1: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}

2: %

3: \usepackage{graphicx}% Include figure files

4: \usepackage{dcolumn}% Align table columns on decimal point

5: \usepackage{bm}% bold math

6: %\documentstyle[aps,twocolumn]{revtex}

7: %\documentstyle[10pt,aps,draft]{revtex}

8: \input{epsf}

9: \begin{document}

10: \title{{Anomalous transport: a deterministic approach.}}

11: 

12: \vskip4mm                      

13: 

14: \author{Roberto Artuso}

15: \email{Roberto.Artuso@uninsubria.it}

16: \altaffiliation{also at

17: Istituto Nazionale di Fisica Nucleare, Sezione di Milano,

18: Via Celoria 16, 20133 Milano, Italy} 

19: \author{Giampaolo Cristadoro}

20: \email{Giampaolo.Cristadoro@uninsubria.it}

21: \affiliation{Center for Nonlinear and Complex Systems and

22: Dipartimento di Scienze Chimiche, Fisiche e Matematiche, Universit\`a

23: dell'Insubria\\

24: Istituto Nazionale di Fisica della Materia, 

25: Unit\`a di Como, Via Valleggio 11, 22100 Como, Italy\\

26: 22100 Como, Italy}

27: \date{\today}

28: 

29: \begin{abstract}

30: 

31: {We 

32: %%establish a deterministic approach to investigate transport 

33: %%exponents of arbitrary order. The theory is applied to the analysis of 

34: %%different kinds of intermittent one-dimensional maps.

35: introduce a cycle-expansion (fully deterministic)

36: technique to compute the asymptotic behavior of

37: arbitrary order transport moments. The theory is applied to different 

38: kinds of one-dimensional

39: intermittent maps, and Lorentz gas with infinite horizon, confirming 

40: the typical appearance of phase transitions in the transport spectrum.

41: % inserisci abstract 

42: 

43: }

44: \end{abstract}

45: 

46: \pacs{05.45.-a }

47: \maketitle

48: %\begin{multicols}{2}

49: \narrowtext

50: 

51: Generic dynamical systems are characterized by the coexistence of chaotic 

52: regions and regular structures, and typical trajectories present 

53: regular segments, due to sticking to the ordered component of the phase 

54: space, separated by erratic behavior, due to wanderings in the chaotic 

55: sea. Though ubiquitous, this mixed behavior still involves hard 

56: theoretical problems, as present theories are tailored to fit

57: the two opposite paradigms of either integrable 

58: or fully chaotic systems. A particularly relevant feature associated to 

59: such weakly chaotic dynamics is anomalous transport:

60: important features of the process are captured by the function 

61: $\nu(q)$ that expresses the asymptotic growth of moments of arbitrary 

62: order:

63: \begin{equation}

64: \langle | x_t-x_0 |^q \rangle \, \sim \, t^{ \nu(q)} \, 

65: \label{nu-mom}

66: \end{equation}

67: ordinary diffusion yields  $\nu(q)=q/2$, while 

68: anomalous transport often leads to a non trivial behavior, which cannot 

69: be encoded by a single exponent, but rather typically exhibits a 

70: phase transition \cite{Ark,V-AD}.

71: Anomalous diffusion has been recognized as an ubiquitous phenomenon in recent years, from intermittency induced anomalous transport in 

72: one-dimensional maps \cite{Gei-AT} to the analysis of area-preserving 

73: maps in the presence of self-similar regular structures around 

74: accelerator modes \cite{Zas-AM}, from passive tracers dynamics in rotating flows \cite{Wee}, to charge carrier transport in amorphous semiconductors \cite{Mont}, to many others physically relevant contexts (see \cite{an-rev} and references therein).

75: In this paper we show how the transport exponents $\nu(q)$ may be 

76: computed in a crisply deterministic manner, by means of periodic orbit 

77: expansions \cite{DB,AACI}: we will then apply the formalism to various 

78: low dimensional settings, where analytic computations may be performed. 

79: These computations accomplish a twofold goal: besides checking the 

80: theory, they will also illustrate how subtle features of the underlying 

81: dynamical system are automatically included in the formalism.

82: 

83: For simplicity, we illustrate the analytic technique for the case of a one-dimensional  map on the real

84: line, even though the method is by no means limited to this context. The key property

85: of the dynamical system under investigation is represented by the symmetry properties

86: \begin{equation}

87: f(-x)  \, = \,   -f(x) \qquad \quad

88: f(x+n) \, = \, n+f(x)

89: \label{symm-P}

90: \end{equation}

91: for any $n \in \mathbf{N}$. These properties guarantee the absence of a net drift, as well as that

92: the map on the real line is obtained by lifting a circle

93: map $\hat{f}(\theta)$  (on the unit torus)

94: \begin{equation}

95: \hat{f}(\theta)\,=\, \left. f(\theta)\right|_{mod\,\,1} \qquad \theta \in {\mathcal{T}}=[0,1)

96: \label{tor-f}

97: \end{equation}

98: We may thus split the $f$ evolution into a box integer plus a fractional part: 

99: $x_n=N_n+\theta_n$,

100: where

101: \begin{equation}

102: N_{n+1} \,=\, N_{n} + \sigma(f(\theta_n)) \qquad \quad

103: \theta_{n+1}\, =\,   \hat{f}(\theta_n) \label{fracf}

104: \end{equation}

105: where $\sigma(y)=[y]$ is the denotes the integer part.  Transport properties

106: are accounted for by the generating function $G_n(\beta)$, defined as

107: \begin{equation}

108: \label{gen-f}

109: G_n(\beta)\,=\langle \,e^{\beta(x_n-x_0)} \rangle_0

110: \end{equation}

111: (the average being taken among initial conditions), whose asymptotic behavior may be

112: associated to the behavior of a generalized

113: \emph{dynamical zeta function}\cite{rD,DB,pD},

114: \begin{equation}

115: \label{gen-z}

116: G_n(\beta)\, \sim \, \frac{1}{2 \pi \imath}\int_{a - \imath \infty}^{a+ \imath \infty} \, ds\, e^{sn}

117: \, \frac{d}{ds} \ln \left[ \zeta_{(0)\beta}^{-1}(e^{-s}) \right]

118: \end{equation}

119: and $\zeta$ is expressed as an infinite product over prime periodic

120: orbits of the torus map $\hat{f}$:

121: \begin{equation}

122: \label{Dz}

123: \zeta_{(0)\beta}^{-1}(z)\,=\, \prod_{\{p\}}\,\left( 1 - \frac{e^{\beta \sigma_p}z^{n_p}}{|\Lambda_p|}

124: \right)

125: \end{equation}

126: The quantities that enter the definition (\ref{Dz}) are the prime period $n_p$ of the orbit $p$, its

127: instability $\Lambda_p=\prod_{i=0}^{n_p-1}\hat{f}'(\hat{f}^i(x_p)$ and the integer factor $\sigma_p$,

128: that accounts for the orbit's behavior once we unfold it on the real line. As a matter of fact, given

129: any point $x_p$ belonging to $p$ (that is $\hat{f}^{n_p}(x_p)=x_p$) we may either have that it is

130: a periodic point of the lift $f$ (i.e. $f^{n_p}(x_p)=x_p$), or it might be a running mode, 

131: $f^{n_p}(x_p)=x_p+\sigma_p$, with $\sigma_p \in \mathbf{Z}$ (see (\ref{tor-f})). If the

132: zeta function (\ref{Dz}) has a simple zero $z(\beta)$, then, from (\ref{gen-z}) it follows that 

133: the second moment of the distribution grows linearly in time (normal diffusion). This is generally the case with fully chaotic systems \cite{DB,rs,AACII,vb}: in the last

134: few years it has been realized \cite{gg,act,per,prisola} that weakly chaotic systems (in particular

135: one dimensional intermittent maps, or infinite horizon Lorentz gas models \cite{perL}) lead to more 

136: complicated analytic structure of the zeta function (\ref{Dz}), which typically exhibit branch 

137: points. In view of the inverse Laplace formula (\ref{gen-z}) the modified analytic structure

138: may induce anomalous behavior (nonlinear diffusion \cite{ACL,DB}).

139: 

140: \begin{figure}[t]

141: \centerline{\epsfxsize=4.cm \epsfbox{figure1a.eps}

142: \hspace{2mm}\epsfysize=4.cm \epsfbox{figure1b.eps}}

143: \vskip4mm

144: \caption{\label{f1}A torus map with intermittent fixed points, and the corresponding lift.}

145: \end{figure}

146: We address here the problem of going beyond diffusion, and characterize the whole spectrum 

147: of transport exponents $\nu(q)$ (\ref{nu-mom}). From small $\beta$ expansion of (\ref{gen-z}),

148: we get

149: \begin{eqnarray}

150: \sigma_k(n)&\,=\,& \langle (x_n -x_0 )^k \rangle_0 \,=\, \left. \frac{\partial^k \,} {\partial \beta^k}

151: G_n(\beta) \right|_{\beta=0} \nonumber \\

152:  &\,\sim\,& \left. \frac{\partial^k \,} {\partial \beta^k} \frac{1}{2 \pi \imath}\int_{a - \imath \infty}^{a+ \imath \infty} \, ds\, e^{sn}

153: \, \frac{d}{ds} \ln \left[ \zeta_{(0)\beta}^{-1}(e^{-s}) \right] \right|_{\beta=0} 

154: \label{gen-mom}

155: \end{eqnarray}

156: The evaluation of the integral on the right hand side of (\ref{gen-mom}) requires dealing with

157: high order derivatives of a composite function: this is accomplished by making use of Fa\`a di Bruno formula:

158: \begin{eqnarray} 

159: \frac{d^n \,}{dt^n}H(L(t))= \sum_{k=1}^n\,\sum_{k_1,\cdots k_n}

160: \, \frac{d^k H}{dt^k}  (L(t))\cdot B_{\vec{k}}(L(t))

161: \label{fdb}\\

162: B_{\vec{k}}(L(t))=\left(\frac{1}{1!}\frac{dL}{dt}\right)^{k_1} \cdots \left(\frac{1}{n!}\frac{d^nL}{dt^n}\right)^{k_n} \, \, \, \,  \quad \, ~\\

163: \vec{k}=\{k_1,\dots k_n\} \,\text{with}\, \sum k_i=k ,\quad \sum i\cdot k_i=n 

164: \end{eqnarray}

165: 

166: When the analytic structure of the zeta function is known, from (\ref{fdb})

167: we may single out the leading singularity in the logarithmic derivative, and then

168: estimate the asymptotic behavior of (\ref{gen-mom}), for instance by employing Tauberian

169: theorems for Laplace transforms \cite{Fell}. 

170: 

171: We will apply the technique to two classes of one dimensional maps, where deviations from 

172: fully chaotic behavior are provided by marginal fixed points, of intermittent type \cite{PM}.

173: %\end{figure}

174: The torus map is shown in fig. (\ref{f1}): it consists of three branches, the central one being hyperbolic

175: (with constant slope), while the other two include a marginal fixed point: to unfold it on the real

176: line it is sufficient to assign jumping numbers to branches. For instance we get the lift of fig. (\ref{f1})

177: once we assign $\sigma_{-}=-1$ to the left branch, $\sigma_{c}=0$ to the central one and 

178: $\sigma_{+}=+1$ to the right branch. 

179: The map is actually

180: taken as a straightforward generalization of the Gaspard-Wang piecewise linear approximation

181: \cite{plapp} of Pomeau-Manneville map: the intermittent behavior is determined by dynamics near

182: the pair of parabolic fixed points, which is accounted for by the intermittency exponent $\gamma > 1$,

183: (the map goes like $x_{n+1}\sim x_n + x_n^{\gamma}$ near the origin, with an analogous behavior

184: at the twin fixed point in $x=1$).

185: The 

186: %major advantage of the piecewise linear approximation is that it allows for an explicit writing of the dynamical zeta

187: %function, without losing important dynamical features \cite{isolapl}: in particular the dynamical

188: zeta function can then be written as

189: \begin{equation}

190: \label{pl-dz}

191: \zeta_{(0)\beta}^{-1}(z)\,=\, 1-az-bz\sum_{k=1}^{\infty}\, \frac{z^k}{k^{\alpha+1}}\cosh (\beta k)

192: \end{equation}

193: where $a$ and $b$ are fixed by specifying the central region slope and the normalization

194: condition $\zeta_{(0)0}^{-1}(1)=0$, and 

195: \begin{equation}

196: \label{stab-z}

197: \alpha=1/(\gamma-1).

198: \end{equation} 

199: The appearance of the Bose function

200: %\begin{equation}

201: %\label{Bf}

202: $g_{\mu}(z)\,=\,\sum_{l=1}^{\infty}\, \frac{z^l}{l^{\mu}}$

203: %\end{equation}

204: is due to sequences of orbits coming closer and closer to the marginal fixed points: their 

205: stability increases only polynomially with the period \cite{plapp,AACII}, a clear signature

206: of local deviation from typical hyperbolic behavior (which is ruled by exponential instability

207: growth). To estimate the various 

208: %singularities appearing once we evaluate the derivative 

209: %expansion 

210: contributions in (\ref{fdb}) we 

211: remind the 

212: %have to remind how Bose functions 

213: behavior as $z \to 1^{-}$

214: \begin{eqnarray}

215: g_{\mu}(z) \sim  \left\{

216: \begin{array}{ll}

217: (1-z)^{\mu -1} \quad & \mu<1 \\

218: \ln (1-z) \quad & \mu=1 \\

219: \zeta(\mu)+C_{\mu}(1-z)^{\mu-1}+D_{\mu}(1-z) \quad & \mu \in (1,2) \\

220: \zeta(2)+C_2(1-z) \ln (1-z) \quad &  \mu=2 \\

221: \zeta(\mu)+C_{\mu}(1-z) \quad & \mu> 2

222: \end{array} \right.

223: \label{B-as}

224: \end{eqnarray}

225: and moreover take into account that

226: \begin{eqnarray}

227: \left. \frac{\partial^i \, }{\partial \beta^i}\zeta_{(0)\beta}^{-1}(z)\right|_{\beta=0}\,\sim \,

228: \left\{

229: \begin{array}{ll}

230: 0 \qquad & i \,\,\mathrm{odd} \\

231: zg_{\alpha+1-i}(z) \quad & i\,\, \mathrm{even}

232: \end{array} 

233: \right.

234: \label{djz}

235: \end{eqnarray}

236: Now take a generic term in (\ref{fdb}): and denote it by ${\cal D}_{k_1 \dots k_n}$: we have

237: in view of (\ref{djz})

238: \begin{equation}

239: \label{typ-fdb}

240: {\cal D}_{k_1 \dots k_n}\,\sim\, \frac{1}{(\zeta_{(0)0}^{-1}(z))^k}\prod_j\,(g_{\alpha+1-j}(z))^{k_j}

241: \,=\,\frac{{\cal D}^{+}_{k_1 \dots k_n}}{{\cal D}^{-}_{k_1 \dots k_n}}

242: \end{equation}

243: where the ${\cal D}^{+}$ picks up the contributions from the product of Bose functions, and

244: all $j$ must be even, due to (\ref{djz}).

245: First we consider the case $\alpha \in (0,1)$, which corresponds to 

246: $\gamma > 2$: we have ${\cal D}^{-}_{k_1 \dots k_n}\sim (1-z)^{k\alpha}$, that,

247: together with (\ref{B-as}), implies that the dominant singularity is of the form

248: \begin{equation}

249: \label{ddom}

250: {{\cal D}_n}\,\sim \, \frac{1}{(1-z)^{\rho}}

251: \end{equation}

252: where $\rho$ is determined by 

253: \begin{equation}

254: \label{rho}

255: \rho=  \sup_{\{k_1 \dots k_n\}} ( k\alpha + \sum_j (j-\alpha)k_j) =n

256: \end{equation} 

257: Once plugged into (\ref{gen-mom}) this leads to the estimate

258: %\begin{equation}

259: $\label{nu-norm}

260: \nu(q)\,=\, q

261: $,

262: %\end{equation}

263: which means that the whole set of moments is ruled by ballistic behavior (at least

264: for even exponents, where the method applies).

265: We now turn to the more subtle case $ \alpha > 1$: since the dynamical zeta function has a

266: simple zero we get ${\cal D}^{-}_{k_1 \dots k_n}\sim (1-z)^k$, while the terms appearing in

267: ${\cal D}^{+}$ modify the singular behavior near $z=1$ only for sufficiently high $j$

268: \begin{equation}

269: g_{\alpha+1-j}(z)\,\sim\,

270: \left\{

271: \begin{array}{ll}

272: (1-z)^{\alpha -j}\quad & j> \alpha \\

273: \zeta(\alpha +1-j) \quad & j< \alpha

274: \end{array}

275: \right.

276: \label{ggg}

277: \end{equation}

278: If all $\{j\}$ are less than $\alpha$ then the singularity is determined by ${\cal D}^{-}$: keeping in

279: mind that the highest $k$ value is achieved by choosing $j=2$ and $k_2=n/2$, we get, by 

280: proceeding as before

281: \begin{equation}

282: \label{mom-}

283: \nu(q)\,=\,\frac{q}{2} \qquad q < \alpha

284: \end{equation}

285: When $q$ exceeds $\alpha$ we have to take into account possible additional singularities in

286: ${\cal D}^{+}$, and thus we get

287: \begin{equation}

288: \label{ddom2}

289: {{\cal D}_n}\,\sim \, \frac{1}{(1-z)^{\rho}}

290: \end{equation}

291: where $\rho$ is determined by 

292: \begin{equation}

293: \rho= \sup_{\{k_1 \dots k_n\}} ( k + \sum_{j>\alpha} (j-\alpha)k_j) =

294: \left\{

295: \begin{array}{ll}

296: n/2 \quad & n < 2(\alpha-1) \\

297: n+1-\alpha \quad & n> 2(\alpha-1)

298: \end{array}

299: \right.

300: \label{rho2}

301: \end{equation} 

302: which, once we take (\ref{mom-}) into account, yields

303: \begin{equation}

304: \label{momfin}

305: \nu(q)\,=\,

306: \left\{

307: \begin{array}{ll}

308: q/2 \quad & q< 2(\alpha-1) \\

309: q+1-\alpha \quad & q> 2(\alpha-1)

310: \end{array}

311: \right.

312: \end{equation}

313: The set of exponents thus has a nontrivial structure, characterized by a sort of phase transition

314: for $q=2(\alpha-1)$, a rather universal feature of many systems exhibiting anomalous transport

315: \cite{V-AD}. We notice that the parameter ruling the presence of a phase transition (and the

316: explicit form of the spectrum) is $\alpha$, the exponent describing the polynomial 

317: instability growth of periodic orbits coming closer and closer to the marginal

318: fixed point, and thus describing the sticking to the regular part of the phase space: in the present

319: example $\alpha=\/(\gamma-1)^{-1}$, and thus the sticking exponent is easily connected to the

320: intermittency index.

321: 

322: \begin{figure}

323: \centerline{\epsfxsize=5.cm \epsfbox{figure2.eps}}

324: \vskip4mm

325: \caption{\label{f3}The intermittent map with constant invariant measure.}

326: \end{figure}

327: We now discuss a further 1-d example: first we recall that the former torus map has non trivial ergodic properties: an absolutely

328: continuous invariant measure only exists for $\alpha > 1$ (see for instance \cite{ci} and references 

329: therein): the ergodic behavior is much more complex when $\alpha <1$. 

330: In the new example

331: %any case sticking induces

332: %peaking of the measure around marginal fixed points (see \cite{plapp}): there exists another

333: %quite interesting system 

334: \cite{Ark}  while the functional form of the map near marginal

335: fixed points is identical to the former case, ergodic properties are completely different (the measure is not singular). 

336: The torus map, again dependent on an intermittency parameter $\gamma$,

337:  is implicitly defined on $\mathcal{T}=[-1,1)$ in the following way \cite{Ark}:

338: \begin{equation}

339: x\,=\,

340: \left\{

341: \begin{array}{ll}

342: \frac{1}{2\gamma}\left(1+ \hat{f}(x)\right)^{\gamma} \qquad & 0 < x < 1/(2\gamma) \\

343: \hat{f}(x) + \frac{1}{2\gamma} \left( 1- \hat{f}(x) \right)^{\gamma} \qquad & 1/(2\gamma) < x < 1

344: \end{array}

345: \right.

346: \label{arkmap}

347: \end{equation}

348: for negative values of $x$ the map is defined as $\hat{f}(-x)=-\hat{f}(x)$ (cfr (\ref{symm-P})):

349: see fig. (\ref{f3}).

350: \begin{figure}[t]

351: \centerline{\epsfxsize=5.cm \epsfbox{figure3.eps}}

352: \vskip4mm

353: \caption{\label{f4}Spectrum of the transport moments for the map (\ref{arkmap}) with $\gamma=1.5$: 

354: the best fit on numerical data is $y=0.50 x +0.04$ for the dotted line, and $y=0.98x+1.82$ for

355: the full line.}

356: \end{figure}

357: The peculiar ergodic features of the map, namely the existence of a {\em constant} invariant

358: measure for {\em any} value of $\gamma$ arise from the property

359: %\begin{equation}

360: %\label{arksumm}

361: $

362: \sum_{y=\hat{f}^{-1}(x)}\, 1/{\hat{f}'(y)}\,=\,1

363: $

364: %\end{equation}

365: Also in this case we can easily identify families of orbits coming closer and closer to the marginal

366: fixed points, but evaluating their instability requires some care, as the slope in the chaotic region

367: is not bounded from above. A piecewise linear approximation (in the same spirit as \cite{plapp})

368: is still possible \cite{acprep}, but we must put particular attention on matching the summation property

369: %(\ref{arksumm})

370: :

371: the corresponding dynamical zeta function is

372: \begin{equation}

373: \zeta_{(0)\beta}^{-1}(z)\,=\,1-\zeta(\alpha+2)z\sum_{k=1}^{\infty}\, \frac{z^k}{k^{\alpha+2}}

374: \cosh(\beta k)

375: \label{zark}

376: \end{equation}

377: where again $\alpha=1/(\gamma-1)$. Thus the continuity of the measure deeply modifies the relationship

378: between the intermittency exponent and the instabilities of periodic orbits shadowing the 

379: marginal fixed points: by repeating the steps of our former calculation we get that we always get a phase transition with

380: \begin{equation}

381: \label{arkspec}

382: \nu(q)\,=\,

383: \left\{

384: \begin{array}{ll}

385: q/2 \quad & q< 2\alpha \\

386: q-\alpha \quad & q> 2\alpha

387: \end{array}

388: \right.

389: \end{equation}

390: which may also be checked numerically (see fig. (\ref{f4})).

391: 

392: Our last example shows how the proposed technique may be applied to higher dimensional systems, where a detailed layout of the full symbolic dynamics is beyond our present understanding. We consider the Lorentz gas with infinite horizon (square lattice of circular scatterers where a test particle freely moves, colliding elastically with the disks). The mechanism that in this case leads to deviation from normal behavior is due to the possibility of arbitrarily long flights of free (collisionless) motion along corridors. In this example the reduced dynamics is that of a Sinai billiard: while a complete description of the full set of periodic orbits of such a system is not known yet a a family of periodic orbits approaching closer and closer the infinite free flights is singled out: it consists (for the unfolded billiard) to orbits that jump $n$ lattice spacings between collisions: their number and instability can be evaluated by geometric arguments\cite{perL}, yielding an istability that grows polynomially with $n$ with a power $3$: so their role is analogous to modes leading to Bose functions in (\ref{pl-dz}) with $\alpha=2$. In view of (\ref{momfin}) this leads to $\nu(q)=q/2$ for $q< 2$ and $\nu(q)=q-1$ for $q > 2$, like recently suggested in \cite{L-Ott}: a careful analysis, based on (\ref{B-as}) moreover gives $\sigma_2(n) \sim n \ln n$ (see \cite{Ble}), so that our method is also capable of picking up logarithmic corrections to the dominating power-law behavior.

393: 

394: We have proposed a crisply deterministic technique to investigate the full spectrum of

395: transport exponent for chaotic systems: in particular this method is capable of explaining

396: the different phase transitions that possibly arise, without any direct information on the invariant

397: measure: the only information to be plugged is local growth of instabilities near the 

398: marginal structures, together with the appropriate jumping factor.

399: 

400: 

401: %

402: %

403: %

404: 	

405: This work was partially supported by INFM PA project {\em

406: Weak chaos: theory and applications}, and by EU contract QTRANS Network 

407: ({\em Quantum transport on an atomic scale}).

408: 

409: %\section{References}

410: \begin{thebibliography}{99}

411: \bibitem{Ark} A.S. Pikovsky, {\em Phys.Rev. \bf A43}, 3146 (1991).

412: 

413: \bibitem{V-AD} P. Castiglione, A. Mazzino, P. Muratore-Ginanneschi and 

414: A. Vulpiani, {\em Physics \bf D134}, 75 (1999)

415: 

416: \bibitem{Gei-AT} T. Geisel, J, Nierwetberg and A. Zacherl, {\em 

417: Phys.Rev.Lett. \bf 54}, 616 (1985).

418: 

419: \bibitem{Zas-AM} R. Ishizaki, T. Horita, T. Kobayashi and H. Mori, {\em 

420: Progr.Theor.Phys. \bf 85}, 1013 (1991); S. Benkadda, S. Kassibrakis, R.B. White and G.M. 

421: Zaslavsky, {\em Phys.Rev. \bf E55}, 4909 (1997).

422: 

423: \bibitem{Wee} E.R. Weeks and H.L. Swinney, {\em Phys.Rev. \bf E57}, 4915 (1998).

424: 

425: \bibitem{Mont} H. Scher and E.W. Montroll, {\em Phys.Rev. \bf B12}, 2455 (1975).

426: 

427: \bibitem{an-rev} J.-P. Bouchaud and A. Georges, {\em Phys.Rep. \bf 195}, 12 (1990); R. Metzler and J. Klafter, {\em Phys.Rep. \bf 339}, 1 (2000).

428: 

429: \bibitem{DB} P. Cvitanovi\'c, R. Artuso, P. Dahlqvist, R. Mainieri, G. 

430: Tanner, G. Vattay, N. Whelan and A. Wirzba, {\sf Chaos: classical and 

431: quantum} ({\tt www.nbi.dk/ChaosBook/} 2003).

432: 

433: \bibitem{AACI} R. Artuso, E. Aurell and P. Cvitanovi\'c, {\em 

434: Nonlinearity \bf 3}, 325 (1990).

435: 

436: \bibitem{rD} R. Artuso, {\em Phys.Lett. \bf A 160}, 528 (1991).

437: 

438: \bibitem{pD} P. Cvitanovi\'c, J.-P. Eckmann and P. Gaspard, {\em Chaos, Solitons and 

439: Fractals, \bf 6}, 113 (1995).

440: 

441: \bibitem{rs} R. Artuso and R. Strepparava, {\em Phys.Lett. \bf A236}, 469 (1997).

442: 

443: \bibitem{AACII} R. Artuso, E. Aurell and P. Cvitanovi\'c, {\em Nonlinearity \bf 3}, 361 (1990).

444: 

445: \bibitem{vb} See for instance 

446: V. Baladi, {\sf A brief introduction to dynamical zeta functions}, in: 

447: DMV-Seminar  {\bf 27},   

448: {\sf Classical Nonintegrability, Quantum Chaos}, by A. Knauf and Ya.G. Sinai, 

449: BirkhŠuser, 3-20 (1997)

450: and references therein.

451: 

452: \bibitem{gg} G. Gallavotti, {\em Rendiconti dell'Accademia Nazionale dei Lincei: \bf 51}, 509 (1977). 

453: 

454: \bibitem{act} R. Artuso, P. Cvitanovi\'c and G. Tanner, {\sf Cycle expansions for intermittent

455: maps}, submitted to {\em Prog.Theor.Phys.} (2002).

456: 

457: \bibitem{per} P. Dahlqvist, {\em Phys.Rev. \bf E60}, 6639 (1999).

458: 

459: \bibitem{prisola} T. Prellberg, {\em J.Phys. \bf A36}, 2455 (2003); S. Isola, {\em Nonlinearity \bf 15}, 1521 (2002).

460: 

461: \bibitem{perL} P. Dahlqvist, {\em Nonlinearity \bf 10}, 159 (1997).

462: 

463: \bibitem{ACL} R. Artuso, G. Casati and R. Lombardi, {\em Phys.Rev.Lett. \bf 71}, 62 (1993).

464: 

465: \bibitem{Fell} W. Feller, {\sf An introduction to probability theory and applications Vol. II},

466: (Wiley, New York, 1966).

467: 

468: \bibitem{PM} P. Manneville and Y. Pomeau, {\em Phys.Lett. \bf 75A}, 1 (1979); Y. Pomeau and

469: P. Manneville, {\em Commun.Math.Phys. \bf 74}, 189 (1980).

470: 

471: \bibitem{plapp} P. Gaspard and X.-J. Wang, {\em Proc.Natl.Acad.Sci. U.S.A. \bf 85}, 4591 (1988);

472: X.-J. Wang,  {\em Phys.Rev. \bf A40}, 6647 (1989).

473: 

474: %\bibitem{isolapl} S. Isola, {\sf Dynamical zeta functions for non-uniformly hyperbolic

475: %transformations}, preprint 1995 \texttt{mp-arc 95-142}

476: 

477: \bibitem{ci} M. Campanino and S. Isola, {\em Forum Math.  \bf 8}, 71  (1996).

478: 

479: \bibitem{acprep} R. Artuso and G. Cristadoro, {\em in preparation}.

480: 

481: \bibitem{L-Ott} D.N. Armstead, B.R. Hunt and E. Ott, {\em Phys.Rev. \bf E67}, 021110 (2003).

482: 

483: \bibitem{Ble} P.M. Bleher, {\em J.Stat.Phys. \bf 66}, 315 (1992).

484: 

485: %\bibitem{QuCh} G. Casati and B.V. Chirikov, {\sf Quantum chaos: between 

486: %order and disorder}, (Cambridge University Press, Cambridge, 1995) and references

487: %therein.

488: 

489: 

490: %\bibitem{ArRu} R. Artuso and M. Rusconi, {\em Phys.Rev. \bf E64}, 

491: %015204(R) (2001).

492: 

493: \end{thebibliography}

494: %\end{multicols}

495: \end{document}

496: