A property where a structure or process repeats its pattern at different scales, like a snowflake.Self-similarity is a property of a system where a pattern or structure repeats itself at different scales. A classic example is a snowflake, where the branching pattern of the whole is mirrored in its smaller parts. In physics, self-similarity appears in phenomena ranging from turbulence and phase transitions to the large-scale structure of the universe. It implies that the same physical laws or processes govern behavior across a wide range of sizes, allowing scientists to scale results from one regime to another.
The concept gained prominence in the 20th century through the work of Benoît Mandelbrot, who coined the term 'fractal' in 1975, and earlier through the self-similar solutions of partial differential equations studied by G.I. Taylor and others in the 1940s.
Self-similarity often arises when a system evolves without a characteristic length scale. For instance, in a turbulent fluid, large eddies break into smaller ones, which break into even smaller ones, following a power-law cascade. Mathematically, this is expressed through scaling laws and renormalization group methods, where the system's properties remain invariant under a change of scale. In fractal geometry, self-similarity is quantified by a fractal dimension that describes how the pattern fills space.
Self-similarity is used in image compression algorithms, in modeling turbulent flows for engineering, and in analyzing the distribution of galaxies in cosmology.
The coastline of Britain is a famous example of self-similarity: its measured length increases as you use a smaller ruler, and it follows a fractal dimension of about 1.25.
No articles yet