Mathematics
Fractal Dimension and the Hausdorff Dimension
Quick fact
The Hausdorff dimension of the Koch snowflake is log 4 / log 3 ≈ 1.2619, meaning it is 'more than a line but less than a plane'.
Why this is interesting
Have you ever wondered whether a coastline has a length? Measure it with a ruler and then with a smaller ruler—you'll get wildly different answers, and the pattern never stops.
Read the full explanation
Understanding Fractal Dimension and the Hausdorff Dimension
When we think of dimension, we usually imagine whole numbers: a line is 1D, a square is 2D, a cube is 3D. But consider the Koch curve: you can draw it without ever lifting your pen, yet it wiggles so much that it covers more area than a straight line but less than a filled rectangle. The idea of fractal dimension captures this 'in-between' complexity. A simple way to think about dimension is through scaling: if you enlarge a line by a factor of 2, its length doubles (factor 2^1). If you enlarge a square by 2, its area quadruples (2^2); a cube becomes 8 times as large (2^3). The exponent in each case is the dimension. For a fractal, that exponent is often a fraction. For the Koch curve, enlarging it by 3 makes it 4 times 'longer' (in a certain sense), so its dimension is log 4 / log 3 ≈ 1.26.
A deeper explanation
The Hausdorff dimension is defined using a mathematical tool called the Hausdorff measure, which generalises length, area, and volume. For a set S, we cover it with small 'balls' of diameter ε. The Hausdorff measure H^d(S) is, roughly, the limit as ε→0 of the sum of (diameter)^d for all the covering balls. For any set, there is a critical value of d—call it D—where H^d(S) is zero for dD and infinite for d<D. That critical value D is the Hausdorff dimension. This mechanism works because, when you scale a shape by a factor k, its d-dimensional 'size' scales by k^d. For self-similar fractals, like the Koch curve or the Cantor set, you can solve for the dimension directly: if a shape is made of N copies of itself, each scaled by a factor of 1/s, then the Hausdorff dimension is log N / log s. For the Cantor set, N=2 and s=3, giving log 2 / log 3 ≈ 0.63. The Hausdorff dimension is powerful because it is a rigorous, measure-theoretic definition that applies to any set—not just self-similar ones. It connects to measure theory and helps us analyse everything from Brownian motion paths to the structure of galaxies. Importantly, it is not the same as the topological dimension: the Cantor set is topologically 0-dimensional, but its Hausdorff dimension is about 0.63. This distinction is crucial in understanding the richness of fractal geometry.