1: \documentstyle[12pt,psfig]{article} 

2: \textwidth=17cm

3: \textheight=22.5cm

4: \oddsidemargin -0.3cm

5: \topmargin -1.5cm

6: \parskip 0.3cm

7: \tolerance=10000

8: \parindent 0pt

9: \newcommand{\ber}{\begin{eqnarray}}

10: \newcommand{\eer}{\end{eqnarray}}

11: \newcommand{\bea}{\begin{equation}}

12: \newcommand{\eea}{\end{equation}}

13: \def\ltap{\raisebox{-.4ex}{\rlap{$\sim$}} \raisebox{.4ex}{$<$}}

14: \def\gtap{\raisebox{-.4ex}{\rlap{$\sim$}} \raisebox{.4ex}{$>$}}

15: 

16: %\renewcommand{baselinestretch}{1.4}

17: 

18: 

19: \begin{document}

20: 

21: \begin{flushright}

22: November 2002\\

23: hep-ph/0303229

24: \end{flushright}

25: \vskip 50pt

26: \begin{center}

27: {\Large \bf A note on the neutrino mass implications of the K2K experiment}\\

28: \vspace{1cm}

29: \renewcommand{\thefootnote}{\fnsymbol{footnote}}

30: {\large \sf Suvadeep Bose$^{1,\!}$\footnote{E-mail address:

31: sb416@cam.ac.uk}},

32: {\large \sf Amitava Raychaudhuri$^{2,\!}$\footnote{E-mail

33: address: amitava@cubmb.ernet.in}}

34: \vskip 10 pt

35: $^{1}${\small \it St Edmund's College, University of Cambridge,

36: Cambridge CB3 0BN, U.K.}\\

37: $^{2}${\small \it Department of Physics, University of Calcutta,\\

38: 92 Acharya Prafulla Chandra Road, Kolkata 700009, India}

39: \normalsize

40: \end{center}

41: 

42: 

43: \begin{abstract}

44: {\small 

45: The K2K experiment has presented the first results on the

46: observation of $\nu_\mu$. They show a depletion compared to

47: the expectations and are consistent with neutrino oscillations with a

48: mass-splitting in the range favoured by the Super-Kamiokande

49: atmospheric neutrino measurements. Here we examine the extent by which

50: the range of $\Delta m^2$ obtained from the K2K measurements can vary

51: due to the uncertainties in the flux, cross-section, and detector

52: efficiency.

53: } 

54: \vskip 5 pt 

55: \end{abstract}

56: 

57: \vskip 20pt

58: 

59: \centerline{\texttt{PACS Nos. 14.60.Pq, 14.60.Lm }}

60: 

61: \vskip 10pt

62: 

63: \centerline{Short title: Neutrino mass implications of the K2K experiment}

64: 

65: 

66: \vskip 10 pt

67: 

68: \renewcommand{\thesection}{\Roman{section}}

69: \setcounter{footnote}{0}

70: \renewcommand{\thefootnote}{\arabic{footnote}}

71: 

72: \section{Introduction}

73: 

74: A non-zero neutrino mass has far-reaching implications in particle

75: physics, astrophysics, and cosmology. Any evidence in its support

76: must be probed from as many different angles as possible.

77: Atmospheric neutrinos provide a strong case for non-degenerate

78: neutrino masses and mixing.  Super-Kamiokande (SK) \cite{sk},

79: MACRO \cite{macro}, and Soudan 2 \cite{soudan}  studying neutrinos

80: produced in interactions of cosmic rays with atmospheric nuclei

81: show a significant suppresion of the ratio of $\nu_{\mu}$ to

82: $\nu_{e}$ with respect to the expectation from standard

83: hadronic shower models. These observations can be explained by the

84: phenomenon of neutrino oscillations in which $\nu_{\mu}$ changes

85: to some other neutrino, {\em e.g.}, $\nu_\tau$. Such oscillations

86: can occur if neutrinos have a non-degenerate mass spectrum and if

87: the neutrino mass eigenstates are not the same as the flavour

88: eigenstates. The atmospheric neutrino data favour a mass

89: splitting $\Delta m^2 = |m_1^2 - m_2^2| \simeq 10^{-3}$ eV$^2$

90: and maximal mixing, $\theta = \pi/4$.  Neutrino oscillations have

91: also been advanced as a natural solution to the long-standing

92: solar neutrino problem, especially when the data from the SK and

93: SNO experiments are taken together.

94: 

95: K2K \cite{k2k} is the first accelerator based long-baseline

96: oscillation experiment. It is designed precisely to probe the

97: mass and mixing parameters favoured by the atmospheric

98: neutrino data. It  uses neutrinos produced by the 12 GeV proton

99: beam of the KEK accelerator laboratory which are subsequently

100: detected by the SK detector at a distance of 250 km.  The average

101: neutrino energy is $\langle E \rangle =1.3$ GeV and the intense beam is

102: $98.0\%$ pure $\nu_{\mu}$.  K2K also uses a near detector at KEK, 300m

103: downstream from the pion production target, to monitor and study the

104: neutrino beam. The experiment started taking data from June, 1999.

105: The primary oscillation search mode is $\nu_{\mu}

106: \rightarrow \nu_{x}$ oscillation (the $\nu_{\mu}$ disappearance

107: mode).

108: 

109: The first results from K2K \cite{k2k2} have been released. With

110: 3.85 $\times 10^{19}$ protons on target (p.o.t.), in the

111: 22.5 kton fiducial volume of the SK detector they observe 44

112: fully contained events in place of the (no oscillation)

113: expectation of $63.9^{+6.1}_{-6.6}$ events. This corresponds to a

114: depletion ratio 

115: \begin{equation}

116: d \equiv \frac{\rm No.~of~observed~events}{\rm No.~of~expected

117: ~events} = 0.689^{+0.079}_{-0.060} \cdot

118: \label{eq:depl}

119: \end{equation}

120: This depletion is consistent with neutrino oscillations with a

121: mass splitting in the atmospheric range and maximal

122: mixing\footnote{Based on 2.6$\times 10^{19}$ p.o.t. \cite{k2ko}

123: the depletion ratio was $d = 0.670^{+0.086}_{-0.070}\cdot$}. In

124: this note, basing ourselves on the available information, we wish

125: to examine to what extent this conclusion could be uncertain. Our

126: finding is that the result is robust.

127: 

128: When oscillations are operative, the survival probability of a

129: neutrino with energy $E$ produced a distance $L$  from a

130: detector is expressed (in the two-flavour approximation) as:

131: 

132: \bea

133: P_{\mu\mu} = P(\nu_{\mu} \rightarrow \nu_\mu) = 1 - \sin^{2}2 \theta

134: \sin^{2}\frac{1.27 \Delta m^{2}({\rm eV}^{2})L({\rm km})}{E({\rm

135: GeV})},

136: \label{eq:prob}

137: \eea

138: 

139: For the purpose of the present work, the nature of the partner

140: neutrino state to which the $\nu_\mu$ oscillates is immaterial so

141: long as it does not mimic a $\nu_\mu$ at the SK detector.

142: 

143: 

144: \section{The K2K results}

145: 

146: 

147: In the presence of neutrino oscillations, the event rate in the

148: K2K experiment, $r$, can be expressed as

149: 

150: \bea

151: r = \int dES(E)P_{\mu\mu}(E),

152: \label{eq:r}

153: \eea

154: where $P_{\mu\mu}$ is given by (\ref{eq:prob}).

155: 

156: The oscillation parameters $\theta$ and $\Delta m^{2}$ will be 

157: determined from the K2K data.  The experiment is sensitive to

158: $(\Delta m^{2},\sin^22\theta)$ around $(10^{-3}$eV$^{2},1)$ favoured

159: by the SuperKamiokande experiment.

160: 

161: 

162: $S(E)$ is the `energy spectrum' of the neutrinos from KEK  

163: as detected at the Kamioka site. This spectrum is

164: given by the convolution of the standard (no oscillation)

165: neutrino energy spectrum $\phi$ at the SK (far) detector,

166: the interaction cross-section $\sigma$, and the

167: detection efficiency $\epsilon$ (all in differential form),

168: integrated over the final state parameters, {\em i.e.}, symbolically,

169: $S = \int \phi \sigma \epsilon$.

170: 

171: In the published literature there is information on $\phi$ but

172: not yet on the product $\sigma . \epsilon$ relevant for the far

173: detector specification. As direct and accurate reconstruction of

174: the spectrum is not possible, some indirect, approximate

175: reconstruction was suggested by the K2K collaboration.

176: 

177: We have two sets of information in hand.

178: 

179: (i) The neutrino spectrum is considered to be practically zero above 5

180: GeV \cite{k2k}.

181: 

182: (ii) According to the K2K Monte Carlo simulation, for an exposure of

183: $3.85\times10^{19}$ p.o.t., assuming maximal mixing, we have the set of

184: predictions for the event rate ($r^{MC}$) for different $\Delta m^{2}$

185: shown in Table 1:

186: 

187: \vskip 5 pt

188: \begin{center}

189: \begin{tabular}{|c|c|c|c|c|}

190: \hline

191: $\Delta m^{2}$ in eV$^{2}$ & $1\times10^{-4}$ & $3\times10^{-3}$ &

192: $5\times10^{-3}$ & $7\times10^{-3}$\\

193: \hline

194: $r^{MC}$ & $63.9$ & $41.5$ & $27.4$ & $23.1$\\

195: \hline

196: \end{tabular}

197: \end{center}

198: \begin{description}

199: \item{\small \sf Table 1:} {\small \sf  The values of $r^{MC}$

200: for different $\Delta m^2$ determined from the K2K Monte Carlo

201: calculations \cite{k2k2} for maximal mixing ($\sin^22 \theta = 1$).}  

202: \end{description}

203: 

204: 

205: \section{Impact of different $S(E)$}

206: In the absence of direct information, a form for the spectrum

207: was suggested \cite{flm} to be $S(E) \propto

208: x^{\alpha}(1-x)^{\beta}$ with $x = E/($5 GeV). It fits the expected

209: behaviour of $S(E)$ at $x$ = 0 and 1 and has a particularly

210: simple form. The overall normalization and the parameters

211: $\alpha$ and $\beta$ were obtained by using a best fit to the

212: data in Table 1.  It was shown using this functional form of

213: $S(E)$ that the preliminary data from K2K are consistent with the

214: $\Delta m^2$ favoured by the SK atmospheric neutrino experiment.

215: 

216: Here, we ask the question: How sensitively does this conclusion

217: depend on the form of $S(E)$? To this end, we consider several

218: different functional forms for the spectrum $S(E)$. For each, the

219: best-fit values of the parameters are determined by minimizing

220: \bea

221: \chi^2 = \sum_{i=2}^4 \frac{(r_i - r_i^{MC})^2}{\sigma_i^2},

222: \eea

223: where $r_i$ is calculated using the chosen form of $S(E)$ and

224: $r_i^{MC}$ is taken from the K2K Monte Carlo results of Table 1

225: and, as in \cite{flm},

226: \bea

227: \sigma_i^2 = 44 + 0.01 r_i^2.

228: \label{eq:sigma}

229: \eea

230: $r^{MC}$ corresponding to $i=1$, {\em i.e.}, $\Delta m^2 = 1.0 \times

231: 10^{-4}$ eV$^2$, is used to determine the overall normalization

232: constant $N$. In this way, in all cases the no oscillation limit is

233: correctly reproduced. The functional forms that we have examined and

234: the best-fit values of the parameters are listed in Table 2.  These

235: functions are plotted in Fig.~1.

236: 

237: \begin{center}

238: \begin{tabular}{|c|c|c|c|c|c|}

239: \hline

240: Case& functional & Normalization & \multicolumn{3}{|c|}{best fit

241: values} \\ \cline{4-6}

242: &form&$N$ & $\alpha$ & $\beta$ & $\chi^2$\\

243: \hline

244: $a$&$N x^{\alpha}(1-x)^{\beta}$ & 1370.7 & 1.14 & 2.78 & 0.044\\

245: $b$&$N x^{\alpha}(1-x^{2})^{\beta}$ & 518.3 & 0.85 & 3.81 & 0.063\\

246: $c$&$N x^{\alpha}(1-x^{3})^{\beta}$ & 307.7 & 0.60 & 5.50 & 0.068\\

247: $d$&$N x^{\alpha}(1-x^{4})^{\beta}$ & 215.7 & 0.42 & 7.48 & 0.061\\

248: \hline

249: \end{tabular}

250: \end{center}

251: \begin{description}

252: \item{\small \sf Table 2:} {\small \sf  The different

253: parametrizations of $S(E)$, the best-fit values of the parameters,

254: and the values of $\chi^2$.}

255: \end{description}

256: 

257: All the above functions have a common nature in that they vanish at

258: $x=$ 0 and 1 and rise to a maximum value at some intermediate energy.

259: The actual neutrino spectrum as a function of the energy is expected to

260: have  rising and falling regions at the two ends and an intermediate

261: region where its variation is rather slow. To mimic this feature, we

262: have also considered two further cases where we have chosen the

263: functional forms of Cases $a$ and $b$ of Table 2 but `chopped' the

264: function at the values of $E$ where it achieves half the maximum value.

265: Between these points, $E_1$ and $E_2$, the function is assumed to be a

266: constant. The points $x_i = E_i$/(5 GeV), ($i$ =1,2) and the best-fit

267: values of the parameters in these cases are presented in Table 3 and

268: the functions exhibited in Fig.~1.

269: 

270: 

271: \begin{center}

272: \begin{tabular}{|c|c|c|c|c|c|c|c|}

273: \hline

274: Case&functional & Normalization &

275: \multicolumn{2}{|c|}{Half-Maxima} & \multicolumn{3}{|c|}{best fit

276: values} \\ \cline{4-8}

277: & form & $N$ & $x_1$ & $x_2$ &$\alpha$ & $\beta$ & $\chi^2$\\

278: \hline

279: $e$&$N x^{\alpha}(1-x)^{\beta}$ & 13136.3&0.128 & 0.529 &

280: 2.00&4.61 & 0.065 \\ \hline

281: $f$&$N x^{\alpha}(1-x^{2})^{\beta}$ & 3392.6 & 0.156 & 0.544 &

282: 1.72 & 6.58 & 0.047 \\

283: \hline

284: \end{tabular}

285: \end{center}

286: \begin{description}

287: \item{\small \sf Table 3:} {\small \sf  

288: Parametrizations of $S(E)$ with a constant region between $x_1$

289: and $x_2$ (see text). The best-fit values of the parameters

290: and the values of $\chi^2$ are shown.}

291: \end{description}

292: 

293: We expect that the six rather different functional forms for

294: $S(E)$ that we have chosen adequately capture the possible

295: uncertainties in the neutrino spectrum, cross-section, and

296: detection efficiency.  We make two remarks. Firstly, the fits are

297: all of very low $\chi^2$. This is due to the comparatively large

298: $\sigma_i^2$ from (\ref{eq:sigma}) associated with each datum

299: point. Secondly, we find that over a rather broad region in the

300: ($\alpha, \beta$) parameter space around the best-fit points,

301: $\chi^2$ varies relatively little. In particular, the best-fit

302: point we find in Case $a$ is somewhat different from the one in

303: \cite{flm} even though the functional forms are the same. For the

304: latter we find a comparable but higher $\chi^2$.

305: 

306: After having obtained the six different functional forms of

307: $S(E)$ which best fit the Monte Carlo results of Table 1 we use

308: them in (\ref{eq:r}) to obtain the predicted value for the number

309: of events for the K2K experiment with $3.85\times10^{19}$ p.o.t.

310: These results are shown in Fig. 2 as a function of $\Delta m^2$

311: for the maximal mixing case ($\sin^22\theta$ = 1). The two

312: horizontal lines are the 1$\sigma$ range of the experimental

313: observation. The intercepts of the curves with this experimental

314: range determine the allowed values of $\Delta m^2$ for the

315: various cases. It is seen from Fig. 2 that the differences

316: between the alternative cases $(a$ - $f)$ are comparatively marginal.

317: There is a small spread in the lowest admissible values of

318: $\Delta m^2$ in the different cases. But as the data improve even

319: this will be removed.  Since we have chosen $S(E)$ with a wide

320: range of forms, it would not be unfair to conclude that the range

321: of $\Delta m^2$ which is favoured by the K2K data can be

322: predicted in a robust fashion.

323: 

324: Motivated by the best-fit to the atmospheric neutrino data, in

325: this work we have used maximal mixing, {\em i.e.},

326: $\sin^22\theta = 1.0$, in (\ref{eq:prob}). The SK data restrict

327: the mixing angle rather tightly and at 90\% C.L. one can allow

328: $\sin^22\theta \gtap 0.9$. How much would our conclusions be

329: affected if the lower limit of $\sin^22\theta$ is used in the analysis?

330: We find, for example, for an $S(E)$ of the form of Case $a$ (see Table

331: 2) the best fit point moves to ($\alpha, \beta$) = (2.39,

332: 6.10)\footnote{This fit has $\chi^2 = 9.74 \times 10^{-3}$.}. The event

333: rate as a function of $\Delta m^2$ for this case is also shown in Fig.

334: 2. Notice that a small region around $2 \times 10^{-2}$ eV$^2$ is

335: allowed for this case.  A large improvement in the K2K systematic and

336: statistical uncertainties will be able to exclude this option.

337: 

338: 

339: 

340: \section{Conclusion}

341: 

342: The first results from the K2K long baseline experiment show a

343: depletion in the number of events compared to the expectation.

344: This is consistent with oscillations of the $\nu_\mu$ to a

345: neutrino of a different flavour with mass and mixing in the range

346: favoured by the atmospheric neutrino results. In this

347: note, we have shown that this conclusion is robust and is not

348: altered even if the uncertainty in the initial neutrino spectrum,

349: interaction cross-section, and detection efficiency are widely

350: varied.  

351: 

352: {\em Note added:} After the submission of the paper for

353: publication, K2K announced results \cite{k2kn} based on $4.8

354: \times 10^{19}$ p.o.t. They observe 56 events with a

355: no-oscillation expectation of 80.1$^{+6.2}_{-5.4}$. This

356: corresponds to the depletion ratio (see Eq.~(\ref{eq:depl})) $d =

357: 0.699^{+0.051}_{-0.050}$, consistent with the earlier result. We

358: find using our analysis that for the different parametrizations

359: of $S(E)$ this corresponds to allowed ranges for $\Delta m^{2}$

360: shown in Table 4. It is gratifying that the best fit point found

361: by the K2K group for maximal mixing ($\Delta m^{2} = 2.8 \times

362: 10^{-3}$ eV$^2$ \cite{k2kn}) lies within the allowed range for all

363: parametrizations considered.

364: 

365: \begin{center}

366: \begin{tabular}{|c|c|c|}

367: \hline

368: Case& \multicolumn{2}{|c|}{$\Delta m^{2}$ Limits (10$^{-3}$ eV$^2$)}

369: \\ \cline{2-3}

370: &Lower& Upper \\

371: \hline

372: $a$& 1.57 & 3.61 \\ \hline

373: $b$&1.60 & 3.72 \\ \hline

374: $c$&1.52 & 3.40 \\ 

375:    &22.48&25.06 \\ \hline

376: $d$&1.47 & 3.73 \\ 

377:    &21.77& 26.38\\ \hline

378: \end{tabular}

379: \end{center}

380: \begin{description}

381: \item{\small \sf Table 4:} {\small \sf  The ranges of $\Delta

382: m^{2}$ allowed at 1$\sigma$ for maximal mixing by the latest K2K

383: results\cite{k2kn} for the alternative parametrizations of $S(E)$ (see

384: Table 2). Notice that in cases (c) and (d) there are two allowed

385: ranges.}

386: \end{description}

387: 

388: 

389: \section*{Acknowledgements}

390: SB has been partially supported by St. Edmund's College,

391: Cambridge, U.K., while the work of AR has been supported in part

392: by C.S.I.R., India and D.S.T., India.

393: 

394: 

395: 

396: 

397: \begin{thebibliography}{99}

398: \bibitem{sk}{Super-Kamiokande Collaboration, Y. Fukuda {\em et

399: al.}, Phys. Rev. Lett. {\bf 81}, 1562 (1998); S. Fukuda {\em et

400: al.}, Phys. Rev. Lett. {\bf 85}, 3999 (2000); T. Toshito, 

401: hep-ex/0105023 (to appear in the proceedings of 36th

402: Rencontres de Moriond on Electroweak Interactions and Unified

403: Theories, Les Arcs, France, 2001).}

404: \bibitem{macro}{MACRO Collaboration, M. Ambrosio {\em et al.}, Phys.

405: Lett. B {\bf 517}, 59 (2001).}

406: \bibitem{soudan}{Soudan-2 Collaboration, W. W. M. Allsion {\em et al.},

407: Phys. Lett. B {\bf 449}, 137 (1999).}

408: \bibitem{k2k}{K. Nishikawa, in Proceedings of the International Europhysics

409: Conference on High Energy Physics, Budapest, Hungary, 2001.

410: Transparencies available at www.hep2001.elte.hu}

411: \bibitem{k2k2} K2K Collaboration, S. H. Ahn {\em et al.}, Phys. Lett.

412: B {\bf 511}, 178 (2001); James E. Hill, hep-ex/0110034.

413: \bibitem{k2ko}{S. Boyd, hep-ex/0011039.}

414: \bibitem{flm}{G. L. Fogli, E. Lisi, and A. Marrone, Phys. Rev. D {\bf

415: 65}, 073028 (2002).}

416: \bibitem{k2kn} K2K Collaboration, M. H. Ahn {\em et al.}, Phys.

417: Rev. Lett. {\bf 90}, 041801 (2003).

418: \end{thebibliography}

419: 

420: \newpage

421: \vskip 30pt

422: \begin{figure}[thb]

423: \vskip -1.00in

424: \psfig{figure=fig1.ps,width=14.0cm,height=22.0cm} 

425: \vskip -5.10in

426: \caption{\sf \small   

427: The  spectrum $S(E)$ as a function of  $E$. The cases to which

428: the different curves correspond are indicated   (see Tables 2 and

429: 3). }

430: 

431: \end{figure}

432: 

433: \newpage

434: \vskip 30pt

435: \begin{figure}[thb]

436: \vskip -1.00in

437: \psfig{figure=fig2.ps,width=14.0cm,height=22.0cm} 

438: \vskip -5.10in

439: \caption{\sf \small   

440: The number of events in the K2K experiment for an exposure of

441: $3.85\times 10^{19}$ p.o.t. as a function of $\Delta m^{2}$ (at

442: $\sin^22\theta =1)$.  The region between the two horizontal lines is

443: the current K2K measured range within 1 standard deviation.  The boxes

444: are the values from the K2K MC simulation (Table 1). The six curves

445: obtained from the  different forms for $S(E)$ considered in this work

446: are shown (same conventions as in Fig. 1). The curve marked `non-max'

447: corresponds to the case of non-maximal mixing ($\sin^22\theta = 0.9$)

448: discussed in the text.}

449: \end{figure}

450: 

451: \end{document}

452: