1: \documentclass[fleqn,twoside]{article}

2: \usepackage{espcrc2}

3: \usepackage{graphicx}

4: \usepackage[figuresright]{rotating}

5: 

6: \newcommand{\ttbs}{\char'134}

7: \newcommand{\AmS}{{\protect\the\textfont2 A\kern-.1667em\lower.5ex\hbox{M}\kern-.125emS}}

8: \newcommand{\nue}{\nu_{e}}

9: \newcommand{\num}{\nu_{\mu}}

10: \newcommand{\nut}{\nu_{\tau}}

11: \newcommand{\nus}{\nu_{sterile}}

12: 

13: \hyphenation{author another created financial paper re-commend-ed Post-Script}

14: 

15: \title{Atmospheric neutrino oscillations with MACRO}

16: 

17: \author{M. Sioli\address[MCSD]{Physics Department of the University and INFN,

18: Viale C. Berti Pichat 6/2, I-40127, Bologna, Italy},

19: for the MACRO Collaboration\thanks{List of authors and institutions: see Ref \cite{macro}}}

20:        

21: \begin{document}

22: 

23: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

24: %\vspace{-6.35cm}

25: %\hspace{9cm}

26: %\bf{DFUB 12/2002}

27: %

28: %\hspace{9cm}

29: %\bf{Bologna, 04/11/02}

30: %\vspace{4.5cm}

31: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

32: 

33: \begin{abstract}

34: \vspace{-3.5cm}

35: \hspace{12cm}

36: \bf{DFUB 12/2002}

37: 

38: \hspace{12cm}

39: \bf{Bologna, 04/11/02}

40: \vspace{2.2cm}

41: \begin{center}

42:   {\bf Contribution to the XII ISVHECRI, CERN, Geneva, 15-20 July 2002}

43: \end{center}

44: \normalfont

45: We present the latest results on the study of atmospheric neutrino oscillations 

46: with the MACRO detector at Gran Sasso. Two sub-samples of events have been analysed,

47: both in terms of absolute flux and zenith angle distribution:

48: high energy events (with $\langle E_{\nu} \rangle \simeq$ 50 GeV) and

49: low energy events (with $\langle E_{\nu} \rangle \simeq$ 4 GeV). 

50: The high energy sample has been used also to check the $\num \leftrightarrow \nu_{sterile}$ 

51: oscillation hypothesis and to estimate neutrino energies using Multiple Coulomb Scattering (MCS) informations.

52: All these analyses are mutually consistent and strongly favour the $\num \leftrightarrow \nut$ 

53: oscillation hypothesis with maximal mixing and $\Delta m^{2} = 2.5 \cdot 10^{-3}$eV$^{2}$.

54: 

55: \vspace{1pc}

56: \end{abstract}

57: 

58: \maketitle

59: 

60: \section{INTRODUCTION}

61: The MACRO detector \cite{macro} was located in the Gran Sasso Laboratory (Italy) and allowed the study

62: of atmospheric neutrinos detecting upgoing muons produced in CC interactions inside or 

63: around the detector. It used scintillation counters for tagging events by 

64: time-of-flight (TOF) measurements, and limited streamer tubes for tracking.

65: Three categories of $\num$-induced muon events were studied (see Fig. \ref{f:topo}): 

66: 1) upthroughgoing muons, produced in the rock below the detector and crossing 

67: the whole apparatus, with an average energy $\langle E_{\nu} \rangle \simeq$ 50 GeV; 

68: 2) internal upgoing (IU) muons, produced inside and leaving the detector from above;

69: 3) upgoing stopping muons (UGS), produced below and stopping inside the detector plus

70: internal downgoing (ID) muons, produced inside and leaving the detector from below.

71: Since at least two scintillation counters are needed to perform a TOF measurement, the events of 

72: samples UGS and ID are studied on a topological basis and are indistinguishable. 

73: Therefore, they are studied together.

74: The average neutrino energy for samples 2 and 3 is about $\simeq$ 4 GeV, and we collectively 

75: refer to these events as to the Low Energy (LE) sample \cite{le}, while events from sample 1 belong to the 

76: High Energy (HE) sample \cite{he}.

77: 

78: \section{HIGH ENERGY SAMPLE}

79: Events in the HE sample have been selected by measuring the TOF between two layers of scintillators.

80: For events crossing three layers of scintillators, a linear fit of the times as a function of 

81: the path lengths has been performed, in order to reduce possible fake events. A detailed tuning of

82: the TOF measurements provided a reduction factor of $\sim 10^{7}$ of the downgoing atmospheric

83: muons. Other sources of background have been reduced by applying a cut on the matching positions 

84: between scintillators and streamer tubes. 

85: Soft upgoing hadrons coming from photonuclear interactions of muons outside the detector

86: could mimic an upgoing muon: to reduce this background we require that each upgoing muon crosses

87: at least 200 g/cm$^{2}$ of absorber in the lower part of the detector \cite{spurio}. 

88: Muons coming from particular

89: azimuthal regions, where the rock amount is not sufficient to reduce the number of downgoing muons, 

90: have been discarded. 863 events survived these cuts (809 after background subtraction), corresponding

91: to a livetime of 6.16 yrs.

92: 

93: Monte Carlo (MC) simulation of upgoing muon events has been performed using the $\nu$ flux computed

94: by the Bartol group \cite{bartol}, neutrino cross sections from Ref. \cite{cs} and muon energy loss

95: in the rock from calculations of Ref. \cite{eneloss}. The overall theoretical uncertainty

96: on the absolute upgoing muon flux is 17\%, while the systematic error on the shape of the angular

97: distribution is 5\%.

98: 

99: \begin{figure}[t!]

100: \includegraphics*[width=17pc]{topo.eps}

101: \vspace{-0.7cm}

102: \caption{Schematic picture of the different topologies of $\nu$-induced events in MACRO.

103: From the left: 1) upward throughgoing muons, 2) internal upgoing (IU) muons, 

104: 3) upgoing stopping muons (UGS) and 4) internal downgoing (ID) muons. 

105: Samples (3+4) are studied together, see text.}

106: \label{f:topo}

107: \end{figure}

108: 

109: Compared to MC predictions the number of measured events is small and the shape of the angular

110: distribution is different, Fig. \ref{f:passanti}.

111: The ratio between the observed number of events and the MC prediction is 

112: $R_{HE} = 0.721 \pm 0.026_{stat} \pm 0.043_{sys} \pm 0.123_{th}$.

113: The shapes of the experimental distribution and MC predictions assuming no oscillations, analysed in 

114: terms of $\chi^2$, give an agreement probability of 0.2\%. 

115: Assuming $\num \leftrightarrow \nut$ oscillations, the angular distribution and the absolute number of events

116: give a best fit for maximal mixing and $\Delta m^{2} = 2.5 \cdot 10^{-3}$eV$^{2}$, 

117: with a $\chi^2$ probability of 66\%. 

118: The result of the fit is shown as a solid line in Fig. \ref{f:passanti}.

119: 

120: \begin{figure}[t!]

121: \includegraphics*[width=17pc]{passanti.eps}

122: \vspace{-0.7cm}

123: \caption{Angular distribution of upthroughgoing muons. Black points are data, with statistical

124: and systematic errors summed in quadrature. Shaded region is Monte Carlo prediction, assuming 

125: no oscillations and with a 17\% error on the scale. The solid line is the Monte Carlo prediction

126: assuming $\num \leftrightarrow \nut$ oscillations with maximal mixing and 

127: $\Delta m^{2} = 2.5 \cdot 10^{-3}$eV$^{2}$.}

128: \label{f:passanti}

129: \end{figure}

130: 

131: \subsection{$\num \leftrightarrow \nu_{sterile}$ oscillations}

132: The weak potential of $\nue$ and $\nus$ with matter is different from the $\num$ and $\nut$ potentials.

133: On the other hand, $\num$ and $\nut$ potentials are equal: this difference translates in a distortion of the oscillation

134: pattern (and hence of the angular distribution) in the $\num \leftrightarrow \nus$ oscillation with respect

135: to the $\num \leftrightarrow \nut$ case. Matter effects become important when 

136: $E_{\nu} / |\Delta m^2| \geq 10^{-3}$ GeV/eV$^{2}$, i.e. for HE events.

137: A detailed analysis \cite{sterile} has shown that the best estimator to disentangle the two hypotheses is the

138: ratio $R$ between the number of events with -1 $< \Theta <$ -0.7 and the number of events with 

139: -0.4 $< \Theta <$ 0, where $\Theta$ is the zenith angle.

140: In this ratio, most of the theoretical uncertainties cancel and the overall uncertainty,

141: combined with the experimental systematic error, is $\sim$ 7\%.

142: The measured ratio is $R = 1.48 \pm 0.13_{stat} \pm 0.10_{sys}$. This result is shown in Fig. \ref{f:sterile}

143: together with the MC prediction, as a function of the $\Delta m^2$. In the best fit point, the expected

144: values for $\num \leftrightarrow \nut$ and $\num \leftrightarrow \nus$ are $R_{\tau} = 1.72$

145: and $R_{\tau} = 2.16$ respectively. The ratio of the probabilities to obtain values as low as the observed

146: one is $R_{prob} = 157$, therefore $\num \leftrightarrow \nus$ can be excluded with respect $\num \leftrightarrow \nus$

147: at 99\% CL.

148: 

149: \begin{figure}[t!]

150: \includegraphics*[width=17pc]{sterile.eps}

151: \vspace{-0.7cm}

152: \caption{ratio of the number of events with -1 $< \theta <$ -0.7 with respect to the number of events with 

153: -0.4 $< \theta <$ 0. The black point is the measured value (with error bar), the solid lines are the

154: prediction for $\num \leftrightarrow \nut$ and $\num \leftrightarrow \nus$ cases.}

155: \label{f:sterile}

156: \end{figure}

157: 

158: \subsection{Energy estimate by means MCS}

159: The limited streamer tube system of the MACRO detector has been used in order to estimate muon energies

160: (and hence neutrino energies) using MCS informations. A first analysis, which used

161: streamer tubes in digital mode (with a spatial resolution of $\sigma \simeq 1$ cm),

162: successfully showed the feasibility of the method \cite{bakari}. In order to improve 

163: the sensitivity of the analysis to higher neutrino energies, we used the streamer tubes in

164: drift mode, improving the resolution by a factor $\sim$ 3.5 \cite{tecnico}. Two dedicated

165: test beams at CERN PS/SPS checked electronics, detector performances and analysis tools. 

166: Starting from MCS sensitive variables, a MC trained neural network was used for 

167: muon (and neutrino) energy reconstruction.

168: These results allowed to increase the sensitivity of the analysis up to neutrino energies of $\sim$ 100 GeV.

169: We separated the HE sample in four different energy regions: the results show that the agreement between data 

170: and MC predictions follows the energy dependence expected in the oscillation hypothesis with the

171: parameters given in the previous section \cite{scapeio}.

172: 

173: The distance travelled by neutrinos inside the Earth (reconstructed with a precision of $\sim$ 3\%)

174: was used to estimate the $L/E_{\nu}$ on a event-by-event basis. This is shown in Fig. \ref{f:lsue},

175: where the ratio $DATA/MC(no osc)$ as a function of the estimated log$_{10} L/E_{\nu}$ is plotted.

176: Errors include uncertainties both on the flux and on the shape. MC prediction is also reported,

177: with the same parameters of the standard analysis, showing that the trend of the data is the one we 

178: expect in the $\num \leftrightarrow \nut$ oscillation hypothesis.

179: 

180: \begin{figure}[t!]

181: \includegraphics*[width=17pc]{lsue.eps}

182: \vspace{-0.7cm}

183: \caption{Ratio $DATA/MC(no-osc)$ as a function of the estimated $log_{10} L/E_{\nu}$ (black points).

184: The solid line is the ratio $MC (sin^{2} 2\theta =1, \Delta m^{2} = 2.5 \cdot 10^{-3}$eV$^{2}$)/$MC(no-osc)$,

185: with corresponding systematic error (see text). The black square refers to IU events.}

186: \label{f:lsue}

187: \end{figure}

188: 

189: \section{LOW ENERGY SAMPLE}

190: IU and ID+UGS events are produced in CC interactions inside the detector, with a small contamination 

191: of NC and $\nue$ events ($\sim$ 13\% and $\sim$ 10\% respectively). IU events have been selected by

192: TOF measurement and by topological criteria, e.g. the requirement of the interaction vertex in the

193: fiducial volume of the detector. The ID+UGS sample has been selected only with topological criteria.

194: Since no timing informations were available for these events, particular attention has been devoted

195: in order to reject events not perfectly confined in the detector. These goals have been reached also

196: by means of visual scanning with the Event Display.

197: After these selections, we remained with 154 and 262 events for the IU and ID+UGS, corresponding

198: to 5.8 yrs and 5.6 yrs of livetime respectively. In Fig. \ref{f:semicont} we show the angular

199: distributions of these two samples and the MC predictions with and without the oscillation hypotheses.

200: A conservative value (25\%) of the theoretical uncertainty on the absolute scale is reported.

201: There is a good agreement between data and MC prediction with the oscillation parameters given by

202: the HE sample analysis.

203: 

204: \begin{figure}[t!]

205: \includegraphics*[width=17pc]{semicont.eps}

206: \vspace{-0.7cm}

207: \caption{Angular distribution of IU events (a) and ID+UGD events (b). Black points are data, with statistical

208: and systematic errors summed in quadrature. Dashed lines are Monte Carlo predictions, assuming 

209: no oscillations and with a 17\% error on the scale (shaded region). Solid lines are the Monte Carlo predictions

210: assuming $\num \leftrightarrow \nut$ oscillations with maximal mixing and $\Delta m^{2} = 2.5 \cdot 10^{-3}$eV$^{2}$.}

211: \label{f:semicont}

212: \end{figure}

213: 

214: \section{CONCLUSIONS}

215: We presented different and independent analyses performed on muon events induced by atmospheric neutrinos.

216: All the analyses are mutually consistent and strongly favour the $\num \leftrightarrow \nut$

217: oscillation hypothesis with maximal mixing and $\Delta m^{2} = 2.5 \cdot 10^{-3}$eV$^{2}$.

218: Absolute flux measurements, shape distributions and energy estimate by MCS were used in order

219: to compute allowed regions in the space of oscillation parameters. This is shown in Fig. \ref{f:contour},

220: where the corresponding contours are computed according to a $\chi^2$ analysis using the prescriptions

221: of Ref. \cite{fc}. A global analysis of all the samples is in progress.

222: 

223: \begin{figure}[t!]

224: \includegraphics*[width=17pc]{contour.eps}

225: \vspace{-0.7cm}

226: \caption{90\% CL contours for allowed region in the ($sin^{2} 2\theta, \Delta m^{2}$) plane.

227: Curves 1 and 2 refer to upthroughgoing muons using flux+angular distribution and energy

228: estimate respectively. The curve 3 refers to the LE sample.}

229: \label{f:contour}

230: \end{figure}

231: 

232: \begin{thebibliography}{9}

233: \bibitem{macro} M. Ambrosio et al. (MACRO Coll.), Nucl. Instr. Meth. A486 (2002) 663.

234: \bibitem{he} S. Ahlen et al. (MACRO Coll.), Phys. Lett. B357 (1995) 481;

235: M. Ambrosio et al. (MACRO Coll.), Phys. Lett. B434 (1998) 451.

236: \bibitem{le} M. Ambrosio et al. (MACRO Coll.), Phys. Lett. B478 (2000) 5.

237: \bibitem{spurio} M. Ambrosio et al. (MACRO Coll.), Astrop. Phys. 9 (1998) 105.

238: \bibitem{bartol} V. Agrawal et al., Phys. Rev. D53 (1996) 1314.

239: \bibitem{cs} M. Gluck et al., Z. Phys. C65 (1995) 433.

240: \bibitem{eneloss} W. Lohmann et al., CERN-EP/85-03 (1985).

241: \bibitem{sterile} M. Ambrosio et al. (MACRO Coll.), Phys. Lett. B517 (2001) 59.

242: \bibitem{bakari} D. Bakari et al., Proc. of the NATO Advanced Research Workshop, Oujda, Morocco, 2001.

243: \bibitem{tecnico} M. Ambrosio et al. (MACRO Coll.), physics/0203018.

244: \bibitem{scapeio} E. Scapparone, Proc. of the NATO Advanced Research Workshop, Oujda, Morocco, 2001;

245: M. Sioli (for the MACRO Coll.), Proc. of the 27th ICRC, Hamburg, Germany, 2001.

246: \bibitem{fc} G. J. Feldman and R. D. Cousin, Phys. Rev. D57 (1998) 3873.

247: \end{thebibliography}

248: 

249: \end{document}

250: