Abstract
A crucial distinction exists between the dissipative systems and conservative Hamiltonian systems. By Liouville’s Theorem, the solution flow for a conservative Hamiltonian system preserves volumes in phase space. Dissipative systems, by contrast, usually give rise to solution flows which contract volumes in phase space. This volume contraction gives rise to a bounded set called an attractor in the phase space, which ultimately contains the solution flow, once the transients associated with initial conditions have died away. Indeed, dissipative systems of dimensions equal to and greater than three can have bounded trajectories, which may be attracted not by a fixed point nor by a periodic/quasi-periodic orbit, but by an object of complicated infinitely many-layered structure called a strange attractor. The trajectories on this attractor diverge continually from each other locally, but remain bounded globally. Besides, the evolution on the attractor is essentially aperiodic. Strange attractors are sometimes modeled by fractals which are geometric objects that are very suited for characterizing roughness and illustrating the difference between the mathematical properties of continuity and differentiability. Adoption of fractal geometry therefore releases one from bondage to smooth surfaces and smooth curves and enables one to come to terms with nature, for as Mandelbrot (The Fractal Geometry of Nature, Freeman, 1983) put it—“clouds are not spheres, mountains are not (cones), and bark is not smooth, nor does lightning travel in a straight line”. (Indeed, the fractal structure of a natural object is closely intertwined with its functionality (like the light-catching capacity of trees or the electrical connectivity of neurons).) Further, fractals are self-similar objects which have the same shape at all scales.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Indeed, the fractal structure of a natural object is closely intertwined with its functionality (like the light-catching capacity of trees or the electrical connectivity of neurons).
- 2.
The existence of an attractor presupposes some kind of minimality property reflecting underlying dynamics and is therefore associated with a contractive mapping on a complete metric space:
$$ a_{n+1}=f(a_n) $$where a 0,a 1,a 2,… is a bounded, monotone sequence of elements from a complete metric space X (which possesses a metric d:X×X→R with the properties: d(x,y)≥0,d(x,y)=0, if and only if x=y,d(x,y)=d(y,x) and d(x,y)≤d(x,z)+d(z,y)—the triangle inequality, ∀(x,y,z)∈X) so that
$$ d\bigl(f(x),f(y)\bigr)\leq c d(x,y),\quad 0\leq c,\ \forall(x,y)\in X, $$(Peitgen et al. 2004). A sequence \(\{a_{n}\}^{\infty}_{n=1}\) of elements in a metric space X is said to converge to a point a∈X if, for a given number ϵ>0, there is an integer N>0 such that, d(a n ,a)<ϵ,∀n>N. The point a∈X is called the limit of the sequence, a=lim n→∞ a n . A metric space X is compact, if every infinite sequence \(\{ a_{n}\}^{\infty}_{n=1}\) in X has a limit a∈X. The uniqueness and invariance of the attractor (or the fixed point of the above contractive mapping) then implies
$$ a=\lim_{n\rightarrow\infty}~ a_n,\quad f(a)=a. $$The contractive nature of the map also implies that
$$ \frac{d(b,a)}{d(b,f(b))}\geq 1,\quad \forall b\in X. $$However, it turns out that this ratio is also bounded from above (Barnsley 1988) because, by noting that the function d(x,y), for fixed x∈X, is continuous in y∈X, we have
$$\begin{aligned} d(b,a) =& d\Bigl( b,\lim_{n\rightarrow\infty} f^{[n]}(b)\Bigr) \\ {}=&\lim_{n\rightarrow\infty} d \bigl( b, f^{[n]}(b)\bigr) \\ {}\leq&\lim_{n\rightarrow\infty} \displaystyle\sum ^n_{m=1}d \bigl(f^{[m-1]}(b),f^{[m]}(b) \bigr),\quad \mbox{by the triangle inequality} \\ {}\leq&\lim_{n\rightarrow\infty} d \bigl( b,f(b)\bigr) \bigl( 1+c+c^2+ \cdots+c^{n-1}\bigr) \\ {}=&(1-c)^{-1}d \bigl(b,f(b) \bigr). \end{aligned}$$So,
$$ \frac{d(b,a)}{d(b,f(b))}\leq(1-c)^{-1},\quad 0\leq c<1. $$ - 3.
The well known Newton’s method for finding zeros of a polynomial g(x) can be viewed as a dynamical system with attracting fixed points corresponding to the zeros of g(x):
$$ a_{k+1}=N(a_k);\quad k=0,1,2,\ldots $$where,
$$ N(a)\equiv a-\frac{g(a)}{g^\prime(a)} $$implying that, if the sequence a 0,a 1,a 2,… converges to some number, then the limit is a zero of g(x). Thus, trajectories from an arbitrary starting point lead to one of these fixed points. From this point of view, the usefulness of Newton’s method is determined by the size of the basins of attraction of the fixed points (a 0 being taken from inside one of these basins of attraction) (Eubank and Farmer 1989).
- 4.
A structure is said to be scale-invariant if its characteristic measures have the following property:
$$ f(b r) \simeq b^{-\alpha} f(r) $$It can be shown that the only function which does not introduce a characteristic scale in the problem and satisfies this property is the one possessing a power-law behavior.
$$ f(r)\sim r^{-\alpha}. $$ - 5.
One picks up more and more of the fragmentation of the object as one increases the resolution of the measurement.
- 6.
It is of interest to note that Richardson (1961) even suggested an index (which is just the fractal dimension) as a measure of the irregularity of the object.
- 7.
For non-uniform fractals, like those in nature, for which the reduction factor r is not identical for all of the N offsprings at a given stage of the fragmentation process, the formula (6.10) can be generalized as follows: The number of boxes needed to cover the complete fractal now is
$$ N(\varepsilon)=\displaystyle\sum^{N(l)}_{i=1}N_i( \varepsilon) $$where, by self-similarity of the structure, the number of boxes on a grid of size ε needed to cover the ith part is given by
$$ N_i(\varepsilon)=N(\varepsilon/r_i). $$If N(l) scales according to
$$ N(l)\sim l^{-D} $$we then have
$$ \sum^{N}_{i=1}r^D_i=1, $$which, for a uniform fractal (with r i =r∼l,i=1,2,…,N), reduces to (6.10)! Observe that the fractal dimension D now appears as a kind of statistical average.
- 8.
The probability density on a Cantor set would correspond to uncountably many delta functions!
- 9.
The Koch curve is an example of a continuous curve which is generated by a non-differentiable function, i.e., it does not possess a tangent at any point on it because, it has a corner almost everywhere (a situation contemplated by Karl Weierstrass already in 1872!).
- 10.
Suppose the initial equilateral triangle has sides of length a. Then, the area of this triangle is \(A_{1}=\frac{\sqrt{3}}{4}a^{2}\). At the first step, n 1=3 triangles of area \(\frac{1}{3^{2}}A_{1}\) are added. At the kth step, n k =3⬝4k−1 triangles of area (1/3)2k A 1 are added. Thus,
$$ A_{k+1}=A_k+3\centerdot 4^{k-1}\biggl( \frac{1}{3}\biggr)^{2k}A_1. $$On iterating this relation, we obtain
$$ A_\infty=A_1+\frac{1}{3}\biggl(1+ \frac{4}{9}+\frac{4^2}{9^2}+\cdots\biggr) A_1= A_1+\frac{3}{5} A_1 = \frac{2}{5} \sqrt{3}a^2. $$ - 11.
f(α) attains practical significance because it can be determined directly in physical experiments (see Chap. 9). However, f(α) can only be determined approximately because, experiments allow a finite number of observed scales while numerical computations allow a finite resolution.
- 12.
Kaplan and Yorke (1979) have conjectured that there is a relationship between the information entropy dimension D 1 and the Liapunov exponents σ j for a typical chaotic attractor (which appears plausible because σ j ’s are associated with unpredictability—see Sect. 5.8.1):
$$ D_1=k+\frac{1}{|\sigma_{k+1}|}\sum ^k_{j=1}\sigma_j $$where k is the largest integer such that
$$ \sum^k_{j=1} \sigma_j\geq0 $$with the ordering
$$ \sigma_1\geq\sigma_2\geq\cdots. $$The Kaplan-Yorke formula implies that k≤D 1≤k+1.
- 13.
In order to prove this, note that for a set of positive numbers a 1,…,a N , we have
$$ \prod^{N(l)}_{i=1}a_i^{p_i} \leq\sum^{N(l)}_{i=1}p_i a_i. $$Putting a i =1/p i , we obtain
$$ \prod^{N(l)}_{i=1}\biggl(\frac{1}{p_i} \biggr)^{p_{i}}\leq \sum^{N(l)}_{i=1}p_i \frac{1}{p_i}=N(l). $$On taking the logarithm on both sides, we obtain
$$ S(l)=\sum^{N(l)}_{i=1}p_i \ln \biggl( \frac{1}{p_i}\biggr)\leq \ln N(l). $$ - 14.
- 15.
Time series data refers to sequences of data representing the time evolution of a variable of a system, sampled at fixed finite time intervals.
- 16.
Grassberger and Proccacia (1983) have shown that the correlation dimension ν represents a lower limit for both the information dimension D 1 and the capacity dimension D 0, i.e.,
$$ \nu\leq D_1 \leq D_0. $$ - 17.
This property can also serve to distinguish a chaotic dynamics from noise, since, for the latter case, ν will continue to increase indefinitely, as m increases without any saturation. The latter is due to the fact that a noise of infinite sample size embedded in an m-dimensional space always fills that space.
- 18.
Thermal convection is the main cause for the circulations of atmosphere and oceans. The latter, in turn, determine the weather changes as well as the continental drift.
- 19.
In order to avoid the appearance of complex spatial structures, experiments are usually performed in a small cell. The boundary conditions limit the number of rolls. The observed dynamical behavior depends sensitively on the liquid chosen and on the linear dimensions of the box.
- 20.
However, the Lorenz equations do a good job of modeling convective flow in a closed circular tube.
- 21.
A supercritical Hopf bifurcation (see Chap. 2), by contrast, typically involves stable limit cycles growing from unstable foci.
- 22.
Such a dimensionality reduction is produced by disparate time scales present in the system (Nicolis 1995). Fast processes eliminate evolution in certain directions while a slow evolution along the most unstable direction persists as time goes on.
- 23.
In order to see this, as we will see below, note that the period-1 cycle bifurcates into a period-2 cycle when
$$ \frac{df_\lambda}{dx}\bigg|_{x=x^\ast}=-1,\qquad \frac{df_\lambda ^{[2]}}{dx} \bigg|_{x=x^\ast}=1 $$and
$$ \frac{d}{dx}\biggl(\frac{df_\lambda^{[2]}}{dx}\biggr)\bigg|_{x=x^\ast}=0\quad \text{and}\quad\frac{d^2}{dx^2}\biggl(\frac{df_\lambda^{[2]}}{dx}\biggr)\bigg|_{x=x^\ast}<0 $$where x=x ∗≠0 is a fixed point of f λ .
Noting that,
$$ \frac{d^3f^{[2]}_\lambda}{dx^3}=f^{\prime\prime\prime}_\lambda (x)f^\prime_\lambda \bigl(f_\lambda(x) \bigr)+3f^{\prime\prime}_\lambda(x)f^{\prime \prime}_\lambda \bigl(f_\lambda(x) \bigr)f^\prime_\lambda(x)+f^{\prime\prime\prime }_\lambda \bigl(f_\lambda(x) \bigr) \bigl(f^\prime_\lambda(x) \bigr)^3 $$we see that
$$\begin{aligned} &\frac{d^2}{dx^2}\biggl(\frac {df^{[2]}_\lambda}{dx}\biggr)\bigg|_{x=x^\ast} \\ &\quad{}=-2\bigl( f^\prime_\lambda\bigl( x^\ast\bigr) \bigr)^3\biggl[-\biggl(\frac {1+( f^\prime_\lambda( x^\ast))^2}{2f^\prime_\lambda( x^\ast )}\biggr) \biggl( \frac{f^{\prime\prime\prime}_\lambda( x^\ast) }{f^\prime_\lambda( x^\ast)}-\frac{3}{2} \biggl(\frac{f^{\prime\prime}_\lambda( x^\ast)}{f^\prime_\lambda( x^\ast)}\biggr)^2 \biggr)\biggr]<0 \end{aligned}$$implies, since \(f^{\prime}_{\lambda}( x^{\ast})=-1\), that (Singer 1978)
$$ \displaystyle\frac{f^{\prime\prime\prime}_\lambda( x^\ast) }{f^\prime_\lambda( x^\ast)}-\displaystyle\frac{3}{2} \biggl( \frac{f^{\prime\prime}_\lambda( x^\ast)}{f^\prime_\lambda( x^\ast)}\biggr)^2<0. $$ - 24.
This is provided for by the Fixed-Point Theorem: If f λ :R→R is continuous and there exists a closed interval I such that I⊃f λ (I), then there exists at least one point x∈I, such that f λ (x)=x.”
- 25.
For λ<1/4, it can readily be shown that \(\{f^{[n]}_{\lambda}( x)\}^{\infty}_{n=0}\) forms a bounded decreasing sequence converging to the fixed point \(\overline{x}=0\).
- 26.
For 1/4<λ<3/4, it can be shown that \(\{f^{[n]}_{\lambda}( x)\}^{\infty}_{n=0}\) forms a bounded increasing sequence if \(0<x<\overline{x}\), and a bounded decreasing sequence if \(\overline{x}<x<1\), both of which converge to the fixed point \(\overline{x}=1-1/4\lambda\).
- 27.
In other words, the sequence (λ 1,λ 2,…) rapidly approaches geometric progression. A geometric progression is a sequence of numbers which converges to a limit point in such a way that the ratio of successive intervals is a constant. So, the sequence of partial sums \(\{s_{n}\}_{n=1}^{\infty}\) of the series \(\sum^{\infty}_{k=1}a_{k}\) is a geometric progression if
$$ \lim_{n\rightarrow\infty} \frac {s_{n+1}-s_n}{s_{n+2}-s_{n+1}}=\mathrm{const}\quad \text{or}\quad\lim_{n\rightarrow\infty}\frac{a_{n+1}}{a_{n+2}}=\mathrm{const} $$i.e., \(\sum^{\infty}_{k=1}a_{k}\) tends to a geometric series as n→∞!
- 28.
In order to give an approximate analytical evaluation of this result, let us do the shifting operation (Hu 1982; Landau and Lifshitz 1987),
$$ x\Rightarrow\frac{x}{2}(2\lambda-1)+\frac{1}{2} $$(I)so that the map (6.93) transforms to
$$ x_{n+1}=1-\varLambda x^2_n $$(II)where,
$$ \varLambda\equiv2 \lambda(2 \lambda-1). $$The map (II) leads to −1<x<1, if 0<Λ<2. Further, the map (II) has a fixed point x ∗ given by
$$ \overline{x}=1-\varLambda\overline{x}^2 $$(III)which becomes unstable and leads to the first bifurcation, when
$$ 2\varLambda\overline{x}=1 $$(IV)i.e., when Λ=3/4 with \(\overline{x}=2/3\).
Next, iterating the map (II), we have
$$ x_{n+2}=1-\varLambda+2\varLambda^2 x_n^2 -\varLambda^3 x^4_n. $$(V)Neglecting the quartic term and rescaling x n , according to x n ⇒x n (1−Λ), (V) becomes
$$ x_{n+2}=1-\varLambda_1 x^2_n $$(VI)where,
$$ \varLambda_1\equiv2\varLambda^2(\varLambda-1 )=\psi(\varLambda),\quad\text{say}. $$Observe that (VI) is of the same form as (II), so becomes unstable when Λ 1=3/4 and leads to the next bifurcation. The succeeding bifurcations occur at Λ 2=ψ(Λ 1)=1.2428, Λ 3=ψ(Λ 2)=1.3440, Λ 4=ψ(Λ 3)=1.3622,… finally accumulating at
$$ \varLambda_\infty=\psi(\varLambda_\infty) $$(VII)giving
$$ \varLambda_\infty=\frac{1+\sqrt{3}}{2}=1.3660 $$(VIIIa)or
$$ \lambda_\infty=0.8856 $$(VIIIb)which is quite close to the exact value λ ∞=0.8924….
Further, observe that after each bifurcation, the mapping interval, which is initially [−1,1], shrinks by a factor (1−Λ). Therefore, after several bifurcations, the iterated map is sensitive only to the properties of the original map near its maximum, thus indicating a universality in the bifurcating sequence!
- 29.
Indeed, if one identifies the period of the period-n cycle by
$$ \tau(\varLambda_n)\equiv2^n $$(I)then we have
$$ \tau(\varLambda_{n+1})=2\tau(\varLambda_n). $$(II)The divergence of the period of the period-n cycle, as n→∞, may then be characterized by a critical exponent á la Landau (1937) via
$$ \lim_{n\rightarrow\infty}\tau(\varLambda_n) \sim(\varLambda_\infty-\varLambda_n)^{-\nu}. $$(III)Approximating,
$$ \lim_{n\rightarrow\infty} \varLambda_n=\lim_{n \rightarrow \infty} 2\lambda_n(2\lambda_n -1) \approx 2\lambda_\infty( 2\lambda_n -1) $$(IV)we have
$$ \lim_{n\rightarrow\infty}\frac{(\varLambda_\infty-\varLambda_n) ^{-\nu}}{(\varLambda_\infty-\varLambda_{n+1})^{-\nu}}\approx\lim_{n\rightarrow\infty} \biggl(\frac{\lambda_\infty-\lambda_n}{\lambda _\infty-\lambda_{n+1}}\biggr)^{-\nu}= \delta^{-\nu} $$(V)from which, on using (III),
$$ \lim_{n\rightarrow\infty}\frac{\tau(\varLambda_n)}{ \tau(\varLambda_{n+1})}\approx \delta^{-\nu}. $$(VI)$$ 1/2 \approx\delta^{-\nu} $$or
$$ \nu\approx\frac{\ln 2}{\ln\delta}\approx 0.42\ldots $$(VII) - 30.
It should be noted that the computer cannot resolve whether the iterates describe a stable cycle of very high period or they are in fact dense on the interval [0,1].
- 31.
In fact, for λ=1, the logistic map is equivalent to the Bernoulli shift map (see Exercise 6.4).
- 32.
λ ∗ may hence determined from the condition
$$ f^{[3]}_\lambda\biggl(\frac{1}{2}\biggr)= \frac{1}{2} $$which leads to
$$ \lambda=\lambda_\ast=\frac{1}{4}(1+\sqrt{8})=0.9571. $$ - 33.
The phenomenon of intermittency appears to be a common feature in chaotic systems because, there are infinitely many periodic windows arising from tangent bifurcations in the chaotic regions. Besides, it is one of the characteristic features of fully-developed turbulence (see Chap. 9).
- 34.
The rational numbers lie densely on any interval of finite length.
- 35.
μ (k)=−1 corresponds to the linearized flow x n+1=−x n which represents a period-2k cycle because x n+2=x n .
References
H.D.I. Abarbanel, Physics of chaotic systems, in The New Physics for the Twenty-First Century, ed. by G. Fraser (Cambridge University Press, Cambridge, 2005)
H.D.I. Abarbanel, R. Brown, J.J. Sidorowich, L.Sh. Tsimring, The analysis of observed chaotic data in physical systems. Rev. Mod. Phys. 65, 1331 (1993)
J.C. Alexander, J.A. Yorke, Z. You, I. Kan, Riddled basins. Int. J. Bifurc. Chaos 2, 795 (1992)
M. Barnsley, Fractals Everywhere (Academic Press, San Diego, 1988)
P. Bergé, Y. Pomeau, C. Vidal, Order Within Chaos: Towards a Deterministic Approach to Turbulence (Wiley, New York, 1984)
P. Collet, J.P. Eckmann, Iterated Maps of the Unit Interval as Dynamical Systems (Birkhäuser, Basel, 1980)
R.L. Devaney, An Introduction to Chaotic Dynamical Systems (Addison-Wesley, Reading, 1989)
P.G. Drazin, Nonlinear Systems (Cambridge University Press, Cambridge, 1992)
J.P. Eckmann, Roads to turbulence in dissipative dynamical systems. Rev. Mod. Phys. 53, 643 (1981)
S. Eubank, D. Farmer, An introduction to chaos and randomness, in Lectures in Complex Systems, ed. by E. Jen (Addison-Wesley, Reading, 1989)
K. Falconer, Fractal Geometry: Mathematical Foundations and Applications, 2nd edn. (Wiley, New York, 2003)
J.D. Farmer, Information dimension and the probabilistic structure of chaos. Z. Naturforsch. A 37, 1304 (1982)
J.D. Farmer, E. Ott, J.A. Yorke, The dimension of chaotic attractors. Physica D 7, 153 (1983)
M.J. Feigenbaum, Quantitative universality for a class of nonlinear transformations. J. Stat. Phys. 19, 25 (1978)
P. Grassberger, I. Proccacia, Characterization of strange attractors. Phys. Rev. Lett. 50, 346 (1983)
J.M. Greene, R.S. MacKay, F. Vivaldi, M.J. Feigenbaum, Universal behavior of area-preserving maps. Physica D 3, 468 (1981)
S. Grossmann, S. Thomae, Invariant distributions and stationary correlation functions of one-dimensional discrete processes. Z. Naturforsch. A 32, 1353 (1977)
D. Gulick, Encounters with Chaos (McGraw-Hill, New York, 1992)
T.C. Halsey, M.H. Jensen, L.P. Kadanoff, I. Proccacia, B.I. Shraiman, Fractal measures and their singularities: characterization of strange sets. Phys. Rev. A 33, 1141 (1986)
H.G.E. Hentschel, I. Proccacia, The infinite number of generalized dimensions of fractals and strange attractors. Physica D 8, 435 (1983)
D. Holton, R.M. May, Chaos and one-dimensional maps, in The Nature of Chaos, ed. by T. Mullin (Clarendon, Oxford, 1993)
B.N. Hu, Introduction to real-space renormalization-group methods in critical and chaotic phenomena. Phys. Rep. 91, 233 (1982)
E.A. Jackson, Perspectives in Nonlinear Dynamics, vols. I and II (Cambridge University Press, Cambridge, 1991)
J.L. Kaplan, J.A. Yorke, Chaotic behavior of multi-dimensional difference equations, in Functional Differential Equations and Approximations of Fixed Points, ed. by H.-O. Peitgen, H.O. Walter. Lecture Notes in Mathematics, vol. 730 (Springer, Berlin, 1979)
L.D. Landau, Theory of phase transitions. Fiz. Sowjetunion 11, 26 (1937)
L.D. Landau, On the nature of turbulence. Sov. Phys. Dokl. 44, 139 (1944)
L.D. Landau, E.M. Lifshitz, Fluid Mechanics (Pergamon, Elmsford, 1987)
A. Libchaber, S. Fauve, C. Larouche, Two parameter study of routes to chaos. Physica D 7, 73 (1983)
E.N. Lorenz, Deterministic nonperiodic flow. J. Atmos. Sci. 20, 130 (1963)
S.K. Ma, Modern Theory of Critical Phenomena (Benjamin/Cummings, Redwood City, 1976)
B. Mandelbrot, Intermittent turbulence in self-similar cascades: divergence of high moments and dimension of the carrier. J. Fluid Mech. 62, 331 (1974)
B. Mandelbrot, The Fractal Geometry of Nature (Freeman, New York, 1983)
B. Mandelbrot, Multifractal measures, in Fractals in Geophysics, ed. by C.H. Scholz, B. Mandelbrot (Birkhäuser, Basel, 1989)
P. Manneville, Y. Pomeau, Intermittency and the Lorenz model. Phys. Lett. A 75, 1 (1979)
R.M. May, Simple mathematical models with very complicated dynamics. Nature 261, 459 (1976)
S.W. McDonald, C. Grebogi, E. Ott, J.A. Yorke, Fractal basin boundaries. Physica D 17, 125 (1985)
J.B. McLaughlin, P.C. Martin, Transition to turbulence in a statically stressed fluid system. Phys. Rev. A 12, 186 (1975)
C. Meneveau, K.R. Sreenivasan, The multifractal nature of turbulent energy dissipation. J. Fluid Mech. 224, 429 (1991)
P.J. Myrberg, Iteration von Quadratwurzeloperationen. Ann. Acad. Sci. Fenn., Ser. A 1 Math. 259, 1 (1958)
A.H. Nayfeh, B. Balachandran, Applied Nonlinear Dynamics (Wiley Interscience, New York, 1995)
G. Nicolis, Introduction to Nonlinear Science (Cambridge University Press, Cambridge, 1995)
V.D. Oseledec, A multiplicative ergodic theorem: Liapunov characteristic numbers for dynamical systems. Trans. Mosc. Math. Soc. 19, 197 (1968)
E. Ott, Strange attractors and chaotic motions of dynamical systems. Rev. Mod. Phys. 53, 655 (1981)
E. Ott, Chaos in Dynamical Systems, 2nd edn. (Cambridge University Press, Cambridge, 2002)
H.-O. Peitgen, H. Jürgens, D. Saupe, Chaos and Fractals, 2nd edn. (Springer, Berlin, 2004)
Y. Pomeau, P. Manneville, Intermittent transition to turbulence in dissipative dynamical systems. Commun. Math. Phys. 74, 189 (1980)
J.W.S. Rayleigh, On convection currents in a horizontal layer of fluid, when the higher temperature is on the underside. Philos. Mag. 32, 529 (1916)
L.F. Richardson, The problem of contiguity: an appendix of statistics of deadly quarrels. Gen. Syst. Yearbook 6, 139 (1961)
G. Rowlands, An introduction to the properties of one-dimensional difference equations, in Order and Chaos in Nonlinear Physical Systems, ed. by S. Lundquist, N.H. March, M.P. Tosi (Plenum, New York, 1988)
B. Saltzman, Finite-amplitude free convection as an initial-value problem. J. Atmos. Sci. 19, 329 (1962)
R.S. Shaw, Strange attractors, chaotic behavior and information flow. Z. Naturforsch. A 36, 80 (1981)
B.K. Shivamoggi, Multi-fractal formulation of compressible fully developed turbulence: parabolic-profile approximation for the singularity spectrum. Ann. Phys. 322, 967 (2007)
D. Singer, Stable orbits and bifurcation maps of the unit interval. SIAM J. Appl. Math. 35, 260 (1978)
S. Smale, Differentiable dynamical systems. Bull. Am. Math. Soc. 73, 747 (1967)
C. Sparrow, The Lorenz Equations: Bifurcations, Chaos and Strange Attractors (Springer, Berlin, 1982)
H.E. Stanley, P. Meakin, Multifractal phenomena in physics and chemistry. Nature 335, 405 (1988)
F. Takens, Detecting Strange Attractors in Turbulence. Lecture Notes in Mathematics, vol. 898 (Springer, Berlin, 1981)
T. Tél, Fractals, multifractals and thermodynamics. Z. Naturforsch. A 43, 1154 (1988)
A.A. Tsonis, Chaos (Plenum, New York, 1992)
D.V. Vassiliades, A.S. Sharma, T.E. Eastman, K. Papadopoulos, Low-dimensional chaos in magnetospheric activity from AE time series. Geophys. Res. Lett. 17, 1841 (1990)
T. Vicsek, Fractal Growth Phenomena (World Scientific, Singapore, 1992)
J. von Neumann, Various techniques used in connection with random digits. J. Res. Natl. Bur. Stand., Appl. Math. Ser. 3, 36 (1951)
Author information
Authors and Affiliations
Appendices
Appendix A: The Hausdorff-Besicovitch Dimension
Let us consider a measure of the size of a set of points, S, in space, corresponding to a resolution of the measurement δ. We take a function h(δ)≡γ(d)δ d, which may represent a line, square, disk, ball or cube, and cover the set S to form the measure:
Here, the geometrical factor γ(d)=1 for lines, squares and cubes; γ=π/4 for disks and γ=π/6 for spheres. In the limit δ→0, the measure M d (S) turns out to be zero or infinite depending on the choice of d—the dimension of M d (S). (Note that for the Koch curve, the 1-dimensional measure, namely, the length, is diverging, while the 2-dimensional measure, namely, the area is 0.) The Hausdorff-Besicovitch dimension D(S) of the set S is the critical dimension for which the measure M d (S) jumps from zero to infinity:
or
which implies N(δ)∼δ −D(S), or that the measure is finite, when d=D(S). Note that this definition makes the Hausdorff-Besicovitch dimension a local property in the sense that it measures properties of sets of points in the limit of a vanishing parameter or size δ of the test function used to cover the set.
Some fundamental properties of the Hausdorff-Besicovitch dimension may be noted:
-
(i)
if S⊂R n, then D(S)≤n;
-
(ii)
if S⊂U, then D(S)≤D(U);
-
(iii)
if S⊂R n, and D(S)<1, then S is totally disconnected.
Equation (6.136) also implies that an object may be called a fractal, if it is not possible to obtain a finite measure for its volume, surface or length when measuring with d,d−1,… dimensional hypercubes of size ε and changing ε over several orders of magnitude. Thus, the length estimate L(δ) of a fractal curve varies with the resolution δ as L(δ)∼δ 1−D, where D is the dimension of the curve; D is clearly equal to unity for classical curves (in the limit of small δ) and is larger than unity for fractal curves.
Example 6.9
(The Cantor set)
Let A be the cover of the Cantor set consisting of 2n intervals of length 1/3′′ each, for n=1,2,…. Then, we have from (6.136),
from which, the Hausdorff-Besicovitch dimension D is
in agreement with the capacity dimension derived in Example 6.2.
Note that, for a given set, the Hausdorff-Besicovitch dimension need not be the same as the capacity dimension because, the latter does not distinguish between a set and its closure (Falconer 2003) and hence leads to counter-intuitive results. For example, the set of rational numbers is a countable union of points (of measure 0) so that its Hausdorff-Besicovitch dimension is 0 while its capacity dimension is 1 because its closure is R.Footnote 34 In general, the Hausdorff-Besicovitch dimension is smaller than (or equal to) the capacity dimension (Farmer et al. 1983).
Appendix B: The Derivation of Lorenz’s Equation
Consider a fluid layer of depth H, with the lower surface maintained at a higher temperature T 0+ΔT than the temperature T 0 of the upper surface. The equation governing the evolution of vorticity in this layer, with the inclusion of buoyancy effects is
where, the flow has been assumed to depend only on x (the horizontal direction), z (the vertical direction) and t. If v=(u,v) is the fluid velocity, the stream function ψ is defined by,
θ is the departure of the temperature profile from the conduction situation in the fluid,
g is the acceleration due to gravity, α is the coefficient of thermal expansion and ν is the kinematic viscosity of the fluid.
The equation governing the evolution of temperature in the fluid layer is
where κ is the thermal diffusivity of the fluid.
As the temperature difference across the fluid layer ΔT is increased, the heat transport process via the conduction mechanism becomes unstable, and gives way to the transport process via the convection mechanism involving a steady cellular flow. (The latter arises due to the fact that the fluid near the warmer lower surface expands and rises because of the buoyancy, while the cooler and heavier fluid near the upper surface descends under gravity.) If ΔT is increased further, the convection process shows a time-dependent behavior. In this regime, taking clues from the previous numerical work of Saltzman (1962), Lorenz (1963) considered the dynamics to be primarily described by three Fourier modes,
where R is the Rayleigh number,
and R c is the value of R, for which conduction gives way to convection. Upon substituting (6.141), equations (6.137) and (6.140) give Lorenz’s equations,
where the dot denotes derivative with respect to a non-dimensionalized time \(\tau\equiv\frac{\pi^{2}( 1+a^{2})\kappa t}{H^{2}}\), and
Appendix C: The Derivation of Universality for One-Dimensional Maps
We will give here an intuitive derivation of universality for one-dimensional maps which have a single, locally quadratic maximum (the following formulation is adapted from Holton and May (1993)).
Consider the composite map
Let \(\overline{x}^{(k)}\) be the fixed point of this map; \(\overline {x}^{(k)}\) is therefore a period-k cycle. Let,
So, a stable period-k cycle appears at μ (k)=1, which becomes unstable, giving birth to a period-doubled period-2k cycle, at μ (k)=−1,Footnote 35 with μ (2k)=1. The latter then becomes unstable when μ (2k)=−1, and μ (k) takes some negative value \(\mu^{(k)}_{c}\), say.
Let \(\lambda^{(k)}_{0}\) be the value of the parameter λ, at which the period-k cycle appears, i.e.,
Writing,
a Taylor expansion then gives
where
Similarly, let \(\lambda^{(2k)}_{0}\) be the value of λ at which the period-k cycle disappears, (and period-2k cycle appears), i.e.,
Writing,
a Taylor expansion then gives
where,
If Δλ(k) denotes the change in λ, as μ (k) varies from 1 to −1, then we have from (6.147),
Similarly, if Δλ(2k) denotes the change in λ, as μ (2k) varies from 1 to −1 or μ (k) varies from −1 to \(\mu^{(k)}_{c}\), then we have from (6.150),
Using (6.151) and (6.152), the Feigenbaum number δ (see (6.106) and (6.107)) is given by
In order to determine \(\mu^{(k)}_{c}\), we need an asymptotic relation between μ (2k) and μ (k). For this purpose, we approximate the map \(f^{[2k]}_{\lambda}\) by using a cubic:
The cubic approximation is necessary to enable \(f^{[2k]}_{\lambda}\) to exhibit a bifurcation into a pair of points (corresponding to a period-2k cycle), as the period-k cycle becomes unstable.
Noting that
we have, on comparison with (6.154),
where,
If we write for the fixed points \(\overline{x}^{(2k)}\) of the map \(f^{[2k]}_{\lambda}\),
then we have
or
On discarding the degenerate period-k solution (ξ ∗=0), (6.158) yields
Next, we have from (6.154),
Using (6.159), (6.160) becomes
Dropping the second term on the right hand side in (6.161) in comparison with the first, we have corresponding to μ (2k)=−1 or \(\mu^{(k)}=\mu^{(k)}_{c}\),
or
Using (6.162), we have from (6.156),
Using (6.163), we have from (6.153),
which differs from the exact numerical result (equations (6.106) and (6.107)) by about 3 %!
Rights and permissions
Copyright information
© 2014 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Shivamoggi, B.K. (2014). Chaos in Dissipative Systems. In: Nonlinear Dynamics and Chaotic Phenomena: An Introduction. Fluid Mechanics and Its Applications, vol 103. Springer, Dordrecht. https://doi.org/10.1007/978-94-007-7094-2_6
Download citation
DOI: https://doi.org/10.1007/978-94-007-7094-2_6
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-007-7093-5
Online ISBN: 978-94-007-7094-2
eBook Packages: Physics and AstronomyPhysics and Astronomy (R0)