1: \section{Analyses}
2:
3: The measurements presented in this paper are mainly based on the following data samples.
4: The analyses of \babar\ are based on $(87.9 \pm 1.0) \times 10^{6}$ $B\bar B$ pairs collected
5: between $1999$ and $2002$ with the \babar\ detector at the PEP-II asymmetric-energy $B$
6: Factory at SLAC. This corresponds to an integrated luminosity of approximately
7: $81 \invfb$ at the $\Upsilon(4S)$ resonance.
8: The results from Belle are based on $78 \invfb$ data sample which corresponds to
9: $85 \times 10^{6}$ $B\bar B$ pairs. The Belle detector\cite{nimbelle} is a large-solid-angle
10: general purpose spectrometer at the KEKB asymmetric-energy $e^+e^-$ storage ring.
11: The current CLEO data set is made up from the full CLEO II and CLEO III data samples totaling
12: $15.3 \invfb$. CLEO\cite{nimcleo} is a general purpose solenoidal magnet detector operated at the
13: Cornell Electron Storage Ring (CESR), a symmetric-energy storage ring tuned to provide center of
14: mass energies near the $\Upsilon(4S)$ resonance.
15: %comment: cleoiii = 6.18 ON, 2.24 OFF, cleoii = 9.13 ON, 4.35 OFF
16:
17: Hadronic events are selected based on track multiplicity and event topology.
18: Tracks are identified as pions or kaons using the various Particle Identification
19: (PID) techniques described below.
20: Candidate $\KS$ mesons are reconstructed from pairs of oppositely charged tracks that
21: form a well-measured vertex and have an invariant mass within a defined window around
22: the nominal \KS\ mass~\cite{PDG}. The candidate must have a displaced vertex and flight
23: direction consistent with a $\KS$ originating from the interaction point.
24: Candidate \piz\ mesons are formed from pairs of photons with an invariant mass within
25: a defined window around the nominal \piz\ mass. The \piz\ candidates are then
26: kinematically fitted with their mass constrained to the nominal \piz\ mass~\cite{PDG}.
27:
28: \input Bona-fig_twobody
29:
30: Reconstructed $B$ candidates are identified using two kinematic variables:
31: the beam-energy substituted mass (or beam-energy constrained mass)
32: $m_{ES} = \sqrt{(E_{\rm beam}^{\rm CM})^2 -(p_{B}^{\rm CM})^2}$ and the energy
33: difference $\Delta E = E_{B}^{\rm CM}-E_{\rm beam}^{\rm CM}$, where $E_{\rm beam}^{\rm CM}$
34: is the beam energy, $p_{B}^{\rm CM}$ and $E_{B}^{\rm CM}$ are the momentum and
35: the energy of the reconstructed $B$ meson in the center-of-mass system (CM).
36:
37: In these charmless modes, the largest source of background comes from random
38: combinations of tracks and neutrals produced in the $e^+e^- \rightarrow q\bar q$
39: continuum (where $q=u$, $d$, $s$ or $c$). These backgrounds are suppressed using
40: the event topology. In the CM frame this background typically exhibits
41: a two-jet structure that can produce two high momentum, nearly
42: back-to-back particles, in contrast to the spherically symmetric
43: nature of the low momentum $\Upsilon(4S) \to b\bar b$ events.
44: This topology difference is exploited by making use of event-shape quantities.
45: Various techniques and variables are used.
46:
47: In the \babar\ analyses, a preliminary requirement is applied on the angle
48: $\theta_{sph}$~\cite{spheric} between the sphericity axes, evaluated in the CM frame,
49: of the $B$ candidate and the remaining tracks and photons in the event.
50: The distribution of the absolute value of $\cos\theta_{sph}$ is strongly peaked near $1$
51: for continuum events and is approximately uniform for \BB\ events.
52:
53: Another similarly powerful variable is the flight direction of the
54: $B$ candidate given by $\cos\theta_B = \hat{\bf p}\cdot \hat{\bf z}$
55: where ${\bf p}$ is the vector sum of the daughter momenta and $\bf z$ is
56: the beam axis. Since the vector $\Upsilon(4S)$ is produced in $e^+e^-$ annihilation
57: it has a polarization $J_z=\pm 1$, and the subsequent flight direction of
58: the pseudo-scalar $B$ mesons is distributed as
59: $|Y^{\pm 1}_1(\theta,\phi)|^2\sim\sin^2\theta = 1-\cos^2\theta$.
60: Background events are flat in this variable.
61:
62: Another quantity largely used in the charmless analyses is a Fisher discriminant
63: ${\cal F}$ which consists of an optimized linear combination of
64: variables that distinguish signal from background~\cite{fisher}.
65:
66: \babar\ analyses use a two-variable Fisher discriminant with $\sum_i p_i$
67: and $\sum_i p_i|\cos \theta_i|^2$ where $p_i$ is the momentum and
68: $\theta_i$ is the angle with respect to the thrust axis of the $B$
69: candidate, both in the CM frame, for all tracks and neutral clusters
70: not used to reconstruct the $B$ meson.
71:
72: Belle's technique consists of forming signal and background
73: likelihood functions, ${\cal L}_S$ and ${\cal L}_{BG}$,
74: from two variables. One is a Fisher
75: discriminant determined from six modified Fox-Wolfram
76: moments~\cite{foxbelle} and the other is $\cos\theta_B$.
77: Requirements are imposed on the likelihood ratio
78: $LR = {\cal L}_S/({\cal L}_S+{\cal L}_{BG})$ for candidate events.
79:
80: CLEO builds its Fisher discriminant as a linear combination of
81: fourteen variables including the direction of the thrust
82: axis of the candidate with respect to the beam axis,
83: $\cos\theta_{thr}$, and the nine conical bins of a ``Virtual Calorimeter''.
84: They are the scalar sum of the momenta of all tracks and photons
85: (excluding the $B$ candidate daughters) flowing into nine concentric
86: cones centered on the thrust axis of the $B$ candidate, in the CM frame.
87: Each cone subtends an angle of $10^\circ$ and is folded to combine
88: the forward and backward intervals.
89: In addition, the momentum of the highest momentum electron, muon, kaon,
90: and proton are used as inputs to the Fisher discriminant taking advantage of
91: the high quality particle identification in CLEO III.
92:
93: In the case of a candidate mode involving one or more charged pions or
94: kaons, such as $B\to K\pi$ or $B\to \pi \pi^0$, each charged track must be
95: positively identified as $K$ or $\pi$.
96:
97: \input Bona-tab_twobody
98:
99: PID in \babar\ analyses is accomplished with the Cherenkov angle measurement from a
100: detector of internally reflected Cherenkov light (DIRC).
101: The final fit to the data includes the normalized Cherenkov residuals
102: $\left( \theta_c - \theta_c^\pi\right)/\sigma_{\theta_c}$ and
103: $\left( \theta_c - \theta_c^K\right)/\sigma_{\theta_c}$, where $\theta_c$ is
104: the measured Cherenkov angle of the charged primary daughter,
105: $\sigma_{\theta_c}$ is its error, and
106: $\theta_c^\pi\,(\theta_c^K)$ is the expected value for a pion(kaon).
107: The latter two quantities are measured separately for negatively and
108: positively charged pions and kaons, from a sample of $\Dz\to \Km\pip$
109: originating from $\Dstarp$ decays.
110:
111: PID in Belle experiment is based on the light yield in the aerogel
112: Cherenkov counter (ACC) and $dE/dx$ measurements on the central drift chamber.
113: For each hypothesis ($K$ or $\pi$), the $dE/dx$ and ACC probability density
114: functions are combined to form likelihoods, ${\cal L}_K$ and ${\cal L}_\pi$.
115: $K$ or $\pi$ mesons are distinguished by requirements on the likelihood ratio
116: ${\cal L}_K/({\cal L}_K+{\cal L}_\pi)$.
117:
118: The $K/\pi$ identification in CLEO relies on the pattern of Cherenkov
119: photon hits in the RICH detector fitted to both a kaon and pion hypothesis,
120: each with its own likelihood ${\cal L}_K$ and ${\cal L}_\pi$.
121: Calibrated $dE/dx$ information from the drift chamber is used to compute a
122: \chisq\ for kaon and pion hypotheses. The RICH and $dE/dx$ results are
123: combined to form an effective \chisq\ difference,
124: $\Delta_{K\pi} = -2\ln{\cal L}_K + 2\ln{\cal L}_\pi + \chisq_K - \chisq_\pi$.
125: Kaons are identified by $\Delta_{K\pi} < \delta_K$ and pions by $-\Delta_{K\pi} <
126: \delta_\pi$, with values of $\delta_K$ and $\delta_\pi$ chosen to yield
127: $(90\pm 3)$\% efficiency as determined in an independent study of tagged
128: kaons and pions obtained from the decay $D^{*+}\to\pi^+D^0~(D^0\to
129: K^-\pi^+)$.
130:
131: Finally PV modes allow for the use of the vector particle helicity angle $\theta_h$
132: defined as the angle between the direction of one of $h$ daughters
133: in the $h$ rest frame and the direction of $h$ in the $B$ rest frame.
134:
135: