1: \documentclass[12pt,a4paper,dvips]{article}
2: \usepackage{a4p}
3: \usepackage{cite,mcite}
4: \usepackage{graphicx}
5: \usepackage{physics }
6: \usepackage{l3_title,ifthen}
7: \usepackage{rotating}
8: \usepackage{epsfig}
9:
10: \journalname{Phys. Lett. B}
11: \date{December 9, 2003}
12:
13: \usepackage{Lep}
14: \preprint{2003-086}
15: \Lep{2}
16:
17: %
18: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
19: %
20: \newlength{\capindent}
21: \setlength{\capindent}{1.0cm}
22: \newlength{\capwidth}
23: \setlength{\capwidth}{\textwidth}
24: \addtolength{\capwidth}{-2\capindent}
25: \newlength{\figwidth}
26: \setlength{\figwidth}{\textwidth}
27: \addtolength{\figwidth}{-2.0cm}
28: \newcommand{\icaption}[2][!*!,!]{\hspace*{\cpaindent}%
29: \begin{minipage}{\capwidth}
30: \ifthenelse{\equal{#1}{!*!,!}}%
31: {\caption{#2}}%
32: {\caption[#1]{#2}}
33: \end{minipage}}
34: %
35: % Shorthand for \phantom to use in tables
36: \newcommand{\pho}{\phantom{0}}
37: \newcommand{\bslash}{\ensuremath{\backslash}}
38: \newcommand{\BibTeX}{{\sc Bib\TeX}}
39: % Upsilion(1S)
40: \newcommand{\UoneS}{\ensuremath{\Upsilon(\mathrm{1S})}}
41: %
42: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
43: %
44: \input{l3at183.def}
45:
46: %
47: \begin{document}
48: %
49: \begin{titlepage}
50: %
51: \title{Measurement of Triple-Gauge-Boson Couplings \\
52: of the W Boson at LEP \\}
53: \vspace*{-20mm}
54: \author{The L3 Collaboration}
55: \vspace*{-10mm}
56:
57: %
58: % The abstract
59: %
60: \begin{abstract}
61: The CP-conserving triple-gauge-boson couplings,
62: $\giZ$, $\kg$, $\Lg$, $\gvZ$, $\kZ$ and $\LZ$
63: are measured using hadronic and semi-leptonic W-pair events
64: selected in 629~pb$^{-1}$ of data collected at LEP with the L3 detector
65: at centre-of-mass energies between $189$ and $209~\GeV$.
66: The results are combined with
67: previous L3 measurements based on data collected at lower
68: centre-of-mass energies and with the results from single-W
69: production and from events with a
70: single-photon and missing energy.
71: Imposing the constraints $\kZ = \giZ - \tan^2\theta_{W}
72: (\kg-1)$
73: and $\LZ = \Lg$, we obtain for the C and P conserving couplings the
74: results:
75: %
76: \begin{center}
77: \parbox{9cm}
78: {
79: \begin{eqnarray*}
80: \giZ & = & \phantom{+} 0.966\pm0.033 \mathrm{(stat.)} \pm0.015 \mathrm{(syst.)} \\
81: \kg & = & \phantom{+} 1.013\pm0.066 \mathrm{(stat.)} \pm0.026 \mathrm{(syst.)} \\
82: \Lg & = & -0.021\pm0.035 \mathrm{(stat.)} \pm0.017 \mathrm{(syst.)}.
83: \end{eqnarray*}
84: }
85: \end{center}
86: %
87: Results from the analysis of fully leptonic W-pair decays are also given.
88: %
89: All results are in agreement with the Standard Model expectations and
90: confirm the existence of self-couplings among electroweak gauge
91: bosons.
92: %
93: \end{abstract}
94: \submitted
95:
96:
97: \end{titlepage}
98: %
99: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
100: % Introduction
101: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
102: %
103: \section{Introduction}
104: \label{sec:intro}
105:
106: The non-Abelian structure of the electroweak
107: theory~\cite{standard_model} implies the existence of trilinear self
108: couplings among gauge bosons.
109: The vertices $\gamma$WW and ZWW are accessible at LEP
110: through W-pair, single-W and single-photon production~\cite{LEP2YRSM}.
111:
112: To lowest order,
113: three Feynman diagrams contribute to W-pair production: the
114: $s$-channel $\gamma$ and Z exchange and the $t$-channel
115: $\nu_{\e}$ exchange.
116: The $s$-channel diagrams contain the $\gamma$WW and ZWW vertices.
117: The $\gamma$WW vertex
118: appears in one of the $t$-channel Feynman diagrams contributing to
119: single-W production, $\EE\rightarrow\PW\EN$; at LEP
120: centre-of-mass energies, $\sqrt{s}$, the contribution from the similar
121: diagram containing the ZWW vertex is negligible.
122: The $\gamma$WW vertex also contributes to the $\rm e^+e^- \rightarrow
123: \nu_{\mathrm{e}}\bar\nu_{\mathrm{e}}\gamma$ process through photon
124: production in W-boson fusion.
125:
126: Assuming only Lorentz invariance, the most general form of the
127: $\gamma$WW and ZWW vertices is parametrised in terms
128: of seven complex triple-gauge-boson couplings (TGC's) each~\cite{Hagiwara87}.
129: Retaining only CP-conserving
130: couplings and assuming electromagnetic gauge invariance, six real TGC's
131: remain, namely $\giZ$, $\kg$, $\Lg$, $\gvZ$, $\kZ$ and $\LZ$.
132: At tree level within the Standard Model, $\giZ=\kg=\kZ=1$ and $\gvZ=\Lg=\LZ=0$.
133: Except $\gvZ$, these TGC's also conserve C and P separately.
134: %
135: The requirement of custodial $\mathrm{SU(2)}$ symmetry leads to the
136: relations $\kZ = \giZ - \tan^2\theta_{W} (\kg-1)$ and $\LZ =
137: \Lg$~\cite{DKLDXYSchild,LEP2YRAC}, where $\theta_{W}$
138: is the weak mixing angle.
139: %
140: When these constraints are applied,
141: $\giZ$, $\kg$ and $\Lg$ correspond to the operators in a linear
142: realisation of a gauge-invariant effective Lagrangian that do not
143: affect the gauge-boson propagators at tree level~\cite{LEP2YRAC}.
144: The $\giZ$, $\kg$ and $\Lg$ couplings are studied assuming these constraints.
145: %
146: The analysis is based on the study of multi-differential cross
147: sections measured in hadronic and semi-leptonic W-pair events.
148: %
149: Measurements at lower $\sqrt{s}$~\cite{l3-189} are included, as
150: well as events selected by the single-W analysis~\cite{l3-252} and
151: events with a single photon and missing energy~\cite{singlePhoton}.
152: %
153: Results from the analyses of fully leptonic W-pair decays are also given.
154: %
155: Results on TGC's were also published by experiments at hadron
156: colliders~\cite{collidertgc} and at LEP~\cite{LEPtgc}.
157:
158: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
159: % Data and MC samples
160: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
161: \section{Data and Monte Carlo Samples}
162: \label{sec:datamc}
163:
164: The data sample collected by the L3 detector~\cite{l3_det} in the
165: years from 1998 through 2000 is used in the W-pair analysis.
166: It corresponds to an integrated luminosity of
167: 629.2~pb$^{-1}$ at $\sqrt{s}=189 -209~\GeV$,
168: detailed in Table~\ref{tab:luminosities}.
169: An additional 76.4~pb$^{-1}$ of data at $\sqrt{s}=161 - 183~\GeV$ is used for the
170: single-W analysis.
171: %
172:
173: The following Monte Carlo
174: event generators are used to simulate the signal and background
175: reactions:
176: KandY~\cite{kandy} and EXCALIBUR~\cite{EXCALIBUR}
177: for $\EEFFFFG$;
178: PYTHIA~\cite{JETSET73} for $\EEQQG, \EE \ra \mathrm{ZZ}(\gamma)$ and
179: $\EE \ra \mathrm{Z}\EE$;
180: KK2f~\cite{KK2f} for $\EEQQG, \EE \ra \MMG$ and $\EE \ra \TTG$;
181: BHAGENE3 \cite{BHAGENE}, BHWIDE~\cite{BHWIDE} and TEEGG~\cite{TEEGG}
182: for $\EEEEG$ and
183: DIAG36~\cite{DIAG36} and PHOJET~\cite{PHOJET} for lepton and hadron
184: production in two-photon collisions, respectively.
185: The KandY program, used to generate
186: W-pair events, combines the four-fermion generator KORALW~\cite{KORALW} with
187: the $\cal{O}(\alpha)$ radiative corrections in the leading-pole
188: approximation~\cite{DPA} implemented in the YFSWW program~\cite{YFSWW3}.
189:
190: The response of the L3 detector is modelled with the
191: GEANT~\cite{geant} program which includes
192: effects of energy loss, multiple scattering and showering in the
193: detector materials and in the beam pipe.
194: Time-dependent detector inefficiencies, as monitored during the data
195: taking period, are included in the simulations.
196:
197:
198:
199:
200: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
201: % Event Selection
202: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
203: \section{Event Selection}
204: \label{sec:evtsel}
205:
206: \subsection{W-pair Events}
207:
208: The event selection is based on that described in
209: Reference~\citen{189paper} and its results are detailed in
210: Reference~\citen{wwxsecpaper}.
211: %
212: The visible fermions in the final state are reconstructed as
213: electrons, muons, jets corresponding to
214: decay products of $\tau$ leptons, and hadronic jets corresponding to
215: quarks.
216: Only events containing leptons with an unambiguous charge assignment are retained.
217: %
218: The numbers of selected hadronic, semi-leptonic and fully leptonic
219: W-pair events and
220: the expected background are given in Table~\ref{tab:events-all}.
221:
222: Kinematic fits are performed
223: to improve the resolution of the measured fermion energies and angles
224: and to determine neutrino momenta in semi-leptonic events.
225: %
226: Four-momentum conservation and equal mass of the two W bosons are
227: imposed as constraints.
228: %
229: In $\QQTN$ events, the energies of the two hadronic jets
230: are rescaled by a common factor so that their sum equals $\sqrt{s}/2$.
231: %
232: The four jets in hadronic events are paired to form W bosons
233: by a neural network based on
234: the difference and sum of the masses of the jet pairs, the sum and the
235: minimum of the angles between paired jets, the energy
236: difference between the jet pairs and between the paired jets,
237: the value of the matrix element for the process $\EE \ra \PW^+\PW^- \ra ffff$
238: as calculated with EXCALIBUR from the jet four-momenta, and
239: the difference between the charges of the jet pairs as determined from the jet
240: charges~\cite{l3-189}. The correct pairing is found
241: for 77\% of the selected Monte Carlo events.
242:
243:
244: \subsection{Single-W Events}
245:
246: The $\EE\rightarrow\PW\EN$ process typically has an electron scattered
247: at very low polar angle, so that only the decay products of the W
248: boson are observed as single-lepton events or acoplanar
249: jets. Single-lepton events are selected by exploiting their peculiar
250: signature in the detector, while a neural network is used to isolate
251: hadronic single-W events from the background~\cite{l3-252}.
252: The hadronic sample consists of 740 events out of which 156
253: are also accepted by the semi-leptonic W-pair selections.
254: %
255: From Monte Carlo studies, about $75\%$ of this overlap consists of
256: W-pair events, mostly $\QQTN$ events, while only $7\%$ consists of
257: single-W events, the remainder being $\EE \ra \QQG$ events.
258: %
259: In order to avoid double counting, these events are
260: considered in the W-pair sample only.
261:
262: The numbers of selected single-W events
263: and the expected background, after the
264: removal of the overlapping events, are reported in
265: Table~\ref{tab:events-all}.
266:
267: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
268: %Event Reconstruction
269: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
270: \section{Event Reconstruction}
271: \label{sec:phasespace}
272:
273: % \subsection{W-pair Analysis}
274:
275: For unpolarised initial states, summing over final-state fermion
276: helicities, fixing the mass of the W
277: boson and neglecting photon radiation, five angles
278: completely describe the four-fermion final state originating
279: from W-pair decay.
280: These angles are the production angle of
281: the W$^-$ boson, $\Theta_{\PW^-}$, and the polar and the azimuthal decay
282: angles of the fermion in W$^-$ decays and the anti-fermion in W$^+$ decays,
283: calculated in the rest frame of the W boson.
284: TGC's affect the total production cross section, the
285: W production angle, and the
286: polarisations of the two W bosons, which in turn determine the
287: W decay angles.
288:
289: For semi-leptonic W-pair events, the W$^-$ production angle is
290: reconstructed from the hadronic part of the event, and the sign
291: of $\cos\Theta_{\PW^-}$ is
292: determined from the lepton charge.
293: %
294: If both W bosons decay into hadrons,
295: the W charge assignment follows from jet-charge technique~\cite{l3-189}.
296: This charge assignment is found to be correct for
297: 69\% of Monte Carlo events with correctly paired jets.
298: %
299: The distributions of $\cos \Theta_{\PW^-}$
300: for hadronic and semi-leptonic events are shown in
301: Figure~\ref{fig:wwctw} where,
302: for illustrative purposes, all data are combined.
303:
304: The charge of the lepton allows the
305: reconstruction of the decay angles $\theta_\ell$ and $\phi_\ell$.
306: %
307: Jet-charge
308: determination is not adequate to determine the quark charge and
309: a two-fold ambiguity arises for the decay angles
310: of W bosons decaying into hadrons,
311: $(\cos\theta_q,\phi_q)\leftrightarrow(-\cos\theta_q,\pi+\phi_q)$.
312: %
313: The $\phi_q$ distribution is restricted to the interval
314: $(0,\pi]$ and the jet with $\phi_q \in (0,\pi]$ is assigned to the quark
315: or the anti-quark originating from the decay of W$^-$ or W$^+$
316: respectively.
317: The absolute value of the cosine of the polar decay angle is considered.
318: %
319: The distributions of the hadronic decay angles for the hadronic
320: channel and the leptonic and hadronic decay
321: angles for the semi-leptonic channels are shown in
322: Figures~\ref{fig:wwdecayqqqq} and \ref{fig:wwdecayqqln},
323: respectively.
324:
325: % lnln
326: Fully leptonic W-pair decay channels with final state
327: muons and electrons are also analysed. The presence of two neutrinos
328: prevents an unambiguous reconstruction of the event. Assuming no
329: initial-state radiation, and fixing the mass of the W boson, the production angle
330: of the latter is kinematically derived with a two-fold
331: ambiguity~\cite{LEP2YRAC}. Due to resolution effects, about 40\% of the
332: events yield complex solutions and are not considered. A weight of one
333: half is given to each solution of the retained events.
334:
335: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
336: %Data Analysis
337: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
338: \section{Data Analysis}
339: \label{sec:fit}
340:
341: \subsection{Fit Method}
342:
343: Binned maximum likelihood fits are used to perform the TGC
344: measurement. Bin sizes are chosen so as to optimise sensitivity for the
345: given Monte Carlo statistics.
346: %
347: For hadronic and semi-leptonic W-pairs, the likelihoods depend on
348: the W production and decay angles.
349: %
350: For $\cos \Theta_{\PW^-}$, 12 bins are considered in the
351: hadronic channel, 10 bins for $\QQEN$ and $\QQMN$ events and 8 bins
352: in the $\QQTN$ channel.
353: %
354: For the leptonic decay angles $\cos \theta_{\ell}$ and $\phi_{\ell}$,
355: 4 bins are used, while 3 bins are considered for the hadronic
356: decay angles $|\cos \theta_{q}|$ and $\phi_{q}$.
357: %
358: For leptonic single-W events, the lepton energy is used in the fit,
359: with bins of $5 \gev$. Its distribution is
360: shown in Figure~\ref{fig:swdist}a.
361: %
362: For hadronic single-W events
363: the neural network output, whose distribution is shown in
364: Figure~\ref{fig:swdist}b, is used in the fit.
365: It is divided in bins of 0.01.
366:
367: For each decay channel and value of $\sqrt{s}$, the likelihood
368: is defined as the product of the Poisson probabilities of occupation
369: in each bin of the phase space as a function of a given set of
370: couplings $\Psi$:
371: %
372: \begin{eqnarray}
373: L(\Psi) & = & \prod_i^{\mathrm{bins}}
374: \frac{e^{-\mu_i(\Psi)} \mu_i(\Psi)^{N_i}}
375: {N_i !}\; ,
376: \end{eqnarray}
377: %
378: where $\mu_i$ is the expected number of signal and background
379: events in the $i$-th bin
380: and $N_i$ is the corresponding observed number of events.
381: The dependence of $\mu_i$ on $\Psi$ is determined
382: by a generator level reweighting procedure
383: applied to fully simulated Monte Carlo events.
384: For any value of $\Psi$,
385: the weight $R$ of the $n$-th event
386: generated with TGC value $\Psi_{\gen}$ is:
387: \begin{eqnarray}
388: R(\Omega_n,\Psi,\Psi_{\gen}) & = &
389: \frac
390: {\left|{\cal M}(\Omega_n,\Psi)\right|^2}
391: {\left|{\cal M}(\Omega_n,\Psi_{\gen})\right|^2}\,,
392: \label{eqn:weight}
393: \end{eqnarray}
394: where ${\cal M}$ is the matrix element of the final
395: state considered, evaluated~\cite{EXCALIBUR} for the generated phase space $\Omega_n$,
396: which includes radiated photons.
397:
398: The expected number of events in the $i$-th bin is:
399: \begin{eqnarray}
400: \mu_{i}(\Psi) & = &
401: \sum_l^{\mathrm{sig+bg}} \left(
402: \frac{\sigma_l^{\gen} \; \cal{L}}{N_l^{\gen}}
403: \sum_{j}^{n_i} R_l(\Omega_j,\Psi,\Psi_{\gen})
404: \right) \,,
405: \end{eqnarray}
406: where the first sum runs over all signal and background samples, and
407: $\sigma_l^{\gen}$
408: denotes the cross section corresponding to
409: the total Monte Carlo sample containing $N_l^{\gen}$
410: events and $\cal{L}$ is the integrated luminosity.
411: The second sum extends over the number $n_i$ of accepted Monte Carlo
412: events in the $i$-th bin.
413: This definition takes properly into account detector effects
414: and $\Psi$-dependent efficiencies and purities.
415: For background sources which are independent of TGC's, $R_l =1$.
416: The fitting method described above determines the TGC's without any
417: bias as long as the Monte Carlo correctly describes photon radiation and
418: detector effects such as resolution and acceptance functions.
419: Different channels and centre-of-mass
420: energies are combined by multiplying together the corresponding likelihoods.
421:
422: The following results are obtained for hadronic and semi-leptonic W-pairs
423: and for their combination, allowing one coupling to vary
424: while fixing the others to their Standard Model
425: values:
426: %
427: \begin{eqnarray}
428: %
429: \!\!\!\!\! \giZ = \phantom{+}0.914\apm{0.065}{0.056} \; (\QQQQ) & \; \; \;
430: \giZ = \phantom{+}0.974\apm{0.039}{0.038} \; (\QQLN) & \; \; \;
431: \giZ = \phantom{+}0.959\apm{0.034}{0.033} \; (\mathrm{combined})
432: \nonumber \\
433: %
434: \!\!\!\!\! \kg = \; \; \phantom{+}0.89\apm{0.12}{0.10} \; \; (\QQQQ) & \; \; \;
435: \kg = \phantom{+}0.918\apm{0.097}{0.085} \; (\QQLN) & \; \; \;
436: \kg = \phantom{+}0.907\apm{0.074}{0.067} \; (\mathrm{combined})
437: \nonumber \\
438: %
439: \!\!\!\!\! \Lg = -0.102\apm{0.069}{0.058} \; (\QQQQ) & \; \; \;
440: \Lg = -0.026\apm{0.040}{0.038} \; (\QQLN) & \; \; \;
441: \Lg = -0.044\apm{0.036}{0.033} \; (\mathrm{combined}) \, .
442: \nonumber
443: %
444: \end{eqnarray}
445: %
446: These couplings are determined under the constraints $\kZ = \giZ -
447: \tan^2\theta_{W} (\kg-1)$ and $\LZ = \Lg$. Relaxing these
448: constraints, and fixing all other couplings to their Standard Model
449: values, yields:
450: %
451: \begin{eqnarray}
452: %
453: \!\!\!\!\! \gvZ = \; \; \phantom{+}0.20\apm{0.21}{0.22} \; \; (\QQQQ) & \; \; \;
454: \gvZ = \; \; -0.10\apm{0.17}{0.17} \; \; (\QQLN) & \; \; \;
455: \gvZ = \; \; \phantom{+}0.00\apm{0.13}{0.13} \; \; (\mathrm{combined})
456: \nonumber \\
457: %
458: \!\!\!\!\! \kZ = \phantom{+}0.856\apm{0.108}{0.091} \; (\QQQQ) & \; \; \;
459: \kZ = \phantom{+}0.957\apm{0.068}{0.066} \; (\QQLN) & \; \; \;
460: \kZ = \phantom{+}0.921\apm{0.059}{0.056} \; (\mathrm{combined})
461: \nonumber \\
462: %
463: \!\!\!\!\! \LZ = -0.179\apm{0.108}{0.085} \; (\QQQQ) & \; \; \;
464: \LZ = -0.038\apm{0.066}{0.063} \; (\QQLN) & \; \; \;
465: \LZ = -0.070\apm{0.060}{0.057} \; (\mathrm{combined})\, .
466: \nonumber
467: %
468: \end{eqnarray}
469: %
470: The fit to fully leptonic W-pair events yields:
471: %
472: \begin{eqnarray}
473: \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \;
474: \giZ = 0.91\apm{0.22}{0.16} \; \; \;\; \; \; &
475: \kg = 1.07\apm{0.61}{0.38} \; \; \;\; \; \; &
476: \Lg = -0.16\apm{0.15}{0.12}
477: \nonumber
478: \end{eqnarray}
479: %
480: Due to the large statistical uncertainties of this channel,
481: compared to the other W-pair decay channels, these
482: results are not considered in the following combinations.
483:
484:
485: \subsection{Cross Checks}
486:
487: The fitting procedure is tested to high
488: accuracy by fitting large Monte Carlo samples, typically a hundred
489: times the size of the data.
490: TGC values are varied in a range corresponding to three
491: times the expected statistical uncertainty
492: and are correctly reproduced by
493: the fit~\cite{villa,dierckxsens}.
494:
495: The fit results are found to be independent of the value
496: $\Psi_{\gen}$ of the Monte Carlo sample subjected to the reweighting
497: procedure.
498:
499: The statistical uncertainties given by the fit are
500: tested by fitting, for each final state, several hundreds of small Monte
501: Carlo samples of the size of the data samples. The width of the
502: distribution of the fitted central values agrees well with the mean of
503: the distribution of the uncertainties.
504:
505: An independent analysis, based on
506: optimal observables technique~\cite{Fanourakisneutraltgc}, is
507: performed for the W-pair events and used as a cross check.
508: %
509: Both the central values and the uncertainties agree
510: with those from the binned maximum likelihood
511: fit.
512:
513:
514: \subsection{Single-Photon Events}
515:
516: Single-photon events
517: are mainly due to initial state radiation (ISR)
518: in neutrino-pair production through $s$-channel
519: Z-boson exchange or $t$-channel W-boson exchange. A small fraction of
520: events is due to W-boson fusion through the WW$\gamma$ vertex, which
521: gives access to $\kappa_{\gamma}$ and $\lambda_{\gamma}$. Data at $\sqrt{s}=189-209~\GeV$ are
522: analysed~\cite{singlePhoton} and 1898 events are selected while 1905
523: are expected from the Standard Model. The KK2f Monte Carlo
524: program~\cite{KK2f} is used to simulate the $\rm e^+ e^-
525: \rightarrow \nu \bar{\nu}\gamma$ process and effects of TGC's are obtained by a
526: reweighting procedure~\cite{was_tgc}.
527:
528: Binned maximum likelihood fits to the photon energy and polar angle
529: yield the results given in Table~\ref{tab:res1d1}.
530: The systematic uncertainties are dominated by uncertainties on the selection
531: efficiency~\cite{singlePhoton}, on the cross section~\cite{KK_theory}
532: and on the TGC modelling~\cite{was_highqed}.
533:
534: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
535: %Systematic Uncertainties
536: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
537: \section{Systematic Uncertainties}
538: \label{sec:syst}
539:
540: The systematic uncertainties for
541: W-pair events are summarised in Table~\ref{tab:sysww}.
542: %
543: The largest contributions are due to the limited Monte Carlo
544: statistics and to uncertainties on the background modelling,
545: the W-pair cross section and the lepton charge reconstruction.
546:
547: Systematic effects typically induce a shift in the position of the
548: maximum of the likelihood as well as a change of sensitivity. For
549: sources of systematic uncertainties evaluated by varying a parameter
550: between two extremes of a range, if the sensitivity loss is larger
551: than the gain, the total uncertainty is evaluated as the sum in
552: quadrature of the difference between the loss and the gain and of the
553: shift in the maximum of the likelihood. If the gain in sensitivity is
554: larger than the loss, only the shift in the maximum is quoted as
555: systematic uncertainty.
556:
557: An uncertainty of 0.5\% on the $\EE\rightarrow \PW^+\PW^-$
558: cross section is assumed~\cite{MC-ffff}, based on the
559: predictions of KandY and RacoonWW~\cite{RACOONWW}.
560: Both programs use either the leading-pole or
561: the double-pole approximation.
562: The $\cos\Theta_{\mathrm{W}^-}$ distribution expected for
563: these $\mathcal{O}(\alpha)$ calculations are compared and found to
564: agree, in average slope, up to $0.4\%$. This value is assigned as
565: systematic uncertainty.
566: Comparable uncertainties were obtained by a dedicated
567: study~\cite{Bruneliere}. Uncertainties from $\mathcal{O}(\alpha)$ corrections on the W-boson decay angles are found to be negligible~\cite{dierckxsens}.
568:
569: Uncertainties in the background cross sections
570: and differential distributions are
571: possible sources of systematic effects. The cross sections of the
572: $\EE \ra \QQ (\gamma)$ and $\EE \ra$~ZZ$(\gamma)$ processes are varied
573: within the theoretical uncertainty~\cite{MC-ffff} of $\pm 2\%$.
574: To reproduce the measured four-jet event rate of the
575: $\EE \ra \QQ (\gamma)$~\cite{wwxsecpaper}, the corresponding Monte Carlo
576: is scaled by 12.7\%. Half of the effect is assigned as an additional
577: systematic uncertainty.
578: Moreover, the $\cos\Theta_{\mathrm{W}^-}$ distributions for these
579: backgrounds are reweighted with a linear function of slope $\pm 5\%$,
580: in order to account for possible
581: inaccuracies of the Monte Carlo predictions, giving a small additional
582: contribution to this systematic uncertainty.
583:
584: The uncertainties on the lepton and jet charge assignment are derived
585: from the statistical accuracy of the two data sets used to check the charge
586: measurement~\cite{villa,dierckxsens}:
587: lepton-pair events in Z-peak calibration data for
588: the measurement of the lepton charge
589: and semi-leptonic W-pair events with muons
590: for the charge of W bosons decaying into hadrons.
591: Uncertainties around 0.2\% are found for single
592: tracks used for electron and tau reconstruction in the barrel and
593: between 1\% and 12\% in the endcaps, uncertainties around 0.06\% for the
594: charge of muons and
595: around 1.3\% for the charge of W bosons decaying into hadrons.
596:
597: The agreement of data and Monte Carlo in the reconstruction of
598: angles and energies of jets and leptons is tested with di-jet and di-lepton
599: events collected during Z-peak calibration runs.
600: The uncertainties on scales and resolutions of
601: energy and angle measurements are propagated in the Monte Carlo and their
602: effect on the TGC results is assigned as a systematic uncertainty.
603:
604: The uncertainty caused by limited Monte Carlo statistics is evaluated
605: by repeating the TGC fit with subsets of the total reference sample,
606: analysing the fit results as a function of the sample size and
607: extrapolating this shift to the full sample.
608:
609: The modelling of initial-state radiation in KandY is included up
610: to $\mathcal{O}(\alpha^3)$ in the leading-logarithm approximation.
611: %
612: The systematic uncertainty is estimated by comparing the fit results when
613: only ISR up to $\mathcal{O}(\alpha^2)$ is considered.
614: A good description of final-state radiation (FSR) is important to
615: properly reconstruct the phase space variables used in the TGC fit.
616: This effect is studied by repeating the TGC fit with Monte Carlo samples
617: from which the events with FSR photons of energy above a
618: cut-off, varied between $100~\MeV$ and $1~\GeV$, are removed.
619: %
620:
621: Systematic effects due to the uncertainty on the measurement
622: of the W mass and width are evaluated by varying these
623: parameters within the uncertainties of the world averages~\cite{PDG02}.
624:
625: High statistics Monte Carlo samples generated with different hadronisation
626: schemes, PYTH\-IA~\cite{JETSET73}, HERWIG~\cite{HERWIG} and
627: ARIADNE~\cite{ARIADNE}, are used to evaluate the effect of
628: hadronisation modelling uncertainties.
629: The average of the absolute value of the TGC shifts observed between
630: different models is assigned as systematic uncertainty.
631:
632: Other final state phenomena which can influence the TGC fit are
633: colour reconnection~\cite{FSI-CR} and
634: Bose-Einstein~\cite{FSI-BE} effects.
635: Monte Carlo samples with implementation of
636: different models of colour reconnection and Bose-Einstein
637: correlations are used to fit TGC's and
638: evaluate the associated systematic uncertainties by comparison with
639: the reference sample.
640: For colour reconnection the following models are tested: model
641: II~\cite{lonnblad} in ARIADNE, the scheme implemented
642: in HERWIG and the SK~I~\cite{sjost} model with full reconnection
643: probability in PYTHIA.
644: Based on a study of compatibility of SK~I with colour flow
645: between jets~\cite{l3-272}, only half the effect is considered.
646: The averages of the absolute values
647: of the shifts obtained using different models are
648: quoted as systematic uncertainties.
649: %
650: For Bose-Einstein correlation, the
651: LUBOEI~\cite{LUBOEI} BE$_{32}$ model as implemented in
652: PYTHIA with and without correlation between jets coming from
653: different W bosons is studied. The difference is taken as systematic
654: uncertainty.
655:
656: Systematic uncertainties for the single-W results are dominated by
657: uncertainties on selection efficiencies and
658: signal cross section~\cite{l3-252} and amount to 0.068 for $\kg$ and 0.08 for
659: $\Lg$.
660:
661: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
662: %Results and Discussion
663: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
664: \section{Results and Discussion}
665: \label{sec:tgcs}
666:
667: The results obtained from the study of W-pair events collected at
668: $\sqrt{s}= 189-209~\GeV$ are
669: combined taking into account correlations
670: of systematic errors between decay channels and between
671: data sets collected at different centre-of-mass energies.
672:
673: Further, they are combined with W-pair results obtained at lower
674: $\sqrt{s}$~\cite{l3-189}, with the single-W
675: results~\cite{l3-252} recalculated after removing the overlap with the W-pair
676: selection and with the results from single-photon events~\cite{singlePhoton}.
677:
678: The results of one-parameter fits, in which only one coupling is
679: allowed to vary while the others are set to their Standard Model
680: values, are given in Table~\ref{tab:res1d1}.
681: Negative log-likelihood curves are shown in Figure~\ref{fig:logl1d}.
682:
683: Multi-parameter fits of TGC's allow a
684: model-independent interpretation of the data.
685: Fits to two of the couplings $\kg$, $\Lg$ and $\giZ$,
686: keeping the third coupling fixed at its Standard Model value, are performed,
687: as well as a simultaneous fit to all these couplings.
688: In each case the constraints
689: $\kZ = \giZ - \tan^2 \theta_{W} (\kg-1)$ and $\LZ =
690: \Lg$ are imposed.
691: The results of these multi-parameter fits
692: are reported in Table~\ref{tab:res2-3d}.
693: The contour curves of 68\% and 95\% confidence level for the two-parameter fits
694: are shown in Figure~\ref{fig:ac-ndlnl}.
695: They correspond to a change
696: in the negative log-likelihood with respect to its minimum of $1.15$ and $3.00$,
697: respectively.
698: Contours derived from three-parameter fits are also shown.
699: They are obtained requiring a log-likelihood
700: change of $1.15$, but leaving the third coupling free to vary in the fit.
701: The comparison of the results derived from fits of different dimensionality
702: shows good agreement.
703:
704: If the W boson were an extended object, \eg \ an ellipsoid of rotation with
705: longitudinal radius $a$ and transverse radius $b$, its size
706: and shape would be related to the TGCs by $R_{\PW} \equiv (a+b)/2 =
707: (\kg+\Lg-1)/m_{\mathrm W}$~\cite{Brodsky:1980} and $\Delta_{\PW} \equiv
708: (a^2-b^2)/2 = (5/4)(\kg-\Lg-1)/m_{\mathrm W}^2$~\cite{quad1},
709: where $m_{\mathrm W}$ is the mass of the W boson.
710: The measurements show no evidence for the W boson to be an extended
711: object:
712: \begin{eqnarray}
713: R_{\PW} & = & (0.3\pm1.9)\times 10^{-19}~\mathrm{m} \\
714: \Delta_{\PW} & = & (0.89\pm0.83)\times 10^{-36}~\mathrm{m}^2 \,,
715: \end{eqnarray}
716: with a correlation coefficient of $-0.63$.
717:
718: In conclusion, TGC's are measured with an accuracy of a few percent.
719: All single- and multi-parameter TGC results show good agreement with
720: the Standard Model expectation and confirm the existence of self-couplings among
721: the electroweak gauge bosons.
722:
723: \clearpage
724:
725: %
726: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
727: % Bibliography
728: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
729: \bibliographystyle{/afs/cern.ch/l3/paper/biblio/l3stylem}
730: \bibliography{tgc}
731:
732: %
733: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
734: % Author List
735: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
736: %
737: \newpage
738: \input namelist281.tex
739: \clearpage
740:
741: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
742: % TABLES
743: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
744:
745: \begin{table}[p]
746: \begin{center}
747: \renewcommand{\arraystretch}{1.2}
748: \begin{tabular}{|c| r r r r r r r r|}
749: \hline
750: $\langle \RS \rangle \; [\GeV]$ &
751: 188.6 & 191.6 & 195.5 & 199.6 & 201.8 & 204.8 & 206.5 & 208.0 \\
752: $\mathcal{L} \; [\mathrm{pb}^{-1}]$ &
753: 176.8 & 29.8 & 84.1 & 83.3 & 37.1 & 79.0 & 130.5 & 8.6 \\
754: \hline
755: \end{tabular}
756: \caption[]{The average centre-of-mass energies, $\langle \RS \, \rangle$,
757: and total integrated luminosities, $\mathcal{L}$, used for the W-pair
758: analysis.}
759: \label{tab:luminosities}
760: \end{center}
761: \end{table}
762:
763:
764: \begin{table}[p]
765: \begin{center}
766: \renewcommand{\arraystretch}{1.2}
767: \begin{tabular}{| l | c | c |}
768: \hline
769: Process & $N_{\mathrm{data}}$ & $N_{\mathrm{bg}}$ \\
770: \hline
771: $\WW\rightarrow\LNLN$ & $ \pz 207$ & $\pz \pz 28.1$ \\
772: $\WW\rightarrow\QQEN$ & $1263$ & $\pz 118.1$ \\
773: $\WW\rightarrow\QQMN$ & $1187$ & $\pz 118.0$ \\
774: $\WW\rightarrow\QQTN$ & $1017$ & $\pz 348.4$ \\
775: $\WW\rightarrow\QQQQ$ & $5219$ & $1109.2$ \\
776: %
777: \hline
778: $\PW\EN,~\PW\rightarrow\LN$ & $ \pz 121$ & $\pz \pz 10.4$ \\
779: $\PW\EN,~\PW\rightarrow\QQ$ & $ \pz 584$ & $\pz 342.2$ \\
780: \hline
781: \end{tabular}
782: \caption[]{
783: Numbers of selected data events, $N_{\mathrm{data}}$, and expected
784: background events, $N_{\mathrm{bg}}$, for the W-pair analysis
785: at $\RS= 189 - 209~\GeV$
786: and for the single-W analysis at $\RS= 161 - 209~\GeV$. }
787: \label{tab:events-all}
788: \end{center}
789: \end{table}
790:
791:
792: \begin{table}[p]
793: \begin{center}
794: \renewcommand{\arraystretch}{1.2}
795: \begin{tabular}{|l| r| r| r| r| r| r|}
796: %
797: \hline
798: %
799: \multicolumn{1}{|c|}{Source}&
800: \multicolumn{6}{c|}{Systematic uncertainty} \\ \cline{2-7}
801: \multicolumn{1}{|c|}{of uncertainty}&
802: \multicolumn{1}{c|}{$\; \giZ$} & \multicolumn{1}{c|}{$\kg$} &
803: \multicolumn{1}{c|}{$\Lg$} & \multicolumn{1}{c|}{$\; \gvZ$} &
804: \multicolumn{1}{c|}{$\kZ$} & \multicolumn{1}{c|}{$\LZ$} \\
805: %
806: \hline
807: %
808: Uncertainty on $\sigma_{\PW\PW}$ & 0.003 & 0.018 & 0.006
809: & 0.03 & 0.009 & 0.014 \\
810: %
811: $\cal{O}(\alpha)$ corrections on $\cos\Theta_{\mathrm{W}^-}$ & 0.004 & 0.004 & 0.003
812: & 0.01 & 0.011 & 0.007 \\
813: %
814: Background modelling & 0.005 & 0.019 & 0.006
815: & 0.02 & 0.009 & 0.014 \\
816: %
817: Jet charge confusion & 0.001 & 0.006 & 0.002
818: & $<0.01$ & 0.002 & 0.005 \\
819:
820: Lepton charge confusion & 0.003 & 0.013 & 0.007
821: & $\pz0.01$ & 0.005 & 0.009 \\
822: %
823: Jet and lepton measurement & 0.001 & 0.003 & 0.002
824: & 0.01 & 0.002 & 0.004 \\
825: %
826: Monte Carlo statistics & 0.012 & 0.010 & 0.014
827: & 0.02 & 0.016 & 0.007 \\
828: %
829: ISR and FSR & 0.001 & 0.016 & 0.001
830: & 0.01 & 0.002 & 0.002 \\
831: %
832: W mass and width & 0.001 & 0.005 & 0.002
833: & 0.01 & 0.002 & 0.004 \\
834: %
835: Fragmentation & 0.003 & 0.002 & 0.001
836: & 0.02 & 0.004 & 0.001 \\
837: %
838: Bose Einstein correlations & 0.001 & 0.001 & 0.001
839: & $<0.01$ & 0.001 & 0.003 \\
840: % %
841: Colour reconnection & 0.001 & 0.004 & 0.001
842: & 0.02 & 0.002 & 0.003 \\
843: % %
844: \hline
845: %
846: Total systematic uncertainty & 0.015 & 0.039 & 0.017
847: & 0.05 & 0.024 & 0.023 \\
848: %
849: \hline
850: \end{tabular}
851: \caption[]{Systematic uncertainties on TGC's determined from semi-leptonic
852: and hadronic W-pairs.
853: For each coupling the uncertainties are obtained in one-parameter fits, by setting all other
854: couplings to their Standard Model values. The constraints
855: $\kZ = \giZ - \tan^2 \theta_{W} (\kg-1)$
856: and $\LZ = \Lg$ are imposed on the first three couplings.
857: }
858: \label{tab:sysww}
859: \end{center}
860: \end{table}
861: %
862:
863: \begin{table}[p]
864: \begin{center}
865: \renewcommand{\arraystretch}{1.3}
866: \begin{tabular}{|c|c|c|c|}
867: \hline
868: Coupling & $\giZ$ & $\kg$ & $\Lg$ \\
869: %
870: \hline
871: % %
872: $\nu_e \bar{\nu}_e\gamma$ 189--209 $\GeV$ & &
873: $0.7\pm 0.5 \pm 0.3$ &
874: $0.3\pm 0.7 \pm 0.4$ \\
875: %
876: %
877: We$\nu$ ~161--209 $\GeV$& & $1.179\amp{0.080}{0.076}\pm 0.068$ &
878: $0.30\amp{0.19}{0.11}\pm 0.08$ \\
879: %
880: WW ~161--209 $\GeV$ & $0.966\amp{0.032}{0.034}\pm 0.015$ &
881: $0.910\amp{0.066}{0.074}\pm 0.039$ &
882: $-0.024\amp{0.033}{0.035}\pm 0.017$ \\
883: %
884: \hline
885: %
886: All channels combined & $0.966\amp{0.032}{0.034} \pm 0.015$ &
887: $1.013\amp{0.064}{0.067} \pm 0.026$ &
888: $-0.021\amp{0.034}{0.035} \pm 0.017$ \\
889: %
890: Standard Model value & 1.0 & 1.0 & 0.0 \\
891: \hline
892: %
893: \multicolumn{4}{}{} \\
894: \hline
895: %
896: Coupling & $\gvZ$ & $\kZ$ & $\LZ$ \\
897: %
898: \hline
899: %
900: %
901: WW 189-209 $\GeV$ & $0.00\pm0.13 \pm 0.05$ &
902: $0.924\amp{0.056}{0.059} \pm 0.024$ &
903: $-0.088\amp{0.057}{0.060} \pm 0.023$ \\
904: %
905: Standard Model value & 0.0 & 1.0 & 0.0 \\
906: %
907: \hline
908: %
909: \end{tabular}
910: \caption{Results of one-parameter fits to the TGC's $\giZ$, $\kg$,
911: $\Lg$, $\gvZ$, $\kZ$ and $\LZ$ based on single-photon events,
912: single-W events and hadronic and semi-leptonic
913: W-pairs, and their combination. The single-W
914: results are obtained after removing events
915: selected as W-pair.
916: All results are at 68\% confidence level.
917: For each TGC fit, all other parameters are set to their Standard Model values;
918: for the set $\giZ$, $\kg$ and $\Lg$ the constraints $\kZ = \giZ -
919: \tan^2 \theta_{W} (\kg-1)$ and $\LZ = \Lg$ are imposed. The first
920: uncertainty is statistical, the second systematic.}
921: \label{tab:res1d1}
922: \end{center}
923: \end{table}
924:
925:
926:
927: \begin{table}[p]
928: \begin{center}
929:
930: \renewcommand{\arraystretch}{1.35}
931: \begin{tabular}{|c|c|c|c|ccc|}
932: \hline
933: Fit & Standard & \multicolumn{2}{|c|}{Results} &
934: \multicolumn{3}{|c|}{correlation coeffs.}\\ \cline{3-7}
935: parameter & Model & \hspace{4mm}(68\% CL)\hspace{4mm}
936: & (95\% CL) & $\giZ$ & $\kg$ & $\Lg$ \\
937: \hline
938: \multicolumn{7}{|c|}{two-parameter fits} \\
939: \hline
940: %
941: $\giZ$ & 1.0 & $\phantom{+}0.912^{+0.054}_{-0.044}$ &
942: $ [\phantom{+}0.83, 1.02]$ & $\phantom{+}1.00$ & $-0.71$ & \\
943: $\kg$ & 1.0 & $\phantom{+}1.162^{+0.124}_{-0.129}$ &
944: $ [\phantom{+}0.94, 1.38]$ & & $\phantom{+}1.00$ & \\
945: %
946: \hline
947: %
948: $\kg$ & 1.0 & $\phantom{+}1.061^{+0.089}_{-0.082}$ &
949: $ [\phantom{+}0.91, 1.24]$ & & $\phantom{+}1.00$ & $-0.42$ \\
950: $\Lg$ & 0.0 & $-0.052^{+0.044}_{-0.042}$ &
951: $ [-0.13, 0.03]$ & & & $ \phantom{+}1.00$ \\
952: %
953: \hline
954: %
955: $\giZ$ & 1.0 & $\phantom{+}0.979^{+0.066}_{-0.065}$ &
956: $ [\phantom{+}0.86, 1.10]$ & $\phantom{+}1.00$ & & $-0.82$ \\
957: $\Lg$ & 0.0 & $-0.025^{+0.071}_{-0.065}$ &
958: $ [-0.14, 0.11]$ & & & $\phantom{+}1.00$ \\
959: %
960: \hline
961: %
962: \multicolumn{7}{|c|}{three-parameter fit} \\
963: %
964: \hline
965: $\giZ$ & 1.0 & $\phantom{+}0.91^{+0.10}_{-0.07}$ &
966: $ [\phantom{+}0.80, 1.08]$ & $\phantom{+}1.00$ & $-0.74$ & $-0.80$ \\
967: $\kg$ & 1.0 & $\phantom{+}1.15^{+0.13}_{-0.14}$ &
968: $ [\phantom{+}0.92, 1.38]$ & & $\phantom{+} 1.00$ & $ \phantom{+}0.44$ \\
969: $\Lg$ & 0.0 & $\phantom{+}0.01^{+0.07}_{-0.08}$ &
970: $ [-0.14, 0.14]$ & & & $\phantom{+}1.00$ \\
971: %
972: \hline
973: \end{tabular}
974: %
975: \caption[]{
976: Results of two- and three-parameter fits of the couplings
977: $\kg$, $\Lg$ and $\giZ$ with the constraints
978: $\kZ = \giZ - \tan^2 \theta_{W} (\kg-1)$ and $\LZ = \Lg$;
979: all other couplings are set to their Standard Model values.
980: Correlation coefficients are also shown.
981: Systematic uncertainties are included.}
982: \label{tab:res2-3d}
983: \end{center}
984: \end{table}
985:
986:
987:
988:
989: \clearpage
990:
991: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
992: % FIGURES
993: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
994:
995: \begin{figure}[p]
996: \begin{center}
997: \epsfig{file=qqqq_cosw.eps,width=0.49\linewidth}\hfill %\hspace*{1.5cm}
998: \epsfig{file=qqln_cosw.eps,width=0.49\linewidth}\\
999: \caption[]{Distributions of the reconstructed W$^-$ production angle,
1000: $\cos\Theta_{\PW^-}$, in
1001: a) hadronic and b) semi-leptonic W-pair events.
1002: Data are shown, together
1003: with the expectations for the Standard Model and for anomalous
1004: values of TGC's.}
1005: \label{fig:wwctw}
1006: \end{center}
1007: \end{figure}
1008:
1009: \begin{figure}[p]
1010: \begin{center}
1011: \epsfig{file=qqqq_cosq.eps,width=0.49\linewidth}\hfill %\hspace*{1.5cm}
1012: \epsfig{file=qqqq_phiq.eps,width=0.49\linewidth}\\
1013: \caption[]{Distributions of the reconstructed W decay angles in
1014: hadronic W-pair events, a) $| \cos\theta_q|$ and b) $\phi_q$.
1015: Distributions for W$^+$ and W$^-$ bosons are
1016: combined.
1017: Data are shown, together with the
1018: expectations for the Standard Model and for anomalous values
1019: of the TGC's.}
1020: \label{fig:wwdecayqqqq}
1021: \end{center}
1022: \end{figure}
1023:
1024: \begin{figure}[p]
1025: \begin{center}
1026: \epsfig{file=qqln_cosl.eps,width=0.49\linewidth}\hfill %\hspace*{1.5cm}
1027: \epsfig{file=qqln_phil.eps,width=0.49\linewidth}\\
1028: \vskip 0.5cm
1029: \epsfig{file=qqln_cosq.eps,width=0.49\linewidth}\hfill %\hspace*{1.5cm}
1030: \epsfig{file=qqln_phiq.eps,width=0.49\linewidth}\\
1031: \caption[]{Distributions of the reconstructed W decay angles in
1032: semi-leptonic events: the production angles of the lepton,
1033: a) $\cos\theta_{\ell}$ and b) $\phi_\ell$, and
1034: the decay angles of W bosons decaying into hadrons,
1035: c) $|\cos\theta_q|$ and d) $\phi_q$.
1036: Data are shown, together with the
1037: expectations for the Standard Model and for anomalous values of the TGC's. }
1038: \label{fig:wwdecayqqln}
1039: \end{center}
1040: \end{figure}
1041:
1042:
1043: \begin{figure}[p]
1044: \begin{center}
1045: \epsfig{file=fig_elep_noww.eps,width=0.49\linewidth}\hfill
1046: \epsfig{file=fig_nnout_noww.eps,width=0.49\linewidth}\\
1047: \caption[]{Distribution of a) the energy spectrum of the lepton in leptonic single-W
1048: events and b) the output
1049: of the neural network used in
1050: the selection of hadronic single-W events.
1051: }
1052: \label{fig:swdist}
1053: \end{center}
1054: \end{figure}
1055:
1056:
1057: \begin{figure}[p]
1058: \begin{center}
1059: \epsfig{file=legend_lik.eps,width=0.92\linewidth} \\
1060: \epsfig{file=gz_lik.eps,width=0.405\linewidth} \hspace*{1cm}
1061: \epsfig{file=kg_lik.eps,width=0.405\linewidth} \\
1062: \epsfig{file=lg_lik.eps,width=0.405\linewidth} \hspace*{1cm}
1063: \epsfig{file=g5z_lik.eps,width=0.405\linewidth} \\
1064: \epsfig{file=kz_lik.eps,width=0.405\linewidth} \hspace*{1cm}
1065: \epsfig{file=lz_lik.eps,width=0.405\linewidth}
1066: \caption[]{
1067: Change in negative log-likelihoods with respect to their minimum
1068: for one-parameter TGC fits. Systematic uncertainties are included.
1069: Contributions from different channels are indicated.
1070: }
1071: \label{fig:logl1d}
1072: \end{center}
1073: \end{figure}
1074:
1075: \begin{figure}[p]
1076: \begin{center}
1077: \epsfig{file=dgz_dkg_cont.eps,width=0.45\linewidth} \hspace*{1cm}
1078: \epsfig{file=lg_dkg_cont.eps,width=0.45\linewidth}\\
1079: \epsfig{file=dgz_lg_cont.eps,width=0.45\linewidth} \hspace*{1cm}
1080: \epsfig{file=legend_cont.eps,width=0.45\linewidth}
1081: \caption[]{Comparison of single- and multi-parameter TGC fits.
1082: The vertical and horizontal lines are the 68\% confidence level
1083: intervals when all couplings but one are fixed to their
1084: Standard Model values, indicated by a star.
1085: The shaded areas represent the 68\% confidence level regions
1086: for the two-parameter fits to the TGC's: a) $\giZ$ and $\kg$ with $\Lg=0$,
1087: b) $\Lg$ and $\kg$ with $\giZ=1$ and c) $\giZ$ and $\Lg$ with $\kg=1$.
1088: The 95\% confidence level contours are also given as solid lines.
1089: The dashed lines represent two-dimensional projections
1090: of the three-parameter log-likelihoods.
1091: The constraints $\kZ = \giZ - \tan^2 \theta_{W} (\kg-1)$ and $\LZ =
1092: \Lg$ are imposed and all other couplings are set to their
1093: Standard Model values. Systematic uncertainties are included.
1094: }
1095: \label{fig:ac-ndlnl}
1096: \end{center}
1097: \end{figure}
1098:
1099: \end{document}
1100:
1101:
1102:
1103:
1104:
1105:
1106:
1107: