1: % sigma_8 NUOVI DISCORSI SU OLIVE NEL \bullet
2: \documentclass[12pt]{article}
3: \usepackage{graphicx}
4: \usepackage{multicol}
5: \usepackage{color}
6: %\usepackage{amsmath}
7: \definecolor{rosso}{cmyk}{0,1,1,0.4}
8: \definecolor{rossos}{cmyk}{0,1,1,0.55}
9: \definecolor{rossoc}{cmyk}{0,0.5,1,0.2}
10: \definecolor{blu}{cmyk}{1,1,0,0.3}
11: \definecolor{blus}{cmyk}{1,1,0,0.6}
12: \definecolor{blucc}{cmyk}{1,0.4,0.2,0}
13: \definecolor{viola}{cmyk}{0,1,0,0.6}
14: \definecolor{viola2}{cmyk}{0,1,0.2,0.6}
15: \definecolor{verde}{cmyk}{0.92,0,0.59,0.25}
16: \definecolor{verdec}{cmyk}{0.92,0,0.59,0.15}
17: \definecolor{verdes}{cmyk}{0.92,0,0.59,0.4}
18: \font\tenrsfs=rsfs10 at 12pt
19: \font\sevenrsfs=rsfs7
20: \font\fiversfs=rsfs5
21: \newfam\rsfsfam
22: \textfont\rsfsfam=\tenrsfs
23: \scriptfont\rsfsfam=\sevenrsfs
24: \scriptscriptfont\rsfsfam=\fiversfs
25: \def\mathscr#1{{\fam\rsfsfam\relax#1}}
26: \def\Lag{\mathscr{L}}
27: \def\Ham{\mathscr{H}}
28:
29: \oddsidemargin -0.5cm \evensidemargin -0.5cm
30: \topmargin -1cm \textwidth 17.2cm \textheight 22.9cm
31: %\oddsidemargin 1cm \evensidemargin 1cm \textwidth 14cm \usepackage{showkeys}
32: \newcommand{\riga}[1]{\noalign{\hbox{\parbox{\textwidth}{#1}}}\nonumber}
33: \newcommand{\gE}{\gamma_{\rm E}}
34: \newcommand{\mub}{\bar{\mu}}
35: %\renewcommand\&{&}
36: %\renewcommand\_{_}
37: \newcommand{\Ord}{{\cal O}}
38: \newcommand{\fig}[1]{~\ref{fig:#1}}
39: \newcommand{\eq}[1]{~{\rm (\ref{eq:#1})}}
40: \newcommand{\sys}[1]{~{\rm (\ref{sys:#1})}}
41: \newcommand{\scatola}[1]{\fbox{$\displaystyle #1$}}
42: \newcommand{\GeV}{\,{\rm GeV}}
43: \newcommand{\TeV}{\,{\rm TeV}}
44: \def\circa#1{\,\raise.3ex\hbox{$#1$\kern-.75em\lower1ex\hbox{$\sim$}}\,}
45:
46: \def\Red {\special{color cmyk 0 1. 1. 0.5}}
47: \def\Black{\special{color cmyk 0 0 0 1.}}
48: \def\Green{\special{color cmyk 0.92 0 0.59 0.5}} % PANTONE 323
49: \def\Purple{\special{color cmyk 0 0.63 0 0.3}} % PANTONE 218
50: \def\Blue {\special{color cmyk 1. 1. 0.3 0}}
51: \def\Cyan{\special{color cmyk 1. 0 0 0.5}} % PANTONE PROCESS-CYAN
52: \newcommand{\NP}{Nucl. Phys.}
53: \newcommand{\PRL}{Phys. Rev. Lett.}
54: \newcommand{\PL}{Phys. Lett.}
55: \newcommand{\PR}{Phys. Rev.}
56: \newcommand{\beq}{\begin{equation}}
57: \newcommand{\eeq}{\end{equation}}
58: \newcommand{\diag}{\hbox{diag}\,}
59:
60: \def\circa#1{\,\raise.3ex\hbox{$#1$\kern-.75em\lower1ex\hbox{$\sim$}}\,}
61: \makeatletter
62:
63: %
64: % formato bibliografico standard
65: %
66: %\art[hep-ph/0403158]{autori}{rivista}{numero}{pagina}{anno}
67: \def\art{\@ifnextchar[{\eart}{\oart}}
68: \def\eart[#1]#2#3#4#5#6{{\rm #2}, {\em #3 \rm #4} {\rm (#6) #5} ({\em #1})}
69: %\def\eart[#1]#2#3#4#5#6{{\rm #2}, {\em #1}}
70: \def\hepart[#1]#2{{\rm #2, \em#1}}
71: \newcommand{\oart}[5]{{\rm #1}, {\em #2 \rm #3} {\rm (#5) #4}}
72: \newcommand{\y}{{\rm and} }
73: %
74: % definizione della macro EQNSYSTEM
75: %
76: \newcounter{alphaequation}[equation]
77: %\def\thealphaequation{\theequation\alph{alphaequation}}
78: \def\thealphaequation{\theequation\hbox to
79: 0.6em{\hfil\alph{alphaequation}\hfil}}
80: % MODIFICATA PER DARE UNA DIMENSIONE UGUALE AD UN 1em AD OGNI LETTERA
81: \def\eqnsystem#1{
82: \def\@eqnnum{{\rm (\thealphaequation)}}
83: %
84: \def\@@eqncr{\let\@tempa\relax \ifcase\@eqcnt \def\@tempa{& & &} \or
85: \def\@tempa{& &}\or \def\@tempa{&}\fi\@tempa
86: \if@eqnsw\@eqnnum\refstepcounter{alphaequation}\fi
87: \global\@eqnswtrue\global\@eqcnt=0\cr}
88: %
89: \refstepcounter{equation} \let\@currentlabel\theequation \def\@tempb{#1}
90: \ifx\@tempb\empty\else\label{#1}\fi
91: %
92: \refstepcounter{alphaequation}
93: \let\@currentlabel\thealphaequation
94: %
95: \global\@eqnswtrue\global\@eqcnt=0 \tabskip\@centering\let\\=\@eqncr
96: $$\halign to \displaywidth\bgroup \@eqnsel\hskip\@centering
97: $\displaystyle\tabskip\z@{##}$&\global\@eqcnt\@ne
98: \hskip2\arraycolsep\hfil${##}$\hfil& \global\@eqcnt\tw@\hskip2\arraycolsep
99: $\displaystyle\tabskip\z@{##}$\hfil
100: \tabskip\@centering&\llap{##}\tabskip\z@\cr}
101: %
102: \def\endeqnsystem{\@@eqncr\egroup$$\global\@ignoretrue} \makeatother
103:
104:
105: \newcommand{\MeV}{\,\hbox{\rm MeV}}
106: \newcommand{\eV}{\,\hbox{\rm eV}}
107:
108: % new commands for the SN part
109: \newcommand{\gcm}{\,\mathrm{g}\,\mathrm{cm}^{-3}}
110: \newcommand{\erg}{\,\mathrm{erg}}
111: \renewcommand{\sec}{\,\mathrm{sec}}
112: \newcommand{\nub}{\bar\nu}
113: \newcommand{\nue}{\nu_e}
114: \newcommand{\nueb}{\bar\nu_e}
115: \newcommand{\num}{\nu_\mu}
116: \newcommand{\numb}{\bar\nu_\mu}
117: \newcommand{\nut}{\nu_\tau}
118: \newcommand{\nutb}{\bar\nu_\tau}
119: \newcommand{\numt}{\nu_{\mu,\tau}}
120: \newcommand{\numtb}{\bar\nu_{\mu,\tau}}
121: \newcommand{\nus}{\nu_{\rm s}}
122: \newcommand{\nusb}{\bar\nu_{\rm s}}
123: \newcommand{\nB}{n_{B}}
124: \newcommand{\GF}{G_{\rm F}}
125: \newcommand{\km}{\,\mathrm{km}}
126: \newcommand{\Feo}{\Phi^0_{\nueb}}
127: \newcommand{\Fmuo}{\Phi^0_{\numb}}
128: \newcommand{\Ftauo}{\Phi^0_{\nutb}}
129: \newcommand{\Fso}{\Phi^0_{\nusb}}
130: \newcommand{\Fe}{\Phi_{\nueb}}
131: \newcommand{\Fmu}{\Phi_{\numb}}
132: \newcommand{\Ftau}{\Phi_{\nutb}}
133: \newcommand{\Fs}{\Phi_{\nusb}}
134:
135: % end of new commands for the SN part
136:
137: \begin{document}
138: \thispagestyle{empty}
139: \setcounter{page}{0}
140:
141: {hep-ph/0403158\hfill IFUP--TH/2004--2}
142: \vspace{1cm}
143:
144: \begin{center}
145: {\LARGE \bf \color{rossos}
146: Probing oscillations into sterile neutrinos\\[3mm]
147: with cosmology, astrophysics and experiments
148: }\\[1cm]
149:
150: {
151: {\large\bf M. Cirelli}$^1$,
152: {\large\bf G. Marandella}$^2$,
153: {\large\bf A. Strumia}$^{3}$, {\large\bf F. Vissani}$^{4}$
154: }
155: \\[7mm]
156: {\it $^1$ Physics Department, Yale University, New Haven, USA} \\[3mm]
157: {\it $^2$ Scuola Normale Superiore and INFN, Pisa,
158: %, Piazza dei Cavalieri 7, Pisa, I-56126,
159: Italia } \\[3mm]
160: {\it $^3$ Dipartimento di Fisica dell'Universit\`a di Pisa,
161: %and INFN,
162: Italia}\\[3mm]
163: {\it $^4$ INFN, Laboratori Nazionali del Gran Sasso, Italia}\\[1cm]
164: \vspace{1cm}
165: {\large\bf\color{blus} Abstract}
166: \end{center}
167: \begin{quote}
168: {\large\noindent\color{blus}
169: We perform a thorough analysis of oscillation signals generated by one extra sterile neutrino,
170: extending previous analyses done in simple limiting cases
171: and including the effects of established oscillations among active neutrinos.
172: We consider the following probes:
173: solar, atmospheric, reactor and beam neutrinos,
174: Big-Bang Nucleosynthesis (helium-4, deuterium), Cosmic Microwave Background,
175: Large Scale Structure, supernov\ae,
176: neutrinos from other astrophysical sources.
177: We find no evidence for a sterile neutrino in present data,
178: identify the still allowed regions, and study which future experiments can best probe them:
179: sub-MeV solar experiments,
180: more precise studies of CMB or BBN,
181: future supernova explosions, etc.
182: We discuss how the LSND hint is strongly disfavoured by the constraints of (standard) cosmology.}
183: \end{quote}
184:
185: \vfill
186:
187: \hrule
188:
189: \smallskip
190:
191: \noindent
192: $^1$ e-mail: Marco.Cirelli@yale.edu\\
193: $^2$ e-mail: Guido.Marandella@sns.it\\
194: $^3$ e-mail: Alessandro.Strumia@df.unipi.it\\
195: $^4$ e-mail: Francesco.Vissani@lngs.infn.it
196:
197: \newpage
198:
199: \tableofcontents
200:
201: \setcounter{page}{1}
202:
203: \section{Introduction}
204:
205: High-energy colliders are the tool to discover new heavy particles with sizable couplings.
206: New light particles with small couplings can be searched for in many different ways.
207: Neutral fermions with eV-scale mass, called `sterile neutrinos' in the jargon,
208: are typically stable enough to give effects in cosmology,
209: and can affect neutrino oscillation experiments.\footnote{Small fermions masses
210: are stable under quantum corrections.
211: Cosmology, astrophysics and neutrino experiments are also sensitive to
212: light scalars or vectors, which give
213: different kinds of signals.}
214: Both fields recently discovered new physics
215: (in both cases around meV energies),
216: but so far no sterile neutrinos.
217:
218: Today the most powerful cosmological probe of sterile effects
219: is standard Big-Bang Nucleosynthesis (BBN)~\cite{BBN}, which constrains
220: the number of thermalized neutrinos present at $T\sim 0.1\MeV$
221: %(assumed to be temperature-independent)
222: to be $N_\nu = 2.5\pm 0.7$.
223: Since the uncertainty is controversial it is not clear
224: how much $N_\nu=4$ is disfavoured.
225:
226: The established solar and atmospheric neutrino anomalies
227: seem produced by oscillations among the three SM neutrinos,
228: with at most minor contributions from possible extra sterile neutrinos.
229: The present 99\% C.L.
230: bounds on sterile mixing,
231: computed {\em assuming} that the initial active neutrino $|\nu_{\rm a}\rangle$
232: oscillates with a large $\Delta m^2$ into
233: an energy-independent mixed neutrino $\cos\theta_{\rm s}|\nu'_{\rm a}\rangle + \sin\theta_{\rm s}|\nu_{\rm s}\rangle$,
234: are $\sin^2\theta_{\rm s} <0.25$ in solar oscillations (where $\nu_{\rm a}=\nu_e$)~\cite{etas}
235: and $\sin^2\theta_{\rm s} <0.21$ in atmospheric oscillations (where $\nu_{\rm a}=\nu_\mu$)~\cite{etaa}.
236:
237: \bigskip
238:
239: We relax these simplifying assumptions and study
240: the more general $4$-neutrino context.
241: This demands a remarkable effort, but it is an
242: important task for the present, for two reasons.
243: First, because recent discoveries
244: in cosmology and neutrino physics
245: are stimulating new experiments that will
246: study the new phenomena with the redundancy necessary to
247: test the minimal models suggested by present data.
248: Second, because oscillations in extra light particles
249: are the natural extension of the emergent
250: massive neutrinos scenario.
251:
252:
253:
254:
255: Although we list hints that might be interpreted as sterile neutrino effects (LSND~\cite{LSND}, certain
256: pieces of solar or atmospheric or BBN data, ...),
257: we do not focus on any of them in particular.
258: Rather, we compute experimental capabilities and constraints on sterile oscillations
259: and compare them with capabilities and constraints
260: from various present and future cosmological and astrophysical probes.
261: Some important constraints are based on untested assumptions
262: and plagued by systematic uncertainties.
263: In such cases, rather than performing global fits, we identify and compute
264: the key observables,
265: trying to explain the basic physics in simple terms emphasizing the
266: controversial issues, so that the reader can judge.
267: %Although this cannot be summarized in plots
268: %we present two kind of bounds: `more aggressive' and `more conservative'.
269:
270: \medskip
271:
272: The paper is organized as follows.
273: In section~\ref{osc} we briefly review theoretical motivations:
274: as in the case of the known light particles, some fundamental reason
275: would presumably be behind the lightness of an extra sterile neutrino.
276: We also describe the non-standard parametrization
277: of active/sterile mixing that we choose (because
278: more convenient and intuitive than standard parametrizations)
279: and describe the qualitatively different kinds
280: of spectra on which we will focus.
281:
282: In section~\ref{cosmo} we study sterile effects in cosmology,
283: comparing the relative sensitivities of
284: two BBN probes (the helium-4 and deuterium abundances),
285: of Cosmic Microwave Background (CMB)
286: and of Large Scale Structures (LSS).
287: In section~\ref{solar} we study sterile oscillations in solar (and KamLAND) neutrinos.
288: In section~\ref{SN} we study sterile oscillations in SN1987A and future supernov\ae.
289: We also briefly discuss other less promising probes
290: (relic SN background, high-energy cosmic neutrinos,\ldots).
291: In section~\ref{atm} we study sterile oscillations in atmospheric and reactor neutrinos,
292: and in short and long-baseline neutrino beams (including LSND).
293:
294: In order to avoid a unreadably long paper
295: each section is written in a concise way and
296: contains a `Results' subsection, which can be read
297: skipping the other more technical parts.
298: From a computational point of view,
299: exploring
300: 4$\nu$ oscillations is $3\div 4$ orders of magnitude more
301: demanding than usual 2 or 3$\nu$ fits
302: and therefore requires significant improvements of usual techniques.
303: Subsections entitled `technical details' describe how this was achieved.
304: In particular we describe how we employ the density matrix formalism
305: in neutrino oscillation computations, and explain why even in simple
306: $3\nu$ situations it gives significant advantages with respect to
307: more direct equivalent approaches.
308: % generic initial flux, average oscillations
309:
310:
311: In section~\ref{all} we conclude by summarizing and comparing
312: the different probes of sterile neutrinos.
313:
314:
315:
316: \section{Active/sterile neutrino mixing}\label{osc}
317: We first introduce a non-standard useful parameterization of
318: the most generic $4\nu$ spectrum, and later present a brief review
319: of theoretical models.
320:
321:
322: \subsection{Parameterization}
323: A generic $4\times 4$ Majorana neutrino mass matrix is described by 4 masses, 6 mixing angles and 6 CP-violating phases;
324: 3 of them affect oscillations.
325:
326: In absence of sterile neutrinos, we denote by
327: $U$ the usual $3\times 3$ mixing matrix that relates
328: neutrino flavour eigenstates $\nu_{e,\mu,\tau}$ to active
329: neutrino mass eigenstates $\nu^{\rm a}_{1,2,3}$
330: as $\nu_\ell = U_{\ell i} \nu_i^{\rm a}$
331: ($i=\{1,2,3\}$, $\ell = \{e,\mu,\tau\}$).
332: The extra sterile neutrino can mix with one arbitrary combination
333: of active neutrinos,
334: $$\vec{n}\cdot \vec{\nu} = n_e \nu_e + n_\mu \nu_\mu + n_\tau \nu_\tau =
335: n_1\nu_1^{\rm a} + n_2 \nu_2^{\rm a} + n_3 \nu_3^{\rm a}\qquad (n_i = U_{\ell i}n_\ell).$$
336: Mixing of the sterile neutrino can be therefore fully described
337: by a complex unit 3-versor $\vec{n}$
338: (containing two CP-violating phases)
339: and by one mixing angle $\theta_{\rm s}$.
340: With this parameterization the 4 neutrino mass eigenstates are
341: \begin{eqnsystem}{sys:param}
342: \label{eq:nu1234}
343: \left\{ \begin{array}{l}
344: \nu_4 = \nu_{\rm s} ~\cos\theta_{\rm s} + n_\ell \nu_\ell ~\sin\theta_{\rm s} \cr
345: \nu_i = U_{\ell i}^* [\delta_{\ell\ell'}- n_{\ell}^* n_{\ell'}(1-\cos\theta_{\rm s}) ] \nu_{\ell'} - \sin\theta_{\rm s}
346: n_\ell^* U_{\ell i}^*\nu_{\rm s}
347: \end{array}\right.\\
348: \riga{i.e.\ the $4\times 4$ neutrino mixing matrix $V$ as
349: that relates flavour to mass eigenstates as $\nu_{e,\mu,\tau,s} = V\cdot \nu_{1,2,3,4}$ is}\\
350: \label{eq:U}
351: V = \pmatrix{1-(1-\cos\theta_{\rm s}) \vec{n}^* \otimes \vec{n} &
352: \sin\theta_{\rm s} \vec{n}^* \cr
353: -\sin\theta_{\rm s} \vec{n} & \cos\theta_{\rm s}} \times \pmatrix{U&0\cr 0 & 1} \\
354: \riga{or, more explicitly,}\\
355: V = \bordermatrix{& \nu_{i} & \nu_4 \cr
356: \nu_{\ell} & U_{\ell i} - n_i n_\ell^*(1-\cos\theta_{\rm s})
357: & n_\ell^* \sin\theta_{\rm s} \cr
358: \nu_{\rm s} & - n_i \sin\theta_{\rm s} & \cos\theta_{\rm s}}.
359: \end{eqnsystem}
360: In summary, $\vec{n}$ identifies which combination of active neutrinos mixes
361: with $\nu_{\rm s}$ with mixing angle $\theta_{\rm s}$.
362:
363:
364: \begin{figure}
365: $$\includegraphics[width=7cm]{spectramus}\qquad
366: \includegraphics[width=7cm]{spectra2s}$$
367: \caption{\label{fig:spettrias}\em {\bf
368: Basic kinds of four neutrino mass spectra}.
369: Left: sterile mixing with a flavour eigenstate ($\nu_\mu$ in the picture).
370: Right: sterile mixing with a mass eigenstate ($\nu_2$ in the picture).
371: }
372: \end{figure}
373:
374:
375:
376: In order to understand how neutrinos oscillate in the generic case,
377: it is convenient to focus on two different kinds of limiting cases,
378: pictorially exemplified in fig.\fig{spettrias}:
379: \begin{itemize}
380: \item {\bf Mixing with a flavour eigenstate} (fig.\fig{spettrias}a):
381: $\vec{n}\cdot\vec{\nu}=\nu_\ell$ ($\ell =e$ or $\mu$ or $\tau$).
382: The sterile neutrino oscillates into a well defined flavour
383: at 3 different $\Delta m^2$
384: (which cannot all be smaller than the observed splittings $\Delta m^2_{\rm sun,atm}$).
385:
386: \item {\bf Mixing with a mass eigenstate} (fig.\fig{spettrias}b):
387: $\vec{n}\cdot\vec{\nu}=\nu_i$ ($i=1$ or 2 or 3).\footnote{From here on,
388: we omit the superscript `a' (that stands for
389: active): it should be clear that whenever we speak of $\nu_2/\nu_{\rm s}$
390: mixing, this is just a short-hand for $\nu_2^{\rm a}/\nu_{\rm s}$.}
391: The sterile neutrino oscillates into a neutrino of mixed flavour
392: at a single $\Delta m^2$, which can be arbitrarily small.
393: \end{itemize}
394: We think that our parametrization of sterile mixing, in eq.~(\ref{sys:param}),
395: makes physics more transparent than
396: other frequently used choices\footnote{When studying sterile mixing with a flavour eigenstate
397: our expression is directly related to the `standard' parameterization
398: $$ V= R_{34}R_{24}R_{14}\cdot U_{23}U_{13} U_{12}$$
399: where $R_{ij}$ represents a rotation in the $ij$ plane by angle $\theta_{ij}$
400: and $U_{ij}$ a complex rotation in the $ij$ plane.
401: % We neglect CP-violating phases, so that $U_{ij} = R_{ij}$ and $V$ becomes
402: % $$V = \bordermatrix{& \nu_1&\nu_2&\nu_3&\nu_4\cr
403: % \nu_e & c_{12}c_{13}c_{14} & s_{12}c_{13}c_{14} & s_{13}c_{14} & s_{14}\cr
404: % \nu_\mu &&&& s_{24}c_{14}\cr
405: % \nu_\tau &&&& s_{34}c_{24}c_{14}\cr
406: % \nu_{\rm s} &&&& c_{34}c_{24}c_{14}}$$
407: $\theta_{14}$ or $U_{e4}$ gives rise to $\nu_e/\nu_{\rm s}$ mixing,
408: $\theta_{24}$ or $U_{\mu 4}$ to $\nu_\mu/\nu_{\rm s}$ mixing,
409: and $\theta_{34}$ or $U_{\tau 4}$ to $\nu_\tau/\nu_{\rm s}$ mixing.
410:
411: The above `standard' parameterization becomes inconvenient when
412: studying mixing with a mass eigenstate.
413: In such a case our parameterization is directly related
414: to the alternative `standard' parameterization appropriate for this case,
415: $$ V = U_{23}U_{13}U_{12}\cdot R_{34}R_{24}R_{14}$$
416: Now $\theta_{i4}$ gives rise to $\nu_i/\nu_{\rm s}$ mixing.
417: Our parameterization instead is convenient because it remains simple in both cases.}.
418:
419:
420:
421: The oscillation probabilities
422: among active neutrinos
423: in the limit where the active/sterile mass splitting dominates,
424: and active/active mass splittings can be neglected, are
425: \beq\label{eq:Pij} P(\nu_\ell \to \nu_{\ell'})=P(\bar\nu_\ell \to\bar\nu_{\ell'})
426: =\left\{\begin{array}{ll}
427: 1-4|V_{\ell 4}^2|(1-|V_{\ell 4}^2|) \sin^2 ({\Delta m^2_{14} L}/{4E_\nu}) & \hbox{for $\ell=\ell'$} \\
428: 4 |V_{\ell 4}^2||V_{\ell' 4}^2| \sin^2 ({\Delta m^2_{14} L}/{4E_\nu} ) & \hbox{for $\ell\neq\ell'$}
429: \end{array}\right.\eeq
430: and in our parametrization $V_{\ell 4} = n_\ell^* \sin\theta_{\rm s}$.
431: %(see eq.\eq{nu1234}; we also have $V_{{\rm s}i} = n_i^* \sin\theta_{\rm s}$).
432:
433: \medskip
434:
435: Older papers studied
436: active/sterile mixing in 2 neutrino approximation.
437: In such a case $\theta_{\rm s} = \pi/2$ gives no oscillation effect.
438: On the contrary, in the full $4$ neutrino case
439: $\theta_{\rm s} =\pi/2$ swaps the sterile neutrino
440: with one active neutrino.
441: (e.g.\ $\nu_\mu$ in fig.\fig{spettrias}a or
442: $\nu_2$ in fig.\fig{spettrias}b,
443: if $\theta_{\rm s}$ were there increased up to $\pi/2$)
444: affecting solar and atmospheric oscillations in an obvious way.
445: Therefore large active/sterile mixing is excluded by
446: experiments for all values of $\Delta m^2_{i4} \equiv m^2_4-m^2_i$
447: (with one exception:
448: the sterile neutrino mixes with a mass eigenstate $\nu_i$
449: and the two states form a quasi-degenerate pair.
450: This structure arises naturally in certain models~\cite{SterileB}).
451:
452: In order to explore a more interesting slice of parameter space
453: when considering sterile mixing with a mass eigenstate $\nu_i$,
454: for $\theta_{\rm s}>\pi/4$ we modify the spectrum of neutrino masses
455: and replace $(m_i^2,m_4^2)$ with $(2m_i^2- m_4^2, m_i^2)$.
456: In such a way, the mostly active state always keeps the same squared mass
457: (that we fix to its experimental value), so that
458: in the limit $\theta_{\rm s}= \pi/2$
459: the sterile neutrino gives no effect rather than giving an already excluded effect.
460: Physically, in our $\nu_{\rm s}/\nu_i$ plots the
461: mostly sterile neutrino is heavier (lighter)
462: than the mass eigenstate $\nu_i$ to which it mixes when
463: $\theta_{\rm s}<\pi/4$ ($\theta_{\rm s}>\pi/4$).
464: When studying mixing with a flavour eigenstate we do not
465: modify the spectra at $\theta_{\rm s}>\pi/4$ in order to obtain some other
466: experimentally allowed configuration.
467: For this reason, we restrict such plots to $\theta_{\rm s}<\pi/4$.
468:
469:
470: \medskip
471:
472:
473:
474: We do not consider `$2+2$' neutrino mixing,
475: namely two neutrino couples separated by a mass splitting much larger than
476: $\Delta m^2_{\rm sun,atm}$.
477: In fact this spectrum does not reduce to
478: active-only oscillations in any limiting case so that
479: sterile effects are always sizable,
480: and present experiments already exclude this possibility~\cite{2+2}.
481: When the separation among the two couples is comparable to
482: $\Delta m^2_{\rm sun,atm}$, `$2+2$' is no longer
483: a special case qualitatively different from `$3+1$'.
484:
485: We assume that active neutrinos have normal hierarchy, $\Delta m^2_{23}>0$.
486: Finally, we assume $\theta_{13}=0$.
487: We verified that using $\theta_{13}\sim 0.2$,
488: the maximal value allowed by present experiments,
489: leads to minor (in some cases) or no (in other cases) modifications,
490: that we do not discuss.
491: Measuring $\theta_{13}$ and
492: discovering sterile effects will likely be two independent issues
493: (however it is curious to note that both could first manifest as disappearance of reactor $\bar\nu_e$).
494:
495: These assumptions are made because we consider three active neutrinos
496: with normal hierarchy as the most plausible spectrum, and view
497: inverted hierarchy, large $\theta_{13}$ and sterile neutrinos as possible surprises:
498: we here study the latter one.
499:
500:
501:
502:
503:
504: \subsection{Theory}
505: The relevant terms in the SU(2)$_L$-invariant effective Lagrangian that describes active
506: neutrinos $\nu$ together with extra light singlet fermions $\nu_R$ are
507: \beq\label{eq:Lsterile} \frac{m_{LL}}{2v^2} (LH)^2+ \frac{m_{RR}}{2}\nu_R^2 + \frac{m_{LR}}{v} \nu_R LH + \hbox{h.c.}\eeq
508: H is the higgs doublet with vacuum expectation value $(0,v)$.
509: The first dimension-5 operator gives Majorana $\nu$ masses $m_{LL}$ and is naturally small
510: if lepton number is broken at a high-energy scale.
511: The second term gives Majorana $\nu_R$ masses $m_{RR}$,
512: and the third term Dirac $\nu_L\nu_R$ masses $m_{LR}$:
513: one needs to understand why $m_{LR}$ and $m_{RR}$ are small.
514:
515:
516: A few theoretically favoured patterns emerge from rather
517: general naturalness considerations.
518: We consider the most generic mass matrix with $LL$, $RR$ and $LR$ mass terms.
519: If $m_{LL}$ dominates one obtains light sterile neutrinos with mass $m_{\rm s}\ll m_{\rm a}$ and
520: active/sterile mixings
521: $\theta_{\rm s}^2 \sim m_{\rm s}/m_{\rm a}$.
522: If $m_{RR}$ dominates sterile neutrinos are heavy with $\theta_{\rm s}^2 \sim m_{\rm a}/m_{\rm s}$.
523: If $m_{LR}$ dominates one obtains quasi-Dirac neutrinos that split into couples.
524: These `more likely' regions can be represented as lines in the
525: logarithmic $(\tan^2\theta_{\rm s}, \Delta m_{i4}^2)$ plane,
526: that we will use to present our results.
527:
528: \medskip
529:
530:
531: A new light particle would probably be a discovery of fundamental importance,
532: because it lightness is likely related to some fundamental principle,
533: as it is the case for the known light particles, the photon, the neutrinos and the graviton.
534: Attempts of guessing physics beyond the SM
535: from first principles motivate a number of fermions which might
536: have $\TeV^2/M_{\rm Pl}$ masses and
537: behave as sterile neutrinos.
538: A few candidates are, in alphabetic order,
539: axino, branino, dilatino, familino, Goldstino, Majorino, modulino,
540: radino.
541: These ambitious approaches so far do not give
542: useful predictions on the flavour parameters in the effective Lagrangian of eq.\eq{Lsterile}.
543: Therefore one needs to consider more specific ad-hoc models~\cite{Sterile}.
544:
545: Unification of matter fermions into SO(10) 16 or $E_6$ 27 representations
546: predicts extra singlets, which however
547: generically receive GUT-scale masses.
548: It is easy to invent ad-hoc discrete or continuous symmetries that keep a fermion light~\cite{Sterile}.
549: One might prefer to use only ingredients already present in the SM.
550: For example, the extra fermions can be forced to be light assuming that they are chiral
551: under some extra gauge symmetry (that could possibly become non perturbative at some
552: QCD-like scale, and give composite sterile neutrinos)~\cite{Sterile}.
553: Alternatively, the extra fermions may be light for
554: the same reason why neutrinos are light in the SM~\cite{Sterile}.
555: Following this point of view up to its extreme, one can add to the SM a set of `mirror particles',
556: obtaining 3 sterile neutrinos~\cite{Sterile}.
557: In presence of Planck-suppressed corrections that mix the two sectors,
558: mirror neutrinos with the same mass as SM neutrinos
559: give rise to quasi-maximal mixing angles between
560: SM neutrino mass eigenstates and sterile neutrinos, splitted by
561: $\Delta m^2 \sim m_i v^2/M_{\rm Pl}\sim 10^{-8}\eV^2$
562: (this kind of $\nu_2/\nu_{\rm s}$ effects are disfavoured by solar data).
563: Mirror neutrinos with different masses from SM neutrinos
564: can give detectable $\nu_1/\nu_{\rm s}$ oscillation effects.
565: The two sectors might instead communicate because
566: coupled to the same heavy see-saw neutrinos:
567: this gives massless states which can be mostly sterile.
568:
569:
570:
571:
572:
573: \section{Sterile effects in cosmology}\label{cosmo}
574: Present cosmological data seem compatible with
575: the following minimal assumption (see e.g.~\cite{WMAP}):
576: primordial perturbations are generated by
577: minimal inflation (flat space, Gaussian perturbations with
578: flat spectral index $n_s=1$)
579: and evolve as dictated by general relativity (with a small cosmological constant)
580: and by the Standard Model (adding some unknown Cold Dark Matter).
581: Global analyses performed under this {\em assumption} give
582: precise determinations of the cosmological parameters
583: and bounds on non standard properties of neutrinos.
584: In particular, the sensitivity to neutrino masses and to oscillations into extra sterile neutrinos
585: is competitive with direct experimental bounds.
586:
587: % Before studying how cosmological signals of sterile neutrinos are
588: % modified by the atmospheric and solar oscillations,
589: % we recall which pieces of data probe what.
590:
591:
592: Present bounds are unsafe because
593: based on assumptions which have been only partially tested.
594: At the level of precision needed to derive bounds on sterile neutrinos stronger than experimental bounds these assumptions stand practically untested.
595: Since these bounds come from a few key measurements
596: (rather than from a redundant set of different observables)
597: compensations among different kinds of new physics are not unnatural.
598: % For example, BBN bounds are today entirely
599: % based on the primordial $^4$He abundancy:
600: % the effects generated by sterile neutrinos
601: % might be compensated by e.g.\ a fermion asymmetry in ordinary neutrinos
602: % (8 orders of magnitude larger than the one in baryons).
603: % The shift in the amount of galaxy clustering
604: % generated by a massive sterile neutrino might be
605: % compensated by e.g.\ a primordial tilt in the spectral index.
606: For instance, the dominant BBN probe~\cite{BBN}, the helium-4 abundancy~\cite{He4},
607: is plagued by controversial systematic uncertainties, and
608: presently it is not clearly incompatible with a fourth thermalized sterile neutrino~\cite{NnuBBN}
609:
610:
611: Therefore we do not try to attribute a precise probabilistic meaning to cosmological bounds
612: by performing global fits.
613: We prefer to identify and compute the key observables.
614: This allows to present the underlying physics in a simple and critical way
615: (while global fits would indirectly involve other cosmological parameters and data).
616: The implications of present or future measurements can be read from our plots.
617: Cosmology is interesting not only because gives indications today,
618: but also because it will allow powerful future searches.
619:
620:
621: \medskip
622:
623: BBN probes the total energy density at $T \sim (0.1 \div 1)\MeV$
624: (dominantly stored in electrons, photons and neutrinos according to the SM)
625: and is also directly sensitive to reactions involving neutrinos (e.g.\ $\bar\nu_e p \leftrightarrow \bar{e} n$). Given a few input parameters
626: (the effective number $N_\nu$ of thermalized relativistic species,
627: the baryon asymmetry $n_B/n_\gamma=\eta$,
628: and possibly the $\nu_\ell /\bar\nu_\ell$ lepton asymmetries)
629: BBN successfully predicts the abundances of several light nuclei~\cite{BBN}.
630: Its non trivial success strongly indicates that primordial BBN really happened.
631: Today $\eta$ is best determined within minimal cosmology by CMB data to be $\eta = (6.15 \pm 0.25) 10^{-10}$~\cite{WMAP}.
632: Thus, neglecting the lepton asymmetries (which is an excellent approximation
633: unless they are much larger than the baryon asymmetry)
634: one can use the observations of primordial abundances to test if $N_\nu=3$
635: as predicted by the SM.
636:
637: Today the $^{4}$He abundancy~\cite{He4} is the most sensitive probe of $N_\nu$.
638: We study also the deuterium abundancy, that might have brighter prospects of future improvements~\cite{DeutQuas}.
639: For arbitrary values of $N_\nu$ around the SM value of 3, BBN predicts~\cite{BBN}
640: \begin{eqnsystem}{sys:HeD}
641: Y_p &\simeq& 0.248+0.0096\ln\frac{\eta}{6.15~10^{-10}} + 0.013 (N_\nu^{^4{\rm He}}-3),\\
642: \frac{Y_{\rm D}}{Y_{\rm H}} &\simeq &(2.75\pm 0.13)~10^{-5}~\frac{1+0.11~(N_\nu^{\rm D}-3)}{(\eta/6.15~10^{-10})^{1.6}}
643: , \end{eqnsystem}
644: where $Y_p \equiv n_{^4{\rm He}}/n_B$.
645: Both $Y_p$ and $Y_{\rm D}$ are plagued
646: by controversial systematic uncertainties.
647: We do not enter into these issues and refer the reader to~\cite{Dolgov,review,NnuBBN,He4,DeutQuas}.
648: Adopting conservative estimates we get
649: \beq\begin{array}{lcl}\label{eq:pD}
650: Y_p = 0.24\pm 0.01 &\Rightarrow& N_\nu^{^4{\rm He}} \simeq 2.4\pm0.7, \\
651: \displaystyle
652: \frac{Y_{\rm D}}{Y_{\rm H}} = (2.8 \pm 0.5)\,10^{-5} &\Rightarrow& N_\nu^{\rm D}\simeq 3\pm 2.
653: \end{array}\eeq
654: In order to safely test and possibly rule out $N_\nu=4$ it is necessary to either
655: $i)$ somewhat improve the determination of $Y_p$, or to
656: $ii)$ improve on $Y_{\rm D}$, somewhat improve on
657: $\eta$ and on the theoretical uncertainty on $Y_{\rm D}$.
658:
659:
660: In our computation, we assume standard cosmology plus a sterile neutrino,
661: and conservatively assume it has zero initial abundancy at $T\gg \MeV$.
662: Oscillations produce sterile neutrinos~\cite{bbnosc}.
663: This is a rather robust phenomenon:
664: it is difficult to modify cosmology in order to avoid
665: production of sterile neutrinos while keeping the success of BBN.
666: In fact sterile neutrinos are dominantly produced at $T\sim \MeV$,
667: simultaneously or after neutrino decoupling.
668:
669:
670: However, a neutrino asymmetry $\eta_\nu$ 8 orders of magnitude larger than $\eta$ in baryons
671: is a significant extra parameter: it affects the $\nu_e$ abundancy and consequently
672: the $n/p$ ratio at freeze-out,
673: and finally the primordial abundances.
674: Also, a relatively large lepton asymmetry, around $10^{-5}$,
675: gives extra MSW effects which can
676: suppress active/sterile oscillations removing cosmological signals~\cite{etarimuove}.
677: More generically, allowing a non vanishing $\eta_\nu$
678: any $N_\nu$ is compatible with the measurement of
679: the helium-4 abundancy,
680: just because a single measurement cannot fix two parameters ($N_\nu$ and $\eta_\nu$).
681: It is important to study how the situation improves when also $Y_{\rm D}$
682: will be precisely measured.
683: For example increasing $N_\nu $ from 3 to 4 increases
684: $Y_{\rm D}$ by $11\%$ and
685: $Y_p$ by $5\%$.
686: If the latter effect were compensated by a large neutrino asymmetry,
687: $Y_{\rm D}$ still remains about $8\%$ higher.
688: Numbers can be somewhat different, depending on how $N_\nu$ gets dynamically
689: increased by sterile oscillations, which are also directly affected by a
690: large neutrino asymmetry in an important way~\cite{etarimuove,asymm}.
691: However the general point remains.
692: Of course one can make any $N_\nu$ allowed by just introducing
693: more than one free parameter;
694: but measuring two or more observables would make hard to believe
695: that new physics effects cancel among each other in all cases.
696:
697:
698:
699:
700: \medskip
701:
702: % Neutrino asymmetries 8 orders of magnitude larger than the one in baryons
703: % can block sterile production and introduce extra unknown parameters in nucleosynthesis,
704: % avoiding bounds and predictions.
705: In the following we stick to standard cosmology.
706:
707:
708: For each choice of oscillation parameters,
709: we follow the evolution with temperature of neutrino abundances
710: and compute how $Y_p$ and $Y_{\rm D}$ are modified.
711: For easy of presentation, we convert their values into effective numbers of neutrinos,
712: $N_\nu^{^4{\rm He}}$ and $N_\nu^{\rm D}$,
713: univocally defined by the inversion of eq.\sys{HeD}.
714: These parameters do not necessarily lie between $3$ and $4$
715: and are employed just as a useful way of presenting our final results.
716:
717:
718: \medskip
719:
720: We also compute another effective number of neutrinos,
721: that parameterizes the total energy density in relativistic species at photon decoupling as\footnote{Including small effects of spectral distortions
722: the precise SM prediction for the effective number of neutrinos is $N_\nu^{\rm CMB} = 3.04$,
723: but we can ignore this subtlety.}
724: \beq\label{eq:CMB}
725: \rho_{\rm relativistic} = \rho_\gamma\bigg[1 + \frac78\bigg(\frac{4}{11}\bigg)^{4/3} N_\nu^{\rm CMB}\bigg].\eeq
726: This quantity (together with other cosmological parameters) determines
727: the pattern of fluctuations of the CMB measured by WMAP (and other experiments).
728: Neutrinos affect CMB in various ways~\cite{NuCDM};
729: in the future studying how
730: neutrino free-streaming shifts the acoustic peaks
731: should offer a clean way of directly counting neutrinos.
732: Global fits at the moment imply~\cite{CMBfits}
733: \beq N_\nu^{\rm CMB} \approx 3\pm2\eeq
734: somewhat depending on which priors and on which data are included in the fit.
735: Future data might start discriminating $3$ from $4$ neutrinos.
736:
737: \medskip
738:
739: Finally, neutrinos can be studied looking at distribution of galaxies because
740: massive neutrinos move without interacting,
741: making galaxies less clustered~\cite{boundMnu}.
742: The effect depends on two parameters, that in absence of sterile neutrinos
743: are both determined by neutrino masses:
744: $1)$ the temperature at which neutrinos become non relativistic,
745: $T_\nu\sim m_\nu/3 $ (so that active neutrinos operated when
746: the horizon of the universe had the size that clusters of galaxies have now);
747: $2)$ the energy density in neutrinos, $\Omega_\nu h^2$,
748: that determines how large is the effect of neutrinos
749: (neutrinos give at most a minor correction).
750: As usual, the parameter $h$ is $H_{\rm today}/(100 \hbox{km/s\,Mpc})$.
751:
752: Global analyses of cosmological data are usually reported as
753: a bound on $\Omega_\nu$, assuming the standard correlation with $T_\nu$~\cite{WMAP,boundMnu}.
754: However, in presence of a non-thermal population of sterile neutrinos these two parameters
755: are no longer universally related.
756: E.g.\ there could be a little number of sterile neutrinos
757: with few eV mass (heavy neutrinos affect also CMB,
758: behaving as cold dark matter).
759: This scenario has not yet been compared with present data (see~\cite{HR} for closely related work).
760: We assume that the bound can be approximated
761: with the standard one on the following quantity:\footnote{We are neglecting CP-violation,
762: and assuming that neutrinos
763: and anti-neutrinos have equal density matrices.}
764: \beq\label{eq:LSS}\Omega_\nu h^2=\frac{\hbox{Tr} [m\cdot \rho]}{93.5\eV}\eeq
765: where $m$ is the $4\times4$ neutrino mass matrix and $\rho$ is the $4\times4$
766: neutrino density matrix, as discussed below.
767: We approximate the present bound with $\Omega_\nu h^2< 0.01$
768: (e.g.\ the WMAP global fit gives $\Omega_\nu h^2<0.76 ~10^{-2}$ at $95\%$ C.L.\
769: in the standard case~\cite{WMAP}.
770: The bound becomes slightly weaker if $N_\nu^{\rm CMB}=4$,
771: or if more conservative priors or estimates of systematic uncertainties are adopted~\cite{boundMnu}).
772: The atmospheric mass splitting guarantees $\Omega_\nu\circa{>} 0.5~10^{-3}$ so that
773: future attempts to reach this level of sensitivity are guaranteed.
774: We assume that a $0.001$ sensitivity in $\Omega_\nu h^2$ will be reached.
775:
776:
777: \subsection{Technical details}
778: Before discussing results, we present the main technical details.
779: In order to compute the observables discussed above one needs to
780: set up the network of relevant Boltzmann equations
781: and study neutrino oscillations in the early universe~\cite{OscUniverse,DolgovReview}.
782: We use our BBN code that includes all main effects,
783: relying on more accurate public codes only to precisely fix the central SM values.
784: This is a complicated computation because many processes
785: proceed at $T\sim \MeV$: active (and maybe sterile)
786: neutrino oscillations, neutrino decoupling,
787: neutron decay, electron decoupling and finally nucleosynthesis.
788: The various processes have been accurately studied in the past.
789: The extension from 2 neutrinos oscillations
790: to 4 neutrinos oscillations
791: does not involve new ingredients,
792: but rewriting and implementing old ones
793: in an appropriate way is not completely trivial
794: due to the presence e.g.\ of oscillations at different frequencies.
795:
796: Brute force alone would not allow
797: to explore the key observables in a vast parameter space:
798: it is necessary to employ accurate approximations
799: that one can invent understanding the physics of BBN.
800: Neglecting spectral distortions
801: possibly induced by active/sterile oscillations,
802: % and we do not run our code for all nuclear species involved in BBN.
803: the Boltzmann equation for the $4\times 4$
804: neutrino (and anti-neutrino) density matrix is~\cite{OscUniverse,DolgovReview}
805: \beq \label{eq:dotrho}\dot\rho = zZH \frac{d\rho}{dz} = i [\Ham,\ \rho] - \{\Gamma,(\rho-\rho^{\rm eq})\}.\eeq
806: where $H$ is the Hubble constant at temperature $T$, $z = m_e/T$.
807: The factor
808: $Z = -3\; \textrm{d} \ln z / \textrm{d} \ln s$ (where $s$ is the entropy density)
809: differs from 1 when the temperature of the universe does not decrease
810: as the inverse of its comoving radius,
811: namely during electron decoupling ($z\sim 1$) and when sterile neutrinos thermalize.
812: In the flavour basis~\cite{OscUniverse}
813: %\beq \Ham =\frac{m m^\dagger}{2E_\nu }-\frac{48\alpha_2 T_\nu}{\pi M_W^4}\bigg[T_\nu^4\cos^2\theta_{\rm W}
814: % \diag(1,1,1,1) + 2T^4\diag(1,0,0,0)\bigg]. \eeq
815: \beq \Ham =\frac{m m^\dagger}{2E_\nu }- \frac{7\pi^3\,T_{\nu }\,{{\alpha }_2}}
816: {15\,M_W^4} \bigg[T_\nu^4\cos^2\theta_{\rm W} \; \diag(\rho_{ee},\rho_{\mu \mu},\rho_{\tau \tau},0) + 2 \; T^4 \; \diag(1,0,0,0)\bigg]\eeq
817: We can neglect contributions from the off-diagonal elements of $\rho$~\cite{OscUniverse}.
818: The usual MSW effect~\cite{MSW} gives an additional subdominant term
819: with different sign for neutrinos and antineutrinos.
820: The average over the neutrino energy spectra is performed
821: using a Fermi-Dirac distributions, and therefore
822: neglecting spectral distortions possibly caused by oscillations.
823:
824: In eq.\eq{dotrho} we use the standard `anticommutator' approximation for the
825: collision terms $\Gamma$, that describe weak $\nu e$ and $\nu\nu$ interactions
826: that tend to thermalize neutrinos, driving
827: their matrix density to its thermal equilibrium value,
828: $\rho^{\rm eq} = {\rm diag}(1,1,1,0)$.
829: A detailed comparison with the full equations~\cite{DolgovReview}
830: reveals that they are accurately mimicked by inserting the following
831: values of the damping coefficients.
832: In the equations for the off-diagonal components of $\rho$, we insert the
833: total scattering rate~\cite{DolgovReview}
834: $$
835: \Gamma_{\rm tot} \approx 3.6 \; G_{\rm F}^2 \; T^5 \quad \textrm{for $\nu_e\qquad$ and}\qquad
836: \Gamma_{\rm tot} \approx 2.5 \; G_{\rm F}^2 \; T^5 \quad\textrm{for $\nu_{\mu,\tau}$}
837: $$
838: because all scatterings damp the coherent interference
839: between different flavours.
840: In the equations for the diagonal components of $\rho_{ii}$,
841: we insert the annihilation rate~\cite{DolgovReview}
842: $$
843: \Gamma_{\rm ann} \approx 0.5 \; G_{\rm F}^2 \; T^5 \quad \textrm{for $\nu_e\qquad$ and}\qquad
844: \Gamma_{\rm ann}\approx 0.3 \; G_{\rm F}^2 \; T^5 \quad\textrm{for $\nu_{\mu,\tau}$}
845: $$
846: since annihilations are needed to change the number of neutrinos.
847: However this procedure~\cite{DolgovReview} introduces an artificial choice of basis,
848: giving equations which are no longer invariant under
849: rotations in the $(\mu,\tau)$ sector.
850: In order to correctly maintain important coherencies among
851: $\nu_\mu$ and $\nu_\tau$, we introduce the distinction
852: between $\Gamma_{\rm tot}$ and $\Gamma_{\rm ann}$
853: in the $\nu_\mu \pm \nu_\tau$ basis
854: (assuming maximal atmospheric mixing; otherwise the generalization is immediate).
855:
856:
857:
858: \medskip
859:
860:
861: After determining neutrino evolution we can study the relative $n/p$ abundancy,
862: that evolves according to~\cite{BBN,bbnosc}
863: $$
864: \dot{r} = zHZ\frac{dr}{dz} = \Gamma_{p\to n} (1-r)- r \Gamma_{n\to p}\qquad
865: r = \frac{n_n}{n_n+n_p} $$
866: where $\Gamma_{p\to n}$ is the total
867: $pe\bar\nu_e\to n$, $pe\to n\nu_e$ and $p\bar\nu_e\to n\bar{e}$
868: reaction rate
869: (in thermal equilibrium the inverse process would satisfy
870: $\Gamma_{n\to p} = \Gamma_{p\to n} e^{(m_n-m_p-m_e)/T}$).
871: The production of sterile neutrinos affects $n/p$ by~\cite{bbnosc}
872: $1)$ increasing the Hubble parameter $H$;
873: $2)$ modifying the $\Gamma_{p\to n},\Gamma_{n\to p}$ rates,
874: if the $\nu_e$ population is depleted by oscillations.
875:
876: \medskip
877:
878: Finally a network of Boltzmann equations describes how
879: electroweak, strong and electromagnetic processes
880: control the evolution of the various nuclei:
881: $p$, $n$, D, T, $^3$He, $^4$He,\ldots
882: Rather than recalling here the main features of these equations,
883: we just state (without explanation) the approximation we use
884: (see also~\cite{BBNapprox}).
885: At a sufficiently low temperature $T^*\sim 0.08\MeV$ almost all neutrons wind up in $^4{\rm He}$,
886: so that its mass abundancy is given by
887: $ Y_p\simeq 2r(T^*)$ with $T^*$ obtained solving
888: $$180 H = \Gamma_{{\rm DD}\to p{\rm T}} (\Gamma_{pn\to {\rm D}\gamma}/\Gamma_{{\rm D}\gamma\to pn})^2.$$
889: The precise numerical value is fixed in such a way that in the SM case our simplified code
890: precisely agrees with state of the art codes
891: (that include thermal, radiative and other corrections corrections,
892: smaller than the present experimental uncertainty).\footnote{A recent paper~\cite{DV}
893: studied, in a four-neutrino context, how active/sterile oscillations affect the
894: $^4$He abundancy.
895: The authors of~\cite{DV} take into account spectral distortions of $\nu_{\rm s}$ (that we neglect),
896: and neglects other minor corrections (that we include), such as those related to the electron mass.
897: They compute an `effective BBN neutrino number' $N_\nu^{\rm BBN}$,
898: which, unlike our $N_\nu^{^4{\rm He}}$ and $N_\nu^{{\rm D}}$,
899: is not directly related to the observable helium-4 and deuterium abundances.
900: However, in the parameter range
901: covered by their plots, their $N_\nu^{\rm BBN}$ should be
902: an approximation to $N_\nu^{^4{\rm He}}$.
903: The plots that can be compared show a reasonable level of agreement.
904:
905: At $\Delta m^2\sim 10^{-8}\eV^2$ spectral distortions of electron neutrinos
906: are not negligible: in the two-neutrino limit they make the helium-4
907: abundancy more sensitive to sterile oscillations
908: (see the papers by Kirilova et al.\ in~\cite{asymm}).}
909: The deuterium abundancy is obtained with a similar technique.
910:
911:
912:
913: \begin{figure}[t]
914: $$\includegraphics[width=17cm]{2nu}$$
915: \caption{\label{fig:2nu}\em Isocurves to the effective
916: number of neutrinos produced by 2 neutrino oscillations in the cases
917: $\nu_e/\nu_{\rm s}$ (left plot) and $\nu_{\mu,\tau}/\nu_{\rm s}$.
918: Solar and atmospheric oscillations are included
919: in the 3 neutrino plots of fig.\fig{BBN}, where the meaning of the various isolines
920: is precisely explained.}
921: \end{figure}
922:
923: \begin{figure}[p]
924: $$\hspace{-8mm}\includegraphics[width=18cm]{BBN}$$
925: \caption{\label{fig:BBN}\em {\bf Cosmological effects of sterile neutrino oscillations}.
926: We compare four different signals.
927: {\color{rosso} The continuous red line refers to the $^4\hbox{\rm He}$ abundancy}
928: (we shaded as `disfavoured' regions where its value corresponds to $N_\nu > 3.8$),
929: {\color{viola} the violet dotted line to the deuterium abundancy}, and
930: {\color{blu} the dashed blue line to the effective number of neutrinos at recombination}.
931: We plotted isolines of these three signals corresponding to an effective number of
932: neutrinos $N_\nu=3.2$ and $3.8$.
933: The precise meaning of the parameter $N_\nu$ in the three cases is explained in the text.
934: {\color{rossoc}
935: The upper (lower) dot-dashed orange lines corresponds to $\Omega_\nu h^2= 10^{-2}$ $(10^{-3})$},
936: where $\Omega_\nu$ is the present energy density in neutrinos.}
937: \end{figure}
938:
939:
940: \subsection{Results}
941: We plot the effective numbers $N_\nu$ of neutrinos defined in terms of
942: the physical observables (the $^4$He and D abundances and the energy density
943: at recombination)
944: from eq.s~(\ref{sys:HeD}) and \eq{CMB}.
945: We also plot the value of the present energy density in neutrinos $\Omega_\nu $,
946: probed by observations of Large Scale Structure
947: together with CMB constraints.
948:
949: The plots have the following meaning:
950: shaded regions have
951: $N^{^4{\rm He}}_\nu> 3.8$
952: or $\Omega_\nu h^2>10^{-2}$ and are therefore `disfavoured' or `excluded'
953: (depending on how conservatively one estimates systematic uncertainties)
954: within minimal cosmology.
955: The other lines indicate the sensitivity that future experiments might reach.
956: More precisely we plot contour-lines corresponding to
957: $N_\nu = 3.2$ and $3.8$
958: and to $\Omega_\nu h^2 = 10^{-2}$ and $10^{-3}$.
959:
960: \medskip
961:
962: It is useful to start discussing the unrealistic but simple cases considered in old papers~\cite{bbnosc}.
963: In fig.\fig{2nu} we show the effects produced by
964: 2 neutrino mixing: $\nu_{\rm s}/\nu_e$ in fig.\fig{2nu}a and
965: $\nu_{\rm s}/\nu_{\mu}$ or $\nu_{\rm s}/\nu_\tau$ mixing in fig.\fig{2nu}b.\footnote{Previous papers
966: studied the $^4$He abundancy and we agree with their results.
967: We however use as a variable $\tan^2\theta$ rather than $\sin^22\theta$,
968: so that we unify in a unique plot the non-resonant ($0<\theta < \pi/4$)
969: and the resonant ($ \pi/4<\theta<\pi/2$) case.}
970: The red dashed line shows the total number of neutrinos, $N_\nu^{\rm CMB}$:
971: it essentially does not depend on which flavour ($\nu_e$, $\nu_\mu$ or $\nu_\tau$)
972: mixes with $\nu_{\rm s}$ and is not affected by oscillations with $\Delta m^2 \circa{<} 10^{-5}\eV^2$
973: that are too slow and start only after neutrino decoupling,
974: when the total number of neutrinos is frozen.\footnote{To be
975: precise we should say `the total entropy in neutrinos per comoving volume
976: remains constant'.
977: For simplicity we will adopt such loose abbreviations.}
978: At this stage neutrinos can still change flavour.
979: The difference between fig.\fig{2nu}a and\fig{2nu}b
980: is due to the fact that only
981: electron neutrinos are involved in the reactions that control the $n/p$ ratio.
982: Therefore $\nu_{\rm s}/\nu_e$ oscillations that occur after neutrino freeze-out and that do not
983: affect the total number of neutrinos ($\nu_{\rm s}$ are created by depleting $\nu_e$)
984: affect $n/p$ and consequently the $^4$He abundancy\footnote{These region
985: are strongly disfavoured because have a helium-4 abundancy
986: corresponding to $N_\nu^{^4{\rm He}} >4$, up to about 5.} (continuous line)~\cite{bbnosc}
987: and, to a lesser extent, the D abundancy.
988: This happens down to $\Delta m^2 \sim 10^{-8}\eV^2$:
989: oscillations with $\Delta m^2 \circa{<} 10^{-8}\eV^2$ occur after decoupling of electroweak scatterings,
990: when the relative $n/p$ abundancy is only affected by neutron decay.
991:
992: Effects are larger at $\theta > \pi/4$ (i.e.\ $\tan\theta>1$) because this corresponds
993: to having a mostly sterile state lighter than the mostly active state,
994: giving rise to MSW resonances in neutrinos and anti-neutrinos
995: (like in cosmology, also
996: supernova $\bar\nu_e$ feel a MSW resonance for $\theta_{\rm s}\circa{>}\pi/4$.
997: On the contrary solar neutrinos feel a resonance for $\theta_{\rm s}\circa{<}\pi/4$).
998: In the past years it has been debated about
999: if a neutrino asymmetry and/or large inhomogeneity
1000: develop as a consequence of non-linear effects,
1001: and this issue has not yet been fully clarified.
1002: Our Boltzmann equations assume that both these effects can be neglected.
1003: At $\tan\theta>1$ the bound from $\Omega_\nu$ holds
1004: even for very small mixing, $\theta\simeq\pi/2$ just because these region correspond
1005: to heavy active neutrinos.
1006:
1007:
1008:
1009: \bigskip
1010:
1011: We now discuss how the above picture changes taking into account oscillations among active
1012: neutrinos.
1013: Our results are shown in fig.\fig{BBN}:
1014: the upper row refers to sterile mixing with mass eigenstates $\nu_{1,2,3}$ and
1015: the lower row to mixing with flavour eigenstates $\nu_e$ and $\nu_{\mu,\tau}$.
1016:
1017: An inspection of the upper row shows that their main features can
1018: be understood in terms of the (unrealistic) results in the case of $2$ neutrino mixing,
1019: fig.\fig{2nu}.
1020: Having assumed $\theta_{13}=0$, $4\nu$ sterile mixing with $\nu_3$
1021: gives no new effects with respect to $2\nu$ sterile mixing with $\nu_{\mu,\tau}$.
1022: Due to solar and atmospheric oscillations, $\nu_e$ depletion due to
1023: oscillations into sterile neutrinos now happens in all other cases
1024: and becomes milder, because
1025: no longer confined to $\nu_e$ but shared among
1026: all active neutrinos.
1027: Fig.\fig{BBN}b shows the effects of
1028: $\nu_{\rm s}/\nu_2$ mixing (this kind of neutrino spectrum is plotted in fig.\fig{spettrias}b):
1029: since $\nu_2$ contains some $\nu_e$ component,
1030: electron neutrinos are in part directly affected.
1031: Fig.\fig{BBN}a shows the effects of
1032: $\nu_{\rm s}/\nu_1$ mixing: $\nu_e$ depletion effects are largest in this last case
1033: because $\nu_1$ is the neutrino eigenstate with the largest
1034: $\nu_e$ component.
1035: In summary: depletion gets transferred to all $\nu$ flavours and diluted.
1036:
1037:
1038: Mixing with flavour eigenstates is qualitatively different,
1039: for the general reasons explained in section~\ref{osc}.
1040: We can see the effects of the solar (atmospheric)
1041: mass splitting as bumps in fig.\fig{BBN}d (\ref{fig:BBN}e)
1042: where cosmological effects of $\nu_{\rm s}/\nu_e$ ($\nu_{\rm s}/\nu_{\mu,\tau}$) mixing are computed.
1043: In the case of $\nu_{\rm s}/\nu_e$ mixing
1044: $N_\nu^{\rm D}, N_\nu^{\rm CMB} =4$ is reached
1045: only if the sterile neutrino has a large enough $\Delta m^2$:
1046: the solar $\Delta m^2_{\rm sun}\approx 0.7~10^{-4}\eV^2$ alone
1047: is not sufficient, as also indicated by the $2\nu$ limit plotted in fig.\fig{2nu}a.
1048: On the contrary, fig.\fig{2nu}b shows that
1049: in the case of $\nu_{\rm s}/\nu_{\mu,\tau}$ mixing a
1050: $\Delta m^2\sim \Delta m^2_{\rm atm}\approx 2~10^{-3}\eV^2$ is large enough
1051: to reach $N_\nu\approx 4$ for any value of the sterile mass.
1052: We have verified that setting $\theta_{13}\sim 0.2$
1053: fig.s\fig{BBN} do not vary in a significant way.
1054:
1055:
1056: \subsection{Hints and anomalies: cosmology}
1057: To conclude we list cosmological data that
1058: do not fit well into the scheme presently considered as standard,
1059: and that can be interpreted as manifestations of sterile neutrino effects:
1060: \begin{itemize}
1061:
1062:
1063: \item
1064: Various determinations of the primordial helium-4 abundancy $Y_p$
1065: try to reduce uncertainties by appropriately choosing and modeling the astrophysical systems
1066: used for the observation~\cite{He4}.
1067: Some analyses find lower values of $Y_p$
1068: corresponding to less than 3 neutrinos
1069: (an effect which could come from active-sterile oscillations),
1070: even after a detailed examination of the systematic uncertainties.
1071: Other determinations give higher values %($Y_p=0.244 \pm 0.002$).
1072: of $Y_p$, more compatible with $N_\nu=3$.
1073:
1074:
1075: \item $X$-ray cluster data seem to prefer a lower value of the parameter $\sigma_8$
1076: than the other sets of CMB and LSS data, considered in~\cite{WMAP,boundMnu}.
1077: A degenerate spectrum of active neutrinos can alleviate this discrepancy~\cite{Allen}.
1078: Otherwise, one can invoke a contribution to $\Omega_\nu$ from sterile neutrinos.
1079:
1080:
1081: \item Decays of sterile neutrinos with mass $\sim 200\MeV$
1082: provide an interpretation of the reionization
1083: at large redshift observed by WMAP,
1084: alternative to the standard one
1085: (early formation of massive stars)~\cite{reionization}.
1086:
1087:
1088: \item Decays of sterile neutrinos with mass $\sim 10\MeV$
1089: and abundancy $\Omega_\nu=10^{-5\div 9}$
1090: can be a non standard source of recently observed galactic positrons~\cite{Picciotto}.
1091:
1092: \end{itemize}
1093:
1094:
1095:
1096:
1097:
1098: \section{Sterile effects in solar (and KamLAND) neutrinos}\label{solar}
1099: We compare sterile effects with present and future solar neutrino experiments
1100: and with KamLAND reactor anti-neutrino data.
1101: While computing sterile effects in reactor $\bar\nu_e$ is straightforward,
1102: solar neutrinos are detected after a long trip from
1103: the center of the sun during which they can
1104: experience sterile effects in several different ways.
1105: This is what makes solar neutrinos a powerful probe of sterile effects.
1106: In section~\ref{suntec} we describe how we precisely compute these effects.
1107: Section~\ref{sunfit} describes how we fit data, dealing with the complication
1108: that these data contain a positive evidence for active/active oscillations.
1109: Results and a qualitative understanding are presented in section~\ref{solarres}.
1110: In section~\ref{compare} we compare our results with previous analyses
1111: performed in limiting cases.
1112:
1113:
1114:
1115: \subsection{Technical details}\label{suntec}
1116: In order to understand what happens when a sterile neutrino is added
1117: and to write a sufficiently fast numerical code,
1118: we must develop an analytical approximation.
1119: It is convenient to employ some more formalism
1120: than in the $2\times 2$ case
1121: and study the evolution of the $4\times4$
1122: neutrino density matrix $\rho_m$ written in the $m$ass basis of instantaneous mass eigenstates.
1123: A $\nu_e$ produced with energy $E_\nu$ at radius $r=r_0$ inside the sun is described by
1124: $\rho_m = V_m^\dagger\cdot
1125: \hbox{diag}\,(1,0,0,0)\cdot V_m $
1126: where $V_m$ depends on $E_\nu$ and $r_0$.
1127: Mixing matrices in matter ($V_m$) and vacuum ($V$)
1128: are computed diagonalizing the Hamiltonian~\cite{MSW}
1129: $$\Ham = \frac{mm^\dagger}{2E_\nu} + \sqrt{2}G_{\rm F}\diag(N_e-\frac{N_n}{2},-\frac{N_n}{2},-\frac{N_n}{2},0)$$
1130: and ordering eigenstates according to their eigenvalues $H_i \equiv m_{\nu_{mi}}^2/2E_\nu$:
1131: $\nu_{m1}$ ($\nu_{m4}$) is the lightest (heaviest) neutrino mass eigenstate in matter.
1132: The evolution up to the detection point is described by
1133: a $4\times4$ unitary evolution matrix $\mathscr{U}$ so that
1134: at detection point the density matrix in the basis of flavour eigenstates is
1135: $\rho =\langle V\cdot \mathscr{U}\cdot\rho_m(r,E_\nu)\cdot \mathscr{U}^\dagger \cdot V^\dagger\rangle$.
1136: where $\langle\cdots\rangle$ denotes average over the production point.
1137: The various observables involve additional averages over neutrino energy
1138: and time (and consequently over different paths in the earth and in vacuum).
1139: The oscillation probabilities are given by $P(\nu_e\to \nu_e)=\rho_{ee}$,
1140: $P(\nu_e\to \nu_{\rm s})= \rho_{ss}$, etc.
1141:
1142:
1143: We briefly describe the main steps in the computation of $\mathscr{U}$, focussing on the subtle ones.
1144: The evolution matrix can be decomposed as
1145: $$
1146: \mathscr{U} = \mathscr{U} _{\rm earth}\cdot \mathscr{U} _{\rm vacuum}\cdot \mathscr{U} _{\rm sun}.$$
1147: Evolution in vacuum is given by
1148: $\mathscr{U} _{\rm vacuum} = \diag\exp (-i L m_{\nu_i}^2/2E_\nu)$.
1149: Combined with average over neutrino energy it suppresses
1150: the off-diagonal element $\rho_m^{ij}$ when
1151: the phase differences among eigenstates $i$ and $j$ are large.
1152:
1153:
1154: Evolution in the earth is computed in
1155: mantle/core approximation, improved by
1156: using the average density appropriate for each
1157: trajectory as predicted in~\cite{PREM}.
1158: Therefore we use $ \mathscr{U} _{\rm earth}=1$ when the earth is not crossed,
1159: $ \mathscr{U} _{\rm earth}
1160: =P\cdot\diag \exp (-i L_{\rm mantle} m_{\nu_{mi}}^2/2E_\nu)\cdot P^\dagger$
1161: when only the mantle is crossed (for a length $L_{\rm mantle}$),
1162: and the obvious generalization
1163: when both mantle and core are crossed.
1164: $P=V^\dagger_B V_A$ is a non-adiabaticity factor
1165: that takes into account the sharp flavour variation of mass eigenstates
1166: when passing from medium $A$ (vacuum) into medium $B$ (the earth mantle)
1167: \label{Psun}.
1168:
1169: Evolution in the sun is more complicated because there can be various
1170: $P$ factors at non-adiabatic level crossings at radii $r_n$ ($n=1,2,\ldots, n_{\rm max}$), that give
1171: $$ \mathscr{U}_{\rm sun} = P_{r_n}\cdots P_{r_2}\cdot
1172: \diag\exp(-i\int_{r_1}^{r_2} ds~\frac{ m_{\nu_{mi}}^2}{2E_\nu})\cdot P_{r_1}\cdot
1173: \diag\exp(-i\int_{r_0}^{r_1} ds~\frac{ m_{\nu_{mi}}^2}{2E_\nu}) $$
1174: The number of level crossings $n_{\rm max}$ ranges between 0 and a few,
1175: e.g.\ a $\nu_e$ produced in the side of the sun
1176: farther from us can experience 4 crossings.
1177: When levels $i$ and $j$ cross in an adiabatic way, $P_{r_n}=1$.
1178: If instead level crossing is fully non adiabatic
1179: $P_{r_n} = V_m^\dagger(r\circa{<}r_n)\cdot V_m(r\circa{>}r_n)$
1180: is a rotation with angle $\alpha = 90^\circ$ in the $(ij)$ plane.
1181: In general the rotation angle is given by $\tan^2\alpha = P_C/(1-P_C)$,
1182: where $P_C$ is the level crossing probability.
1183:
1184:
1185:
1186: \medskip
1187:
1188: So far we only presented the well known formalism~\cite{MSW,Petcov,KPreview}
1189: in a non standard way appropriate for applying it in numerical computations with
1190: multiple and overlapping level crossings.
1191: We now need to compute $P_C$.
1192: This last step turns out to be non trivial.
1193: %Let us begin with an important remark.
1194: By generalizing well known results valid in the simpler $2\times2$ case~\cite{Petcov,KPreview},
1195: we find that in all the parameter range $P_C$ can be accurately
1196: approximated analytically (i.e.\ no need of numerically solving the
1197: differential neutrino evolution equation $i\,d\rho/ds=[\mathscr{H},\rho]$).
1198: Such a simple result is possible because,
1199: in the relevant neutrino energy range, level crossings are non adiabatic only for
1200: $\theta^2_{\rm s}\cdot \Delta m^2/10^{-8}\eV^2\gg1$
1201: i.e.\ when either $\theta$ is small or the active/sterile
1202: $\Delta m^2$ is much smaller than the LMA splitting.
1203:
1204:
1205: % It is convenient to follow the evolution of the $4\times4$
1206: % neutrino density matrix and use in the sun the MSW level crossing approximation,
1207: % which is adequate in all the relevant parameter space.
1208:
1209: % At sufficiently large active/sterile mixing, ,
1210: % active/sterile resonances are adiabatic in the relevant neutrino energy range.
1211: % In such a case the $4\times 4$ neutrino evolution matrix is $U_{\rm sun} (E_\nu)= V_m(E_\nu)\cdot P\cdot V^\dagger$
1212: % where $P$ is the unity matrix and
1213: % $V_m$ is the mixing matrix at the production point.
1214: % At smaller $\theta_{\rm s}$ or $\Delta m^2$ active/sterile
1215: % resonances are non adiabatic and one
1216: % must compute level-crossing factors, $P_C$.
1217: % When levels $i$ and $j$ cross,
1218: % the factor $P$ becomes a rotation in the $(ij)$ plane
1219: % with angle $\tan^2\alpha = P_C/(1-P_C)$.
1220: % In presence of multiple level crossings
1221: % one multiplies the evolution matrices for each crossing
1222:
1223:
1224:
1225:
1226: % We are interested in
1227: % active/sterile resonances, where $H_{ss}$ is constant and $H_{aa}$ is a non trivial function
1228: % of $r$, as shown in fig.\fig, due to overlappiing with LMA.
1229:
1230:
1231: To compute $P_C$ it is convenient to consider the
1232: basis of mass eigenstates in absence of active/sterile mixing
1233: (i.e.\ $\theta_{\rm s}\to 0$ or $\theta_{\rm s}\to \pi/2$, depending
1234: on which limit is closer to the value of $\theta_{\rm s}$ under examination.
1235: Here we focus on $\theta_{\rm s}\to 0$.).
1236: This limit allows to precisely define level-crossings.
1237: When $\nu_{\rm s}$ crosses one of the active eigenstates, $\nu_{a}^m$ ($a=1,2,3$)
1238: (see fig.\fig{spettrias}a for one example),
1239: the level crossing probability $P_C$ is well approximated by
1240: \begin{equation}\label{eq:PC}
1241: P_C = \frac{e^{\tilde\gamma \cos^2\theta^m_{as}}-1}{e^{\tilde\gamma} -1}\qquad
1242: \gamma =\frac{ 4 \mathscr{H}_{as}^2}{d H_a/dr}
1243: \equiv
1244: \tilde\gamma \cdot \frac{\sin^2 2\theta^m_{as}}{2\pi| \cos2\theta_{as}^m|}
1245: \quad\hbox{where}\quad \sin\theta_{as}^m = \vec{n}\cdot \vec{\nu}_a^m \sin\theta_{\rm s} .
1246: \end{equation}
1247: The above equation might seem a complicated way of rewriting
1248: the well known
1249: expression for $P_C$ valid in the simpler $2\nu$ case~\cite{Petcov,KPreview},
1250: but there is one important difference.
1251: In the $2\nu$ case one can write the result in an analogous way,
1252: which contains the mixing angle in vacuum
1253: and the off-diagonal elements of the Hamiltonian in vacuum.
1254: On the contrary,
1255: our $\gamma$ and $\theta^m_{as}$ must be computed
1256: around the resonance, where $\mathscr{H}_{aa}=\mathscr{H}_{ss}$ (or around the point where adiabaticity is maximally violated, in cases where there is no resonance).
1257: We emphasize that reducing the full $4\times 4$ Hamiltonian to the effective
1258: $2\times 2$ Hamiltonian of the 2 states that cross
1259: and computing $\mathscr{H}_{as}$ is non trivial,
1260: since sterile mixing sometimes redefines
1261: the flavour of the active neutrino involved in the crossing.\footnote{Neglecting
1262: this subtlety
1263: would give a qualitatively wrong result e.g.\ in the following situation:
1264: $\nu_{\rm s}$ is mixed with $\nu_e$ and is quasi degenerate to $\nu_1$.}
1265: In these situations it is useful to know the
1266: physical meaning of $2\mathscr{H}_{as}$: it is the minimal difference between the eigenvalues
1267: of the two states that cross.
1268: In order to elucidate the physical meaning of
1269: the crossing angle $\theta^m_{as}$, we emphasize that
1270: it can also be extracted from
1271: the scalar product between the flavour vectors of the two matter eigenstates
1272: $i$ and $i+1$
1273: that cross: $\sin \theta^m_{as} = \nu_{m_i}^*(r\circa{<} r_n)\cdot \nu_{m_i}(r\circa{>} r_n)$
1274: and $\cos \theta^m_{as} = \nu_{m_i}^*(r\circa{<} r_n)\cdot \nu_{m_{i+1}}(r\circa{>} r_n)$.
1275:
1276:
1277:
1278:
1279: \smallskip
1280:
1281: When the active/sterile mixing is large,
1282: $P_C$ is sizable only if $\Delta m^2_{as}\ll \Delta m^2_{\rm LMA}$:
1283: in this case our expression\eq{PC} reduces to the standard 2-neutrino formula.
1284: In the narrow resonance limit (e.g.\ for $\theta_{as}^m\ll 1$),
1285: eq.\eq{PC} reduces to a Landau-Zener form
1286: $P_C \simeq e^{-\pi\gamma/2}$:
1287: our expression for the adiabaticity factor $\gamma$
1288: holds for a generic $2\times2$ Hamiltonian.
1289: In the sun, resonances with $\theta^m_{as}$ close to $\pi/4$
1290: happen only in the quasi-vacuum region at the border of the sun.
1291: This will be no longer true when studying supernova neutrinos,
1292: that will necessitate an extension of eq.\eq{PC}.
1293:
1294:
1295: %\footnote{More generically we find that the that controls the
1296: %level crossing probability
1297: %$P_C \simeq e^{-\pi\gamma/2}$
1298: %at a narrow resonance between two
1299: %states $n$ and $m$, whose evolution is described by a {\em generic}
1300: %Hamiltonian $i\dot \psi = H(t)\psi$
1301: %is
1302: %$$ \gamma = \bigg|\frac{4 H_{mn}^2}{\dot{H}_{mm}-\dot{H}_{nn}}\bigg|_{\rm res}$$
1303: %evaluated at the resonance point where $H_{nn}=H_{mm}$.}.
1304:
1305:
1306:
1307:
1308: \begin{figure}
1309: $$\includegraphics[width=8cm]{sunfits}\hspace{1cm}
1310: \includegraphics[width=8cm]{sunKLfits}$$
1311: \caption{\label{fig:sunfits}\em
1312: We compare the usual `active only'
1313: global fit of solar (shaded regions in the left plot) and of
1314: solar plus KamLAND data (shaded regions in the right plot),
1315: with the Gaussian approximation employed in this paper (dashed lines)}
1316: \end{figure}
1317:
1318:
1319:
1320: \subsection{Experimental data and fit procedure}\label{sunfit}
1321: We fit all available latest data:
1322: \begin{itemize}
1323:
1324: \item The SNO CC+NC+ES spectra~\cite{SNOlast}, divided in 34 bins (17 energy bins times 2 day/night bins).
1325:
1326: \item The total CC, NC and ES rates measured by SNO with enhanced NC sensitivity~\cite{SNOsalt}.
1327:
1328: \item The Super-Kamiokande ES spectra~\cite{SKlast}, divided
1329: in 44 zenith-angle and energy bins.
1330:
1331:
1332: \item The Gallium rate~\cite{Galliumlast}, $R_{\rm Ga} = (68.0\pm3.8)\,{\rm SNU}$, obtained averaging
1333: the most recent SAGE, Gallex and GNO data.
1334:
1335: \item The Chlorine rate~\cite{Chlorinelast},
1336: $R_{\rm Cl} = (2.56 \pm 0.23)\,{\rm SNU}$.
1337:
1338: \item The KamLAND reactor anti-neutrino
1339: data, divided in 13 energy bins with prompt energy higher than $2.6\MeV$~\cite{KamLAND}.
1340:
1341: % \item The CHOOZ reactor anti-neutrino data, divided in 14 bins and fitted as in analysis ``A'' of~\cite{CHOOZlast}.
1342: \end{itemize}
1343: We revised solar model predictions and uncertainties~\cite{BP}
1344: including the recent measurement of the
1345: $^{14}$N$(p,\gamma)^{15}$O nuclear cross section~\cite{LUNA},
1346: which reduces the predicted CNO fluxes by roughly $50\%$.
1347:
1348:
1349:
1350: Data are compared with predictions forming a $\chi^2$ that takes
1351: into account statistical, systematic and theoretical uncertainties
1352: (on the total solar neutrino fluxes and on the $^8$B spectrum)
1353: and their correlations~\cite{BP}.
1354: We plot the $\chi^2$ as function of the 2 parameters that
1355: describe sterile oscillations,
1356: marginalizing the full $\chi^2$ with respect to
1357: all other sources of uncertainty {\em including the LMA parameters}
1358: $\Delta m^2_{\rm sun}$ and $\theta_{\rm sun}$.
1359: This step is of course not performed in usual analyses.
1360: Proceeding in a fully numerical way, it would be too demanding for present computers.
1361: We can however approximate all observables with a first order Taylor expansion
1362: around the best-fit LMA point,
1363: since experiments allow only relatively minor shifts from it.
1364: In this way marginalization over $\Delta m^2_{12}$ and $\theta_{\rm sun}$
1365: can be performed analytically, using the same Gaussian techniques
1366: commonly employed for other `systematic' parameters.
1367: Fig.a\fig{sunfits} shows that performing this linearization (dashed lines) we obtain
1368: a satisfactory approximation to the usual active-only global fit of solar neutrinos
1369: (continuous lines).
1370: With more data this approximation will become more and more accurate.
1371:
1372: Solar $\nu$ and reactor $\bar\nu$ data cannot be fit independently,
1373: since both sets of data depend on solar oscillation parameters.
1374: Inclusion of KamLAND data presents a slight complication:
1375: due to poor statistics, the Poissonian distribution must be used.
1376: Nevertheless small deviations from the LMA best fit
1377: (due to sterile effects and to systematic uncertainties)
1378: can be taken into account in Gaussian approximation, obtaining
1379: $$\chi^2 = 2(t-r\ln t)\simeq 2(t_0-r\ln t_0)+2\epsilon(t_0-r) +r\epsilon^2$$
1380: where $r$ are the observed rates and $t = t_0(1 + \epsilon)$ the predicted rates.
1381: In our analysis $\epsilon$ is the correction due to sterile neutrinos.
1382: Fig.\fig{sunfits}b shows that we can accurately analytically approximate the
1383: usual active-only global fit of solar plus KamLAND data.
1384:
1385: %\medskip
1386:
1387: %The approximation presented above does not allow to study
1388: %if a good fit can be obtained with active oscillation parameters
1389: %much outside the best-fit LMA region.
1390: %We have therefore performed a dedicated study of the only particular case
1391: %which looks theoretically interesting~\cite{AllDmm} and potentially successful.
1392: %We studied if it is possible to successfully describe data with a large
1393: %splitting among active neutrinos,
1394: %$\Delta m^2_{\rm sun}$ around $ 7~10^{-4}\eV^2$.
1395: %Without sterile neutrinos this gives a quasi-energy-independent
1396: %survival probability of solar neutrinos,
1397: %$P_{ee}\simeq 1 - \frac12 \sin^2 2\theta_{\rm sun} \ge 1/2$,
1398: %which is excluded by SNO data.
1399: %We find that, although sterile neutrinos can (for various specific values of
1400: %its oscillation parameters) reduce the disagreement
1401: %by lowering $P_{ee}$ at higher Boron energies,
1402: %the global fit remains significantly worse than LMA.
1403:
1404:
1405:
1406:
1407:
1408: \begin{figure}[t]$$
1409: \includegraphics[width=8cm]{NuSun2}\hspace{1cm}
1410: %\includegraphics{NuCosmo}
1411: \raisebox{1mm}{\includegraphics[width=8cm]{NuSN}}$$
1412: \caption{\label{fig:levels}\em {\bf Level crossing schemes}.
1413: The right plot shows qualitatively the effective anti-neutrino masses in a supernova.
1414: The left plot shows the effective neutrino masses in the sun.
1415: We assumed hierarchical active neutrinos
1416: (i.e.\ $m_1 = 0$ and $m_2 = (\Delta m^2_{\rm sun})^{1/2}$)
1417: and plotted the two matter eigenstates that give rise to LMA oscillations
1418: for two different values of $E_\nu$:
1419: $10\MeV$ and $0.86\MeV$, the energy of the main Beryllium line.
1420: Colors indicate the flavour composition.
1421: An extra sterile neutrino with small mixing is represented by an horizontal line
1422: with height equal to its mass.
1423: %and in the early universe (right plot, for $E_\nu = 3T$) [ORA SN] .
1424: }
1425: \end{figure}
1426:
1427:
1428:
1429: \begin{figure}[p]
1430: $$\hspace{-5mm}\includegraphics[width=8cm]{sun1}\hspace{5mm}
1431: \includegraphics[width=8cm]{sun2}$$
1432: $$\hspace{-5mm}\includegraphics[width=8cm]{sune}\hspace{5mm}
1433: \includegraphics[width=8cm]{sunmu}$$
1434: \caption{\label{fig:sun}\em {\bf Sterile mixing:
1435: effects in solar neutrinos.} No statistically significant evidence is found.
1436: Shaded regions: excluded at $90,\,99\%$ C.L.
1437: Coloured lines are iso-curves of a few promising signals.
1438: {\color{rossos} Continuous red line: $A_{\rm d/n}^{\rm ES}$ differs from LMA by 0.005}.
1439: {\color{blu} Continuous blue line: 0.02 day/night asymmetry at Borexino}.
1440: {\color{viola} Dashed violet line: seasonal variation at Borexino with 0.02 amplitude}.
1441: {\color{verdes} Short-dashed green line: $P_{ee}$ at sub-MeV energies differs from LMA by 0.02}.
1442: Other signals are discussed in the text.
1443: The letters {\rm A},\ldots,{\rm F} indicate sample points, studied in detail in fig.\fig{samples}.
1444: }
1445: \end{figure}
1446:
1447:
1448: \begin{figure}
1449: $$\hspace{-5mm}\includegraphics[width=18cm]{samples}$$
1450: \caption{\label{fig:samples}\em
1451: A few samples of still allowed sterile effects in solar neutrinos.
1452: We plot, as function of the neutrino energy,
1453: {\color{rossos} $P(\nu_e\to\nu_e)$ (decreasing red curve)},
1454: {\color{blus} $P(\nu_e\to\nu_{\mu,\tau}$) (increasing blue curve)} and
1455: $P(\nu_e\to \nu_{\rm s})$ (lower black curve).
1456: The continuous (dotted) curve are the values during day (night).
1457: The sample points {\rm A},\ldots, {\rm F} are drawn in fig.\fig{sun} as dots.
1458: }
1459: \end{figure}
1460:
1461: \subsection{Results}\label{solarres}
1462: We start recalling how LMA oscillations behave in absence of sterile neutrinos~\cite{MSW}.
1463: Fig.\fig{levels}a shows the composition of the two neutrino mass eigenstates.
1464: For sake of illustration we assumed normal hierarchy,
1465: $m_1^2\ll m_2^2 \simeq \Delta m^2_{\rm sun}
1466: \ll m_3^2\simeq\Delta m^2_{\rm atm}$.
1467: At higher neutrino energies,
1468: $$E_\nu\gg E_* \approx
1469: \Delta m^2_{\rm sun}/G_{\rm F}N_e^{\rm sun}\sim\hbox{few MeV},$$
1470: matter effects dominate around the center of the sun, $r\circa{<}0.2 R_{\rm sun}$
1471: where neutrinos are produced as $\nu_e \approx \nu_{2m}$.
1472: The LMA level crossing at $r\approx 0.2 R_{\rm sun}$ is adiabatic
1473: (in fact the solar mixing angle $\theta_{\rm sun}$ is so large that
1474: fig.\fig{levels} does not look like a level crossing)
1475: so that neutrinos produced as $\nu_{2m}$ exit from the sun as
1476: $\nu_2 = \sin\theta_{\rm sun} ~\nu_e + \cos\theta_{\rm sun}~\nu_{\mu,\tau}$
1477: i.e. $P_{ee} = \sin^2 \theta_{\rm sun}$.
