonance_ChL_cuts1_topo4}}
215: \includegraphics[width=0.48\textwidth,angle=0]{\nonSUSYBSMfigure{WZ_resonance_QCDLike_topo4}}
216: \end{center}
217: \caption{Reconstructed resonances and backgrounds for the ChL model,
218: 1.2 TeV (left) and for the Higgsless model, 700 GeV (right), in the
219: 2-lepton channel. Although no $t\bar t$ nor $Wj$ background remains,
220: they cannot be statistically excluded.}
221: \label{restopo4}
222: \end{figure}
223:
224:
225:
226: \subsection{$qqWZ \to qq\ell\nu \ell \ell$}
227: %\subsection{$qqWZ \to qq\ell\nu \ell \ell$}
228:
229: This channel is relatively clean, because of the presence of three
230: leptons, but it is suppressed by the branching ratios. We will therefore
231: consider a signal for an integrated luminosity of 300 fb$^{-1}$,
232: for an integrated luminosity of 100 fb$^{-1}$.
233:
234: We apply similar cuts to the case above, with the difference that we
235: require the presence of 3 leptons with $p_T> 40$ GeV, as well as
236: transverse momentum of $\slashed{p}_T > 40$ GeV. The transverse
237: momentum of the neutrino is assumed to be the measured $\slashed{p}_T$
238: and the longitudinal momentum is constrained by requiring that
239: $m_{\ell\nu}= m_W$. We also require that two opposite sign, same
240: flavour leptons have the mass of the $Z$ within 15 GeV. With these
241: cuts, no events remain from $t\bar t$ and other backgrounds (except
242: the irreducible SM $qqWZ$ background), although the statistics are
243: insufficient to claim that they are completely
244: eliminated. Figs.~\ref{res3l} shows preliminary results for the ChL
245: and Higgsless models studied here.
246:
247: \begin{figure}[!ht]
248: \begin{center}
249: \includegraphics[width=0.48\textwidth,angle=0]{\nonSUSYBSMfigure{WZ_ResonanceChL3lep}}
250: \includegraphics[width=0.48\textwidth,angle=0]{\nonSUSYBSMfigure{WZ_ResonanceHless3lep}}
251: \end{center}
252: \caption{Reconstructed resonances and backgrounds for the ChL model,
253: 1.2 TeV (left) and for the Higgsless model, 700 GeV (right), in the
254: 3-lepton channel, for an integrated luminosity of 300
255: fb$^{-1}$.Although no $t\bar t$ nor $Wj$ background remains, they
256: cannot be statistically excluded.}
257: \label{res3l}
258:
259: \end{figure}
260:
261: \section{CONCLUSION}
262: %\section{Conclusion}
263:
264: The reconstruction of high mass $WZ$ resonances arising from a
265: Chiral Lagrangian model and from a Higgsless model have been studied
266: using full detector simulation. Although insufficient statistics were
267: available for background estimation, preliminary results show that,
268: with appropriate cuts, and depending on the parameters of the models,
269: significant signals can be obtained within 1-3 years of data taking at
270: the LHC at nominal luminosity (corresponding to 100-300 fb$^{-1}$).
271:
272: %\bibliography{WZ}
273:
274: %\end{document}
275: