Gray-Scott Reaction-Diffusion
Watch Turing patterns self-organise from chemical noise — the same mathematics that creates animal stripes, coral branches, and snowflake arms.
Presets
Parameters
Controls
Click/drag canvas to paint V seeds
Keys: Space=pause R=reset 1-7=presets
The Gray-Scott Equations
The Gray-Scott model describes two chemical species U and V diffusing across a 2D surface and reacting:
∂v/∂t = Dv ∇²v + u·v² − (F + k)v
The term u·v² represents the reaction: two V molecules and one U molecule combine to produce three V molecules (autocatalytic). U is continuously replenished at feed rate F. V is removed at rate F + k. The two species diffuse at different rates (Du = 0.2097, Dv = 0.105), and this differential diffusion is the engine of pattern formation.
Pattern Regimes
Spots
F=0.035, k=0.065. Circular blobs of V emerge and stabilise into hexagonal arrays, like leopard spots or fish scales.
Stripes
F=0.060, k=0.062. Elongated ridges reminiscent of zebra stripes and tiger markings. Breaks spot symmetry.
Coral
F=0.0545, k=0.062. Branching dendritic structures that resemble brain coral or fern fronds. Edge-driven growth.
Spirals
F=0.014, k=0.054. Rotating spiral waves, chemically similar to the Belousov-Zhabotinsky reaction in oscillating chemistry.
Maze
F=0.029, k=0.057. Interconnected labyrinthine channels with no preferred orientation, like fingerprint ridges.
Mitosis
F=0.028, k=0.053. Spots that spontaneously elongate and split in two, mimicking cell division dynamics.
U-Skate
F=0.062, k=0.0609. Propagating soliton-like waves that travel without dissipating, nicknamed for their shape.
Why Differential Diffusion Creates Patterns
Alan Turing’s 1952 paper “The Chemical Basis of Morphogenesis” identified diffusion-driven instability as the mechanism. A uniform state that is stable without diffusion can become unstable when species diffuse at different rates. Intuitively: V (activator) produces itself from U, but diffuses slowly — so where V is slightly elevated, it keeps amplifying locally. U (substrate) diffuses faster, resupplying V-depleted regions and creating boundaries. This creates a spatial wavelength where the feedback is strongest, locking in a characteristic length scale for the pattern.
The Gray-Scott model (P. Gray & S. K. Scott, 1983) is a simple but rich two-variable reaction scheme. John Pearson’s 1993 Science paper mapped the full (F, k) parameter space and identified the menagerie of patterns now named after his study. The simulation you are seeing runs a standard 5-point finite-difference discretisation on a 256×256 periodic grid.
🧪 Unlock All 32 Simulations
Premium members get unlimited access to all 32 interactive simulations including Lorenz attractor, quantum mechanics, neural networks, and more.
View Premium PlansRelated Simulations
Key Equations
| Concept | Formula | Notes |
|---|---|---|
| GrayScott (U) | ∂U/∂t = Du∇²U − UV² + F(1−U) | U: activator; F: feed rate; Du: diffusion |
| GrayScott (V) | ∂V/∂t = Dv∇²V + UV² − (F+k)V | V: inhibitor; k: kill rate; Dv: diffusion |
| Turing instability | d = Du/Dv < 1 | Pattern forms when activator diffuses slower than inhibitor |
| Laplacian (2D) | ∇²U = ∂²U/∂x² + ∂²U/∂y² | Discretised as 9-point stencil in simulation |
| Discretised update (U) | U[n+1] = U[n] + Δt(Du∇²U − UV² + F(1−U)) | Euler forward step; typical Δt = 1 |
Curriculum Relevance
| Level | Topic | Relevance |
|---|---|---|
| A-Level Chemistry/Biology | Reaction kinetics | Concentration changes, autocatalytic reactions |
| IB / AP | Differential equations | Coupled PDEs, diffusion, numerical methods |
| Undergraduate | Applied mathematics, biophysics | Turing patterns, pattern formation theory, morphogenesis |
| Postgraduate | Mathematical biology | Reaction-diffusion systems, activator-inhibitor models |
🧬 Turing Patterns
Alan Turing’s 1952 paper “The Chemical Basis of Morphogenesis” proposed that reaction-diffusion systems of an activator and inhibitor could spontaneously generate stable spatial patterns. The Gray–Scott model is a canonical example: by tuning feed rate F and kill rate k, it produces spots, stripes, labyrinthine mazes, and self-replicating “bugs” — mimicking biological coat patterns, coral textures, and embryonic segmentation.
🧪 Parameter Space
The (F, k) parameter space of Gray–Scott is rich: small changes in feed or kill rate switch between qualitatively different regimes. Spots appear around F≈0.035, k≈0.065; labyrinthine stripes near F≈0.06, k≈0.062; waves and self-replication at F≈0.025, k≈0.055. Experiments with the Belousov–Zhabotinsky reaction confirm that real chemistry obeys similar dynamics.