Fractal dimension describes how complicated shapes fill space as you zoom in. Many natural and mathematical shapes, such as coastlines, branching trees, lightning, and the Koch snowflake, are too irregular to be described well as only 1 dimensional lines or 2 dimensional areas. A fractal can have a dimension between whole numbers because its detail repeats across scales.
This idea matters because it gives a way to measure roughness, branching, and space filling structure.
One common method is box counting, where a grid of boxes is placed over the shape and the number of boxes touching the shape is counted. The grid size is then made smaller, and the count is repeated to see how quickly the number of needed boxes grows. If N is the number of boxes and s is the box side length, the fractal dimension is estimated from how N changes as s gets smaller.
A steeper growth rate means the shape fills space more strongly.
Key Facts
- A smooth line has dimension 1, a flat filled region has dimension 2, and a fractal curve can have dimension between 1 and 2.
- Box-counting dimension is estimated by D = log(N) / log(1/s) when a scaled copy needs N boxes of side length s.
- For repeated self-similar fractals, D = log(N) / log(k), where N is the number of smaller copies and k is the scale factor.
- For the Koch curve, N = 4 and k = 3, so D = log(4) / log(3) ≈ 1.262.
- A larger fractal dimension means the shape fills more of the surrounding space at small scales.
- Box-counting uses multiple grid sizes because one grid size alone cannot reveal scaling behavior.
Vocabulary
- Fractal
- A fractal is a shape with detailed structure that repeats or remains complex at many different scales.
- Fractal dimension
- Fractal dimension is a number that measures how strongly a shape fills space as it is viewed at smaller scales.
- Self-similarity
- Self-similarity means a shape contains smaller parts that resemble the whole shape.
- Box-counting
- Box-counting is a method that estimates fractal dimension by counting how many grid boxes contain part of the shape at different box sizes.
- Scale factor
- The scale factor is the number by which lengths are divided or multiplied when making smaller or larger copies of a shape.
Common Mistakes to Avoid
- Treating every fractal as exactly self-similar is wrong because many natural fractals only show approximate self-similarity over a limited range of scales.
- Using only one grid size in box counting is wrong because fractal dimension depends on the pattern of change across several scales.
- Thinking a dimension of 1.3 means the shape is physically 1.3 dimensional is wrong because dimension here describes scaling behavior, not a separate direction in space.
- Confusing perimeter with fractal dimension is wrong because two shapes can have long or even infinite perimeter but different rates of space filling.
Practice Questions
- 1 A self-similar fractal is made of 5 smaller copies, each scaled down by a factor of 3. Find its fractal dimension using D = log(N) / log(k).
- 2 In a box-counting experiment, a curve touches 18 boxes when the box side length is 1/4 of the original length. Estimate D using D = log(N) / log(1/s).
- 3 Explain why a Koch curve has a dimension greater than 1 but less than 2, using the idea of how it fills space as you zoom in.