1478: This limit roughly holds at energies probed by SNO and SK:
1479: e.g.\ their total rate is $P_{ee} \approx 1.15 \sin^2 \theta_{\rm sun}$.
1480: At $E_\nu\circa{<} E_*$ $P_{ee}$ increases and
1481: $\nu_{1m}$
1482: contains some $\nu_e$ because matter effects are no longer dominant.
1483: At $E_\nu \ll E_*$ matter effects are negligible and
1484: one gets averaged vacuum oscillations,
1485: $P_{ee} = 1 - \frac12 \sin^22\theta_{\rm sun}$.
1486: This energy range has been explored by Gallium experiments.
1487:
1488: \medskip
1489:
1490: We now discuss how to understand qualitatively the sterile/active mixing effects~\cite{MSW}.
1491: If sterile/active mixing is small, the mostly $\nu_{\rm s}$ state is represented by adding
1492: one quasi-horizontal line to fig.\fig{levels}a.
1493: Depending on its height (determined by the mass of the sterile neutrino)
1494: the mostly sterile level
1495: crosses one or none of the two mostly active neutrinos
1496: (for all relevant neutrino energies the sterile state does not cross both active neutrinos)
1497: after or before the LMA resonance (or after and before).
1498: In each case one can understand the behavior of the survival probabilities
1499: from the level-crossing scheme:
1500: in the example plotted in fig.\fig{levels}a
1501: a neutrino produced at $r_0\approx 0.2$ experiences a single level
1502: crossing at $r_1\approx 0.3$.
1503: This example corresponds to the case considered in~\cite{SmirnovSterile}:
1504: a sterile neutrino weakly mixed with $\nu_1$,
1505: with mass splitting $\Delta m^2_{14}$ somewhat smaller than $\Delta m^2_{12}$.
1506: A solar $\nu_e$ produced at $r\sim 0.2 R_{\rm sun}$ contains a $\nu_{1m}$ component that
1507: crosses the $\nu_{\rm s}$ state once, getting partially converted into $\nu_{\rm s}$.
1508: This gives a dip in the survival probability at intermediate energies:
1509: at low energies one has averaged vacuum oscillations (negligibly affected by the small sterile mixing angle), at high energies matter effects dominate so that
1510: $\nu_e \simeq\nu_{2m}$ that does not cross the sterile level.
1511: Fig.\fig{samples}A,B show examples of this behavior.
1512: Even sticking to the case of $\nu_{\rm s}/\nu_1$ mixing,
1513: qualitatively different effects are present for other values of the oscillation parameters,
1514: studied in fig.\fig{sun}a.
1515: E.g.\ the example in fig.\fig{samples}D illustrates the case discussed in~\cite{freq,Sterile}:
1516: large $\theta_{\rm s}$ and small $\Delta m^2_{14}\sim 10^{-12}\eV^2$.
1517: In all these examples {\em sterile neutrinos manifest at low energy},
1518: $E_\nu \circa{<} E_*$.
1519: There is a general reason for this behavior:
1520: in absence of sterile effects, only at such energies
1521: LMA oscillations allow a $\nu_1$ component
1522: in the solar neutrino flux.
1523: %Solar effects of $\nu_{\rm s}/\nu_1$ mixing are studied,
1524: %for generic values of the oscillation parameters, in fig.\fig{sun}a.
1525:
1526: \smallskip
1527:
1528: We also consider a sterile neutrino mixed with $\nu_2$ or $\nu_3$.
1529: $\nu_{\rm s}/\nu_1$ mixing and $\nu_{\rm s}/\nu_2$ mixing
1530: affect solar neutrinos in similar ways.
1531: In fact the $\nu_1^m$ and $\nu_2^m$ neutrino eigenstates in matter
1532: both typically contain significant $\nu_1$ and $\nu_2$ components.
1533: There are some differences in ${\cal O}(1)$ factors, which produce
1534: the difference between fig.\fig{sun}a ($\nu_{\rm s}/\nu_1$) and b
1535: ($\nu_{\rm s}/\nu_2$).
1536: The most evident difference in the shape of the excluded region
1537: is due to our choice of the parameterization:
1538: when studying $\nu_{\rm s}/\nu_2$ mixing
1539: we produce plots with $\Delta m^2_{24}$ on the vertical axis,
1540: so that small values of $\Delta m^2_{24}$
1541: correspond to $\nu_{\rm s}$ quasi-degenerate with $\nu_2$
1542: (rather than with $\nu_1$).
1543:
1544: A sterile neutrino mixed with $\nu_3$ gives much smaller effects in solar neutrinos,
1545: so that we do not show the corresponding plot.
1546: This happens because matter effects negligibly mix $\nu_3$ with $\nu_{1m}$ or $\nu_{2m}$,
1547: so that MSW resonances are highly non adiabatic i.e.\ ineffective.\footnote{Narrow dips in $P_{ee}(E_\nu)$ are possible close to specific energies such that
1548: the mostly sterile state crosses active neutrinos
1549: when their position-dependent eigenvalues are maximal.
1550: A look at fig.\fig{levels}a might help to understand this issue.}
1551:
1552: \medskip
1553:
1554: As discussed in section~\ref{osc},
1555: sterile mixing with one active mass eigenstate is a special configuration.
1556: Therefore we also consider a sterile neutrino mixed
1557: with a flavour eigenstate: $\nu_e$, $\nu_\mu$ or $\nu_\tau$.
1558: In such a case there are active/sterile oscillations at multiple
1559: $\Delta m^2$ values, which (in view of the observed mass differences
1560: among active neutrinos) cannot be all small.
1561: This is the main difference with respect to the previous case,
1562: and implies that sterile oscillation effects are present even for $\Delta m^2_{i4}=0$.
1563: To understand better this point let us consider the case $\Delta m^2_{14}\to 0$.
1564: In this limit neutrinos exit from the sun as an incoherent mixture of $\nu_2$
1565: and $\nu_1$, which both contain some sterile component.
1566: At $E_\nu\gg E_*$ LMA is fully effective and neutrinos exit as pure $\nu_2$,
1567: so that vacuum oscillations related to the small $\Delta m^2_{14}$ have no effect.
1568: On the contrary, at $E_\nu \circa{<} E_*$ there is a $\nu_1$ component
1569: which experiences vacuum oscillations
1570: with the mostly sterile state.
1571: These vacuum oscillations affect the $\nu_e$ flux
1572: (if $\nu_{\rm s}$ is mixed with $\nu_e$ as in fig.\fig{sun}c),
1573: or the $\nu_{\mu,\tau}$ flux
1574: (if $\nu_{\rm s}$ is mixed with $\nu_{\mu,\tau}$ as in fig.\fig{sun}d).
1575: In general, a mostly sterile state significantly mixed with $\nu_1$
1576: and almost degenerate to it,
1577: $\Delta m^2_{14}\sim 10^{-10}\eV^2$,
1578: gives significant spectral distortions, {\em but only
1579: below the energy threshold of SK and SNO}.
1580: This is illustrated in fig.\fig{samples} by the sample points D, E and F.
1581:
1582:
1583:
1584: Fig.\fig{sun}c studies $\nu_{\rm s}/\nu_e$ oscillations for generic values of $\Delta m^2_{14}$,
1585: and is similar to fig.\fig{sun}a,b apart from the difference discussed above.
1586: %\footnote{The stupid island
1587: %seems phsyical and due to crossing at $r\ll R_{\rm sun}$ with $P_C\sim1/2$.??????}
1588: $\nu_{\rm s}/\nu_{\mu}$ mixing
1589: affects solar neutrinos in the same way
1590: as $\nu_{\rm s}/\nu_\tau$ mixing, and both cases
1591: give relatively mild effects, as shown in fig.\fig{sun}d.
1592:
1593: It is interesting to notice that there are sizable
1594: and specific earth matter effects
1595: if the sterile neutrino has a large mixing e.g.\ with $\nu_e$,
1596: and is quasi-degenerate e.g.\ to $\nu_1$.
1597: In fact, the two quasi-degenerate neutrinos have different amounts of
1598: active and sterile components;
1599: due to matter effects, in the earth the two states are no longer quasi-degenerate,
1600: giving oscillations with wave-length
1601: $\sqrt{2}\pi/G_{\rm F} N_e^\oplus\sim 6000\,\hbox{km}$,
1602: comparable to the size of the earth
1603: %{\bf DA VERIFICARE E RIPENSARCI BENE. CHI LO FA?}
1604:
1605:
1606: \bigskip
1607:
1608: In all cases fig.s\fig{sun} show the regions already excluded
1609: at 90 and $99\%$ C.L. (2 dof) by present data
1610: (these regions are precisely defined by $\chi^2>\chi^2_{\rm min}+4.6$ and $9.2$).\footnote{Using
1611: instead $\chi^2>\chi^2_{\rm LMA}+4.6$ and $9.2$
1612: would give slightly weaker constraints.
1613: One might think that this latter procedure should be preferred
1614: because gives more `robust' bounds.
1615: It gives instead under-constraining bounds.
1616: More robust bounds should be obtained
1617: demanding a higher C.L.\ to the correct statistical test,
1618: rather than inventing `more robust' tests.
1619: A correct $90\%$ C.L.\ bound does not hold with more than $90\%$ probability.}
1620: We emphasize that most of the excluded regions are
1621: disfavoured only by combining high-energy (SNO, SK) with low-energy (Gallium) solar data:
1622: each kind of data-sets alone tests only a minor region.
1623: This means that present data are already able of fixing oscillation parameters
1624: with some redundancy, partially testing the LMA hypothesis.
1625: Only little regions are excluded if we fit only high-energy data, or only low-energy data.
1626: In fact, sterile oscillations that crucially involve $\nu_1$
1627: affect solar neutrinos only at $E_\nu \circa{<}E_*$ because only in this energy range
1628: the sun emits $\nu_1$.
1629:
1630:
1631: \medskip
1632:
1633: The other contours in fig.s\fig{sun} show which regions can be explored by
1634: some future experiments.\begin{itemize}
1635: \item In view of the previous comment it should be not surprising
1636: that the relatively more powerful future measurement (dotted green line) is an
1637: improved measurement of $P_{ee}$ at sub-MeV energies.
1638: Taking into account solar model uncertainties and
1639: plans of future experiments,
1640: we assumed --- maybe optimistically ---
1641: that it will be possible to see 0.02 shifts from the LMA value of $P_{ee}$.\footnote{As
1642: in~\cite{subMeV} we averaged $P_{ee}$ assuming that $pp$ neutrinos
1643: will be detected by elastic scattering on electrons;
1644: similar results hold for other possible techniques.
1645: Assuming that LMA is the end of the story,
1646: ref.~\cite{subMeV} found that
1647: feasible sub-MeV measurements would not have a significant impact
1648: on the determination of the oscillation parameters.
1649: We here find that sub-MeV measurements are instead crucially important
1650: for probing sterile oscillations.}
1651:
1652: \item
1653: The red continuous line shows what can be achieved by measuring the
1654: day/night asymmetry
1655: (normalized as $2(R_{\rm day}-R_{\rm night})/(R_{\rm day}+R_{\rm night})$)
1656: at Boron energies with $0.01$ precision
1657: in a future Mton water \v{C}erenkov experiment.
1658: We have considered this test because the needed experiment
1659: seems highly motivated by various other considerations.
1660:
1661: \item
1662: Coming to near-future experiments,
1663: the other lines show what Borexino (and/or possibly KamLAND) can do
1664: by studying (mainly) Beryllium neutrinos.
1665: We do not show the impact of a measurement of the total
1666: rate, which has a relatively large theoretical error~\cite{BP}\footnote{For values of the oscillation parameters
1667: around the `vertex' of the MSW `triangle' (at $\Delta m^2_{14}\approx 10^{-5.5}\eV^2$
1668: and $\theta_{\rm s}\approx 10^{-2}$)
1669: sterile neutrinos can give a deep dip in $P_{ee}$
1670: at $E_{\rm Be}=0.863\MeV$.
1671: Such effects can significantly reduce the total
1672: rate at Borexino compatibly with present data and without giving
1673: different kind of signals, such as
1674: day/night or seasonal variations.}
1675: and instead focus on signals that LMA predicts
1676: to be unobservable.
1677: The region inside the continuous blue line has a
1678: day/night asymmetry in the Beryllium rate larger than $0.02$.
1679: In the region inside the dot-dashed blue line the Beryllium rate
1680: shows anomalous seasonal variations with amplitude larger than $0.02$.
1681: The physics behind these effects is similar
1682: to the one well known from discussions of
1683: LOW and (Q)VO $\nu_e\to \nu_{\mu,\tau}$ oscillations (now excluded).
1684:
1685: \item
1686: We do not show results for a few other signals, that seem
1687: less promising than the ones discussed above.
1688: One can measure better the energy spectrum of Boron neutrinos,
1689: measure Beryllium neutrinos both in NC and CC scatterings,
1690: and look for day/night or seasonal variations in $pp$ neutrino rates.
1691: Active/sterile oscillations with
1692: $\Delta m^2\sim 10^{-5}$
1693: and $\tan^2\theta_{\rm s}\sim 0.1$ can distort the
1694: $\bar\nu_e$ energy spectrum
1695: in KamLAND or in future reactor experiments.
1696:
1697: \end{itemize}
1698: In table~\ref{tab:sintesi} we qualitatively classify
1699: how MSW resonances with a sterile neutrino affect solar neutrinos.
1700: The main variables are: which state is crossed, how large is the mixing.
1701: Fig.\fig{samples} shows a few examples of specific oscillation patterns
1702: compatible with present data.
1703:
1704:
1705: \begin{table}
1706: \centering
1707: \begin{tabular}{c|cc}
1708: & $\theta_{\rm ad}\ll \theta_{\rm s}\ll 1$ & $\theta_{\rm s}\sim 1$ \\ \hline
1709: $m_{\rm s}\gg m_2$ &no effect & $P_{es}\sim 1$\\
1710: %
1711: $m_{\rm s}\circa{>} m_2$ &
1712: $P_{es}=1$ at large $E_\nu$, 0 below&
1713: $P_{es}\sim 1$ increases at $E_\nu>E_*$\\
1714: %
1715: $m_{\rm s}\circa{<} m_2$ &no effect&
1716: $P_{es}\sim 1$ decreases at $E_\nu >E_*$\\
1717: %
1718: $m_{\rm s}\gg m_1$ &peak $P_{es}\sim 1$ at $E\sim E_*$
1719: & $P_{es}\sim 1$ decreases at $E_\nu >E_*$\\
1720: %
1721: $m_{\rm s}\circa{>} m_1$ & $P_{es}\sim 1$ decreases at $E_\nu >E_*$
1722: & $P_{es}\sim 1$ decreases at $E_\nu >E_*$\\
1723: \end{tabular}
1724: \caption{\label{tab:sintesi}\em Rough
1725: classification of possible sterile MSW resonances in the sun.
1726: $m_{\rm s}$, $m_1$ and $m_2$ are respectively the masses
1727: of the mostly $\nu_{\rm s}$, $\nu_1^{\rm a}$ and $\nu_2^{\rm a}$ states;
1728: $E_* \sim \hbox{few MeV}$ is the LMA critical energy and
1729: $\theta_{\rm ad} \sim (\Delta m^2/10^{-9}\eV^2)^{1/2}$.
1730: No effect is present if $\theta_{\rm s}\ll \theta_{\rm ad}$.
1731: }
1732: \end{table}
1733:
1734:
1735:
1736:
1737: \subsection{Hints and anomalies: solar data}
1738: Finally, we have studied if present solar data contain hints of sterile neutrino effects.
1739: There are two ways of searching for new effects:
1740: data-driven or theory-driven.
1741:
1742: In the data-driven approach one performs a goodness-of-fit test,
1743: that should reveal if data contain some generic indication
1744: for new physics beyond LMA oscillations.
1745: In practice, this means carefully looking if data contain hints of anomalous results.\footnote{In the past,
1746: global analyses of solar data reported the result of a Person $\chi^2$ test.
1747: Today it would tell that even the LOW solution (which has been excluded) is perfectly acceptable.
1748: Therefore the fact that, according to the $\chi^2$ test, LMA gives a good fit
1749: is not a useful information.
1750: In fact, as discussed in~\cite{freq}, the $\chi^2$ test is only sensitive to
1751: variations in the $\chi^2$ larger than $\sqrt{N}$, where
1752: $N\sim 100$ is the number of data points.
1753: In the present case $N\gg1$: this explains why the $\chi^2$ test is so inefficient
1754: and one has to perform less standard more efficient tests.}
1755: Present neutrino data contain the following non-statistically-significant
1756: hints: \begin{itemize}
1757: \item The rate measured by the Chlorine experiment is $1.4$ experimental standard
1758: deviations lower than the best-fit value predicted by active-only oscillations.\footnote{Before
1759: revising solar model predictions including the recent LUNA data~\cite{LUNA},
1760: the Chlorine rate was lower by $1.7\sigma$ (according to our results,
1761: in agreement with other similar analyses, see e.g.~\cite{SmirnovSterile}).
1762: The inclusion of LUNA results negligibly affects
1763: the global fit of solar and KamLAND data in terms of active/active oscillations.}
1764: % scende a 1.32 se uno non shifta il best-fit, ed a 1.37 se lo shifta
1765:
1766: \item The lack of an up-turn in present $\nu_e$ energy spectra
1767: does not give a statistically significative
1768: hint for new effects additional to LMA oscillations.
1769: To get a feeling of how accurately SNO and (mainly) SK data
1770: test the energy spectrum around $E_\nu\sim 10\MeV$ we fit these data
1771: assuming
1772: $$P(\nu_e\to \nu_e)=\sin^2\theta_{\rm sun} + \alpha (\hbox{MeV}/E_\nu)^2 =
1773: 1-P(\nu_e\to \nu_{\mu,\tau}).$$
1774: At energies probed by SK and SNO
1775: LMA oscillations have this form with $\alpha \approx 2$.
1776: Data give $\alpha = 1\pm 2.5$.
1777:
1778:
1779: \item A $2\div 3\sigma$ indication for sterile neutrinos might appear if
1780: analyses that suggest higher values of the
1781: nuclear factors $S_{17}(0)$ and $S_{34}(0)$
1782: (and therefore a Boron flux higher than what measured by SNO)
1783: were confirmed~\cite{S17}.
1784:
1785: \item According to some reanalyses,
1786: solar neutrino rates of various experiments
1787: show some time modulation at specific frequencies.
1788: However subsequent analyses performed by some experimental collaborations
1789: do not corroborate these claims~\cite{Pee(t)}.
1790:
1791:
1792:
1793:
1794: \end{itemize}
1795: The data-driven approach cannot see new physics that manifests
1796: giving small corrections to many observables.
1797: These diffuse minor effects can be seen if one knows what one is looking for.
1798: More precisely, one performs a fit assuming a theory with $n$ extra parameters
1799: and looks if the best-fit improves by more than $n$.
1800: Our plots in fig.s\fig{sun}
1801: explore a few $n=2$ representative slices of the full 4-dimensional
1802: parameter space of relevant active/sterile oscillation parameters:
1803: various regions (not shown) are favoured in a non statistically significant way.
1804: The best fit values, indicated by the green dots in fig.s\fig{sun},
1805: have a $\chi^2$ lower than best-fit LMA oscillations at most by 3 units.
1806: In order to study the general $4\nu$ case we also computed local minima of the $\chi^2$ using numerical minimization techniques.
1807: We have not found statistically significant indications for an extra sterile neutrino.
1808:
1809:
1810:
1811:
1812:
1813: \subsection{What is the bound on the sterile fraction in solar oscillations?}\label{compare}
1814: Assuming that solar neutrinos oscillate into
1815: $\nu_e\to \sin\alpha~\nu_{\rm s} + \cos\alpha ~\nu_{\mu,\tau}$
1816: the question is: what is the bound on the `sterile fraction' $\eta_{\rm s} \equiv \sin^2\alpha$?
1817: We here compare with previous results.
1818: Oscillation effects at Boron energies can be parameterized as
1819: \beq\label{eq:PeeEta}
1820: \Phi_{\nu_e} = \Phi_{\rm B} P_{ee},\qquad
1821: \Phi_{\nu_{\mu,\tau}}=\Phi_{\rm B}(1-P_{ee})(1-\eta_{\rm s}),\qquad
1822: \Phi_{\nu_{\rm s}} = \Phi_{\rm B}(1-P_{ee})\eta_{\rm s}.\eeq
1823: SK and SNO have measured $\Phi_{\nu_e} $ and the total
1824: flux of active neutrinos $\Phi_{\nu_{e,\mu,\tau}}$: these two measurements alone
1825: cannot determine the three unknown quantities
1826: $P_{ee}$, $\eta_{\rm s}$ and $\Phi_{\rm B}$ (unoscillated total Boron flux).
1827: Adding the solar model prediction for $\Phi_{\rm B}$ gives
1828: %\footnote{We combine
1829: %uncertainties in Gaussian approximation, that under-estimates
1830: %how much fully sterile oscillations, $\eta_{\rm s}=1$ are excluded.}
1831: \beq\label{eq:etas}%P_{ee}\approx \frac{\Phi_{\nu_e}}{\Phi_{\rm B}}\approx 0.33\pm0.06,\qquad
1832: \eta_{\rm s} \approx \frac{\Phi_{\rm B} - \Phi_{\nu_{e,\mu,\tau}}}{\Phi_{\rm B} - \Phi_{\nu_e}}\approx 0\pm 0.2\eeq
1833: where $\eta_{\rm s}$ is small because $\Phi_{\rm B}$ agrees with the measured total $\nu_{e,\mu,\tau}$ flux.
1834: In line of principle this agreement might be accidental:
1835: $\Phi_{\rm B}$ could be larger than what predicted by solar models,
1836: and a fraction of it could oscillate into sterile neutrinos leaving a reduced $\Phi_{\nu_{e,\mu,\tau}}$.
1837: In view of all other constraints (on spectral distortions, Gallium rates,\ldots)
1838: it seems difficult to realize this scenario with oscillations.
1839:
1840:
1841: Most previous analyses prefer to obtain constraints on $\eta_{\rm s}$ only from this second argument.
1842: They add one sterile neutrino mixed only with $\nu_{\mu,\tau}$
1843: and separated by a large mass splitting: $\Delta m^2_{14} \gg \Delta m^2_{\rm sun}$.
1844: This special configuration gives an energy-independent $\eta_{\rm s}$
1845: and an energy dependence of $P_{ee}$ similar to what obtained for $\eta_{\rm s} =0$.
1846: This choice does not allow to fit low energy Gallium data together with
1847: high energy SNO and SK data: the resulting bound on $\eta_{\rm s}$
1848: is~\cite{etas} \beq\label{eq:etas2}\eta_{\rm s} \approx 0 \pm 0.1,\eeq
1849: somewhat stronger than the bound in eq.\eq{etas}
1850: obtained using only the solar model prediction for the Boron flux.
1851: We stress that this constraint on $\eta_{\rm s}$ can be relaxed by making milder assumptions on the
1852: active/sterile oscillation parameters.
1853: For example, even still maintaining $\Delta m^2 \gg \Delta m^2_{\rm sun}$
1854: $\eta_{\rm s}$ is energy-dependent if the heavy $\nu_{\rm s}$ mixes
1855: with $\nu_e$, or with the mass eigenstate $\nu_{1}$
1856: ($\eta_{\rm s}$ is about 2 times larger at lower energy, see fig.\fig{samples})
1857: or with $\nu_2$ ($\eta_{\rm s}$ is about 2 times larger at higher energy).
1858: %Sterile effects present dominantly at Boron energies
1859:
1860:
1861: When we focus on the special configuration considered by previous analyses
1862: we get $\chi^2(\eta_{\rm s} = 1/4) - \chi^2(\eta_{\rm s} =0) = 8.3$
1863: (including in the fit the BP00 prediction for the total Boron flux)
1864: and $6.9$ (not including the prediction for the Boron flux: this value
1865: of the $\Delta \chi^2$ agrees with other analyses~\cite{etas}).
1866: In fact, in the special configuration
1867: our parameter $\theta_{\rm s}$
1868: is related to the sterile fraction as $\sin^2\theta_{\rm s}=2\eta_{\rm s}$
1869: (for maximal atmospheric mixing).
1870: We produced fig.\fig{sun}
1871: including in our data-set the solar model prediction for the total Boron flux.
1872: Dropping it would give only minor modifications,
1873: as a comparison of eq.\eq{etas} with eq.\eq{etas2} indicates.
1874:
1875:
1876:
1877: % Bounds on a sterile component of solar neutrinos are usually
1878: % studied assuming $\nu_e\to \sin\alpha~\nu_{\rm s} + \cos\alpha ~\nu_{\mu,\tau}$
1879: % oscillations, which means assuming that the following ratio is energy-independent:
1880: % \beq \label{eq:alpha}\frac{P(\nu_e \to \nu_{\rm s}) }{P(\nu_e\to \nu_{\mu,\tau})}=\tan^2\alpha .\eeq
1881: % In a four neutrino scheme this special case corresponds to $\nu_{\rm s}/\nu_{\mu,\tau}$ mixing
1882: % with a heavy sterile.
1883:
1884:
1885:
1886:
1887:
1888:
1889:
1890:
1891:
1892: \section{Sterile effects in supernov\ae{} and other astrophysical sources of neutrinos}\label{SN}
1893: Having discussed sterile effects in the sun, it is useful to discuss the analogous effects in
1894: core-collapse supernov\ae{}
1895: (SN) focussing on the differences between the two cases and emphasizing what one loses and
1896: what one gains studying SN neutrinos.
1897: \begin{itemize}
1898:
1899: \item Present detectors can only study SN neutrinos from our galaxy,
1900: and possibly from nearby galaxies: such
1901: SN neutrinos have a duty time of ${\cal O}(10)$ seconds every ${\cal O}(10^9)$ seconds~\cite{SNgeneral,raffelt book}.
1902: This makes backgrounds less problematic, but allowed to detect so far
1903: only ${\cal O}(10)$ SN1987A events~\cite{SN1987a signal}.
1904: Running solar neutrino experiments could detect thousands of events
1905: from a future SN exploding at distance $D\sim 10$~kpc.
1906: An even more impressive harvest of data could come from a future Mton water-\v{C}erenkov detector or from other more SN-oriented future projects~\cite{Cei}.
1907:
1908:
1909:
1910: \item In a SN the matter density grows
1911: from zero up to nuclear density ($\rho \simeq 10^{14} \gcm$) in the stiff inner core:
1912: matter effects are important for all the mass range that we consider (up to about $10^2 \eV$).
1913: In this mass range, and given the typical SN neutrino energy of $\sim 10 \MeV$, active/sterile MSW resonances occur outside the neutrino-sphere (roughly defined as the regions after which neutrinos freely stream, $\rho \ll 10^{12}\gcm$). In particular, conversions inside the neutrino-spheres (which could drain almost all the SN energy) do not take place.\footnote{The resonances with the sterile state would enter in the neutrino-spheres (in the inner core) for $\Delta m^2 \circa{>} 10^5 \eV^2$ ($\circa{>} 10^7 \eV^2$ respectively).}
1914: We only receive neutrinos that come out of the neutrino-spheres on the side of the SN
1915: closer to us.
1916: Unlike the case of the sun, it is not necessary to average
1917: over the production point,
1918: because the production and the oscillation regions
1919: are spatially separated.
1920:
1921: %(very energetic (cooling phase) electron neutrinos are produced in the inner core and then redistribute their energy to all families in the scattering processes up to the neutrinospheres);
1922:
1923: \item While only $\nu_e$ are produced in the sun,
1924: SN produce all active $\nu$ and $\bar\nu$,
1925: roughly in similar amounts~\cite{raffelt book}. They mix and convert among themselves,
1926: and possibly with sterile neutrinos~\cite{SNsterile2,SNsterile1,ShiSigl,SN3+1}.
1927: Present experiments can accurately study $\bar\nu_e$. In fact
1928: in the energy range $m_e \ll E_\nu \ll m_p$ relevant for SN neutrinos,
1929: $\bar\nu_e p\to \bar{e} n$ scatterings allows to detect $\bar\nu_e$ and to measure their energy.
1930: Detecting other neutrinos (e.g.\ via $\nu e$ scattering, deuterium dissociation, $\nue$ absorption on carbon) is possible and future projects could give significant information.
1931: In the following, we focus on the $\bar\nu_e$ flux.
1932:
1933: \item $\bar\nu_{e,\mu,\tau}$ experience no active/active matter resonance
1934: (unlike $\nu_{e,\mu,\tau}$;
1935: we are assuming normal hierarchy, and equal initial fluxes of muon and tau-neutrinos):
1936: their effective $\Delta m^2$ in matter increases in a monotonous way
1937: when the matter density grows.
1938: The sterile effect that can more strongly affect the $\bar\nu_e$ rate
1939: is a low-density $\nu_{\rm s}/\nu_1$ MSW resonance, possible
1940: when the mostly sterile neutrino is lighter than $\nu_1$
1941: (i.e.\ $\theta_{\rm s}>\pi/4$ in our parameterization).
1942:
1943: Furthermore, due to the peculiar composition of the inner part of the mantle (deleptonized matter)
1944: %is the origin of a peculiar shape of the matter potentials experienced by the neutrinos.
1945: the $\nu_e$ potential changes its sign in the deep region of the mantle, and it does so in a very steep manner~\cite{raffelt book, bethe,burrowlattimer,burrows}
1946: (see fig.\fig{levels}b), adding a MSW resonance also for $\theta_{\rm s}<\pi/4$.
1947: Although details are uncertain, this is a robust prediction.
1948: It implies that $\nu_{\rm s}$ always meets a ``sharp'' resonance with $\bar\nu_e$ in the deep region of the mantle~\cite{SNsterile1,SN3+1}.
1949:
1950:
1951: \item SN neutrinos are emitted from galactic distances,
1952: allowing to probe vacuum oscillations with $\Delta m^2$ as low as $10^{-18}\eV^2$
1953: (the precise value depend on the SN and on its distance from the earth: different SN probe different ranges).
1954:
1955: SN neutrino reach energies higher than solar $\nu$, perhaps up to $100\MeV$,
1956: and could therefore probe earth matter effects and spectral distortions.
1957:
1958:
1959:
1960:
1961: \item While the sun is essentially static,
1962: an exploding SN is a dynamical environment (neutrino light-curve evolution, passage of the shock wave...); including the time dependence in the neutrino fluxes and in the matter density profile is too much demanding and probably useless for our purposes,
1963: given the poor knowledge of the details. We focus on a typical SN configuration,
1964: which includes all the characteristic features of the SN cooling phase.
1965:
1966:
1967: A reliable prediction of the emitted fluxes is still lacking and the explosion mechanism is not yet under control.
1968: However, it is relatively easier to predict the $\nu_{e,\mu,\tau},\bar{\nu}_{e,\mu,\tau}$ energy spectra,
1969: which in thermal approximation only depend on their cross sections.
1970: On the contrary their total fluxes also depend on the SN density profile;
1971: in particular the thermal approximation does not imply equipartition among different flavours.
1972: The underlying complexity of SN explosions makes hard to reduce
1973: theoretical uncertainties, posing a threat on the usefulness of
1974: precise SN $\nu$ experiments as tools for studying oscillations.
1975:
1976: As a last warning, we have to recall the
1977: reluctance of the neutrino data from the only SN we know
1978: to fit into a simple and straightforward interpretation.
1979: The main reasons are: the energy distribution of Kamiokande
1980: and IMB look different; both experiments have an excess of forward
1981: events; the 5 LSD events cannot be accounted for.
1982:
1983: \end{itemize}
1984: As for the future, there are various interesting observables, and it is
1985: difficult to guess on which signals we should focus.
1986: This will probably depend on compromises among future experimental and theoretical capabilities.
1987:
1988:
1989: Several SN-related bounds on the active/sterile neutrino mixing have been considered in the literature~\cite{SNsterile1,SNsterile2, ShiSigl,fuller r-process}.
1990: Focussing on the cooling phase neutrinos, they include ({\bf i}) the detection of $\nueb$ from SN1987a in the Kamiokande and IMB experiments~\cite{SN1987a signal},
1991: which sets a constraint on the portion of neutrinos that oscillate into $\nu_{\rm s}$;
1992: ({\bf ii}) r-process nucleosynthesis: a fraction of the heavy elements in nature is supposed to be synthesized in the region surrounding the core of the exploding stars, provided that the electron fraction $Y_e < 0.5$; even small modifications of neutrino fluxes affect the process, so that the request of a successful nucleosynthesis has been used to set limits (see e.g.~\cite{fuller r-process, SNsterile1});
1993: ({\bf iii}) re-heating of the shock: in the delayed-shock/neutrino-driven picture of SN explosion, the flux of neutrinos and antineutrinos from the early stages of the accretion phase (which should carry $10\div20 ~\%$ of the total flux) are responsible for the actual explosion of the star, pushing from below the stalling shock wave; since $\nu_e$
1994: and $\bar\nu_e$ are the most effective in this (interacting with charged and neutral
1995: currents with baryonic matter), a cut of their flux would prevent this mechanism from working~\cite{SNsterile1,ShiSigl,sterileshock};
1996: ({\bf iv})...
1997: We will only consider the bound from direct observation, which we consider
1998: robust enough.\footnote{Indeed, for instance, alternative sites for effective nucleosynthesis have been repeatedly proposed. Moreover, the position of such a bound could depend quite heavily on the complicated details of SN dynamics and, in addition, on the interplay of sterile oscillations with it, see below. On the same footing, since the details and the nature
1999: itself of the neutrino-driven explosion mechanism are still to be fully understood, we do not
2000: consider it here as a robust constraint, not mentioning that it would require a demanding
2001: understanding and simulation of the evolution of the SN mantle.}
2002:
2003:
2004:
2005: \subsection{Technical details}
2006:
2007: The analysis of SN neutrinos can be carried on in a way similar to that of solar neutrinos,
2008: after adapting the computational procedure to SN peculiarities,
2009: which introduce some complications and allow some simplifications.
2010: We must follow the fate of the neutrinos emitted from neutrino-spheres along their travel through the star matter, the vacuum and the earth.
2011: The density matrix formalism, already described in the `solar' section~\ref{suntec},
2012: automatically handles
2013: the complication that the SN initial neutrino flux has a mixed flavour composition.
2014: In the SN case one should follow the evolution of two $4\times 4$
2015: density matrices: $\rho(E_\nu)$ for neutrinos,
2016: and $\bar\rho(E_\nu)$ for anti-neutrinos.
2017: We focus on the total $\bar\nu_e$ rate, as measured
2018: by $\bar\nu_e p$ scatterings with the cuts and efficiency of the KamiokandeII experiment.
2019: The cross section is taken from~\cite{IBD}.
2020:
2021:
2022: \smallskip
2023:
2024: MSW resonances affect the neutrino density matrix, possibly introducing off-diagonal elements
2025: which are however averaged to zero by large oscillation phases.
2026: Therefore we can combine probabilities rather than quantum amplitudes:
2027: the $4\times 4$ density matrix can be replaced by a $4$-vector containing its diagonal elements,
2028: $\Phi = (\rho_{11},\rho_{22},\rho_{33},\rho_{44})$
2029: and the evolution equation
2030: $\rho = \mathscr{U}\cdot \rho^0\cdot \mathscr{U}^\dagger$
2031: ($ \mathscr{U}= \mathscr{U}_n\cdots \mathscr{U}_1$)
2032: by~\cite{SNsterile2}
2033: \beq\Phi = \mathscr{P}_n\cdots \mathscr{P}_1\cdot \Phi^0
2034: \qquad\hbox{where}\qquad (\mathscr{P}_n)_{ij} = |(\mathscr{U}_n)_{ij}|^2 \eeq
2035: are $4\times 4$ matrices of conversion probabilities.
2036: At the production region, matter effects are dominant, so that
2037: matter eigenstates coincide with flavour eigenstates
2038: (up to a trivial permutation):
2039: $\Phi^0 = ( \Fso,\Feo, \Ftauo,\Fmuo)$.
2040: $\Phi$ are the fluxes of neutrino vacuum eigenstates reaching the earth surface,
2041: related to the fluxes of flavour eigenstates by an incoherent sum
2042: weighted by the the neutrino mixing matrix $V$:
2043: \begin{equation}
2044: \left( \Fe,
2045: \Fmu ,
2046: \Ftau ,
2047: \Fs \right)_\alpha
2048: = \sum_i |V_{\alpha i}|^2 \Phi_i
2049: \end{equation}
2050: If neutrinos cross the earth, earth matter effects~\cite{SNmattereffects} may reintroduce
2051: coherencies among the various fluxes,
2052: and one must return to the general quantum expression.
2053: Similarly, if two states have $\Delta m^2\circa{<} 10^{-18}\eV^2$
2054: vacuum oscillations do not give large phases:
2055: evolution in the outer region of the SN and in vacuum
2056: must be described keeping the off-diagonal components
2057: of the matrix density.
2058: In practice, we found convenient to evolve the density matrix
2059: in the instantaneous mass eigenstate basis,
2060: setting to zero its off diagonal elements when they accumulate large phases.
2061: This procedure is described in greater detail in the `atmospheric' section~\ref{AtmDetails}.
2062:
2063:
2064: \smallskip
2065:
2066: The effect of active/active oscillations is well known~\cite{SNmattereffects}:
2067: the LMA resonance is adiabatic and partially swaps active neutrinos, giving\footnote{We are
2068: assuming normal hierarchy and $\theta_{13}=0$.
2069: The situation becomes more complicated if instead $\theta_{13}\circa{>} 1^\circ$,
2070: because the atmospheric resonance starts changing the result in eq.\eq{SN3}.
2071: For $1^\circ\ll \theta_{13}\ll 1 $ it is adiabatic.
2072: If $\Delta m^2_{23}>0$ (normal hierarchy) it affects only neutrinos
2073: by giving a full conversion, $\Phi_{\nu_e}^{\rm after} = \Phi_{ \nu_{\mu,\tau}}^{\rm before}$.
2074: If instead $\Delta m^2_{23}>0$ (inverted hierarchy) it affects only anti-neutrinos,
2075: by giving a full conversion,
2076: $\Phi_{\bar\nu_e}^{\rm after} = \Phi_{ \bar\nu_{\mu,\tau}}^{\rm before}$.}
2077: \beq\label{eq:SN3}\Phi_{\bar \nu_e}^{\rm after} = \cos^2\theta_{\rm sun} \Phi_{\bar \nu_e}^{\rm before} + \sin^2\theta_{\rm sun}
2078: \Phi_{\bar \nu_{\mu,\tau}}^{\rm before}\qquad
2079: \Phi_{ \nu_e}^{\rm after} = \sin^2\theta_{\rm sun} \Phi_{ \nu_e}^{\rm before} + \cos^2\theta_{\rm sun}
2080: \Phi_{ \nu_{\mu,\tau}}^{\rm before}\eeq
2081: Active/sterile MSW resonances can occur after of before the LMA resonance.
2082:
2083:
2084: \bigskip
2085:
2086:
2087: Concerning the initial fluxes, the accurate results of simulations are usually empirically approximated by a so called ``pinched'' Fermi-Dirac spectrum for each flavor $\alpha=\nue,\nueb, \nu_x$
2088: ($\nu_x$ collectively denotes $\nu_{\mu,\tau},\bar\nu_{\mu,\tau}$)~\cite{jankahillebrandt, raffelt book}
2089: %\stackrel{\mbox{{\tiny (--)}}}{\nu}_{\mu,\tau}$
2090: \begin{equation}\label{eq:SNspettro}
2091: \frac{d\Phi_{\nu_\alpha}}{dt\,dE_\nu}(E_\nu,t) = N(\eta_\alpha)
2092: \frac{120}{7 \pi^4} \frac{L_\alpha}{T^4_\alpha}\frac{E_\nu^2}{e^{E_{\nu}/T-\eta_{\alpha}}+1}
2093: \end{equation}
2094: where the pinching parameter $\eta_{\alpha}$ takes the typical values $\eta_{\nue} \sim 5 - 3$, $\eta_{\nueb} \sim 2.5 - 2$, $\eta_{\nu_x} \sim 0 - 2$.
2095: For $\eta_\alpha = 0$ the normalization factor is $N=1$ and $T=\langle E_\nu\rangle/3.15$.
2096: Based on the recent results of~\cite{keil}, we adopt the following average energies and
2097: total luminosities for the various neutrino components
2098: at the time of the snapshot of fig.\fig{profiles} (see below)
2099: \beq\label{eq:Tnu}\begin{array}{lll}
2100: \langle E_{\nu_e}\rangle\simeq12 \MeV,\qquad&
2101: \langle E_{\bar\nu_e}\rangle\simeq14 \MeV,\qquad &
2102: \langle E_{\nu_x}\rangle\simeq14 \MeV\cr
2103: L_{\nu_e}\simeq30 \cdot 10^{51} \erg \sec^{-1},&
2104: L_{\bar\nu_e}\simeq30 \cdot 10^{51} \erg \sec^{-1}, &
2105: L_{\nu_x}\simeq20 \cdot 10^{51} \erg \sec^{-1}.
2106: \end{array}
2107: \eeq
2108: In accordance with numerical calculations, we shall assume
2109: that the ratios of luminosities do not vary much during the
2110: whole emission.
2111: The initial flux of sterile neutrinos is assumed to be vanishing, as a consequence of the fact that matter oscillations only take place out of the neutrinosphere.\footnote{To be precise, shortly after the collapse the electron neutrino matter potential in the very center of the core is positive, due to the contribution of the trapped neutrinos themselves. This configuration lasts for a short transient period, until the neutrino diffusion depletes their abundance and carries the potential to negative values, where it stays. A small fraction of the electron (anti)neutrinos produced in the deep core could then oscillate into sterile states and constitute a non vanishing flux injected in the mantle.}
2112:
2113:
2114: \begin{figure}[t]
2115: \begin{center}
2116: \includegraphics[width=17cm]{figuraSN}
2117: \caption{\label{fig:profiles}\em {\bf Supernova profile}.
2118: Fig.\fig{profiles}a:
2119: {\color{rosso} Density $\rho(r)$} and {\color{verdes} electron fraction $Y_e(r)$} from~\cite{burrows}.
2120: Fig.\fig{profiles}b: matter potentials in the SN, for {\color{rossos} $\bar\nu_e$} (red solid line) and for {\color{blu} $\bar{\nu}_{\mu,\tau}$} (blue solid line).
2121: The dashed lines are the analytic modelization that we adopt.
2122: }
2123: \end{center}
2124: \end{figure}
2125:
2126:
2127:
2128: \bigskip
2129:
2130: Neutrinos traveling in the SN matter experience the MSW potentials~\cite{MSW,KPreview}
2131: \beq\label{sys:potentials}\begin{array}{ll}
2132: V_e =\sqrt{2} \GF \nB \left( 3 Y_e - 1\right)/2,\qquad & V_\tau =V_\mu + V_{\mu\tau}, \\
2133: \displaystyle
2134: V_\mu = \sqrt{2} \GF \nB \left({Y_e} -1 \right)/2, & V_{\rm s} = 0, \end{array}
2135: \eeq
2136: where $\nB$ is the baryon number density ($\nB =
2137: {\rho}/{m_N}$ where $m_N \approx 939 \MeV$ is the nucleon mass\footnote{For the very high densities closer to the core, $m_N$ should be replaced by a (quite different) effective nucleon mass.
2138: We can neglect this refinement in the regions of our interest.}) and $Y_e=(N_{e^-} - N_{e^+})/\nB$
2139: is the electron fraction per baryon.
2140: Antineutrinos experience the same potentials with opposite sign.
2141:
2142: The difference $V_{\mu\tau}$ in the $\num$ and $\nut$ potentials, which appears at one
2143: loop level due to the different masses of the muon and tau leptons \cite{mutau}, is,
2144: according to the SM
2145: \begin{equation}
2146: V_{\mu \tau}=\frac{3\GF^2 m^2_{\tau} }{2 \pi^2} \left[ 2 (n_p + n_n) \ln \left(
2147: \frac{M_W}{m_\tau} \right) -n_p -\frac{2}{3}n_n \right].
2148: \end{equation}
2149: The effect is not irrelevant in the inner dense regions: for densities above $\rho \sim 10^{8} \gcm$, the $\mu\tau$ vacuum mixing is suppressed.
2150:
2151:
2152:
2153: A crucial point concerns the characteristic of the matter density and of the electron fraction in the mantle of the star. We adopt the profiles represented in fig.\fig{profiles}~\cite{burrows} and we model them with analytic functions that preserve their main features.\footnote{We thank Adam Burrows for having provided us with the data, and for useful discussions.}
2154: Namely, the density profile decreases according to a power law $r^{-4}$ out of the $\sim 10 \km$ inner core (where instead it has a roughly constant, nuclear density value).
2155: At much larger distances the density profile gets modified in a time-dependent way by the
2156: passage of the shock wave.
2157: Present simulations have difficulties in reproducing this phenomenon
2158: and therefore cannot reliably predict the density profile in the outer region.
2159: Therefore for $r\circa{>}500\km$
2160: we assume a power law $\rho = 1.5~10^4(R_\odot/r)^{3}{\rm g/cm}^3$, which roughly
2161: describes the static progenitor star.
2162:
2163: The peculiar $Y_e$ profile in fig.\fig{profiles}
2164: is inevitably dictated by the deleptonization process~\cite{burrowlattimer,bethe}:
2165: behind the shock wave which has passed in the mantle matter, the electron capture on the newly liberated protons is rapid, driving $Y_e$ to low values ($\sim 1/4$).
2166: In the outer region, where the density is sensibly lower, the efficiency of the capture is much lower, so that $Y_e$ essentially maintains the value $\sim 1/2$ typical of normal matter.
2167: This is important because the matter potential $V_e$ of electron (anti)neutrinos
2168: flips sign, see eq.~(\ref{sys:potentials}),
2169: when, in the deep region of the mantle $Y_e$ steeply decreases below $1/3$.
2170: On the contrary, at this point the matter potentials of muon and tau (anti)neutrinos are
2171: marginally affected.
2172: Both are plotted in fig.\fig{profiles}.
2173:
2174:
2175: The data refer to $\sim$0.3 sec after bounce for a typical star of $\sim 11$ solar masses.
2176: The subsequent evolution is supposed to move the wave of the $Y_e$ profile slightly outwards, maintaining, however, its characteristic shape.
2177: The slight dependence on the progenitor mass, in turn, is not really relevant~\cite{burrows 2}.
2178:
2179:
2180: This SN density profile has been computed~\cite{burrows} in absence of sterile neutrinos effects.
2181: Adding a sterile neutrino, $\nu_e\to \nu_{\rm s}$ conversions can reduce $Y_e$ due to
2182: a non-trivial feedback mechanism on the MSW potential
2183: experienced by neutrinos~\cite{SNsterile1,fuller r-process}.
2184: This could even create a new intermediate region with $Y_e<1/3$,
2185: thus introducing two more level-crossing in the $\bar\nu_e$ channel.
2186: We neglect these possible extra MSW resonances because they
2187: do not affect the $\bar\nu_e$ rate when they are both
2188: adiabatic, or both fully non adiabatic.
2189:
2190:
2191:
2192:
2193: In summary, although the profiles that we adopt come from a specific computation and refer to a specific instant in time, they incorporate the peculiar features that are important for our purposes. A more refined treatment of this point (later times behavior of the profiles, fine structures connected with the passage of the shock wave...) would of course require to obtain first a complete simulation of the SN evolution, including the explosion. This could be needed
2194: to describe the signal from a future supernova, but seems unnecessarily complicated for SN1987A.
2195:
2196:
2197: \subsection{Results}
2198:
2199: Active/sterile mixing significantly affects SN neutrinos for $\theta_{\rm s}\sim1$.
2200: Due to MSW resonances, significant effects can also be present for
2201: small mixing, i.e.\ $\theta_{\rm s}\to 0$ and $\theta_{\rm s}\to \pi/2$.
2202: The recently established active/active mixings are not taken into
2203: account in older studies~\cite{SNsterile2, ShiSigl} (see however~\cite{SN3+1}).
2204: To understand the main features it is useful to look at the pattern of possible
2205: level crossings, qualitatively depicted in fig.\fig{levels}b at page~\pageref{fig:levels}.
2206: The three mostly active anti-neutrino eigenstates depend on the radius $r$ as dictated
2207: by their measured masses and mixings (we assume $\theta_{13}=0$ and normal hierarchy)
2208: and by the predicted SN density profiles,
2209: and are represented in fig.\fig{levels}b by the three colored curves.
2210: The mostly sterile neutrino is generically
2211: represented by an almost-horizontal line, which is
2212: plotted in fig.\fig{levels}b in the specific case of small $\bar\nu_e/\bar\nu_{\rm s}$ mixing
2213: and $\Delta m^2_{14}\gg \Delta m^2_{\rm atm}$.
2214: There are three possible kinds of active/sterile MSW resonances~\cite{SNsterile1,SN3+1,SNsterile2, ShiSigl}:
2215: \begin{itemize}
2216: \item[1.] The mostly $\bar\nu_{\rm s}$ eigenstate crosses the mostly $\bar\nu_e$ eigenstate
2217: at $r\sim 100\km$, where $V_e$ flips sign.
2218: At this point matter effects dominate over active neutrino masses,
2219: so that active mass eigenstates coincide with flavour eigenstates.
2220: Since $V_e$ flips sign in a steep way this resonance is effective
2221: only if $\Delta m^2_{14}\circa{>} 10^{-1\div0}\eV^2$
2222: (different SN simulations gives values in this range).
2223:
2224:
2225: \item[2.] If the mostly sterile eigenstate is the lightest one
2226: (in our parameterization this needs $\theta_{\rm s}\circa{>} \pi/4$)
2227: the two eigenstates in 1.\ cross again at larger $r$.
2228: Pictorially, this second resonance is present when
2229: the sterile black line is lower than what assumed in fig.\fig{levels}b.
2230: This MSW resonance occurs at large $r$ where $V_e$ is smooth,
2231: so that it is effective down to $\Delta m^2_{14}\circa{>} 10^{-6\div 8}\eV^2$.
2232: Again, the significant uncertainty is due to uncertainties on the SN density gradient.
2233: %fig.\fig{SN} applies with a conservative profile.
2234:
2235:
2236: \item[3.] If instead the mostly sterile eigenstate is
2237: the heaviest or the next-to-heaviest state, it crosses
2238: one or both of two mostly $\bar\nu_{\mu,\tau}$ eigenstates.
2239: This is the case illustrated in fig.\fig{levels}b.
2240: The values of $\Delta m^2_{24}$ and $\Delta m^2_{34}$
2241: determine at which $r$ these crossings takes place,
2242: and consequently the flavour composition of the mostly active states at the resonance.
2243: Entering in the SN, the small $\bar\nu_e$ component of $\bar\nu^m_{2,3}$
2244: disappears as soon as
2245: $V_e-V_\mu$ dominates over $\Delta m^2_{\rm sun}$.\footnote{When
2246: $V_{\mu}-V_{\tau}$ dominates over $\Delta m^2_{\rm atm}$
2247: the flavour composition varies from $\bar\nu_\mu\pm\bar\nu_\tau$ to
2248: $\bar\nu_\mu$ and $\bar\nu_\tau$.
2249: This active/active resonance
2250: happens at so high densities that is not relevant for our purposes.}
2251: The color of mass eigenstates in fig.\fig{levels}b illustrates these phenomena.
2252: In any case, active/sterile MSW resonances with the mostly $\bar\nu_{\mu,\tau}$ states
2253: do not significantly affect the $\bar\nu_e$ rate, see eq.\eq{SN3}.
2254: \end{itemize}
2255: These considerations allow to understand fig.s\fig{SN},
2256: where we plot our results for the reduction of the $\bar\nu_e$ rate
2257: due to sterile mixing.
2258:
2259:
2260:
2261: \begin{figure}
2262: $$\includegraphics[width=17cm]{SN}$$
2263: \caption{\label{fig:SN}{\bf Sterile effects in supernov\ae.}
2264: \em
2265: The iso-contours correspond to a ${\color{verdes}10}$, ${\color{blucc}20},
2266: {\color{blu}30}$, ${\color{blus}40},{\color{viola}50}$, ${\color{viola2}60},{\color{rossos}70}\,\%$ deficit
2267: of the SN $\bar\nu_e$ total rate due to oscillations into sterile neutrinos.
2268: The deficit is measured with respect to the rate in absence of
2269: active/sterile oscillations and in presence of active/active oscillations
2270: (which reduce the no-oscillation rate by $\sim 10\%$).
2271: We shaded as disfavoured by SN1987A data regions with
2272: a deficit larger than $70\%$.
2273: While the qualitative pattern is robust,
2274: regions with MSW resonances can shift by one order of
2275: magnitude in $\Delta m^2$ using different SN density profiles.
2276: $\bar\nu_3/\bar\nu_{\rm s}$ mixing (not plotted) does not give significant effects.
2277: Fig.\fig{SNsample} studies in detail the sample points here marked as {\rm A, B, C},
2278: and the regions 1, 2, 12, 3 are discussed in the text.}
2279: \end{figure}
2280:
2281:
2282: \begin{figure}
2283: $$\hspace{-7mm}\includegraphics[width=18cm]{SNdist}$$
2284: \vspace{-1.5cm}
2285: \caption{\label{fig:SNsample}{\bf Sterile effects in supernov\ae.}
2286: \em Distortion of the $\bar\nu_e$ flux at sample points {\rm A, B, C}.}
2287: \end{figure}
2288:
2289:
2290: \smallskip
2291:
2292: Let us start from fig.\fig{SN}a that studies $\nu_{1}/\nu_{\rm s}$ mixing.
2293: Resonance 1 gives a sizable reduction in region 1 of fig.\fig{SN}a
2294: and resonance 2 gives a sizable reduction in region 2.
2295: Had we ignored solar mixing the maximal deficit would have been $100\%$,
2296: while in presence of solar oscillations the
2297: maximal effect is a $\sim 80\%$ deficit (see also~\cite{SN3+1}).
2298: More precisely, in the interior of region 1 one obtains
2299: $\Phi_{\bar\nu_e} = \sin^2\theta_{\rm sun} \Phi^0_{\bar\nu_e}$
2300: because resonances 1 and 3 are fully adiabatic.
2301: In the interior of region 2 one obtains
2302: $\Phi_{\bar\nu_e} = \sin^2\theta_{\rm sun} \Phi^0_{\bar\nu_{\mu,\tau}}$
2303: because resonance 2 is fully adiabatic and resonance 1 irrelevant.
2304: Therefore, given the assumed initial fluxes,
2305: the $\bar\nu_e$ rate gets reduced slightly more strongly in region 2
2306: than in region 1.
2307:
2308: In region 12 both resonances 1 and 2 are effective,
2309: and tend to compensate among each other:
2310: resonance 1 converts $\bar\nu_e$ into $\bar\nu_{\rm s}$
2311: and resonance 2 reconverts $\bar\nu_{\rm s}$ into $\bar\nu_e$.
2312: In region 3, resonances 3 gives a $20\%$ suppression of the $\bar\nu_e$ rate,
2313: that sharply terminates when $\Delta m^2_{14}< \Delta m^2_{\rm sun}$.
2314: This is due to a strong suppression of the mostly $\bar\nu_{\mu,\tau}$ eigenstates,
2315: which due to solar oscillations would give a $20\%$ contribution to the $\bar\nu_e$ rate
2316: (ignoring solar mixing, there would be no suppression of the $\bar\nu_e$ rate in region 3).
2317: This reduction of the $\bar\nu_{\mu,\tau}$ fluxes induced by resonances 3
2318: could be better probed by measuring the NC rate (which gets a $\circa{<}40\%$ reduction)
2319: and, if neutrinos cross the earth, by distortions of the $\bar\nu_e$ energy spectrum.
2320: At smaller $\Delta m^2$ and around maximal mixing
2321: vacuum oscillations can reduce the $\bar\nu_e$ rate by $\circa{<}50\%$:
2322: their effect persists down to $\Delta m^2\sim E_\nu/D\sim 10^{-18}\eV^2$.
2323: The precise value depends on the distance $D$
2324: (we assumed $D=10$ kpc).
2325:
2326:
2327: The other mixing cases are understood in similar ways.
2328: We remind that our parametrization is discontinuous at $\theta_{\rm s}=\pi/4$:
2329: this is reflected in fig.\fig{SN}b, where we consider $\nu_{2}/\nu_{\rm s}$ mixing.
2330: Resonance 2 sharply terminates when $\Delta m^2_{24}<\Delta m^2_{\rm sun}$.
2331: Up to these differences, this case is quite similar to the previous one,
2332: because $\bar\nu_1$ and $\bar\nu_2$ get strongly mixed by matter effects.
2333:
2334: On the contrary $\nu_{3}/\nu_{\rm s}$ mixing (not shown) does not give a significant reduction
2335: of the $\bar\nu_e$ rate.
2336: Mixing with the flavour eigenstates behaves in a similar way.
2337: Namely,
2338: the $\nu_{\mu,\tau}/\nu_{\rm s}$ figure shows the reduction in region 3 due to resonances 3,
2339: and the $\nu_e/\nu_{\rm s}$ figure shows the reduction in region 1 due to resonance 1.
2340: The additional feature at $\Delta m^2_{14} \sim \Delta m^2_{\rm{sun}}$ is
2341: due to adiabatic conversion (for large sterile angles) between the mostly-sterile state and $\nu_2$,
2342: that are almost degenerate in this condition.
2343: For larger $\Delta m^2_{14}$, there is no $\bar\nu_e$ component in $\bar\nu^m_2$ so that the crossing is totally non adiabatic, while for smaller $\Delta m^2_{14}$ the two states are separated.
2344: %There is no such feature for $\Delta m^2_{14} \sim \Delta m^2_{\textrm{atm}}$,
2345: %consistently with the assumption of negligible $\theta_{13}$.
2346: As in the solar case, sterile effects persists at all values of $\Delta m^2_{14}$, even if it is small.
2347: Unlike in the solar case, such effects are not sensitive to vacuum oscillations
2348: (i.e.\ $\Delta m^2\sim E_\nu/D\sim 10^{-18}\eV^2$) because in a SN
2349: neutrinos of all energies experience the adiabatic LMA resonance.
2350: Therefore in the lower row of fig.s\fig{SN}
2351: we only show results at $\Delta m^2_{14}>10^{-6}\eV^2$;
2352: nothing changes at smaller values.
2353:
2354:
2355:
2356: \medskip
2357:
2358:
2359: SN1987A data can be precisely compared with expectations
2360: doing an event by event fit~\cite{SNfit}; however the result strongly depends
2361: on the assumed average energies and total luminosities of the initial fluxes.
2362: The data do not permit detailed studies of the energy distribution.
2363: Even more, they cannot discriminate in a significant way between a larger
2364: total flux with smaller average energy and a smaller flux with higher
2365: average energy.
2366: %
2367: %The total $\bar\nu_e$ rate is the main observable measured by SN1987A experiments,
2368: %that could not discriminate in a significant way between
2369: %a larger total flux with smaller average energy
2370: %and a smaller flux with higher average energy.
2371: In absence of a quantitative estimation of theoretical uncertainties,
2372: and in view of the doubtful aspects of SN1987A data discussed previously,
2373: today we cannot derive precise constraints.
2374: Therefore we simply shaded as `disfavoured' regions where sterile effects reduce the
2375: $\bar\nu_e$ rate by more than $70\%$.
2376: This is nothing more than a reasonable arbitrary choice.
2377:
2378:
2379: \medskip
2380:
2381:
2382: Future data will permit to know precisely the total rate of
2383: $\bar\nu_e$ events, and also its distribution in energy and time.
2384: The most important single observable could be the average $E_{\bar\nu_e}$ energy,
2385: or more generically the energy spectrum.
2386: Rather than showing iso-contour plots of $\langle E_{\bar\nu_e}\rangle$
2387: we describe their main features.
2388: In absence of sterile oscillations we expect $\langle E_{\bar\nu_e}\rangle \approx 15\MeV$.
2389: Along the `diagonal sides of the MSW triangles' 1 and 2
2390: (e.g. around our sample point B)
2391: the average energy increases up to $\langle E_{\bar\nu_e}\rangle \approx 18\MeV$.
2392: Along the `diagonal sides of the MSW triangle' 12
2393: it can decrease down to $\langle E_{\bar\nu_e}\rangle \approx 11\MeV$.
2394: Vacuum oscillations can give the well known distortions of the spectrum,
2395: as exemplified in fig.\fig{SNsample}C.
2396: These effects seem larger than experimental and theoretical uncertainties.
2397: In all other cases sterile effects give a quasi-energy-independent suppression of
2398: the $\bar\nu_e$ rate, and therefore negligibly affect $\langle E_{\bar\nu_e}\rangle$.
2399: Fig.\fig{SNsample}A gives an example.
2400: Of course, the average energy of positrons generated by $\bar\nu_e p \to n \bar{e}$ scatterings
2401: is higher than $\langle E_{\bar\nu_e}\rangle$ because the cross section increases with energy.
2402:
2403:
2404: \bigskip
2405:
2406: We focused on supernova
2407: antineutrinos.
2408: The analogous plots for neutrinos cannot obtained from
2409: our $\bar\nu$ plots by flipping $\tan\theta_{\rm s}\to 1/\tan\theta_{\rm s}$,
2410: because we are taking into account the effects of active/active oscillations.
2411: Effects in neutrinos are more similar to what happens to solar $\nu_e$.
2412:
2413:
2414: %With large mixing $\nu_{\rm s}$ can carry out some ${\cal O}(1)$ fraction of the total neutrino flux.
2415: %However this signal could be not detectable due to too large theorethical uncertainties.
2416: %It is not clear how accurate they can be, using all astronomical observation that
2417: %can be made before and after the next explosion.
2418:
2419: %In the purely thermal limit neutrinos of different flavour would have
2420: %a well defined pattern of average energies (directly related to their cross sections)
2421: %but in no way this limit implies an equipartion of their fluxes,
2422: %which instead depend on the density profile.
2423: %More recent computations suggest.,......
2424:
2425: %Due to limited statistics in 1987 experiments could not discriminate between
2426: %a larger neutrino with less energy or smaller neutrinos with more energy.
2427:
2428: %More energetic: earth matter effects?
2429: %Matter effects are probably best studied looking at spectral distrtions
2430: %with one Mton
2431:
2432: \medskip
2433:
2434: \subsection{Hints and anomalies: supernov\ae}
2435: To conclude, we list hints
2436: of anomalous effects possibly related to sterile effects in SN neutrinos:
2437: \begin{itemize}
2438: \item In the past, the initial SN neutrino fluxes in different flavours
2439: were believed to follow an
2440: almost exact equipartition of the energy and have
2441: gaps in neutrino temperatures larger than in eq.\eq{Tnu}
2442: (typical values were $13, 16, 23 \MeV$~\cite{raffelt book}).
2443: Under these assumptions SN1987A data, which point to a lower average energy,
2444: disfavor solar oscillations with large mixing angle
2445: ($\sin^2 2\theta_{\rm sun}<0.9$ at $99\%$ C.L.~\cite{LMAdead})
2446: while solar data established the relatively large mixing $\sin^2 2\theta_{\rm sun}\approx 0.8$.
2447: The tension strongly depends on the initial temperatures,
2448: and disappears assuming
2449: the more recent values in eq.\eq{Tnu}~\cite{LMAdead}.
2450:
2451: \item Pulsar motion. It has been proposed that the resonant (or non resonant~\cite{PulsarMotion}) conversion of sterile neutrinos could explain the observed large proper velocity of newly born neutrons stars~\cite{kusenko segre}. Indeed, in presence of very strong and axially oriented magnetic fields, which could be plausible in the NS environment, the MSW potential includes a contribution (at one loop, from neutrino scattering on the polarized medium) which depends on the relative angle between the neutrino momentum and the magnetic field; this leads to an asymmetric neutrino emission, which could be enough to account for the observed velocities.
2452: This mechanism needs sterile neutrinos with keV-scale masses and $\theta_{\rm s}\circa{<}10^{-7}$.
2453:
2454: \item r-process nucleosynthesis~\cite{fuller r-process}: $\nu_e \to \nu_{\rm s}$ conversions in the mantle of the star could drive $Y_e$ to low values and have thus been proposed to provide a favorable environment for r-process nucleosynthesis. The relevant range of $\nu_e/\nu_{\rm s}$
2455: mixing parameters spans $\Delta m^2 \sim (1 \div 10^2) \eV^2$ and
2456: $\sin^2 2 \theta_{\rm s} \sim 10^{-3} \div \textrm{few} 10^{-1}$,
2457: which should be checked against the bounds from cosmological probes that we discussed in the present paper.
2458: In the same region, the ``conversion plus re-conversion'' of antineutrinos instead guarantees that the SN $\bar\nu_e$ is not dramatically affected.
2459:
2460:
2461: \end{itemize}
2462:
2463: \subsection{Neutrinos from other astrophysical sources}
2464: Different kinds of experiments will try to
2465: detect neutrinos emitted by extragalactic sources,
2466: such as active galactic nuclei.
2467: There are no firm expectations.
2468: In particular we do not know if the fluxes of these neutrinos
2469: will be detectably large, and eventually in which energy range.
2470: However, since these neutrinos are presumably
2471: mostly generated by $\pi$ decays,
2472: like atmospheric neutrinos, one expects
2473: similar fluxes of $\nu$ and $\bar\nu$ with
2474: flavour ratio $e:\mu:\tau\sim1:2:0$
2475: (this expectation might be wrong).
2476: Atmospheric oscillations then
2477: convert the flavour ratio into $1:1:1$, which is blind
2478: to other active/active oscillations.
2479:
2480: In presence of an extra sterile neutrino,
2481: the $e:\mu:\tau:{\rm s} \sim 1:1:1:0$ flavour ratio
2482: is not blind to extra active/sterile oscillations.
2483: Sterile oscillations can reduce the fluxes of active neutrinos,
2484: and vary the relative flavour proportion.
2485: Since the total initial fluxes and energy spectra are unknown, and
2486: since experiments will probably be able of
2487: tagging $\mu$ and maybe $\tau$ neutrinos well~\cite{UHE},
2488: we focus only on the observable $\Phi_\mu/\Phi_\tau$.
2489: %\footnote{It might be possible to reconstruct $\Phi_e/\Phi_{\mu,\tau}$ if the energy spectra are known.}
2490:
2491:
2492: We assume that the baseline, about 100 Mpc,
2493: is much longer than all oscillation lengths.
2494: In such a limit the oscillation probability reduces to
2495: multiplication of probabilities (rather than of quantum amplitudes):
2496: \beq \label{eq:degosc}P(\nu_\ell\to \nu_{\ell'})=
2497: P(\bar\nu_\ell\to \bar \nu_{\ell'}) = \sum_{i=1}^4 |V_{\ell i}|^2 |V_{\ell'i}|^2
2498: %= \sum_{i=1}^3 |U_{\ell i}|^2 |U_{\ell'i}|^2 - 2n_\ell n_{\ell'}\sin^2\theta_{\rm s} + \frac32
2499: %n_\ell^2 n_{\ell'}^2\sin^4\theta_{\rm s}
2500: \eeq
2501: One can verify that sterile mixing with a flavour eigenstate
2502: $\nu_\ell$ ($\ell = e$ or $\mu$ or $\tau$)
2503: mainly gives a depletion of $\Phi_\ell$.
2504: However, as discussed in the rest of this paper,
2505: in such a case large active/sterile
2506: mixing angles are already disfavoured by other data,
2507: so that it is not possible to get a sizable effect.
2508:
2509: Present data allow large active/sterile mixing with mass eigenstates
2510: $\nu_i$ ($i=1$ or $2$ or $3$)
2511: if $\nu_{\rm s}$ and $\nu_i$ are quasi-degenerate,
2512: with $\Delta m^2 \ll 10^{-9}\eV^2$.
2513: In view of the long base-line, neutrinos from cosmic sources
2514: are affected by oscillations with $\Delta m^2 \circa{>}10^{-17}\eV^2$.
2515: However each of the mostly active mass eigenstates contains roughly equal component
2516: of $\nu_\mu$ and $\nu_\tau$ (unless the atmospheric mixing angle is non-maximal,
2517: or unless $\theta_{13}$ and the CP-phase significantly differ from zero):
2518: therefore $\nu_{\rm s}/\nu_i$ oscillations do not significantly affect
2519: $\Phi_\mu/\Phi_\tau$.
2520:
2521: In conclusion, near-future experiments that
2522: will try to discover neutrinos from cosmic sources
2523: do not seem to allow promising searches of sterile neutrino oscillations.
2524: See also~\cite{SterileUHE}.
2525:
2526: \medskip
2527:
2528: Eq.\eq{degosc} also allows to study vacuum oscillation effects in other
2529: kinds of cosmological neutrinos:
2530: \begin{itemize}
2531:
2532: \item In a near future it seems possible to
2533: detect the $\bar\nu_e$ emitted by {\bf past core-collapse supernov\ae}.
2534: At the moment we only have order-of-magnitude predictions,
2535: but future SN experiments and studies might allow to
2536: better predict their total flux or spectrum.
2537: SK almost reached the apparently necessary sensitivity,
2538: and it seems possible to improve the efficiency
2539: of tagging neutrons emitted in $\bar\nu_e p \to \bar{e} n$ scatterings~\cite{SKbarnu}.
2540: Relic SN $\bar\nu_e$ are affected by oscillations in the SN (as discussed
2541: in the previous section) and by oscillations in vacuum (down to $\Delta m^2 \circa{>} 10^{-25}\eV^2$).
2542:
2543:
2544:
2545: \item In a far future, it might be possible to directly detect
2546: {\bf CMB neutrinos} by coherent scatterings.
2547: This would allow to probe oscillations down to $\Delta m^2\sim 10^{-30}\eV^2$.
2548:
2549:
2550:
2551:
2552:
2553: \end{itemize}
2554:
2555:
2556:
2557: \section{Sterile effects in atmospheric, reactor and beam neutrinos}\label{atm}
2558: In this section we discuss how SK, K2K, MACRO, {\sc Chooz}, {\sc Bugey},
2559: CDHS, CCFR, {\sc Karmen}, {\sc Nomad}, {\sc Chorus}
2560: and future experiments of these kinds probe sterile oscillations.
2561: KamLAND data have been studied in section~\ref{solar},
2562: together with solar data.
2563:
2564:
2565:
2566:
2567: \subsection{Technical details}\label{AtmDetails}
2568: It is convenient to compute oscillation probabilities using the neutrino
2569: and anti-neutrino density matrices.
2570: We convert the initial value
2571: (e.g.\ $\rho(E_\nu) = \diag(\Phi_e(E_\nu), \Phi_\mu(E_\nu),0,0)$
2572: for atmospheric neutrinos)
2573: to the instantaneous mass eigenstate basis, $\rho_m = V^\dagger \rho V$.
2574: In this basis evolution in each medium (air, mantle, core)
2575: is simply given by a diagonal matrix of phases,
2576: $\mathscr{U} =\diag \exp (-2 i \delta)$
2577: where $\delta_i = m_{\nu_{mi}}^2L/4E_\nu$.
2578: Using
2579: the matrix density in the mass eigenstate basis, we can
2580: analytically average `fast' oscillations to their mean value
2581: by appropriately inserting a small imaginary
2582: part $\epsilon$ in the oscillation phases $\delta$.
2583: This is achieved by evolving the matrix density as
2584: $\rho^m_{ij}(L)=\rho^m_{ij}(0) e^{2i \delta_{ij} - \epsilon|\delta_{ij}|}$
2585: where $\delta_{ij} = \delta_i - \delta_j$ and $\epsilon$ is an arbitrary positive small number
2586: (we choose $\epsilon =0.01|\delta_{ij}|$).
2587: This makes computations much faster than usual techniques
2588: which require lengthy numerical averages of the oscillation factors.
2589: In the simplest case of vacuum oscillations of 2 neutrinos
2590: this amounts to modify the oscillation factor as
2591: $\sin^2 \delta_{12}\to (1-e^{-|\delta_{12} \epsilon|}\cos 2\delta_{12})/2$.
2592: %The density matrix formalism allows to interpret this
2593: %replacement as having added a decoherence effect and to properly
2594: %generalize it to evolution of $n$ neutrinos in presence of matter effects.
2595:
2596: At the air/mantle and mantle/core boundaries
2597: eigenstates change in a non-adiabatic way:
2598: this effect is accounted by the `level-crossing' flavour matrices $P$
2599: already described at page~\pageref{Psun}.
2600:
2601: \smallskip
2602:
2603: Our fit of atmospheric data takes into account all most recent results:
2604: SK atmospheric data~\cite{SKatm}, the K2K spectrum and total rate~\cite{K2K},
2605: and MACRO~\cite{MACRO} data about trough-going muons.
2606: These events arise from neutrino scatterings in the rock below the detector:
2607: MACRO is competitive with SK because the important parameter
2608: is the surface of the detector, rather than its mass.
2609: We do not include data from
2610: older atmospheric experiments, which studied essentially
2611: the same kind of observables better measured by SK.
2612: Data are fitted by forming a global $\chi^2$.
2613: Uncertainties are taken into account following~\cite{SKatm},
2614: and systematically working in Gaussian approximation.
2615: More na\"{\i}ve definitions of the $\chi^2$ (e.g.\ fitting only the
2616: zenith-angle spectra of the single classes of events)
2617: do not give significantly different final results.
2618:
2619: \smallskip
2620:
2621: Atmospheric data contain the evidence for the atmospheric
2622: anomaly. We assume it is due to $\nu_\mu\to\nu_\tau$ oscillations and
2623: take into account the uncertainty on
2624: $\Delta m^2_{\rm atm}$ and on $\theta_{\rm atm}$
2625: using the same technique developed in section~\ref{sunfit}
2626: for marginalizing
2627: over $\Delta m^2_{\rm sun}$ and on $\theta_{\rm sun}$.
2628: Even in the atmospheric case, this technique now gives an accurate
2629: analytical approximation to the usual
2630: active-only atmospheric fit (not shown).
2631: As previously discussed, we assume $\theta_{13}=0$.
2632:
2633:
2634:
2635: Finally, we include $\bar\nu_e$ disappearance data from the {\sc Chooz}~\cite{CHOOZ} (14 bins) and {\sc Bugey}~\cite{Bugey} (60 bins) reactor experiments.
2636: We include $\nu_\mu$ disappearance data from
2637: CDHS~\cite{CDHS} (15 bins) and CCFR~\cite{CCFR} (15 bins).
2638: {\sc NuTeV} data could give additional information,
2639: if certain anomalous features will be understood~\cite{NuTeV}.
2640: Although disappearance experiments give the dominant constraint on sterile effects,
2641: we also include $\bar\nu_\mu\to\bar\nu_e$ data from {\sc Karmen}~\cite{Karmen},
2642: using the likelihood computed by the {\sc Karmen} collaboration
on an event-by-event basis.\footnote{We thank K. Eitel and M. Steidl
for giving us the table of the {\sc Karmen} likelihood,
and B. Louis and
G. Mills for the LSND likelihood.}
2643: LSND results are not included and separately discussed in section~\ref{LSND}.
2644: Finally, we include {\sc Nomad} and {\sc Chorus} data on $\nu_\mu\to\nu_\tau$~\cite{NOMADCHORUS},
2645: which are relevant in the case of $\nu_3/\nu_{\rm s}$ mixing
2646: at $\Delta m^2\sim 10\eV^2$.
2647:
2648:
2649:
2650: \subsection{Results}
2651: Since there are many relevant experiments,
2652: it is useful to divide them into
2653: two classes, and separately study their impact:
2654: \begin{itemize}
2655: \item[1)] Fig.s\fig{SBL} show the constraints from experiments that are
2656: {\rm not} sensitive to the atmospheric anomaly ({\sc Chooz}, {\sc Bugey}, CDHS, CCFR,
2657: {\sc Karmen}, {\sc Nomad}, {\sc Chorus}).
2658: Sterile oscillation effects in such experiments are simply described by eq.\eq{Pij},
2659: so that we only had to compile their results.
2660: Disappearance experiments provide the dominant constraints.
2661:
2662:
2663: \item[2)]
2664: Atmospheric neutrinos are a powerful probe of neutrino oscillations
2665: with large mixing angle and $\Delta m^2 \circa{>}10^{-4}\eV^2$.
2666: Fig.s\fig{atm} show how the experiments that see the atmospheric anomaly
2667: (SK, MACRO and K2K) constrain extra oscillations into sterile neutrinos.
2668: \end{itemize}
2669: Analyzing the second class of experiments requires
2670: a non trivial work: these experiments probe
2671: sterile mixing in a significant but indirect way.
2672:
2673:
2674: \begin{figure}[t]
2675: \vspace{-9mm}
2676: $$\hspace{-5mm}\includegraphics[width=18cm]{SBfits}$$
2677: \caption{\label{fig:SBL}\em {\bf Sterile mixing:
2678: effects in short base-line experiments} {\sc Chooz}, {\sc Bugey}, {\sc CDHS}, {\sc CCFR}, {\sc Karmen},
2679: {\sc Nomad}, {\sc Chorus}.
2680: Shaded regions: excluded at $90,99\%$ C.L.
2681: The {\color{blue} blue dashed lines} estimate the region that seems
2682: explorable by a future short-baseline reactor experiment.
2683: The plot is symmetric under
2684: $\tan\theta_{\rm s}\leftrightarrow 1/\tan\theta_{\rm s}$
2685: so that we only show $\tan\theta_{\rm s}\le 1$.}
2686: \end{figure}
2687:
2688: \begin{figure}[t]
2689: \vspace{-9mm}
2690: $$\hspace{-9mm}\includegraphics[width=18cm]{atmfits}$$
2691: \caption{\label{fig:atm}\em {\bf Sterile mixing:
2692: effects in atmospheric neutrinos (SK, MACRO) and K2K.}
2693: No statistically significant evidence is found.
2694: Shaded regions: excluded at $90,99\%$ C.L.
2695: Coloured lines are iso-curves of a few promising signals.
2696: {\color{blue} Blue dashed lines: $5\%$ and $1\%$ reduction of the NC rate at MINOS.}
2697: {\color{verdes} Green dot-dashed lines: $P_{e\mu} = 0.01$.}
2698: {\color{rossos} Red dotted lines: $|\Delta P_{\mu\tau}| = 0.01$ at CNGS.}}
2699: \end{figure}
2700:
2701:
2702:
2703:
2704:
2705: The upper limit on the `sterile fraction involved in atmospheric oscillations'
2706: conventionally quoted in the literature corresponds to the following `minimal' configuration:
2707: $\nu_{\rm s}/\nu_{\tau}$ mixing with $\Delta m^2_{34}\gg\Delta m^2_{\rm atm}$.
2708: In this limit, the `sterile fraction' is related to our $\theta_{\rm s}$ angle by
2709: $\eta_{\rm s} = \sin^2\theta_{\rm s}$ and
2710: our analysis reasonably agrees with previous computations.
2711: Atmospheric neutrinos oscillate at a single detectable frequency.
2712: Matter effects become relevant at higher neutrino energies
2713: and suppress active/sterile mixing
2714: without affecting $\nu_\mu\to \nu_\tau$ oscillations.
2715: This effect allows to indirectly discriminate the two channels.
2716: Sterile oscillations are mostly disfavoured by the zenith-angle
2717: spectra of $\mu$-like events with TeV-scale energies.
2718:
2719: % a 99% ok. A 90% quasi. A bassi CL io non trovo un piccolo etas favorito.
2720:
2721: More general active/sterile oscillation schemes manifest in different ways.
2722: A sterile neutrino mixed with different flavours can give appearance signals.
2723: Active/sterile oscillations with $\Delta m^2$ comparable to $\Delta m^2_{\rm atm}$
2724: give oscillations at multiple frequencies, distorted by matter effects.
2725: When $\Delta m^2_{34}<0$ matter effects
2726: increase the oscillation length and give a non-standard
2727: energy and path-length dependence of $P(\nu_\mu\to\nu_\mu)$.
2728: SK can best probe such effects with its multi-GeV $\mu$-like sample.
2729: However, SK cannot safely test if an oscillation dip is present.
2730: This kind of studies needs a dedicated future detector~\cite{Monolith}.
2731:
2732:
2733: The SK collaboration probes sterile effects
2734: in another more direct way: selecting a sample of NC-enriched events~\cite{SKatm}.
2735: In absence of a precise description of the cuts performed
2736: to obtain the NC-enrichment, we have not included this sample
2737: (as also done in previous reanalyses of SK data).
2738: The `minimal' active/sterile configuration
2739: is more strongly constrained indirectly by the zenith-angle spectra.
2740: We estimate that this remains true in most of the parameter space,
2741: and consequently we do not include in our final results our
2742: approximate reanalysis of NC-enriched data.
2743:
2744:
2745: \medskip
2746:
2747: As in the solar case, we looked if present data contain some evidence for
2748: sterile effects which correct in a minor way many observables,
2749: by searching for local minima of the global $\chi^2$.
2750: No statistically significant hint is found: since subleading sterile
2751: effects do not improve the global fit in a significant way
2752: (at most by $\Delta \chi^2\approx 4$)
2753: % da 39.8 a 36.1
2754: our plots only show excluded regions.
2755: The excluded region in fig.s\fig{atm}d, e
2756: (which correspond to $\nu_\mu/\nu_{\rm s}$ and
2757: to $\nu_\tau/\nu_{\rm s}$ mixing) extends down to
2758: $\Delta m^2_{41}=0$ because even in this limit there are
2759: sterile oscillations at the atmospheric and solar frequencies.
2760:
2761:
2762: \medskip
2763:
2764:
2765: It is useful to compare the sensitivity of the two classes of experiments, 1) and 2).
2766: Since there are no MSW resonances,
2767: all these experiments are sensitive only to relatively large sterile mixing,
2768: $\theta_{\rm s} \circa{>} 0.1$.
2769: Sterile mixing with $\nu_e$ (and with the $\nu_1$ and $\nu_2$ mass eigenstates
2770: that contain a sizable $\nu_e$ fraction)
2771: is better probed by reactor experiments,
2772: although $e$-like events at SK
2773: extend the sensitivity down to smaller values of $\Delta m^2$.
2774: On the contrary atmospheric experiments give more stringent tests
2775: of $\nu_{\rm s}/\nu_\tau$ mixing and of $\nu_{\rm s}/\nu_\mu$ mixing.
2776: %It is well known that a sterile interpretation of the atmospheric anomaly
2777: %is excluded only by atmospheric esperiments, not by SBL.
2778: Within standard cosmology,
2779: the sterile effects detectable by the experiments discussed in this section
2780: are already disfavoured by measurements of the primordial $^4$He abundancy,
2781: and can be fully tested by future CMB or BBN data.
2782:
2783:
2784:
2785: %Only these experiments have studied neutrinos at
2786: %the large energy (up to more than $10^2\GeV$)
2787: %and with the long-baseline (up to about $10^4$ km)
2788: %necessary to test matter effects, which
2789: %discriminate $\nu_\mu\to\nu_\tau$ from $\nu_\mu\to \nu_{\rm s}$.
2790:
2791:
2792:
2793: \bigskip
2794:
2795: A detailed analysis of capabilities of future
2796: beam or reactor neutrinos as probes of sterile neutrinos
2797: seems not necessary.
2798: In fact, there are many proposals motivated by other considerations,
2799: and in each case it is easy to compute sterile effects.
2800: We only make a few general comments.
2801:
2802: The blue dashed line in fig.s\fig{SBL} shows what can achieved by
2803: a future high-precision short-baseline reactor experiment
2804: able of detecting a $2\%$ deficit in the $\bar\nu_e$ flux~\cite{SterileReactor}.
2805: Sterile oscillations give a $\bar\nu_e$ deficit,
2806: which might be energy-dependent if the sterile oscillation length at $E_\nu \sim \hbox{few}\MeV$
2807: is comparable to the base-line $L$.
2808: Both are unknown; we assumed $L\sim 2\km$.
2809: Most of the explorable region at small $\Delta m^2$ is already excluded
2810: by solar and atmospheric experiments.
2811: The region with large $\Delta m^2\sim \eV^2$ is more difficult
2812: because even a near detector only sees averaged oscillations;
2813: one has to rely on theoretical predictions for the total flux of reactor $\bar\nu_e$~\cite{ReTh}.
2814:
2815:
2816:
2817: Sterile oscillations with large $\Delta m^2\sim \eV^2$
2818: have a wave-length comparable to the earth radius at energies
2819: $E_\nu \sim \TeV$: experiments such as
2820: AUGER and {\sc IceCube} can study atmospheric neutrinos
2821: in this energy range, and could see an anomalous
2822: zenith-angle dependence if $\theta_{\rm s}$ is large enough~\cite{nuno}.
2823:
2824: Future atmospheric experiment such as {\sc Monolith}~\cite{Monolith} could test if
2825: the first oscillation dip is present as predicted by $\nu_\mu\to \nu_\tau$ oscillations.
2826: Sterile effects with $\Delta m^2 \sim \Delta m^2_{\rm atm}$
2827: can significantly distort the expected oscillation pattern.
2828:
2829: Distortions not related to earth matter effects
2830: can also be searched by the planned
2831: $\nu_\mu$-beam experiments.
2832: Furthermore, the MINOS detector can statistically distinguish NC/CC events
2833: that are tagged as short/ long tracks, and it should be possible to measure the
2834: NC rate with $5\%$ precision~\cite{MinosNC}.
2835: The blue dashed lines in fig.s\fig{atm} correspond to a
2836: $5\%$ and $1\%$ anomaly in the NC total rate
2837: in a $\nu_\mu$ beam experiment with $L=730\km$
2838: and {\sc Minos}-like parent energy spectrum.
2839:
2840: A sterile neutrino can also manifest as $\nu_\tau$ appearance:
2841: considering a CNGS-like beam~\cite{CNGS}
2842: in fig.s\fig{atm} we show iso-curves corresponding to
2843: an average $\nu_\mu\to\nu_\tau$ conversion probability $\pm 0.01$
2844: different from atmospheric oscillations only
2845: (which give a conversion probability of about $0.02$, depending
2846: on the precise value of $\Delta m^2_{\rm atm}$ and on the
2847: average energy of the beam).
2848:
2849: Constraints from reactor experiments make more difficult,
2850: but not impossible, to have detectable $\nu_e$ appearance
2851: in a $\nu_\mu$ beam (or $\nu_\mu$ appearance in a $\nu_e$ beam)
2852: as signals of active/sterile oscillations.
2853: In experiments sensitive to atmospheric oscillations
2854: a non zero $\theta_{13}$ gives such effects.
2855: Assuming $\theta_{13}=0$, in fig.\fig{atm} we show iso-curves corresponding to
2856: a $\nu_\mu\to\nu_e$ conversion probability of $0.01$.
2857:
2858:
2859:
2860:
2861: \bigskip
2862:
2863: \subsection{Hints and anomalies}
2864: To conclude, we list some possibly anomalous features of present data:
2865: \begin{itemize}
2866: \item The total rate of $e$-like events at SK
2867: is higher than the expected central value.
2868: The $\mu/e$ ratio seems to be too low
2869: (see e.g.~\cite{LoSecco}).
2870: Constraints from solar, atmospheric and short-baseline data show that
2871: sterile neutrinos can reduce the $\mu$ rate
2872: (e.g.\ in the case of $\nu_\mu/\nu_{\rm s}$ mixing with $\Delta m^2_{14}\sim 10^{-4}\eV^2$)
2873: but cannot increase the $e$ rate.
2874:
2875:
2876:
2877:
2878: \item The total rate of trough-going muons at SK
2879: is higher than the expected central value.
2880: Predictions and direct measurements of cosmic ray primaries at
2881: the relevant energies are difficult~\cite{Battistoni}.
2882:
2883: \item The LSND result~\cite{LSND}, discussed below.
2884: \end{itemize}
2885:
2886:
2887: \begin{figure}
2888: $$\includegraphics[width=7cm]{LSND}$$
2889: \caption{\label{fig:LSND}\em
2890: {\bf The LSND anomaly interpreted as oscillations of 3+1 neutrinos}.
2891: Shaded region: suggested at 99\% C.L.\ by LSND.
2892: Black dotted line: 99\% C.L.\ global constraint from other neutrino experiments
2893: (mainly Karmen, Bugey, SK, CDHS).
2894: {\color{rossos} Continuos red line: $N_\nu = 3.8$ thermalized neutrinos.}
2895: {\color{rossoc} Dot-dashed orange line: $\Omega_\nu h^2 =0.01$.}}
2896: \end{figure}
2897:
2898: \subsection{LSND}\label{LSND}
2899: The LSND experiment~\cite{LSND} finds an evidence for $\bar\nu_\mu\to\bar\nu_e$,
that ranges between 3 to $7\sigma$
depending on how
data are analyzed.
2900: It is difficult to conceive new physics that can explain the LSND
2901: result compatibly with all other constraints.
2902: Among reasonably conservative interpretations,
2903: a sterile neutrino with eV-scale mass
2904: seems to be the less disfavoured possibility~\cite{LSND?,SN3+1,2+2}.
2905: According to this interpretation
2906: the LSND anomaly arises as $\nu_\mu\to\nu_{\rm s}\to \nu_e$
2907: so that the effective $\theta_{\rm LSND}$ $\nu_\mu/\nu_e$ mixing angle
2908: is predicted to be $\theta_{\rm LSND}\approx \theta_{es}\cdot\theta_{\mu s}$.
2909: This formula is valid only for small mixing angles, and eq.\eq{Pij} gives
2910: the general expression.
2911: This prediction gives rise to 3 problems:
2912: \begin{itemize}
2913: \item[1)] $\nu_e$ and $\nu_\mu$ disappearance experiments
2914: imply that $\theta_{es}$ are $ \theta_{\mu s}$ are
2915: somewhat smaller than what suggested by LSND~\cite{LSND?,SN3+1,2+2};
2916: \item[2)] standard BBN predicts that the sterile neutrino thermalizes,
2917: so that primordial abundances should have values corresponding to $N_\nu = 4$~\cite{LSNDBBN};
2918: \item[3)] according to standard cosmology, the sterile neutrino
2919: gives a contribution to the neutrino density $\Omega_\nu$
2920: somewhat larger than what suggested by global fit of CMB and LSS data
2921: (see e.g.~\cite{instant}).
2922: \end{itemize}
2923: Concerning points 2) and 3),
2924: there is not yet general consensus that the sterile neutrino thermalizes,
2925: maybe because this LSND issue has never been analyzed by authors
2926: that performed, at the same time,
2927: a precise study of neutrino data and of cosmology
2928: with mixed neutrinos.
2929: Estimates indicate that the region favoured by LSND lies
2930: well inside the region where the sterile neutrino is thermalized~\cite{LSNDBBN}.
2931: This is confirmed by our analysis, shown in fig.\fig{LSND}.
2932: In the relevant region the constraint on $\Omega_\nu$
2933: is well approximated by the horizontal line
2934: corresponding to $\Omega_\nu h^2 = m_4/93.5\eV$,
2935: as assumed in previous analyses~\cite{instant,2+2}.
2936: The accurately computed constraint starts to be weaker only
2937: at much smaller values of the effective $\theta_{\rm LSND}$ mixing angle.
2938:
2939: In fig.\fig{LSND} the BBN constraint has been minimized (when allowed by neutrino data)
2940: setting $\theta_{e\rm s}\approx \theta_{\mu\rm s}\approx \theta_{\rm LSND}^{1/2}$.
2941: We see that short-baseline experiments sensible to a $P(\nu_\mu\to \nu_e)$
2942: about $2$ orders of magnitude smaller than the value suggested by LSND
2943: are needed to probe regions compatible with standard BBN.
2944:
2945:
2946:
2947:
2948: \section{Summary}\label{all}
2949: A few years ago active/sterile oscillations were studied as an alternative to active/active oscillations.
2950: For example, %assuming pure $\nu_{\rm a}\leftrightarrow\nu_{\rm s}$ oscillations,
2951: it was shown that neither the solar
2952: nor the atmospheric anomaly can be produced by oscillations into sterile neutrinos
2953: compatibly with standard BBN.
2954: These active vs sterile issues have now been firmly solved by experiments,
2955: and the new relevant questions become:
2956: \begin{quote}
2957: How large can be the subdominant sterile component possibly present in
2958: solar or atmospheric oscillations?
2959: How can we discover new anomalies due to sterile effects?
2960: \end{quote}
2961: In order to address these issues we systematically compared oscillation effects
2962: generated by one sterile neutrino
2963: (including the effects of the now established solar and atmospheric oscillations)
2964: with present experiments
2965: and studied capabilities of future probes,
2966: extending previous analyses in several ways.
2967: Almost all previous analyses that studied these new questions considered
2968: only the peculiar sterile oscillation pattern that gives the simplest physics:
2969: the sterile neutrino mixes with a specific flavour
2970: ($\nu_{\mu,\tau}$ in solar analyses, $\nu_\tau$ in atmospheric ones)
2971: and has a large mass, $\Delta m^2\gg \Delta m^2_{\rm atm,sun}$.
2972: Dropping each one of these assumptions gives quite different physics.
2973: We explored the whole parameter space
2974: (three sterile mixing parameters and one sterile mass)
2975: producing precise results for six representative slices,
2976: that span the spectrum of the various possibilities.
2977: We considered one extra sterile neutrino with arbitrary mass $m_4$, and allowed it to mix
2978: with
2979: $$\nu_e\quad\hbox{or}\quad \nu_\mu\quad\hbox{or}\quad\nu_\tau\quad\hbox{or}\quad
2980: \nu_1\quad\hbox{or}\quad\nu_2\quad\hbox{or}\quad\nu_3$$
2981: where $\nu_{1,2,3}$ are the mass eigenstates in absence of sterile mixing.
2982: The spectrum of active neutrinos is not yet fully known:
2983: rather than studying all possible cases we focussed
2984: on what we believe is most plausible case:
2985: we assumed normal hierarchy of active neutrinos (i.e. $m_1\ll m_2\approx m_3/6$)
2986: and that $\theta_{13}$ is small enough that we can neglect its effects.
2987: If experiments will contradict these assumptions, it will be easy to update
2988: our results; most of them do not depend on these assumptions.
2989:
2990: \medskip
2991:
2992:
2993: \begin{figure}[p]
2994: $$\hspace{-7mm}\includegraphics[width=18cm]{all}$$
2995: \caption{\label{fig:all}\em {\bf Summary of sterile neutrino effects}.
2996: The shaded region is excluded at $99\%$ C.L.\ (2 dof)
2997: by solar or atmospheric or reactor or short base-line experiments.
2998: We shaded as excluded also regions where sterile neutrinos
2999: suppress the SN1987A $\bar\nu_e$ rate by more than $70\%$.
3000: This rate is suppressed by more than $20\%$ inside the {\color{blue} dashed blue line},
3001: that can be explored at the next SN explosion if it will be possible to
3002: understand the collapse well enough.
3003: Within standard cosmology,
3004: the region above the {\color{rossos} red continuous line} is disfavoured (maybe already excluded) by BBN and LSS.
3005: Plots in the various sections show how much each probe can be improved
3006: by future experiments.}
3007: \end{figure}
3008:
3009: We considered the most promising ways to probe the existence of
3010: eV-scale sterile neutrinos.
3011: Most probes are based on a careful study of
3012: natural sources of neutrinos (the universe, the sun, supernov\ae, cosmic rays,...)
3013: which have their own peculiar capabilities and limitations.
3014: We studied how one extra sterile neutrino affects BBN (helium-4, deuterium), CMB, LSS and solar, atmospheric, reactor, beam experiments.
3015: The sensitivity of some of these probes is enhanced by MSW resonances~\cite{MSW}.
3016: In cosmology, active $\nu$ and $\bar\nu$ encounter a
3017: MSW resonance with sterile neutrinos lighter than active ones.
3018: Roughly the same happens to supernova $\bar\nu_e$
3019: (that also experience less important MSW resonances in the opposite situation).
3020: On the contrary, solar $\nu_e$ encounter a MSW resonance with sterile neutrinos heavier than
3021: active ones.
3022:
3023:
3024: Fig.s\fig{all} combines present constraints.
3025: Each probe is described in greater detail in its specific series of figures:
3026: fig.s\fig{BBN} for cosmology,
3027: fig.s\fig{sun} for solar experiments,
3028: fig.s\fig{SN} for supernov\ae,
3029: fig.s\fig{SBL} for short base-line experiments,
3030: fig.s\fig{atm} for atmospheric experiments.
3031: These figures also show the capabilities of some future experiments,
3032: that we now try to summarize in words.
3033: \begin{itemize}
3034: \item Compatibility with standard BBN constrains
3035: sterile oscillations occurred at temperatures $T\circa{>} 0.1\MeV$.
3036: It is very important to
3037: improve measurements of the {\bf helium-4} primordial abundancy
3038: (that we parameterize in terms of an effective number of neutrinos $N_\nu^{^4{\rm He}}$,
3039: see eq.~(\ref{sys:HeD}))
3040: until $N_\nu^{^4{\rm He}}=4$ will be safely tested.
3041: This requires overcoming `systematic' uncertainties.
3042: % as discussed in the relevant literature
3043: % and summarized in reviews.
3044: The helium-4 abundancy is sensitive to two different sterile effects:
3045: increase of the total neutrino density, and depletion of electron-neutrinos.
3046: The second effect makes
3047: $N_\nu^{^4{\rm He}}$
3048: sensitive to sterile oscillations down to $\Delta m^2\sim 10^{-8}\eV^2$,
3049: while the first effect becomes negligible at $\Delta m^2\circa{<} 10^{-5}\eV^2$.
3050:
3051: However, if BBN were non-standard, a modified density of electron neutrinos
3052: could compensate the sterile corrections to $N_\nu^{^4{\rm He}}$:
3053: for example helium-4 constraints on sterile oscillations can be evaded by
3054: allowing a large neutrino asymmetry.
3055:
3056:
3057:
3058: \item For all these reasons it is important to measure a second BBN effect.
3059: The {\bf deuterium} primordial abundancy is affected by milder systematic problems:
3060: in the future it might be possible to improve its measurement obtaining an
3061: uncertainty on the effective parameter $N_\nu^{\rm D}$ (precisely defined in eq.~(\ref{sys:HeD})))
3062: significantly below 1,
3063: possibly making deuterium the most significant BBN probe.
3064: We have computed the ranges of active/sterile oscillation parameters that significantly affect $N_\nu^{\rm D}$:
3065: it is less sensitive than helium-4 to $\nu_e$ depletion and therefore
3066: to values of $\Delta m^2$ below $10^{-5}\eV^2$ (fig.\fig{BBN}).
3067: %{\bf COME DIPENDE DA DEG?}
3068:
3069:
3070:
3071: \item Future studies of {\bf Cosmic Microwave Background} acoustic oscillations
3072: should allow to precisely measure the total neutrino density $N_\nu^{\rm CMB}$
3073: at recombination ($T\sim \eV$) with $\pm0.2$ ({\sc Planck}) or
3074: maybe $\pm0.05$ ({\sc CMBpol}) error~\cite{NuCDM}.
3075: Neutrinos affect CMB in various ways;
3076: neutrino free-streaming offers a clean signature that allows to count neutrinos.
3077: Sterile neutrinos affect $N_\nu^{\rm CMB}$ only if $\Delta m^2\circa{>}10^{-5}\eV^2$;
3078: in such a case
3079: $N_\nu^{\rm CMB}\approx N_\nu^{^4{\rm He}}\approx N_\nu^{\rm D}$.
3080: Therefore CMB will cover only a part of the region that
3081: BBN could probe with fully reliable measurements of
3082: the helium-4 or deuterium abundances.
3083:
3084:
3085:
3086:
3087: %\footnote{Present
3088: %experiments are far from being able of directly detecting CMB neutrinos.}
3089:
3090: \item CMB also allows to probe $\eV$-scale $\nu$ masses.
3091: Smaller $\nu$ masses can be probed by measuring how much galaxies are clustered,
3092: because neutrinos become non relativistic when the observable universe had a size comparable
3093: to present cluster of galaxies.
3094: Relativistic neutrinos freely move and tend to
3095: reduce the amount of clustering.
3096: Recently, {\bf Large Scale Structure} data (together with precise CMB measurements)
3097: gave a bound on the present energy density in neutrinos
3098: $\Omega_\nu < 0.0076$ at 95 \% C.L.~\cite{WMAP,boundMnu}, dominated by neutrino masses
3099: (rather than neutrino energy).
3100: With only active neutrinos this implies a significant bound on their mass, $m_\nu < 0.23\eV$~\cite{WMAP,boundMnu}.
3101:
3102: Sterile neutrinos contribute to $\Omega_\nu$
3103: (a small non-thermal population of relatively heavy sterile neutrinos also
3104: modifies the way $\Omega_\nu$ manifests).
3105: For small active/sterile mixing
3106: LSS constrains sterile oscillations more strongly than BBN (fig.\fig{BBN}).
3107:
3108:
3109:
3110: \item {\bf Solar} $\nu_e$ experiments have explored sterile oscillations not testable by BBN,
3111: thanks to two different effects.
3112: (1) MSW resonances make solar $\nu_e$ sensitive to small active/sterile mixing
3113: and $\Delta m^2\circa{>}10^{-8}\eV^2$.
3114: (2) With large mixing, solar $\nu_e$ are sensitive down to $\Delta m^2\sim 10^{-12}\eV^2$.
3115: Future experiments will explore new aspects of the solar neutrino anomaly,
3116: allowing to measure in a redundant way the active oscillation parameters
3117: or to discover a new anomaly.
3118: We emphasize one qualitative point.
3119: Due to LMA oscillations, neutrinos exit from the sun as almost pure $\nu_2$
3120: at energies $E_\nu \circa{>} \hbox{few MeV}$.
3121: Neutrinos with these energies have been precisely studied by SK and SNO,
3122: but are almost unaffected by sterile oscillations if they involve mostly $\nu_1$.
3123: This could happen either when $\nu_{\rm s}$ mixes with $\nu_1$,
3124: or when $\nu_{\rm s}$ experiences a level-crossing with $\nu_1$.
3125: Therefore {\em there is a whole class of sterile effects which
3126: manifest only at $E_\nu \circa{<} \hbox{\rm few MeV}$} ---
3127: an energy range explored so far only by Gallium experiments.
3128: Future precise measurement of solar $\nu_e$ at sub-MeV energies
3129: will allow to significantly
3130: extend searches for active/sterile effects.
3131: Part of these extended region can be soon
3132: tested by Borexino, where a sterile neutrino can manifest
3133: as day/night variations, or as seasonal variations,
3134: or even by reducing the total rate.
3135:
3136:
3137:
3138:
3139: \item {\bf Supernova} neutrinos
3140: will be good probes of sterile oscillations because
3141: have a different pattern of MSW resonances
3142: and a longer base-line than solar $\nu_e$.
3143: Consequently supernova $\bar\nu_e$ are more sensitive than solar $\nu_e$
3144: in two main cases:
3145: (a) small $\Delta m^2\circa{>}10^{-18}\eV^2$ with large $\theta_{\rm s}$;
3146: (b) $\nu_{\rm s}$ lighter than $\nu_1$ with small mixing.
3147: Oscillations into one sterile neutrino can reduce the $\bar\nu_e$ rate by up to $80\%$
3148: (see fig.s\fig{SN})
3149: and, in a more restricted range of oscillation parameters,
3150: vary the average $\bar\nu_e$ energy by $30\%$.
3151: SN1987A data agreed with expectations.
3152: Future SN experiments can perform quantitative test, but
3153: it is not clear how to deal with theoretical uncertainties.
3154: We also discussed other less promising astrophysical probes.
3155:
3156:
3157: \item {\bf Atmospheric} experiments (SK, MACRO, K2K) indirectly exclude
3158: active/sterile oscillations with $\Delta m^2\circa{>}10^{-3\div 4}\eV^2$
3159: and $\tan^2\theta_{\rm s}\circa{>}0.1\div0.2$.
3160: Up to minor differences, this applies to all flavours (fig.\fig{atm}).
3161: {\bf Terrestrial experiments} that mainly probed disappearance of
3162: $\bar\nu_e$ and $\nu_\mu$ ({\sc Chooz}, CDHS,\ldots)
3163: exclude active/sterile mixings with these flavours with $\tan^2\theta_{\rm s}\circa{>}0.03$
3164: and $\Delta m^2\circa{>}10^{-3}\eV^2$ (fig.\fig{SBL}).
3165: Therefore future short-baseline experiments can search for sterile effects with smaller $\theta_{\rm s}$.
3166: Possible signals are $\bar\nu_e$ disappearance in reactor experiments,
3167: a deficit of NC events or $\nu_\tau$ appearance in beam experiments.
3168: %(not with smaller $\Delta m^2$).
3169: Within standard cosmology these effects can be probed by CMB and BBN,
3170: which already disfavour them.
3171:
3172:
3173: \end{itemize}
3174: We listed present anomalies that can be interpreted as due to sterile neutrinos.
3175: None looks particularly significant, with the possible exception of
3176: the LSND anomaly, that can be due to oscillations of `$3+1$' neutrinos.
3177: Our precise study confirms that the extra sterile neutrino suggested by LSND
3178: thermalizes almost completely before BBN
3179: (in agreement with estimations of most previous analyses).
3180: In fig.\fig{LSND} we plot, as function of the effective LSND oscillation parameters
3181: $\theta_{\rm LSND}$ and $\Delta m^2_{\rm LSND}$, the constraints on
3182: $N_\nu^{^4{\rm He}}\simeq N_\nu^{\rm D}\simeq N_\nu^{\rm CMB}$
3183: (mainly from helium-4 data) and $\Omega_\nu$ (mainly from LSS data):
3184: in the relevant region our precise plot negligibly differs from previous estimates~\cite{instant,2+2}.
3185: The 3+1 interpretation of LSND is not compatible with standard BBN,
3186: and gives a $\Omega_\nu$ which is only marginally compatible with standard cosmology.
3187:
3188: \bigskip
3189:
3190:
3191: \paragraph{Acknowledgments}
3192: We thank S.~Woosley, A.~Burrows, P.~Di Bari, S.~Sarkar, S.~Pascoli, A.~Romanino, K.~Olive.
3193: The work of M.C. is supported in part by the USA department of energy
3194: under contract DE-FG02-92ER-40704.
3195: Part of the work of M.C.\ was done while at scuola normale superiore (Pisa).
3196: The authors thank the CERN theory division, where part of the work was done.
3197: %\newpage
3198:
3199: \footnotesize
3200: \begin{multicols}{2}
3201: \begin{thebibliography}{99}
3202:
3203: \bibitem{BBN}
3204: For an old but still useful review see
3205: \art{R.V. Wagoner, W.A. Fowler, F. Hoyle}{The Astrophys. J.}{148}{3}{1967}.
3206: For a recent review see
3207: \art[hep-ph/9602260]{S.~Sarkar}{Rept.\ Prog.\ Phys.}{59}{1493}{1996}.
3208: For precision computations, see
3209: \art{D.A. Dicus et al.}{\PR}{D26}{2694}{1982}.
3210: \art[astro-ph/9807279]{R.E.~Lopez, M.S.~Turner}{\PR}{D59}{103502}{1999}.
3211: \art{S. Esposito et al.}{\NP}{B568}{2000}{421}.
3212: \hepart[astro-ph/0307213]{A.~Cuoco et al.}.
3213:
3214:
3215:
3216: \bibitem{etas}
3217: \hepart[hep-ph/0309174]{A. Bandyopadhyay et al.}.
3218: \hepart[hep-ph/0311217]{T. Schwetz}.
3219:
3220:
3221: \bibitem{etaa}
3222: See e.g.\ the talk by K. Okumura at
3223: the NO-VE workshop, Venice 3--5 dec.\ 2003,
3224: available at the web page
3225: axpd24.pd.infn.it/NO-VE/NO-VE.html.
3226:
3227:
3228:
3229: \bibitem{LSND} \hepart[hep-ex/0104049]{The LSND collaboration}.
3230:
3231:
3232:
3233:
3234: \bibitem{SterileB} V. Berezinsky et al., as cited in~\cite{Sterile}.
3235:
3236:
3237:
3238:
3239:
3240: \bibitem{2+2}
3241: \art[hep-ph/0201134]{A. Strumia}{\PL}{B539}{91}{2002}.
3242: Other authors finally agree
3243: that `2+2' oscillations are no longer viable,
3244: see e.g.\ \art[hep-ph/0207157]{M. Maltoni, T. Schwetz, M.A. Tortola, J.W.F. Valle}{\NP}{B643}{321}{2002}.
3245: However some controversy remains, see e.g.
3246: \art[hep-ph/0007065]{R. Foot}{\PL}{B496}{169}{2000},
3247: \art[hep-ph/0210393]{R. Foot}{Mod. Phys. Lett.}{A18}{207}{2003} and
3248: \art[hep-ph/0209373]{H. P\"as, L.G. Song, T. J. Weiler}{\PR}{D67}{073019}{2003}.
3249: According to
3250: \hepart[hep-ph/0303005]{R. Foot}.
3251: the NC-enriched sample of SK atmospheric data
3252: might be less enriched than what claimed by SK.
3253: The K2K near detector is measuring the relevant cross sections,
3254: with results reported by S. Nakayama at the NuInt04 conference
3255: (web site: nuint04.lngs.infn.it).
3256:
3257: In any case, exclusion of `2+2' oscillations follows from global fits,
3258: not directly from a few simple measurements.
3259:
3260:
3261:
3262: \bibitem{Sterile}
3263: {Models with sterile neutrinos}.
3264: Light fermions from a discrete symmetry:
3265: \art{E. Ma, P. Roy}{\PR}{D52}{4780}{1995}.
3266: From a continuos symmetry:
3267: \art{E. Ma}{Mod. Phys. Lett.}{A11}{1893}{1996}.
3268: % From decomposing a $27=16+10+1$ $E_6$ multiplet into SO(10) fragments:
3269: % \art{E. Ma}{\PL}{B380}{286}{1996}.
3270: From a supersymmetric $R$-symmetry:
3271: \art{E.J. Chun, A.S. Joshipura, A.Yu. Smirnov}{\PL}{B357}{608}{1995}.
3272: As Goldstone particles:
3273: \art{E.J. Chun, A.S. Joshipura, A.Yu. Smirnov}{\PR}{D54}{4654}{1996}.
3274: As modulinos:
3275: \art{K. Benakli, A.Yu. Smirnov}{\PRL}{79}{4314}{1997}.
3276: From a mirror world:
3277: \art{S.I. Blinikov, M. Yu Khlopov}{Sov. Astron.}{27}{371}{1983}.
3278: \art[hep-ph/9503481]{Z. Silagadze}{Phys. Atom. Nucl.}{60}{272}{1997}.
3279: \art[hep-ph/9505359]{R. Foot, R. Volkas}{\PR}{D52}{6595}{1995}.
3280: \art[hep-ph/9505385]{Z.G. Berezhiani, R.N. Mohapatra}{\PR}{D52}{6607}{1995}.
3281: \hepart[hep-ph/0210204]{V. Berezinsky, M. Narayan, F. Vissani}.
3282: From compositness:
3283: \hepart[hep-ph/9806223]{N. Arkani-Hamed, Y. Grossman}.
3284: From warped extra dimensions or compositness:
3285: \hepart[hep-ph/0312392]{T. Gherghetta}.
3286: From mirror fermions:
3287: \art[hep-ph/9806400]{M. Bando, K. Yoshioka}{Prog.\ Theor.\ Phys.}{100}{1239}{1998}.
3288: From supersymmetry breaking:
3289: \art[hep-ph/0008110]{K.S. Babu, T. Yanagida}{\PL}{B491}{148}{2000}.
3290: \art[hep-ph/0011141]{F.~Borzumati, K.~Hamaguchi, T.~Yanagida}{\PL}{B497}{259}{2001}.
3291: Small sterile masses from supersymmetric flat directions:
3292: \art[hep-ph/9805281]{P. Langacker}{\PR}{D58}{093017}{1998}.
3293: Models with 2 sterile neutrinos:
3294: \hepart[hep-ph/0312285]{K.S. Babu, G. Seidl}.
3295: \hepart[hep-ph/0401241]{K.L. McDonald, B.H.J. McKellar, A. Mastrano}.
3296: \hepart[hep-ph/0402183]{W. Krolikowski}.
3297:
3298:
3299: \bibitem{WMAP}
3300: \hepart[astro-ph/0302207]{C.L. Bennett et al.}
3301: and
3302: \hepart[astro-ph/0302209]{D.N. Spergel et al.}.
3303:
3304:
3305:
3306: \bibitem{He4} { Recent measurements of the helium-4 primordial abundancy}.
3307: %\art{B.~E.~J.~Pagel, E.~A.~Simonson, R.~J.~Terlevich and M.~G.~Edmunds}{Mon.\ Not.\ Roy.\ Astron.\ Soc.\}{255}{325}{1992}.
3308: %\art{K.~A.~Olive and G.~Steigman}{Astrophys.\ J.\ Suppl.\}{97}{49}{1995}.
3309: \art{Y.I. Izotov, T.X. Thuan}{Astrophys.\ J.}{500}{188}{1998}.
3310: \art{K.A. Olive, E. Skillman, G. Steigman}{Astrophys.\ J.}{483}{788}{1997}.
3311: \art{Y.I. Izotov, F.H. Chaffee, C.B. Foltz, R.F. Green, N.G. Guseva. T.X. Thuan}{Astrophys.\ J.}{527}{757}{1999}.
3312: \hepart[astro-ph/0003234]{T.X. Thuan, Y.I. Izotov}.
3313: \art{M. Peimbert, A. Peimbert, M. T. Ruiz}{Astrophys.\ J.}{541}{688}{2000}.
3314: \art{A. Peimbert, M. Peimbert, V. Luridiana}{Astrophys.\ J.}{565}{668}{2002}.
3315: \hepart[astro-ph/0310421]{Y.I. Izotov, T.X. Thuan}.
3316: \hepart[astro-ph/0405588]{K.A. Olive, E.D. Skillman}.
3317:
3318:
3319:
3320: \bibitem{NnuBBN}
3321: \art{P. Di Bari}{Phys. Rev. D} {65}{043509}{2002} and addendum:
3322: \art{P. Di Bari}{Phys. Rev. D} {67}{127301}{2003};
3323: \art{R. H. Cyburt, B. D. Fields, K. A. Olive}{Phys. Lett. B} {567}{227}{2003};
3324: S. Hannestad as cited in~\cite{boundMnu};
3325: \hepart[hep-ph/0305075]{V. Barger et al.};
3326: \art[astro-ph/0205238]{K.N. Abazajian}{Astropart.\ Phys.}{19}{303}{2003}.
3327:
3328:
3329:
3330: %
3331: \bibitem{DeutQuas}
3332: { Recent measurements of the deuterium primordial abundancy}.
3333: \art{D. Kirkman, D. Tytler, S. Burles, D. Lubin, J.M. O'Meara}{Astrophys.\ J.}{529}{665}{2000}.
3334: \art{J. M. O'Meara, D. Tytler, D. Kirkman, N. Suzuki, J.X. Prochaska, D. Lubin, A.M. Wolfe}{Astrophys.\ J.}{552}{718}{2001}.
3335: \hepart[astro-ph/0302006]{D. Kirkman, D. Tytler, N. Suzuki, J.M. O'Meara, D. Lubin}.
3336: \art{M. Pettini, D.V. Bowen}{Astrophys.\ J.}{560}{41}{2001}.
3337: \art{S. D'Odorico, M. Dessauges-Zavadsky, P. Molaro}{Astronomy and Astrophysics}{368}{L21}{2001}.
3338: \art{S.A. Levshakov, P. Molaro, M. Dessauges-Zavadsky, S. D'Odorico}{Astrophys.\ J.}{565}{696}{2002}.
3339:
3340:
3341:
3342:
3343: \bibitem{Dolgov} \hepart[hep-ph/0310138]{F.L. Villante, A.D. Dolgov}.
3344:
3345:
3346: \bibitem{review} A. Strumia, F. Vissani, a full review on neutrino masses and mixings,
3347: IFUP-TH/2004-01,
3348: hopefully to appear.
3349:
3350: %
3351: \bibitem{bbnosc}
3352: { Sterile/active oscillations and BBN}:
3353: D. Kirilova, Dubna preprint JINR E2-88-301.
3354: \art{R. Barbieri, A. Dolgov}{\PL}{B237}{440}{1990}.
3355: \art{K. Enqvist et al.}{\PL}{B249}{531}{1990}.
3356: \art{K. Kainulainen}{\PL}{B244}{191}{1990}.
3357: \art{R. Barbieri, A. Dolgov}{\NP}{B349}{743}{1991}.
3358: \art{K. Enqvist et al.}{\NP}{B373}{498}{1992}.
3359: \art{J.M. Cline}{\PRL}{68}{3137}{1992}.
3360: \art{X. Shi, D.N. Schramm, B.D. Fields}{\PR}{D48}{2563}{1993}.
3361: \art{R. Foot, M.J. Thomson, R.R. Volkas}{\PR}{D53}{5349}{1996}.
3362: \art[hep-ph/9901404]{E. Lisi, S. Sarkar, F.L. Villante}{\PR}{D59}{123520}{1999}.
3363: \art[hep-ph/9707282]{D.P. Kirilova, M.V. Chizhov}{\PR}{D58}{073004}{1998}.
3364: \art[hep-ph/9909408]{D.P. Kirilova, M.V. Chizhov}{\NP}{B591}{457}{2000}.
3365: \art{A.D. Dolgov et al.}{\NP}{B632}{363}{2002}.
3366: \hepart[astro-ph/0205238]{H.N. Abazajian}.
3367:
3368:
3369: \bibitem{etarimuove}
3370: \art[hep-ph/9508275]{R. Foot, R.R. Volkas}{\PRL}{75}{4350}{1995}.
3371:
3372:
3373: \bibitem{asymm}
3374: The rising of a large neutrino asymmetry (in specific regions of the parameters space)
3375: and its effects (for specific mixing patterns) are discussed e.g.\ in
3376: \art[hep-ph/9509325]{R. Foot, M.J. Thomson, R.R. Volkas}{\PR}{D53}{5349}{1996},
3377: \art[hep-ph/9706242]{R. Foot and R.R. Volkas}{\PR}{D56}{6653}{1997}
3378: and Erratum {\it ibid.} D59 (1999) 029901,
3379: \art[hep-ph/9805259]{N. F. Bell, R. Foot and R.R. Volkas}{\PR}{D58}{105010}{1998},
3380: and also debated in
3381: \art[hep-ph/9910444]{A.D. Dolgov, S.H. Hansen, S. Pastor, D.V. Semikoz}{Astropart. Phys.}{14}{79}{2000},
3382: \art[hep-ph/0008245]{P. Di Bari, R. Foot, R.R. Volkas, Y.Y.Y. Wong}{Astropart. Phys.}{15}{391}{2001},
3383: \art[hep-ph/9909408]{D.P. Kirilova, M.V. Chizhov}{\NP}{B591}{457}{2000}.
3384: The role of a primordial, non-dynamical neutrino asymmetry is also addressed in several papers of \cite{NnuBBN} and e.g. in
3385: \art[hep-ph/0306061]{V. Barger et al.}{\PL}{B569}{123}{2003}.
3386:
3387:
3388: \bibitem{NuCDM}
3389: For a useful review see
3390: \hepart[astro-ph/0310198]{S. Bashinsky, U. Seljak}.
3391:
3392:
3393: \bibitem{CMBfits}
3394: \hepart[astro-ph/0302337]{P. Crotty, J. Lesgourges, S. Pastor};
3395: \art{E. Pierpaoli}{Mon. Not. Roy. Astron. Soc.}{342}{L63}{2003};
3396: \art{V. Barger, J. P. Kneller, H. S. Lee, D. Marfatia, G. Steigman}{Phys. Lett. B}{566}{8}{2003};
3397: S. Hannestad as cited in~\cite{boundMnu}.
3398:
3399: %
3400: \bibitem{boundMnu}
3401: The cosmological bound on neutrino masses
3402: is obtained by combining WMAP data
3403: with large-scale structure (and possibly other) data, see the papers in~\cite{WMAP}
3404: and references therein.
3405: This is similar to pre-WMAP analyses, e.g.
3406: \art{A. Lewis, S. Bridle}{\PR}{D66}{103511}{2002}.
3407: Other (sometimes more conservative) analysis find similar (sometimes weaker) bounds:
3408: \art[astro-ph/0303076]{S. Hannestad}{JCAP} {0305}{004}{2003},
3409: \art[astro-ph/0306386]{S.W. Allen, R.W. Schmidt, S.L. Bridle}{Mon.\ Not.\ Roy.\ Astron.\ Soc.}{346}{593}{2003},
3410: \art[astro-ph/0310723]{M.~Tegmark et al. [SDSS Collaboration]}{\PR}{D69}{103501}{2004}
3411: \art[hep-ph/0312065]{V.Barger et al.}{\PL}{B595}{55}{2004},
3412: \art[hep-ph/0402049]{P. Crotty, J. Lesgourgues, S. Pastor}{\PR}{D69}{123007}{2004},
3413: \hepart[astro-ph/0407372]{U.~Seljak et al.}.
3414: %\hepart[hep-ph/0408045]{G.~L.~Fogli et al.}.
3415: The robustness of the bound in presence of
3416: non-adiabatic incoherent fluctuations was studied in
3417: \art[hep-ph/0401239]{R.H. Brandenberger, A. Mazumdar, M. Yamaguchi}{\PR}{D69}{081301}{2004}.
3418:
3419:
3420:
3421:
3422: \bibitem{HR}
3423: \hepart[hep-ph/0312154]{S. Hannestad, G. Raffelt}.
3424: \hepart[hep-ph/0402049]{P. Crotty et al.}.
3425:
3426:
3427:
3428: %
3429: \bibitem{OscUniverse}
3430: { Neutrino oscillations in the early universe}.
3431: \art{A. Dolgov}{Sov. J. Nucl. Phys.}{33}{700}{1981}.
3432: \art{L. Stodolsky}{\PR}{D36}{2273}{1987}.
3433: \art{A. Manohar}{\PL}{B186}{370}{1987}.
3434: \art{M.J. Thomson, B.H.J. McKellar}{\PL}{B259}{113}{1991}.
3435: \art{J. Pantaleone}{\PL}{B287}{128}{1992}.
3436: \hepart[hep-ph/0304055]{A. Friedland, C. Lunardini}.
3437: The formalism is clearly summarized in
3438: \art{G. Sigl, G. Raffelt}{\NP}{B406}{423}{1993}.
3439: The dominant contribution to the refraction index was discussed in
3440: \art{D. N\"otzold, G. Raffelt}{\NP}{B307}{924}{1988}.
3441:
3442:
3443:
3444: \bibitem{DolgovReview} \art[hep-ph/0202122]{A.D. Dolgov}{Phys. Rept.}{370}{333}{2002}.
3445:
3446:
3447:
3448: \bibitem{MSW}
\art{L. Wolfenstein}{\PR}{D17}{2369}{1978};
\art{S.P. Mikheyev, A. Yu Smirnov}{Sovietic Journal
Nucl. Phys.}{42}{913}{1986}.
3449:
3450:
3451:
3452:
3453: \bibitem{BBNapprox}
3454: \art{J. Bernstein et al.}{Rev. Mod. Phys.}{61}{25}{1989}.
3455: The paper
3456: \art{R. Esmailzadeh, G.D. Starkman, S. Dimopoulos}{Astrophysical Journal}{378}{504}{1991},
3457: {\em `Primordial nucleosynthesis without a computer'},
3458: contains rough analytical estimates.
3459: With a computer their methods can be used to develop
3460: accurate approximations that do not involve the slow integration
3461: of the network of Boltzmann BBN equations.
3462:
3463:
3464: \bibitem{DV} \hepart[hep-ph/0308083]{A.D. Dolgov, F.L. Villante}.
3465:
3466:
3467: \bibitem{Allen}
3468: S.W. Allen et al., as cited in~\cite{boundMnu}.
3469: See however
3470: \hepart[hep-ph/0312122]{S. Hannestad}
3471: and
3472: SDSS collaboration as cited in~\cite{boundMnu}.
3473:
3474: %
3475: \bibitem{reionization} \hepart[hep-ph/0305126]{S.H. Hansen, Z. Haiman}.
3476:
3477:
3478: \bibitem{Picciotto} \hepart[hep-ph/0402178]{C. Picciotto, M. Pospelov}.
3479:
3480:
3481: \bibitem{PREM} \art{A.M. Dziewonski, D.L. Anderson}{Phys. Earth Planet. Interior}{25}{297}{1981}.
3482:
3483:
3484: \bibitem{Petcov}
3485: {Neutrino propagation in a varying density}.
\art{S. Parke}{\PRL}{57}{1275}{1986}.
\art{P. Pizzochero}{\PR}{D36}{2293}{1987}.
\art{S.T. Petcov}{\PL}{B200}{373}{1988}.
3486: \art{T.K. Kuo, J. Pantaleone}{\PR}{D39}{1930}{1989}.
3487:
3488:
3489: \bibitem{KPreview}
3490: For a review see
\art{T.K. Kuo, J. Pantaleone}{Rev. Mod. Phys.}{61}{937}{1989}.
3491:
3492:
3493:
3494:
3495: \bibitem{SNOlast} \hepart[nucl-ex/0204008]{SNO collaboration}.
3496: \hepart[nucl-ex/0204009]{SNO collaboration}.
3497:
3498:
3499: \bibitem{SNOsalt}
3500: \hepart[nucl-ex/0309004]{SNO collaboration}.
3501: See also ``{\em HOWTO use the SNO salt flux results}'' available at the SNO web site
3502: www.sno.phy.queensu.ca.
3503:
3504:
3505: \bibitem{SKlast}
3506: \hepart[hep-ex/0205075]{Super-Kamiokande collaboration}.
3507:
3508:
3509: \bibitem{Galliumlast} \art{Gallex collaboration}{\PL}{B447}{127}{1999}.
3510: \hepart[astro-ph/0204245]{SAGE collaboration}.
3511: Latest data have been presented in a talk by C. Cattadori
3512: at the `Neutrino 2004' conference (Paris, 14-19 June),
3513: web site neutrino2004.in2p3.fr.
3514:
3515:
3516:
3517: \bibitem{Chlorinelast}
3518: The results of the Homestake experiment are reported in
3519: \art{B.T. Cleveland et al.}{Astrophys. J.}{496}{505}{1998}.
3520:
3521:
3522: \bibitem{KamLAND}
3523: \art[hep-ex/0212021]{KamLAND collaboration}{\PRL}{90}{021802}{2003}.
3524: \hepart[hep-ex/0406035]{KamLAND collaboration}.
3525:
3526:
3527:
3528:
3529: \bibitem{BP}
3530: \art[astro-ph/0010346]{J.N. Bahcall, S. Basu, M.H. Pinsonneault}{Astrophys. J.}{555}{990}{2001}.
3531: A $\chi^2$ for solar data was constructed in
3532: \art{G.L. Fogli, E. Lisi}{Astropart. Phys.}{3}{185}{1995}.
3533:
3534:
3535:
3536:
3537: \bibitem{LUNA} \hepart[nucl-ex/0312015]{LUNA collaboration}.
3538: See also \hepart[astro-ph/0312559]{S. degl'Innocenti et al.}.
3539:
3540:
3541:
3542:
3543: \bibitem{SmirnovSterile} \hepart[hep-ph/0307266]{P.C. de Holanda, A. Yu. Smirnov}.
3544: See also \hepart[hep-ph/0211264]{P.C. de Holanda, A. Yu. Smirnov}.
3545:
3546:
3547:
3548: \bibitem{freq} \art[hep-ph/0102234 updated version 5]{P. Creminelli et al.}{J. HEP}{05}{2001}{052}.
3549:
3550: %
3551: \bibitem{subMeV}
3552: Sub-MeV solar $\nu_e$ experiments were regarded as the tool
3553: for discriminating LMA from the alternative solutions,
3554: which have been already excluded.
3555: The impact of sub-MeV on still open questions was discussed in
3556: \art[hep-ph/0109172]{A. Strumia, F. Vissani}{J.HEP}{048}{0111}{2001}.
3557: Recent analyses find similar results, see e.g.\
3558: \art{J.N. Bahcall, C. Pe\~na-Garay}{J.HEP}{0311}{004}{2003}.
3559:
3560:
3561:
3562:
3563:
3564:
3565:
3566: \bibitem{S17}
3567: \art{E.G. Adelberger et al.}{Rev. of Modern Phys.}{70}{1265}{1998}.
3568: Recent measurements of $S_{17}(0)$:
3569: \art{N. Iwasa et al.}{\PRL}{83}{2910}{1999}.
3570: \art{B. Davids et al.}{\PRL}{86}{2750}{2001}.
3571: \art{F. Hammache et al.}{Phys. Rev. Lett.}{86}{3985}{2001}.
3572: \art{F. Strieder et al.}{Nucl. Phys.}{A696}{219}{2001}.
3573: \art[nucl-ex/0212011]{ISOLDE collaboration}{Phys.Rev. C}{67}{065805}{2003}.
3574: \art[nucl-ex/0308003]{A.R. Junghans et al.}{Phys.\ Rev.\ C}{68}{065803}{2003}.
3575: \hepart[nucl-ex/0311012]{K.A. Snover et al.}.
3576: \art{F. Schumann et al.}{\PRL}{90}{232501}{2003}.
3577: For a recent summary see
3578: \hepart[nucl-ex/0312003]{M. Gai}.
3579:
3580:
3581: \bibitem{Pee(t)}
3582: \art{P.A. Sturrock, G. Walther, M. S. Wheatland}{Astrophys. J.}{491}{409}{1997}.
3583: \art{P.A. Sturrock, M. A. Weber}{Astrophys. J.}{565}{1366}{2002}.
3584: \hepart[hep-ph/0305303]{D. O. Caldwell, P. A. Sturrock}.
3585: \hepart[hep-ph/0301252]{A. Milsztajn}.
3586: \hepart[hep-ph/0309239]{P.A. Sturrock}.
3587: \hepart[hep-ph/0402194]{B.C. Chauhan, J. Pulido}.
3588: See however
3589: \art[hep-ex/0307070]{J. Yoo et al., SK collaboration}{\PR}{D68}{092002}{2003} and
3590: \hepart[hep-ph/0406248]{L. Pandola (GNO collaboration)}.
3591:
3592:
3593:
3594: \bibitem{SNgeneral}
3595: See e.g.\
3596: \art[hep-ph/0201099]{G.G. Raffelt}{\NP Proc. Suppl.}{110}{254}{2002};
3597: \hepart[astro-ph/0310859]{E.~Cappellaro, R.~Barbon, M.~Turatto}.
3598:
3599:
3600:
3601: \bibitem{raffelt book}
3602: G. G. Raffelt, ``{\em Stars as laboratories for fundamental physics}'', Chicago Univ. Press, 1996.
3603:
3604:
3605:
3606: \bibitem{Cei}
3607: For a recent review see e.g.\
3608: \art[hep-ex/0202043]{F. Cei}{Int. J. Mod. Phys.}{A17}{1765}{2002}.
3609:
3610:
3611: \bibitem{bethe}
3612: \art{H.A. Bethe}{Rev. Mod. Phys.}{62}{801}{1990} and references therein.
3613:
3614:
3615: \bibitem{burrowlattimer}
3616: \art{A. Burrows, J.M. Lattimer}{Astrophys. J.}{307}{178}{1986}.
3617:
3618:
3619:
3620:
3621: \bibitem{burrows}
3622: \art[astro-ph/0211194]{T.A. Thompson, A. Burrows, P.A. Pinto}{Astrophys.\ J.}{592}{434}{2003}
3623:
3624:
3625:
3626: \bibitem{SNsterile1}
3627: \art[hep-ph/9702372]{H. Nunokawa, J. T. Peltoniemi, A. Rossi, J. W. F. Valle}{\PR}{D56}{1704}{1997}.
3628:
3629:
3630:
3631:
3632: \bibitem{SN3+1}
3633: \art[hep-ph/0011054]{O.L.G. Peres, A.Yu. Smirnov}{\NP}{B599}{3}{2001}.
3634:
3635:
3636:
3637: \bibitem{SNsterile2}
3638: Neutrino oscillations in SN were discussed in
3639: \art{S.P. Mikheev, A.Yu. Smirnov}{Sov. Phys. JETP}{64}{4}{1986}.
3640: Bounds on sterile oscillations from the SN 1987A rate were discussed in
3641: \art{S.P. Mikheev, A.Yu. Smirnov}{JETP Lett.}{46}{10}{1987}.
3642: \art{T.K.~Kuo, J.~Pantaleone}{\PR}{D37}{298}{1988}.
3643: \art{S.P. Mikheev, A.Yu. Smirnov}{Prog. Nucl. Phys.}{23}{41}{1989}.
3644: \art{K.Kainulainen, J.Maalampi and J.T. Peltoniemi}{\NP}{B358}{435}{1991}.
3645: \art{H.Athar, J.T. Peltoniemi}{\PR}{D51}{5785}{1995}.
3646: %%CITATION = PHRVA,D51,5785;%%
3647: \art[hep-ph/0008138]{A. D. Dolgov, S. H. Hansen, G. Raffelt, D. V. Semikoz}{\NP}{B590}{562}{2000}.
3648: \hepart[hep-ph/0401082]{P. Keranen, J. Maalampi, M. Myyrylainen, J. Riittinen}.
3649:
3650:
3651:
3652: \bibitem{ShiSigl}
3653: \art[hep-ph/9312247]{X. Shi, G. Sigl}{\PL}{B323}{360}{1994}.
3654:
3655:
3656:
3657:
3658:
3659:
3660: \bibitem{fuller r-process}
3661: \art[astro-ph/9902106]{G.C. McLaughlin, J.M. Fetter, A.B. Balantekin, G.M. Fuller}{\PR}{C59}{2873}{1999}.
3662: \art[astro-ph/9910175]{D. O. Caldwell, G. M. Fuller, Y. Z. Qian}{\PR}{D61}{123005}{2000}.
3663: \art[hep-ph/0205029]{J. Fetter, G. C. McLaughlin, A. B. Balantekin, G. M. Fuller}{Astropart. Phys.}{18}{433}{2003}.
3664:
3665:
3666:
3667: \bibitem{SN1987a signal}
3668: \art{Kamiokande-II collaboration}{\PRL}{58}{1490}{1987};
3669: \art{R.M. Bionta et al.}{\PRL}{58}{1494}{1987}.
3670:
3671:
3672:
3673: \bibitem{sterileshock}
3674: \art{J.T. Peltoniemi}{Astronomy and Astrophysics}{254}{121}{1992} first identified the region of large $Delta m^2_s$ (now excluded by the upper bound on the masses by cosmology)
3675: where the effect on the shock can be positive.
3676:
3677:
3678:
3679: \bibitem{IBD}
3680: \art[hep-ph/9903554]{P. Vogel, J.F. Beacom}{\PR}{D60}{053003}{1999};
3681: \art[astro-ph/0302055]{A. Strumia, F. Vissani}{\PL}{B564}{42}{2003}.
3682:
3683:
3684:
3685: \bibitem{SNmattereffects}
3686: See e.g.\
3687: \art[hep-ph/0106149]{C.~Lunardini, A.~Y.~Smirnov}{\NP}{B616}{307}{2001},
3688: \art{E. K. Akhmedov, C. Lunardini, A. Yu. Smirnov}{\NP}{B643}{339}{2002}
3689: and ref.s therein.
3690:
3691:
3692:
3693: \bibitem{jankahillebrandt}
3694: \art{H.-T. Janka, W. Hillebrandt}{Astronomy and Astrophysics Supplement Series}{78, 3}{375}{1989}.
3695: \art{H.-T. Janka, W. Hillebrandt}{Astronomy and Astrophysics}{224}{49}{1989}.
3696:
3697:
3698:
3699: \bibitem{keil}
3700: \art[astro-ph/0208035]{M. Th. Keil, G. G. Raffelt, H.-T. Janka}{Astrophys. J.}{590}{971}{2003}.
3701:
3702:
3703:
3704:
3705: \bibitem{mutau}
3706: \art{F.J. Botella, C.S. Lim, W.J. Marciano}{\PR}{D35}{896}{1987}.
3707:
3708:
3709:
3710: \bibitem{burrows 2}
3711: \art[hep-ph/0306056]{K. Takahashi, K. Sato, A. Burrows and T. A.~Thompson}{\PR}{D68}{113009}{2003}.
3712:
3713:
3714: \bibitem{SNfit}
3715: \art[astro-ph/9601111]{B. Jegerlehner, F. Neubig, G. Raffelt}{\PR}{D54}{1194}{1996}.
3716:
3717:
3718:
3719: \bibitem{LMAdead} See e.g.\
3720: \art[hep-ph/9305204]{A.Y. Smirnov, D.N. Spergel, J.N. Bahcall}{\PR}{D49}{1389}{1994}.
3721: Even though the title of the paper raises the doubt that large solar mixing
3722: is disfavored by SN1987A,
3723: it is pointed out that the effect of
3724: oscillations in the star can be partially undone by subsequent
3725: oscillations in the Earth. This happens for certain values of
3726: $\Delta m^2_{\rm sun}$ which are now excluded.
3727: Dependence on the initial temperatures has been studied in
3728: \art{M. Kachelriess et al.}{JHEP}{0101}{030}{2001};
3729: \art[hep-ph/0108100]{M. Kachelriess et al.}{\PR}{D65}{73016}{2002}.
3730:
3731:
3732: \bibitem{PulsarMotion}
3733: See e.g.\
3734: \art[astro-ph/0307267]{G.M. Fuller, A. Kusenko, I. Mocioiu, S. Pascoli}{\PR}{D68}{103002}{2003}.
3735: %\hepart[hep-ph/0402259]{M. Barkovich et al.}.
3736:
3737:
3738: \bibitem{kusenko segre}
3739: \art[hep-ph/9701311]{A. Kusenko, G. Segre}{\PL}{B396}{197}{1997}.
3740:
3741:
3742: \bibitem{UHE}
3743: \art[hep-ph/0307025]{J.F. Beacom et al.}{\PR}{D68}{093005}{2003}.
3744:
3745:
3746:
3747: \bibitem{SterileUHE} \art[hep-ph/030704]{P. Keranen, J. Maalampi, M. Myyrylainen, J. Riittinen}{\PL}{B674}{162}{2003}.
3748:
3749:
3750: \bibitem{SKbarnu}
3751: \art[hep-ex/0209028]{Super-Kamiokande collaboration}{\PRL}{90}{061101}{2003}
3752: and references therein.
3753: Future improvements are proposed in
3754: \hepart[hep-ph/0309300]{J.F. Beacom, M.R. Vagins}.
3755:
3756:
3757: \bibitem{SKatm}
3758: \art{Super-Kamiokande collaboration}{\PRL}{81}{1562}{1998}.
3759: Recent detailed results can be found in the thesis by J. Kameda available from the SK
3760: web page: www-sk.icrr.u-tokyo.ac.jp/doc/sk/pub.
3761: Latest data: % AGGIORNARE
3762: talk by M. Shiozawa at the Neutrino 2002 conference, web site
3763: neutrino2002.ph.tum.de.
3764:
3765:
3766:
3767:
3768:
3769:
3770: \bibitem{K2K}
3771: See e.g.\ \hepart[hep-ex/0306043]{K2K collaboration}.
3772:
3773:
3774: \bibitem{MACRO}
3775: \art[hep-ex/0304037]{MACRO collaboration}{\PL}{B566}{35}{2003}.
3776:
3777:
3778: \bibitem{CHOOZ}
3779: \art[hep-ex/9907037]{CHOOZ collaboration}{\PL}{B466}{415}{1999}.
3780: \hepart[hep-ex/0301017]{CHOOZ collaboration}.
3781: See also
3782: \art[hep-ex/9912050]{{\sc Palo Verde} collaboration}{\PRL}{84}{3764}{2000}.
3783:
3784:
3785: \bibitem{Bugey}
\art{{\sc Bugey} collaboration}{\NP}{B434}{503}{1995}.
3786:
3787:
3788: \bibitem{CDHS}
\art{CDHS collaboration}{\PL}{134B}{281}{1984};
3789:
3790:
3791: \bibitem{CCFR}
\art{CCFR collaboration}{\PRL}{52}{1384}{1984}.
3792:
3793:
3794: \bibitem{NuTeV}
3795: \hepart[hep-ex/0110059]{{\sc NuTeV} collaboration}.
3796: Talks available at
3797: www-e815.fnal.gov/NuTeV.html
3798: discuss how well simulated spectra agree with data.
3799:
3800:
3801: \bibitem{Karmen}
\hepart[hep-ex/0203021]{{\sc Karmen} collaboration}.
For a combined analysis of {\sc Karmen} and LSND data
see \hepart[hep-ex/0203023]{E.D. Church et al.}.
3802:
3803:
3804:
3805: \bibitem{NOMADCHORUS}
3806: \art[hep-ex/0106102]{NOMAD collaboration}{\NP}{B611}{3}{2001}.
3807: \art{CHORUS collaboration}{\PL}{B497}{8}{2001}.
3808:
3809: \bibitem{Monolith}
The {\sc Monolith } proposal is available at
www.desy.de/\~{}hoepfner/Neutrino/Monolith.
3810: See also \hepart[hep-ph/0402246]{G. Rajasekaran}.
3811:
3812: \bibitem{SterileReactor}
3813: See e.g.\
3814: \hepart[hep-ph/0310246]{V. Kopeikin, L. Mikaelyan, V. Sinev}.
3815: See also www.hep.anl.gov/ndk/hyper\-text/white.html.
3816:
3817:
3818: \bibitem{ReTh}
3819: \art[hep-ph/0107277]{C.~Bemporad, G.~Gratta, P.~Vogel}{Rev.\ Mod.\ Phys.}{74}{297}{2002}.
3820:
3821:
3822: \bibitem{nuno}
3823: \art[hep-ph/0302039]{H.~Nunokawa, O.L.G.~Peres, R.~Zukanovich Funchal}{\PL}{B562}{279}{2003}.
3824:
3825:
3826:
3827:
3828: \bibitem{MinosNC}
3829: \art{D. Michael}{Nucl. Phys. Proc. Suppl.}{118}{189}{2003}.
3830:
3831:
3832: \bibitem{CNGS}
3833: The CNGS project is described at its web site,
3834: proj-cngs.web.cern.ch/proj-cngs.
3835:
3836:
3837: \bibitem{LoSecco}
3838: \hepart[hep-ph/0305022]{J.M. LoSecco}.
3839:
3840:
3841: \bibitem{Battistoni}
3842: \hepart[hep-ph/0305208]{{\sc Fluka} collaboration}.
3843: For a review see
3844: \art[hep-ph/0203272]{T.K. Gaisser, M. Honda}{Ann.\ Rev.\ Nucl.\ Part.\ Sci.}{52}{153}{2002}.
3845:
3846:
3847:
3848: \bibitem{LSND?}
3849: \art[hep-ph/9606411]{N. Okada, O. Yasuda}{Int. J. Mod. Phys.}{A12}{3669}{1997}.
\art[hep-ph/9607372]{S.M. Bilenky, C. Giunti, W. Grimus}{Eur. Phys. J.}{C1}{247}{1998}.
3850: \art[hep-ph/0010009]{C. Giunti, M. Laveder}{JHEP}{102}{1}{2001}.
3851: For recent desperate attempts see
3852: \hepart[hep-ph/0305255]{M. Sorel et al.}.
3853: \art[hep-ph/0308299]{V.~Barger, D.~Marfatia, K.~Whisnant}{\PL}{B576}{303}{2003}.
3854:
3855:
3856:
3857:
3858: \bibitem{LSNDBBN}
3859: See e.g.\ some of the papers in~\cite{NnuBBN}
3860: and references therein.
3861:
3862:
3863: \bibitem{instant}
3864: \hepart[hep-ph/0302131]{A. Pierce, H. Murayama}.
3865: \hepart[hep-ph/0302173]{C. Giunti}.
3866:
3867:
3868:
3869:
3870:
3871:
3872: \end{thebibliography}
3873: \end{multicols}
3874:
3875: \end{document}
3876: