Turing Diffusion Instability Band
Find the spatial modes that diffusion destabilizes despite stable local reactions.
By Riano · Published · Updated · Report a correction
Turing Diffusion Instability Band
The Turing instability (or diffusion-driven instability) is a phenomenon in which a system that is stable to homogeneous perturbations becomes unstable to spatially inhomogeneous perturbations when diffusion is introduced. Specifically, we study a linear two-species reaction-diffusion system. Find the spatial modes that diffusion destabilizes despite stable local reactions.
Worked Example: Unequal Diffusion and Instability
Consider a system with reaction Jacobian J = [[1, -2], [2, -3]] and diffusion matrix D = diag(1, 16). The homogenous state without diffusion is stable since its trace is -2 and determinant is 1. When diffusion is introduced and we test a Fourier mode with squared wavenumber q = 1/4, we evaluate the eigenvalues of J - qD.
Given D = 16, q = 1/4 Trace = -2 - (1 + D)q = -2 - (1 + 16)(1/4) = -2 - 17/4 = -6.25 Det = 1 + (3 - D)q + Dq^2 = 1 + (3 - 16)(1/4) + 16(1/4)^2 = 1 - 13/4 + 1 = -1.25 Growth (Max Eigenvalue) = (Trace + sqrt(Trace^2 - 4*Det)) / 2 = (-6.25 + sqrt(39.0625 - 4(-1.25))) / 2 = (-6.25 + sqrt(44.0625)) / 2 = 0.19397951183793838
Since the maximum growth rate is positive (approximately 0.194), this mode is growing, placing the system in an unstable band. The bounds of the unstable band can also be independently determined: q = (D - 3 ± sqrt((D - 3)2 - 4D)) / (2D). For D = 16, the bounds are roughly 0.086 and 0.726.
Experiment: Changing Wavenumbers and Diffusivity
In this experiment, we explore how changing the diffusion ratio and the spatial mode impacts stability. Follow these steps in the interactive Turing Diffusion Instability Band module:
- Set the Diffusivity ratio control to
16. - Set the Squared wavenumber control to
1/4. Observe the Mode stability indicator showing "Growing". - Change the Squared wavenumber to
0. The system behaves like the homogeneous case without diffusion, and the mode decays. - Change the Diffusivity ratio to
9and the Squared wavenumber to1/3. You will find that the system reaches a "Critical touch" with a maximum growth rate of exactly 0. - Finally, change the Diffusivity ratio to
1. With equal diffusion, the system cannot destabilize, and there is "No unstable band" regardless of the wavenumber.
Limitations
This linear stability analysis determines whether tiny spatial perturbations will initially grow. It is a local linear instability and does not predict the final nonlinear pattern saturation or guarantee that a finite domain will support a specific growing eigenmode. Additionally, the analysis applies to an infinite or periodic domain where Fourier modes are exact; bounded domains may restrict the available wavenumbers.
References
For a detailed mathematical treatment of reaction-diffusion systems and the Turing instability, see:
- UCSD Physics 221A Lectures on Nonlinear Dynamics, which outlines the conditions required for diffusion-driven instabilities, including unequal diffusion rates.
Experience it interactively
Adjust parameters, observe in real time, and build deep intuition with Riano’s interactive Turing Diffusion Instability Band module.
Try Turing Diffusion Instability Band on Riano →