Linear spreading speed Created 2026-09-24 Updated 2026-11-05
The linear spreading speed is the propagation speed predicted by the linearized dynamics near an unstable state. For
insert an exponentially decreasing leading-edge mode , . This gives
The minimum occurs at and equals . Equivalently, the spatial equation has a repeated positive root at that speed.
For Fisher's equation, this is the Fisher-KPP minimum wave speed. It is a leading-edge calculation, not a universal speed theorem for every reaction-diffusion equation. A pulled travelling front is selected by that linear leading edge, while a sufficiently strong nonlinear interior can select a larger speed. Initial-data decay and nonlinear assumptions determine whether an observed invasion actually travels at the linear prediction.
With the travelling wave coordinate , the reaction–diffusion system becomes
At the advancing leading edge the healthy-tissue density is small. Put with . The linearization at gives
so a nonoscillatory decaying front requires real positive roots,
The pulled travelling front selects the Fisher-KPP minimum wave speed, and therefore
For the initial wound , two mirror-image fronts enter from its endpoints. The left front moves right from and the right front moves left from , each with speed approximately . A sequence of sketches would show the initial rectangular depression becoming two smooth monotone transition layers, the zero-density plateau shrinking, and the two layers meeting near before the profile relaxes to . Their meeting time is
There is an additional local relaxation time of order if “close to one” is assigned a fixed numerical tolerance, so the displayed estimate is the dominant healing time when is large compared with the front width .