1: \hfuzz 2pt
2:
3: \font\titlefont=cmbx10 scaled
4: \magstep1\magnification=\magstep1\null\vskip 1.5cm
5:
6: \centerline{\titlefont DISSIPATIVE NEUTRINO OSCILLATIONS}
7: \medskip
8: \centerline{\titlefont IN RANDOMLY FLUCTUATING MATTER}
9: \vskip 2.5cm
10:
11: \centerline{\bf F. Benatti}\smallskip
12:
13: \centerline{Dipartimento di Fisica Teorica, Universit\`a di Trieste}
14: \centerline{Strada Costiera 11, 34014 Trieste, Italy}
15: \centerline{and}
16: \centerline{Istituto Nazionale di Fisica Nucleare, Sezione di Trieste}
17: \vskip 1cm
18:
19: \centerline{\bf R. Floreanini}
20: \smallskip
21:
22: \centerline{Istituto Nazionale di Fisica Nucleare, Sezione di Trieste}
23: \centerline{Dipartimento di Fisica Teorica, Universit\`a di Trieste}
24: \centerline{Strada Costiera 11, 34014 Trieste, Italy}
25: \vskip 2.5cm
26:
27: \centerline{\bf Abstract}\smallskip
28: \midinsert\narrower\narrower
29: \noindent
30: The generalized dynamics describing the propagation of neutrinos
31: in randomly fluctuating media is analyzed: it takes into
32: account matter-induced, decoherence phenomena that go beyond the
33: standard MSW effect. A widely adopted density fluctuation pattern
34: is found to be physically untenable: a more general model needs
35: to be instead considered, leading to flavor changing effective neutrino-matter
36: interactions. They induce new, dissipative effects that modify the
37: neutrino oscillation pattern in a way
38: amenable to a direct experimental analysis.
39: \endinsert
40: \bigskip
41: \vfill\eject
42:
43:
44: \noindent
45: {\bf 1. INTRODUCTION}
46: \medskip
47:
48: When a neutrino propagates in a constant distribution of matter, coherent
49: forward scattering phenomena can affect its time evolution.
50: Despite the smallness of the cross-section induced by the neutrino
51: interaction with the medium, these matter effects can significantly modify
52: the oscillation pattern, through the so-called MSW mechanism.\hbox{[1, 2]}
53:
54: However, forward scattering phenomena are just the simplest matter induced
55: effects that can occur to a neutrino when the medium is allowed to fluctuate.
56: In this case, the neutrino can be viewed as an open system, {\it i.e.}
57: a subsystem immersed in an external environment (the medium);[3-7] its time
58: evolution, obtained from the total neutrino+matter dynamics by
59: eliminating ({\it i.e.} integrating over) the matter degrees of freedom,
60: is no longer unitary: it takes into account possible exchanges of
61: entropy and energy between the neutrino and the fluctuating medium.
62:
63: In many physical situations, one can safely ignore the details of
64: the matter dynamics and use an effective description of the medium as
65: a classical, random external field. Quite in general, any environment
66: can be modeled in this way, provided the characteristic decay time
67: of the associated correlations is sufficiently small with respect to the
68: typical evolution time of the subsystem. In the case of
69: relativistic neutrinos, this time scale can be roughly identified with the
70: vacuum oscillation length: we shall therefore consider media that fluctuate on
71: time scales shorter than this. It has been recently pointed out that
72: the interior of the sun could indeed satisfy such a condition,[8] as
73: likely as the earth mantel. Thus, a neutrino created in the sun or moving through the
74: earth would effectively see a random fluctuating distribution of scattering
75: centers and therefore be subjected to stochastic, incoherent interaction
76: with the medium.
77: In this situation, correlations in the medium play a fundamental role:
78: they are responsible for the generation of new matter effects, beyond the
79: MSW ones, leading to irreversibility and loss of quantum coherence.
80:
81: The effects of fluctuating matter on neutrino propagation have been
82: first discussed in [9-11], and recently reconsidered in [12-14]. However,
83: all these analysis deal with a simple density fluctuation pattern,
84: naturally suggested by the standard MSW treatment. Further, these fluctuations
85: are assumed to be exactly $\delta$-correlated; this is a highly
86: idealized description of the environment, that {\it e.g.}
87: for heat baths can be attained only in the limit
88: of infinite temperature.
89:
90: Instead, in the following a more realistic exponentially damped form for
91: the correlation functions in the medium will be adopted.
92: Limiting for simplicity the discussion to the analysis of the oscillations
93: of two species of neutrinos, we shall see that the effects induced
94: by matter fluctuations can be fully described in terms of a limited number of
95: phenomenological parameters. They affect the oscillation pattern
96: in a very distinctive way, that is amenable to a direct experimental
97: study.
98:
99: On the other hand, when the simplified density fluctuation hypothesis
100: considered in [9-14] is adopted, a single constant is sufficient to
101: parametrize the new matter effects. However, this approximation appears
102: physically untenable, since by adopting it certain
103: transition probabilities take unacceptable
104: negative values; this serious inconsistency can be cured only by allowing
105: more general matter fluctuations, pointing towards the presence
106: of flavor changing neutrino-matter interactions.%
107: \footnote{$^\dagger$}{Although within
108: a mean-field (MSW) approach, this possibility
109: has recently been reconsidered in [15-17]
110: and found compatible with present experimental data.}
111:
112: As a final remark, it is interesting to point out that the dissipative effects
113: induced by a randomly fluctuating medium on neutrino oscillations
114: involve in general the $CP$-violating phase that is present in the
115: mixing matrix for Majorana neutrinos. Therefore, contrary to the
116: vacuum case, matter oscillation experiments can provide, at least in
117: principle, a way to distinguish between Dirac and Majorana neutrinos.
118:
119: \vskip 1cm
120:
121: \noindent
122: {\bf 2. MASTER EQUATION}
123: \medskip
124:
125: In discussing the mixing of two neutrino species, we shall adopt
126: the familiar effective description in terms of a two-dimensional
127: Hilbert space;[18-22] the flavor states, that we shall conventionally call
128: $|\nu_e\rangle$ and $|\nu_\mu\rangle$, will be chosen as basis states.
129: With respect to this basis, the physical neutrino states
130: are then represented by density matrices $R$, {\it i.e.} by
131: hermitian $2\times2$ matrices, with non-negative eigenvalues and unit trace.
132: Their time evolution equation can be cast in a standard
133: Liouville -- von Neumann form:[23, 24]
134: $$
135: {\partial R(t)\over \partial t}= -i\big[ H_0, R(t)\big]+ L_t\big[R(t)\big]\ .
136: \eqno(2.1)
137: $$
138: The first piece on the r.h.s. describes the propagation of the neutrinos
139: in vacuum; in the chosen basis, the effective hamiltonian $H_0$
140: takes the standard form:
141: $$
142: H_0=\omega\, \vec n\cdot\vec\sigma\ ,
143: \eqno(2.2)
144: $$
145: where $\omega=\Delta m^2/4E$, $\Delta m^2$ being the square mass difference
146: of the two mass eigenstates and $E$ the average neutrino energy,
147: while the unit vector $\vec n=(\sin2\theta, 0, -\cos2\theta)$ contains
148: the dependence on the mixing angle, $\vec\sigma=(\sigma_1,\sigma_2,\sigma_3)$
149: being the vector of Pauli matrices.
150: The additional contribution $L_t[R]$ takes into account the presence
151: of matter. As explained above, we shall consider the case of
152: a rapidly fluctuating medium, which can be
153: described by classical stochastic fields. Its action on the travelling
154: neutrinos can then be expressed via the commutator with a time-dependent
155: hermitian matrix $V(t)$,
156: $$
157: L_t\big[R(t)\big]=-i\big[ V(t), R(t) \big]\ ,\qquad V(t)=\vec V(t)\cdot\vec\sigma\ ,
158: \eqno(2.3)
159: $$
160: whose components $V_1(t)$, $V_2(t)$, $V_3(t)$ form a real, stationary Gaussian
161: stochastic field $\vec V(t)$; they are assumed to have in general a nonzero
162: constant mean and translationally invariant correlations:
163: $$
164: \widehat{W}_{ij}(t-s)\equiv\langle V_i(t)\, V_j(s)\rangle
165: -\langle V_i(t)\rangle\, \langle V_j(s)\rangle\ ,\quad i,j=1,2,3\ .
166: \eqno(2.4)
167: $$
168:
169: Since the generalized hamiltonian $V(t)$ in (2.3) involves stochastic
170: variables,
171: the density matrix $R(t)$, solution of the equation of
172: motion (2.1), is also stochastic. Instead, we are interested in the behaviour
173: of the reduced density matrix $\rho(t)\equiv\langle R(t)\rangle$
174: which is obtained by averaging over the noise; it is $\rho(t)$ that
175: describes the effective evolution of the neutrinos in the medium and allows
176: the computation of relevant transition probabilities.
177: By making the additional assumption that neutrinos
178: and noise be decoupled at $t=\,0$, so that the initial state is
179: $\rho(0)\equiv\langle R(0)\rangle=R(0)$, a condition very well satisfied
180: in typical situations, an effective master equation for $\rho(t)$
181: can be derived by going to the interaction representation,
182: where we set:
183: $$
184: \widetilde{R}(t)=e^{it\, H_0}\ R(t)\ e^{-it\, H_0}\ ,\quad
185: \vec\sigma(t)=e^{it\, H_0}\ \vec\sigma\ e^{-it\, H_0}\ ,\quad
186: \widetilde{L}_t[\ \ ]\equiv-i\Big[\vec V(t)\cdot\vec\sigma(t),\ \ \Big]\ .
187: \eqno(2.5)
188: $$
189: By averaging $\widetilde{R}(t)$ over the noise, we get the reduced density
190: matrix $\tilde{\rho}(t)\equiv\langle\widetilde{R}(t)\rangle$
191: in the interaction representation: it is convenient to operate on the
192: standard series expansion of $\widetilde{R}(t)$, so that:
193: $$
194: \tilde\rho(t)= {\cal N}_t[\tilde\rho(0)]\equiv
195: \sum_{k=0}^\infty N_t^{(k)}[\tilde\rho(0)]
196: \eqno(2.6a)
197: $$
198: where the terms $N_t^{(k)}$ are explicitly given by
199: $$
200: N_t^{(k)}[\tilde\rho(0)]=
201: \int_0^t ds_1\int_0^{s_1}ds_2\cdots \int_0^{s_{k-1}}ds_k\
202: \langle \widetilde{L}_{s_1} \widetilde{L}_{s_2}\cdots
203: \widetilde{L}_{s_k}\rangle[\tilde\rho(0)\,]\ ,
204: \eqno(2.6b)
205: $$
206: with $N_t^{(0)}=1$ the identity, $N_t^{(0)}[\tilde\rho(0)]=\tilde\rho(0)$.
207: The resulting series is a sum over multiple integrals of
208: correlators $\langle \widetilde{L}_{s_1} \widetilde{L}_{s_2}\cdots
209: \widetilde{L}_{s_k}\rangle$, that is of averages
210: over the noise of successive commutators with respect to the
211: stochastic operators $V(t)=\vec V(t)\cdot\vec\sigma$ at different times.
212: The density matrix $\widetilde{\rho}(0)$ is not averaged over due to the
213: assumption on the initial state $\rho(0)=\widetilde{\rho}(0)$.
214:
215: In order to arrive at a more manageable time-evolution, we use a
216: techique [23], of which we give a brief account below, that leads to a
217: so-called convolution-less master equation and
218: is based on the hypothesis of weak coupling between system and stochastic
219: environment.
220: The first step is to write the formal inverse of the map ${\cal N}_t$ in $(2.6a)$,
221: {\it i.e.} $\tilde{\rho}(0)={\cal N}^{-1}_t[\tilde{\rho}(t)]$,
222: so that:
223: $$
224: {\cal N}^{-1}_t=\Bigl(1+\,N_t^{(1)}\,+\, N_t^{(2)}+\cdots\Bigr)^{-1}=
225: 1\,-\,N_t^{(1)}\,-\,N_t^{(2)}-(N_t^{(1)})^2-\cdots\ ,
226: \eqno(2.7)
227: $$
228: where only terms containing up to two-point correlation functions have
229: been indicated. Further, denoting
230: with $\dot{N}^{(k)}_t$ the time-derivative of $N^{(k)}_t$, it follows that
231: the reduced density matrix in the interaction representation satisfies the
232: equation of motion
233: $$
234: {\partial \tilde{\rho}(t)\over\partial t}
235: ={\partial{\cal N}_t\over\partial t}[\tilde{\rho}(0)]
236: ={\partial{\cal N}_t\over\partial t}
237: \, {\cal N}_t^{-1}[\tilde\rho(t)]=
238: \Big\{\dot{N}_t^{(1)}\,+\,\big(\dot{N}_t^{(2)}
239: -\dot{N}_t^{(1)}\, N_t^{(1)}\big)+\ldots\Big\}[\tilde\rho(t)]\ .
240: \eqno(2.8)
241: $$
242:
243: Since the interaction of the travelling neutrinos with the medium is weak,
244: one can focus the attention on the dominant terms of the previous expansion,
245: neglecting all contributions higher than the second-order ones.
246: Further, since the characteristic decay time of correlations in the medium
247: is by assumption much smaller than the typical time scale of the system,
248: the memory effects implicit in (2.8) should not be physically relevant
249: and the use of the Markovian approximation justified. This is implemented
250: in practice by extending to infinity the upper limit of the integrals
251: appearing in $\dot{N}^{(2)}$ and $N^{(1)}$.[3-5]
252:
253: By returning to the Schr\"odinger representation, one finally obtains [24]
254: $$
255: {\partial\rho(t)\over\partial t}=-i\big[ H,\, \rho(t)\big]
256: +L\big[\rho(t)\big]\ ,
257: \eqno(2.9a)
258: $$
259: where
260: $$
261: \eqalignno{
262: &H=H_0 + H_1 + H_2\equiv\vec\Omega\cdot\vec\sigma\ , &(2.9b)\cr
263: &L[\rho]={1\over2}\sum_{i,j=1}^3 {\cal C}_{ij}
264: \Big[2\sigma_i \rho\, \sigma_j
265: - \{\sigma_j\sigma_i\, ,\, \rho\}\Big]\ . &(2.9c)
266: }
267: $$
268: The effective hamiltonian in matter, $H$, differs from the one in vacuum,
269: $H_0$, by first order terms (coming from the piece $\dot{N}^{(1)}$
270: in (2.8)) depending on the noise mean values:
271: $$
272: H_1= \langle \vec V(t)\rangle\cdot \vec\sigma\ ,
273: \eqno(2.10)
274: $$
275: and by second-order contributions (coming from the second-order terms
276: in (2.8)),
277: $$
278: H_2=\sum_{i,j,k=1}^3\,\epsilon_{ijk}\ C_{ij}\, \sigma_k\ ,
279: \eqno(2.11)
280: $$
281: involving the noise correlations (2.4) through the
282: time-independent combinations:
283: $$
284: C_{ij}=\sum_{k=1}^3\int_0^\infty dt\ \widehat{W}_{ik}(t)\ U_{kj}(-t)\ ,
285: \eqno(2.12)
286: $$
287: where the $3\times3$ orthogonal matrix $U(t)$ is defined by the following
288: transformation rule:
289: $e^{it\, H_0}\ \sigma_i\ e^{-it\, H_0}=\sum_{j=1}^3 U_{ij}(t)\ \sigma_j$.
290: On the other hand, the contribution $L[\rho\,]$ in $(2.9c)$
291: is a time-independent, trace-preserving linear map
292: involving the symmetric coefficient matrix ${\cal C}_{ij}\equiv C_{ij}+C_{ji}$.
293: It introduces irreversibility, inducing
294: in general dissipation and loss of quantum coherence.
295: Altogether, equation (2.9) generates a semigroup of linear maps,
296: ${\mit\Gamma}_t:\ \rho(0)\mapsto\rho(t)\equiv{\mit\Gamma}_t[\rho(0)]$,
297: for which composition is defined only forward in time:
298: ${\mit\Gamma}_t\circ{\mit\Gamma}_s={\mit\Gamma}_{t+s}$, with $t,s\geq0$;
299: this is a very general physical requirement that should be
300: satisfied by all Markovian open system dynamics.%
301: \footnote{$^\dagger$}{Notice that the procedure of averaging transition probabilities over
302: random matter profiles as performed in [25] is not compatible
303: with this basic evolution law, and therefore it is a pure phenomenological
304: device, hardly amenable to a rigorous theoretical treatment.}
305: The set of maps ${\mit\Gamma}_t$
306: is usually referred to as a quantum dynamical semigroup.\hbox{[3-7]}
307:
308: The typical observable that is accessible to the experiments is the
309: probability ${\cal P}_{\nu_e\to\nu_\mu}(t)$ for having a transition to
310: a neutrino of type $\nu_\mu$ at time $t$, assuming that the neutrino
311: has been generated as $\nu_e$ at $t=\,0$. In the language of density
312: matrices, it is given by:
313: $$
314: {\cal P}_{\nu_e\to\nu_\mu}(t)\equiv
315: {\rm Tr}\big[\rho_{\nu_e}(t)\ \rho_{\nu_\mu}\big]\ ,
316: \eqno(2.13)
317: $$
318: where $\rho_{\nu_e}(t)$ is the solution of (2.9) with the
319: initial condition given by the matrix
320: $\rho_{\nu_e}(0)=\rho_{\nu_e}\equiv |\nu_e\rangle\langle\nu_e |$,
321: while $\rho_{\nu_\mu}=1-\rho_{\nu_e}$. By expanding the neutrino
322: density matrix in terms of the Pauli matrices and the identity $\sigma_0$,
323: $\rho=\big[\sigma_0+\vec\rho\cdot\vec\sigma\big]/2$,
324: the linear equation $(2.9a)$ reduces to a diffusion equation
325: for the components $\rho_1$, $\rho_2$, $\rho_3$ of the vector
326: $\vec\rho$:
327: $$
328: {\partial \vec\rho(t)\over\partial t}=-2{\cal H}\, \vec\rho(t)\ ;
329: \eqno(2.14)
330: $$
331: the entries of $3\times3$ matrix $\cal H$ can be expressed
332: in terms of the coefficients $\Omega_i$ and ${\cal C}_{ij}$
333: appearing in the hamiltonian and noise contribution in $(2.9b)$, $(2.9c)$:[26]
334: $$
335: {\cal H}=\left[
336: \matrix{a&b+\Omega_3&c-\Omega_2\cr
337: b-\Omega_3&\alpha&\beta+\Omega_1\cr
338: c+\Omega_2&\beta-\Omega_1&\gamma\cr}
339: \right]\ ,
340: \eqno(2.15)
341: $$
342: with $a={\cal C}_{22}+{\cal C}_{33}$, $\alpha={\cal C}_{11}+{\cal C}_{33}$,
343: $\gamma={\cal C}_{11}+{\cal C}_{22}$, $b=-{\cal C}_{12}$,
344: $c=-{\cal C}_{13}$, $\beta=-{\cal C}_{23}$.
345: The solution of (2.14) involves the exponentiation of the matrix $\cal H$,
346: $$
347: \vec\rho(t)={\cal M}(t)\ \vec\rho(0)\ ,\qquad {\cal M}(t)=e^{-2{\cal H} t}\ ,
348: \eqno(2.16)
349: $$
350: so that the transition probability in (2.13) can be rewritten as
351: $$
352: {\cal P}_{\nu_e\to\nu_\mu}(t)=
353: {1\over2}\Big[1+\sum_{i,j=1}^3\rho^i_{\nu_e}\rho^j_{\nu_\mu}\ {\cal M}_{ij}(t)\Big]
354: ={1\over2}\Big[1-{\cal M}_{33}(t)\Big]\ .
355: \eqno(2.17)
356: $$
357: Indeed, taking the standard form of the Pauli matrices with respect to the
358: orthonormal basis $\vert\nu_e\rangle=\pmatrix{1\cr 0}$ and
359: $\vert\nu_\mu\rangle=\pmatrix{0\cr 1}$, then
360: $\displaystyle \rho_{\nu_e}={1+\sigma_3\over2}$ and
361: $\displaystyle \rho_{\nu_\mu}={1-\sigma_3\over2}$.
362:
363: When correlations in the medium are negligible,
364: {\it i.e.} the combination in (2.4) are vanishingly small,
365: equation (2.9) describes standard (MSW) matter effects, for
366: the presence of matter is signaled
367: solely by the shift $H_1$ in the effective hamiltonian.
368: In this case, the neutrino-medium interaction is dominated
369: by coherent forward scattering, and, in absence of flavor changing effects,
370: the stochastic vector field in (2.3) results oriented along the third axis,
371: whence $H_1=A\sigma_3$, where $A\equiv\langle V_3(t)\rangle=G_F n_e/\sqrt{2}$
372: gives the extra energy contribution that electron neutrinos receive
373: when travelling in ordinary matter
374: ($G_F$ is the Fermi constant, while $n_e$ represents the electron
375: number density in the medium).
376: As a consequence,
377: the transition probability in (2.17) can be expressed in terms
378: of a modified frequency $\omega_M$ and mixing angle $\theta_M$
379: in matter,
380: $$
381: \omega_M=\omega \Big[\sin^22\theta +(1-A/A_R)^2 \cos^2 2\theta\Big]^{1/2}
382: \ ,\qquad
383: \sin 2\theta_M={\omega\over\omega_M}\sin 2\theta\ ,
384: \eqno(2.18)
385: $$
386: $A_R=\omega \cos2\theta$ being the value of $A$ at resonance.
387: In fact, the assumption of negligible correlations amounts to considering
388: in equation (2.14) a matrix $\cal H$ of the form
389: $$
390: {\cal H}=\pmatrix{0&\Omega_3&0\cr-\Omega_3&0&\Omega_1\cr0&-\Omega_1&0}\ ,\quad
391: \Omega_1=\omega\sin(2\theta)\ ,\ \Omega_3=A-\omega\cos(2\theta)\ .
392: \eqno(2.19)
393: $$
394: This matrix can be easily exponentiated as in (2.16),
395: $$
396: {\cal M}(t)=\pmatrix{{\Omega^2_1+\Omega_3^2\cos(2\omega_M t)\over\omega_M^2}&
397: -{\Omega_3\over\Omega_1}\sin(2\omega_M t)&
398: {\Omega_1\Omega_3\over\omega_M^2}(1-\cos(2\omega_M t))\cr
399: -{\Omega_3\over\Omega_1}\sin(2\omega_M t)&
400: \cos(2\omega_M t)&
401: -{\Omega_1\over\Omega_1}\sin(2\omega_M t)\cr
402: {\Omega_1\Omega_3\over\omega_M^2}(1-\cos(2\omega_M t))&
403: -{\Omega_1\over\Omega_1}\sin(2\omega_M t)&
404: {\Omega^2_3+\Omega_1^2\cos(2\omega_M t)\over\omega_M^2}
405: }\ ,
406: \eqno(2.20)
407: $$
408: whence the explicit form of the element ${\cal M}_{33}(t)$ yields
409: the familiar expression:
410: $$
411: {\cal P}_{\nu_e\to\nu_\mu}(t)=\sin^2 2\theta_M\ \sin^2 \omega_M t\ .
412: \eqno(2.21)
413: $$
414:
415: The situation can significantly change for neutrinos immersed
416: in a fluctuating medium; while travelling in it, they encounter
417: matter fluctuations, whose correlations $\widehat W_{ij}(t-s)$ determine
418: the dissipative contribution in $(2.9c)$.
419: In a typical bath at finite temperature, the correlation functions
420: assume an exponentially damped form; therefore, one can generically
421: write:
422: $$
423: \widehat{W}_{ij}(t-s)=W_{ij}\ e^{-\lambda_{ij} |t-s|}\ ,
424: \eqno(2.22)
425: $$
426: with $W_{ij}$ and $\lambda_{ij}$ time-independent, real coefficients,
427: with $\lambda_{ij}\geq0$. Further, as discussed before, the
428: stochastic medium fluctuates on time intervals much shorter than
429: the typical neutrino ``free'' evolution time scale $1/\omega$,
430: so that the decay parameters $\lambda_{ij}$ must be much larger
431: than the vacuum frequency $\omega$. This fact allows neglecting
432: all contributions higher than the first-order one in
433: the ratio $\omega/\lambda_{ij}$
434: while evaluating the coefficients $C_{ij}$ in (2.12).
435: For generic correlations as in (2.22), these coefficients,
436: and therefore the entries of the matrix $\cal H$ in (2.15),
437: are all nonvanishing. However, the parameters
438: $a$, $b$, $c$, $\alpha$, $\beta$, $\gamma$ describing
439: matter decoherence effects are not all free:
440: as we shall see, physical consistency requires them
441: to satisfy certain inequalities; in turn, these constraints
442: reflect some fundamental characteristics of the
443: matter-neutrino interactions.
444:
445: We shall now discuss some interesting cases
446: of the master equation (2.9), corresponding to specific physical
447: realizations of the medium through which the neutrinos propagate.
448:
449: \vskip 1cm
450:
451: \noindent
452: {\bf 3. GENERALIZED MSW DYNAMICS}
453: \medskip
454:
455: The simplest instance of a stochastic
456: medium corresponds to ordinary matter with density fluctuations,
457: where only the propagation of electron neutrinos is affected.
458: It generalizes the familiar MSW mean field treatment by
459: adding to it decoherence effects.
460: In this case, the stochastic hamiltonian in (2.3) becomes
461: diagonal and, without loss of generality,
462: only the stochastic field $V_3(t)$ can be taken to be non-vanishing;
463: the neutrinos are still forward scattered by the medium,
464: although no longer in a coherent way.
465: This is situation discussed in [9-14], where however the density
466: fluctuations in the medium are taken to be exactly $\delta$-function
467: correlated. This is a highly idealized assumption, that can hardly
468: be reproduced in ordinary conditions. Instead, the much more realistic
469: exponential ansatz (2.22) will be used here, where the only nonvanishing
470: correlation strength and decay constant are $W_{33}\equiv W$ and
471: $\lambda_{33}\equiv\lambda$, respectively.
472:
473: The noise contributions in (2.9) can be explicitly computed;
474: within our approximation,
475: one finds that only the entries ${\cal C}_{23}$ and ${\cal C}_{33}$
476: of the coefficient matrix in $(2.9c)$ are nonvanishing,
477: $$
478: a={\cal C}_{33}={2W\over\lambda}\ ,\qquad
479: \beta=-{\cal C}_{23}={\omega W\over\lambda^2}\sin2\theta\ ,
480: \eqno(3.1)
481: $$
482: while the hamiltonian contribution $H_1$ is proportional to $\sigma_3$
483: (the standard MSW piece) and $H_2$ to $\sigma_1$:
484: $$
485: \Omega_1=\omega\Bigg(1+{W\over\lambda^2}\Bigg)\sin2\theta\ ,\quad
486: \Omega_2=\, 0\ ,\quad
487: \Omega_3=-\omega\Bigg(1-{A\over A_R}\Bigg)\cos2\theta\ .
488: \eqno(3.2)
489: $$
490: Surprisingly, the dynamics generated by (2.9), or equivalently (2.14),
491: with these coefficients appears to be physically unacceptable.
492:
493: As mentioned at the beginning, any density matrix must be a positive operator
494: ({\it i.e.} its eigenvalues should be non-negative) in order to
495: represent a physical state: its eigenvalues have the physical meaning
496: of probabilities. Therefore, any time evolution needs to preserve this property,
497: otherwise an initial state would not be mapped to another state at
498: a later time. This is precisely what happens when the neutrino evolution
499: in the medium is modeled by (2.9) with dissipative parameters as in (3.1).
500: In fact, the probability ${\cal P}(t)$ for having a transition from
501: an initial neutrino state $\rho(0)$ to its orthogonal state
502: $\rho_\perp\equiv 1-\rho(0)$ at a later time $t$ is given by
503: the first equality in (2.17), with the substitutions
504: $\rho_{\nu_e}\to\rho(0)$, $\rho_{\nu_\mu}\to\rho_\perp$.
505: Since ${\cal P}$ is initially zero,
506: its time derivative must be positive at $t=\,0$, otherwise
507: we would have physically unacceptable negative transition
508: probabilities as soon as $t>0$. A simple computation gives:
509: $\dot{\cal P}(0)=\sum_{i,j=1}^3\rho(0)_i\, {\cal H}_{ij}\, \rho(0)_j\geq0$,
510: and since this must be true for any initial state, physical
511: consistency requires the symmetric part of the matrix $\cal H$ in (2.15)
512: to be positive. One easily sees that this is impossible with
513: the assignment in (3.1).%
514: %
515: \footnote{$^\dagger$}{An example of the emergence of negative transition
516: probabilities is explicitly provided in the Appendix.}
517: %
518:
519: In the case of a $\delta$-correlated medium,
520: the parameter $\beta$ identically vanishes and no inconsistencies arise;
521: however, as mentioned before, this choice is not supported by strong
522: physical motivations and appears just a
523: mathematically convenient simplification.
524: By naively relaxing the $\delta$-correlated assumption, one ends up
525: with the simple stochastic system discussed above,
526: which turns out to be seriously flawed. As a consequence,
527: modelling matter fluctuations only in terms
528: of electron density is physically untenable and indicates
529: that in order to consistently describe neutrino oscillations
530: in random matter more complex situations need to be analyzed,
531: involving a richer covariance structure than with
532: $\vec V(t)=(0,0,V_3(t))$.
533:
534: Alternatively, instead of the random matter model one may question
535: the approximations used in deriving the master equation (2.9), and precisely
536: the weak coupling hypothesis and the markovian limit.
537: However, the first assumption appears rather well satisfied in the case
538: of the neutrinos, as they interact very weakly with matter,
539: while the markovian approximation is justified by the physically motivated
540: choice
541: of rapidly decaying matter correlations: $\lambda\gg\omega$.
542: In reality, once a slightly generalized model of
543: random medium is adopted, the master equation (2.9)
544: results perfectly adequate to consistently treat decoherence phenomena
545: in neutrino matter oscillations.
546:
547: \vskip 1cm
548:
549: \noindent
550: {\bf 4. DIAGONAL CORRELATIONS}
551: \medskip
552:
553: When the components of the stochastic field $\vec V(t)$
554: are all nonvanishing, the noise hamiltonian in (2.3) is no longer diagonal:
555: in this case, while travelling in the medium,
556: all neutrino species undergo incoherent scatterings, in general involving
557: not exclusively the forward direction; this may happen only in presence
558: of flavor changing interactions.
559: However, as a minimal extension of the previously
560: treated case, we shall assume $V_1(t)$ and $V_2(t)$ to have zero mean,
561: so that the hamiltonian correction $H_1$ contains only the standard MSW
562: contribution, and further take the correlation functions in (2.22)
563: to be diagonal:
564: $$
565: \widehat{W}_{ij}(t-s)=W_i\ e^{-\lambda_i |t-s|}\, \delta_{ij}\ .
566: \eqno(4.1)
567: $$
568: In addition, for simplicity we shall consider situations for which
569: the ratios $W_i/\lambda_i$ are all equal to a common factor ${\cal W}>0$;
570: in this case, the parameters appearing in (2.15) take the form:
571: $$
572: a=\alpha=\gamma=4{\cal W}\ ,\quad
573: b=2\omega {\cal W}\Bigg({1\over\lambda_1}-{1\over\lambda_2}\Bigg)\cos2\theta\ ,\quad
574: c=\,0\ ,\quad
575: \beta=2\omega {\cal W}\Bigg({1\over\lambda_3}-{1\over\lambda_2}\Bigg)\sin2\theta\ ,
576: \eqno(4.2a)
577: $$
578: %
579: $$
580: \Omega_1=\omega\Bigg[1+2{\cal W}\Bigg({1\over\lambda_2}+{1\over\lambda_3}\Bigg)
581: \Bigg]\sin2\theta\ ,\ \
582: \Omega_2=\,0\ ,\ \
583: \Omega_3=-\omega\Bigg[1-{A\over A_R}+2{\cal W}\Bigg({1\over\lambda_1}+{1\over\lambda_2}\Bigg)
584: \Bigg]\cos2\theta\ ,
585: \eqno(4.2b)
586: $$
587: and the master equation (2.9) can be exactly integrated. Notice that
588: the request of positivity of $\rho(t)$ for any $t\geq0$ now requires
589: $\alpha^2\geq b^2 +\beta^2$, condition that is always satisfied
590: by the original hypothesis of fast decaying matter
591: correlations: $\lambda_i\gg\omega$. Even more, this inequality
592: guarantees not only the positivity of the evolution generated by (2.9),
593: but actually a stronger attribute, that of ``complete positivity''.[3-5]
594: This property is crucial in assuring the consistency of any
595: generalized, dissipative dynamics in all possible physical conditions
596: and should always be imposed in place of simple positivity
597: to avoid possible inconsistencies in the treatment;[26]
598: it is reassuring that it emerges naturally from our simple
599: model of random matter, without the need of further assumptions.
600:
601: The transition probability ${\cal P}_{\nu_e\to\nu_\mu}$ in (2.17)
602: can be explicitly computed and cast in the simple form:
603: $$
604: {\cal P}_{\nu_e\to\nu_\mu}(t)={1\over2}\Big(1-e^{-2\alpha t}\Big)
605: +e^{-2\alpha t}\,\sin^2 2\tilde\theta\, \sin^2\Omega t\ ,
606: \eqno(4.3)
607: $$
608: where $\Omega=[\Omega_1^2+ \Omega_3^2 -b^2-\beta^2]^{1/2}$ is the modified
609: oscillation frequency, while
610: $\sin^2 2\tilde\theta=(\Omega^2_1-\beta^2)/\Omega^2\leq 1$ defines a new
611: mixing angle (notice that the absence of the parameters $c$ and $\Omega_2$
612: is due to the assumptions that led to $(4.2a,b)$).
613: In comparison with the standard result in (2.21), one sees that
614: the presence of a random medium introduces exponential
615: damping terms and further modifies the neutrino effective masses and
616: mixing properties; a resonance enhancement is still present
617: for $A=A_R$, but its effectiveness is reduced by the damping factors.
618: This is even more dramatic at large times, where the decoherence
619: effects dominate: the neutrino state $\rho$ is driven to
620: the totally mixed state $\sigma_0/2$ and the transition probability approaches
621: its asymptotic $1/2$ value.
622:
623: These conclusions apply to neutrinos travelling in uniform random media.
624: When this is not the case,
625: the neutrino total time evolution results from the composition
626: of arbitrarily many partial evolutions corresponding to media with uniform
627: properties,
628: but in general of different thicknesses; then, the complete evolution matrix
629: ${\cal M}(t)$
630: as defined in (2.16) will be the result of the composition
631: of the corresponding ones pertaining to the various media (a simple example
632: is given in the Appendix).
633: Nevertheless, for slowly varying conditions, this composition
634: can be well approximated by its adiabatic expression,
635: obtained by the instantaneous diagonalization of the
636: now time-dependent matrix
637: $\cal H$ in (2.15) and the assumption that the neutrino states
638: evolve as one of its eigenstates.%
639: %
640: \footnote{$^\dagger$}{Possible hoppings among the instantaneous eigenstates
641: can also be easily included; for simplicity, we ignore them here.}
642: %
643: Within this approximation and neglecting fast oscillating terms,
644: the averaged transition probability
645: can be cast in the following form:
646: $$
647: {\cal P}_{\nu_e\to\nu_\mu}(t)={1\over2}
648: \Bigg[1-{e^{-8{\cal W} t}\over{\cal R}}\,
649: \Bigg(1-{A\over A_R}+{4{\cal W}\over\lambda_2}\Bigg)\
650: \cos 2\theta \Bigg]\ ,
651: \eqno(4.4)
652: $$
653: with
654: $$
655: {\cal R}=\Bigg\{ \Bigg[1+{4{\cal W}\over\lambda_2}
656: \Bigg(1+{\lambda_2\over\lambda_3}+{4{\cal W}\over\lambda_3}\Bigg)\Bigg]\tan^2 2\theta
657: +\Bigg(1-{A\over A_R}\Bigg)\Bigg[ 1-{A\over A_R} +
658: {4{\cal W}\over\lambda_2}
659: \Bigg(1+{\lambda_2\over\lambda_1}+{4{\cal W}\over\lambda_1}\Bigg)\Bigg]
660: \Bigg\}^{1/2}\ .
661: \eqno(4.5)
662: $$
663:
664: With respect to standard, familiar expressions, the action of the stochastic
665: medium is signaled by the presence in the second term of a modified weight and
666: a damping factor; these additional contributions depend on the ratios of the three
667: matter-correlations decay constants $\lambda_i$ and the
668: corresponding strength $\cal W$. Although in the weak-coupling regime
669: one expects $\cal W\ll\omega$,
670: the decay constant ratios need not be small. Therefore, the behaviour of (4.4)
671: as a function of the neutrino energy can sensibly differ from the
672: one obtained in absence of decoherence effects (concrete examples
673: are shown in Fig.1).
674:
675: Of particular interest
676: is the application of (4.4) to the solar neutrino case, where
677: $\Delta m^2$ and $\theta$ can be taken to assume the
678: best fit values obtained in recent data analysis
679: ({\it e.g.} see [27, 28] and references therein);
680: thanks to the availability of a larger decoherence parameter space,
681: the electron surviving probability
682: ${\cal P}_{\nu_e\to\nu_e}\equiv 1-{\cal P}_{\nu_e\to\nu_\mu}$
683: is found to differ not only from the
684: standard, noiseless expression, but also from those obtained with
685: $\delta$-correlated fluctuating matter
686: as reported in [13, 14] (for a comparison, see Fig.1).
687: These results, together with the still present uncertainties in the
688: fluctuating behaviour
689: of the solar matter, appear to open concrete possibilities for
690: an experimental study of matter induced effects in neutrino oscillations
691: that go beyond the standard MSW phenomenology.
692:
693: \vskip 1cm
694:
695: \noindent
696: {\bf 5. DISCUSSION}
697: \medskip
698:
699: In the most general situation, the correlations in
700: the stochastic medium have the form (2.22) and thus all the entries of
701: the matrix $C_{ij}$ in (2.12) result nonvanishing; as a consequence,
702: all second order pieces in the effective hamiltonian $(2.9b)$
703: as well as in the dissipative part $(2.9c)$ will contribute
704: to the master equation $(2.9a)$. Further, the first order
705: mean field approximation in (2.10) will no longer be
706: diagonal, taking into account the presence of possible flavor changing
707: interactions.[15-17]
708:
709: Nevertheless, even in this very general case,
710: the corresponding matrix $\cal H$ in (2.15) can not result
711: totally generic: as already pointed out,
712: the positivity of the evolved state $\rho(t)$ must be preserved
713: under all circumstances; this is guaranteed by the
714: mentioned condition of complete positivity of the evolution
715: generated by (2.9). This property requires
716: the positivity of the matrix ${\cal C}_{ij}$
717: in $(2.9c)$ and as a consequence imposes certain inequalities
718: among the dissipative parameters in (2.15)
719: (see [29, 30] for explicit expressions). These conditions
720: are certainly of help in restricting the parameter space
721: needed to describe a totally generic random medium.
722:
723: Even with these constraints, no simple, exact analytic expressions for
724: the transition probability
725: ${\cal P}_{\nu_e\to\nu_\mu}(t)$ in (2.17) can in general be given.
726: However, as discussed before,
727: second-order matter contributions to $\cal H$
728: are small with respect to the vacuum frequency $\omega$;
729: therefore, in solving (2.14) one can integrate the hamiltonian
730: dynamics exactly,
731: while treating the dependence on
732: $a$, $b$, $c$, $\alpha$, $\beta$, $\gamma$
733: in perturbation theory.%
734: %
735: \footnote{$^\dagger$}{The effects of the hamiltonian corrections
736: to the free motion are in general not small, in particular near resonance;
737: this is why no approximation is allowed in the evolution
738: generated by the effective hamiltonian $(2.9b)$.}
739: %
740: In this way manageable, approximate
741: expressions for the transition probabilities can be obtained.
742: Having now at disposal a larger parameter space, their
743: form involves multiple damping factors and oscillation phases,
744: showing possible larger deviations from the standard behaviour.
745: In addition, notice that in order to describe neutrino mixing
746: in a generic random medium
747: two mixing angles, $\hat\theta$ and $\hat\varphi$, are in general needed:
748: they parametrize the components of the unit vector
749: $\Omega_i/\Omega\equiv\big(\cos\hat\varphi\sin2\hat\theta,
750: \sin\hat\varphi\sin2\hat\theta, -\cos2\hat\theta\big)$, with
751: $\Omega=|\vec\Omega|$,
752: which identifies the effective hamiltonian
753: $H=\vec\Omega\cdot\vec\sigma$ in $(2.9b)$.
754:
755: Actually, in presence of Majorana neutrinos,
756: also in vacuum the most general mixing matrix
757: involves two angles, $\theta$ and $\varphi$, so that
758: the explicit expression of the free effective hamiltonian
759: $H_0$ in terms of these angles is as for $H$ above.
760: Although for oscillations in vacuum involving only
761: two species of neutrinos the angle $\varphi$
762: disappears from all observables, this is no longer
763: true in presence of matter induced decoherence effects.
764: Indeed, one can directly check that the transition probabilities
765: explicitly depend
766: on $\varphi$, unless the dissipative parameters
767: $a$, $b$, $c$, $\alpha$, $\beta$, $\gamma$ are all zero;
768: at least in principle, it is therefore possible
769: to distinguish between Dirac and Majorana neutrinos
770: by studying their oscillations in random matter.
771: The detailed analysis of such of dependence
772: is certainly beyond the scope of the present
773: investigation and thus, in order to keep the
774: treatment as simple as possible,
775: in the previous discussions we have tacitly assumed the neutrinos travelling
776: in matter to be of Dirac type, setting
777: $\varphi=\,0$ from the beginning.
778:
779: As a final remark, let us mention that master
780: equations of the type (2.9)
781: generate the most general open system dynamics compatible
782: with a semigroup composition law and the requirement of complete positivity,
783: and as such can be applied to model in a physically consistent way
784: a wide range of phenomena.[3-7]
785: In particular, they have been recently used in order to describe dissipative
786: effects induced at low energies by the dynamics of fundamental objects
787: (strings and branes) at a very high scale, typically the Planck mass.[29, 30]
788: These string induced decoherence effects may modify the pattern
789: of neutrino oscillations,
790: and in principle interfere with the phenomena described above.
791: Nevertheless, besides being very small, they
792: affect in equal manner all types of neutrinos, so that they can be
793: isolated from the matter-induced effects by analyzing data taken
794: in different experimental conditions.
795:
796:
797:
798: \vskip 1.5cm
799: %\vfill\break
800:
801:
802: \noindent
803: {\bf APPENDIX}
804: \medskip
805:
806: In order to show that negative probabilities arise
807: in experimental accessible observables once the naive model
808: of density fluctuating matter discussed in the text is adopted,
809: one needs to combine neutrino propagation in vacuum with
810: that in the medium. Consider a neutrino, created as
811: $\nu_e$, that propagates for time $t'$ in vacuum, then enters
812: the random medium in which stays for a time $t$, and is finally
813: detected after having travelled again in vacuum for a further time $t''$.
814: The probability ${\cal P}_{\nu_e\to\nu_\mu}(\tau)$
815: of finding a neutrino of type $\nu_\mu$
816: at the final time $\tau=t'+t+t''$ can be expressed as in (2.17),
817: where the total transition matrix ${\cal M}(\tau)$
818: is now the product of three terms,
819: ${\cal M}(\tau)={\cal M}_0(t')\cdot{\cal M}(t)\cdot{\cal M}_0(t'')$,
820: the middle representing the propagation in the medium
821: with parameters as in (3.1) and (3.2), while the outer two
822: the ``free'' motion in vacuum, generated by the hamiltonian (2.2).
823:
824: When the vacuum evolution time $t'$ is chosen to be very short, such that
825: $\sin\omega t'=\beta/(2a\sin2\theta)=\omega/4\lambda$, the state
826: of the neutrino entering the medium is
827: $\rho_-=[\sigma_0-\vec\rho_-\cdot\vec\sigma]/2$,
828: where $\vec\rho_-=-{\cal M}_0(t') \vec\rho_{\nu_e}(0)$
829: coincides with the eigenvector of the dissipative part
830: of $\cal H$ relative to its negative eigenvalue.
831: Similarly, with the same choice also for $t''$,
832: one finds ${\cal M}_0(-t'') \vec\rho_{\nu_\mu}(0)=\vec\rho_-$, so that
833: when exiting the medium the neutrino is found in the state $\rho_+=1-\rho_-$,
834: orthogonal to $\rho_-$.
835: With these conditions, one has:
836: ${\cal P}_{\nu_e\to\nu_\mu}(\tau)={\cal P}_{-\to +}(t)$,
837: and near resonance, one explicitly finds:
838: $$
839: {\cal P}_{\nu_e\to\nu_\mu}(\tau)={1\over 2}\Bigg[1-e^{- a t}
840: \Bigg(\cos 2\Omega t + D\ {\sin 2\Omega t\over\Omega}\Bigg)\Bigg]\ ,
841: $$
842: where $D=[a^2/4+\beta^2]^{1/2}$ and $\Omega=[\omega^2-D^2]^{1/2}$;
843: this expression
844: indeed assumes unphysical negative values for sufficiently small times:
845: ${\cal P}_{\nu_e\to\nu_\mu}(\tau)\simeq (a/2-D)t$.
846:
847:
848: %\vskip 2cm
849: \vfill\eject
850:
851: \line{}
852: \centerline{\bf ACKNOWLEDGMENT}
853: \bigskip
854:
855: We thank A. Yu. Smirnov for reading the manuscript and for very useful discussions.
856:
857: \vskip 3.5cm
858:
859: \centerline{\bf REFERENCES}
860: \vskip 1cm
861:
862: \item{1.} L. Wolfenstein, Phys. Rev. D {\bf 17} (1978) 2369;
863: {\it ibid.} {\bf 20} (1979) 2634
864: \smallskip
865: \item{2.} S.P. Mikheyev and A.Yu. Smirnov, Sov. J. Nucl. Phys.
866: {\bf 42} (1985) 913; Nuovo Cim. {\bf 9C} (1986) 17
867: \smallskip
868: \item{3.} R. Alicki and K. Lendi, {\it Quantum Dynamical Semigroups and Applications}, Lect. Notes Phys. {\bf 286}, (Springer, Berlin, 1987)\smallskip\item{4.} V. Gorini, A. Frigerio, M. Verri, A. Kossakowski andE.C.G. Surdarshan, Rep. Math. Phys. {\bf 13} (1978) 149 \smallskip\item{5.} H. Spohn, Rev. Mod. Phys. {\bf 52} (1980) 569\smallskip
869: \item{6.} H.-P. Breuer and F. Petruccione, {\it The Theory of Open
870: Quantum Systems}, (Oxford University Press, Oxford, 2002)
871: \smallskip
872: \item{7.} {\it Dissipative Quantum Dynamics}, F. Benatti and R. Floreanini eds.,
873: Lect. Notes Phys. {\bf 612}, (Springer, Berlin, 2003)
874: \smallskip
875: \item{8.} C.P. Burgess, N.S. Dzhalilov, T.I. Rashba, V.B. Semikoz and J.W.F. Valle,
876: Mon. Not. Roy. Astron. Soc. {\bf 348} (2004) 609
877: \smallskip
878: \item{9.} F.N. Loreti and A.B. Balantekin, Phys. Rev. D {\bf 50} (1994) 4762
879: \smallskip
880: \item{10.} H. Nunokawa, A. Rossi, V.B. Semikoz and J.W.F. Valle,
881: Nucl. Phys. {\bf B472} (1996) 495
882: \smallskip
883: \item{11.} E. Torrente-Lujan, Phys. Rev. D {\bf 59} (1999) 073001
884: \smallskip
885: \item{12.} C.P. Burgess, N.S. Dzhalilov, M. Maltoni, T.I. Rashba, V.B. Semikoz,
886: M. Tortola and J.W.F. Valle, Astrphys. J. {\bf 588} (2003) L65
887: \smallskip
888: \item{13.} A.B. Balantekin and H. Y\"uksel, Phys. Rev. D {\bf 68} (2003)
889: 013006
890: \smallskip
891: \item{14.} M.M. Guzzo, P.C. de Holanda and N. Reggiani,
892: Phys. Lett. {\bf B569} (2003) 45
893: \smallskip
894: \item{15.} A. Friedland, C. Lunardini and C. Pena-Garay,
895: Phys. Lett. {\bf B594} (2004) 347
896: \smallskip
897: \item{16.} M.M Guzzo, P.C. de Holanda and O.L.G. Peres,
898: Phys. Lett. {\bf B591} (2004) 1
899: \smallskip
900: \item{17.} K.M. Zurek, JHEP {\bf 04} (2004) 058
901: \smallskip
902: \item{18.} S.M. Bilenky and S.T. Petcov, Rev. Mod. Phys.
903: {\bf 59} (1987) 671
904: \smallskip
905: \item{19.} T.K. Kuo and J. Pantaleone, Rev. Mod. Phys. {\bf 61} (1989) 937
906: \smallskip
907: \item{20.} C.W Kim and A. Pevsner, {\it Neutrinos in Physics and
908: Astrophysics}, (Harwood Academic Press, 1993)
909: \smallskip
910: \item{21.} R.N. Mohapatra and P.B. Pal, {\it Massive Neutrinos in Physics
911: and Astrophysics}, 2nd ed., (World Scientific, Singapore, 1999)
912: \smallskip
913: \item{22.} M. Fukugita and T. Yanagida, {\it Physics of Neutrinos and
914: Applications to Astrophysics}, (Springer, Berlin, 2003)
915: \smallskip
916: \item{23.} J. Budimir and J.L. Skinner, J. Stat. Phys. {\bf 49}
917: (1987) 1029
918: \smallskip\item{24.} F. Benatti, R. Floreanini and R. Romano, J. Phys. A
919: {\bf 35} (2002) 4955
920: \smallskip
921: \item{25.} F.N. Loreti, Y.ÐZ. Qian, G.M. Fuller and A.B. Balantekin,
922: Phys. Rev. D {\bf 52} (1995) 6664
923: \smallskip
924: \item{26.} F. Benatti and R. Floreanini, Mod. Phys. Lett. {\bf A12} (1997) 1465;
925: Banach Center Publications, {\bf 43} (1998) 71;
926: Phys. Lett. {\bf B468} (1999) 287;
927: Chaos, Solitons and Fractals {\bf 12} (2001) 2631
928: \smallskip
929: \item{27.} V. Barger, D. Marfatia and K. Whisnant,
930: Int. J. Mod. Phys. {\bf E12} (2003) 569
931: \smallskip
932: \item{28.} R.D. McKeown and P. Vogel, Phys. Rep. {\bf 394} (2004) 315
933: \smallskip
934: \item{29.} F. Benatti and R. Floreanini, JHEP {\bf 02} (2000) 032
935: \smallskip
936: \item{30.} F. Benatti and R. Floreanini, Phys. Rev. D {\bf 64} (2001)
937: 085015
938:
939: \vfill\eject
940:
941: \vskip 3cm
942:
943: \input epsf
944: \centerline{
945: \epsfxsize=10cm
946: \epsfbox{fig1.eps}
947: }
948: \midinsert
949: \narrower\noindent
950: {\bf Figure 1.} Behaviour of electron neutrino mean survival
951: probability ${\cal P}_{\nu_e\to\nu_e}$ as a function of the neutrino energy
952: (through the ratio $A/A_R$), for $\sin^2 2\theta\simeq 0.8$,
953: density dominated matter fluctuations,
954: $\lambda_1,\ \lambda_2\gg\lambda_3$, and different correlation strengths,
955: ${\cal W}/\lambda_2\simeq 10^{-4}\div 10^{-3}$. The lower starting (black) curve corresponds to
956: the case of noisless matter (standard MSW effect), while the remaining
957: (colored) ones show the effect of the stochastic fluctuations.
958: The initial gap among the group of curves is due to the presence
959: of the decoherence driven damping factor.
960:
961:
962: \endinsert
963:
964:
965: \bye
966: