Mathematics.

chaos and fractals

Fractal Geometry

Dynamical Systems55 minDifficulty6 out of 10

You should know: chaos theory, lorenz system

Overview

Fractal geometry studies objects with non-integer dimension, self-similar structure at all scales, and intricate detail that cannot be captured by classical Euclidean geometry. Fractals arise naturally as attractors of chaotic dynamical systems, boundaries of basins of attraction, and limit sets of iterated function systems. Key measures include Hausdorff dimension, box-counting dimension, and the relationship between dimension and chaos.

Intuition

Classical geometry asks: what is the length of a coastline? As you use a finer and finer ruler, you find more and more detail, and the measured length keeps growing. Fractals formalize this idea: a coastline is not 1-dimensional (a line) nor 2-dimensional (an area), but something in between -- perhaps dimension 1.25. Self-similarity means zooming in reveals copies of the whole: the Mandelbrot set is endlessly complex at every scale.

Formal Definition

Definition

The Hausdorff dimension of a set E is defined via s-dimensional Hausdorff measure: dim_H(E) = inf{s >= 0 : H^s(E) = 0}. The box-counting (Minkowski) dimension is: dim_B(E) = lim_{epsilon->0} log N(epsilon)/log(1/epsilon), where N(epsilon) is the number of boxes of size epsilon needed to cover E. For self-similar sets satisfying the open set condition, both coincide.

dim⁡H(E)=inf⁡{s≥0:Hs(E)=0}\dim_H(E) = \inf\{s \geq 0 : \mathcal{H}^s(E) = 0\}
Hausdorff dimension
dim⁡B(E)=lim⁡ε→0log⁡N(ε)log⁡(1/ε)\dim_B(E) = \lim_{\varepsilon \to 0} \frac{\log N(\varepsilon)}{\log(1/\varepsilon)}
Box-counting dimension
dim⁡H(Cantor set)=log⁡2log⁡3≈0.6309\dim_H(\text{Cantor set}) = \frac{\log 2}{\log 3} \approx 0.6309
Dimension of the Cantor set
dim⁡H(Koch snowflake boundary)=log⁡4log⁡3≈1.2619\dim_H(\text{Koch snowflake boundary}) = \frac{\log 4}{\log 3} \approx 1.2619
Dimension of the Koch curve
dim⁡H(Sierpinski triangle)=log⁡3log⁡2≈1.585\dim_H(\text{Sierpinski triangle}) = \frac{\log 3}{\log 2} \approx 1.585
Dimension of the Sierpinski triangle

Notation

NotationMeaning
dim⁡H\dim_HHausdorff dimension
dim⁡B\dim_BBox-counting (Minkowski) dimension
Hs\mathcal{H}^ss-dimensional Hausdorff measure
N(ε)N(\varepsilon)Number of boxes of size epsilon needed to cover the set

Theorems

Theorem 1: Dimension of the Cantor Set
Themiddle−thirdsCantorsetC(obtainedbyrepeatedlyremovingthemiddlethirdofintervals)hasHausdorffdimensionlog(2)/log(3).Ateachstage,2nintervalsoflength(1/3)narepresent;thedimensionfollowsfromthescalingN∗rd=1withN=2,r=1/3.The middle-thirds Cantor set C (obtained by repeatedly removing the middle third of intervals) has Hausdorff dimension log(2)/log(3). At each stage, 2^n intervals of length (1/3)^n are present; the dimension follows from the scaling N * r^d = 1 with N=2, r=1/3.
Theorem 2: Dimension of the Koch Curve
The Koch snowflake boundary is obtained by replacing each segment with four segments of length 1/3 of the original. So N=4, r=1/3, and dimension d = log(4)/log(3) approximately 1.2619. The curve is continuous everywhere but differentiable nowhere.
Theorem 3: Self-Similarity Dimension Formula
ForaniteratedfunctionsystemofNcontractionseachwithratior(satisfyingtheopensetcondition),theHausdorffdimensionsatisfiesN∗rd=1,givingd=log(N)/log(1/r).For an iterated function system of N contractions each with ratio r (satisfying the open set condition), the Hausdorff dimension satisfies N * r^d = 1, giving d = log(N)/log(1/r).
Theorem 4: Julia Sets and the Mandelbrot Set
Forthequadraticfamilyfc(z)=z2+c,theJuliasetJcistheboundaryofthesetofpointsthatdonotescapetoinfinityunderiteration.TheMandelbrotsetM=cinC:Jcisconnected.Botharefractals;theMandelbrotsethasHausdorffdimension2(provedbyShishikura,1998).For the quadratic family f_c(z) = z^2 + c, the Julia set J_c is the boundary of the set of points that do not escape to infinity under iteration. The Mandelbrot set M = {c in C : J_c is connected}. Both are fractals; the Mandelbrot set has Hausdorff dimension 2 (proved by Shishikura, 1998).

Worked Examples

  1. 1

    At stage n, the Cantor set is covered by 2^n intervals of length epsilon = (1/3)^n.

  2. 2

    So N(epsilon) = 2^n and epsilon = (1/3)^n.

    N(ε)=2n,ε=(1/3)nN(\varepsilon) = 2^n, \quad \varepsilon = (1/3)^n
  3. 3

    Compute the box-counting dimension:

    d=lim⁡n→∞log⁡2nlog⁡3n=nlog⁡2nlog⁡3=log⁡2log⁡3d = \lim_{n\to\infty} \frac{\log 2^n}{\log 3^n} = \frac{n \log 2}{n \log 3} = \frac{\log 2}{\log 3}

✓ Answer

dim_B(Cantor set) = log(2)/log(3) approximately 0.631.

Practice Problems

Mediumapplication

Compute the Hausdorff dimension of the Sierpinski triangle (each step replaces a triangle with 3 half-sized copies).

Mediumapplication

A self-similar fractal consists of 5 copies of itself, each scaled by factor 1/4. What is its Hausdorff dimension?

Mediumfree response

Explain why the boundary of the Mandelbrot set is considered a fractal, and what property characterizes points on it.

MediumMultiple choice

What is the Hausdorff dimension of the Koch snowflake curve (boundary only)?

Common Mistakes

Common Mistake

A fractal must be self-similar (exactly the same at every scale).

Exact self-similarity is a special case. Many fractals (like the Mandelbrot set boundary) are statistically self-similar or have approximate self-similarity under magnification, but not exact copies.

Common Mistake

Hausdorff dimension and topological dimension are the same thing.

Topological dimension is always an integer (0 for points, 1 for curves, 2 for surfaces). Hausdorff dimension can be any non-negative real number. For a fractal, Hausdorff dimension strictly exceeds topological dimension.

Quiz

The Hausdorff dimension of the Cantor set (middle thirds) is:
A fractal with Hausdorff dimension between 1 and 2 is best described as:
The Mandelbrot set M contains parameter c if and only if:
The Cantor set has Lebesgue measure zero because:

Historical Background

Benoit Mandelbrot coined the term 'fractal' in 1975, unifying previously isolated mathematical curiosities (Cantor set, Koch curve, Sierpinski triangle) under a common framework. However, these objects had been studied decades earlier: Georg Cantor described his set in 1883, Helge von Koch created his snowflake in 1904, and Waclaw Sierpinski his triangle in 1915. Mandelbrot's 1982 book 'The Fractal Geometry of Nature' demonstrated that fractals appear throughout nature in clouds, coastlines, ferns, and galaxy distributions.

  1. 1883

    Cantor describes the Cantor set (middle-thirds construction)

    Georg Cantor

  2. 1904

    Koch introduces the snowflake curve (continuous but nowhere differentiable)

    Helge von Koch

  3. 1915

    Sierpinski describes his triangle (gasket)

    Waclaw Sierpinski

  4. 1918

    Hausdorff introduces Hausdorff measure and fractional dimension

    Felix Hausdorff

  5. 1975

    Mandelbrot coins 'fractal' and demonstrates fractals in nature

    Benoit Mandelbrot

  6. 1979

    Mandelbrot discovers the Mandelbrot set during exploration of quadratic maps

    Benoit Mandelbrot

  7. 1982

    Mandelbrot publishes The Fractal Geometry of Nature

    Benoit Mandelbrot

Summary

  • Fractals are sets with non-integer Hausdorff dimension, self-similarity at multiple scales, and intricate structure.
  • Hausdorff dimension is defined via measure theory; box-counting dimension is computed as lim log N(eps)/log(1/eps).
  • Key examples: Cantor set (dim = log2/log3 ~ 0.63), Koch curve (dim = log4/log3 ~ 1.26), Sierpinski triangle (dim = log3/log2 ~ 1.58).
  • Julia sets and the Mandelbrot set arise from complex quadratic dynamics and are canonical examples of fractals in dynamical systems.

References

  1. BookMandelbrot, B. B. (1982). The Fractal Geometry of Nature. W. H. Freeman.
  2. BookFalconer, K. (2003). Fractal Geometry: Mathematical Foundations and Applications (2nd ed.). Wiley.
  3. WebsiteWeisstein, E. W. Fractal Dimension. MathWorld.