The phenomenon by which a system's relaxation time toward equilibrium diverges or increases markedly as a control parameter approaches a critical point or bifurcation, causing slower recovery from perturbations.

Principle

As a dominant stability eigenvalue approaches zero at criticality, the linear return rate vanishes and the corresponding timescale (inverse rate) grows, producing slow decay of perturbations and increased correlation times.

Demonstration

Approaching a saddle-node bifurcation, the leading eigenvalue of the linearized system tends to zero and the recovery time scales like the inverse of that eigenvalue, so small disturbances persist longer as the bifurcation is neared.

Misapplication

Interpreting any observed slow recovery as evidence of an approaching critical point without ruling out alternative causes such as seasonal forcing, external drift, finite-size effects, or measurement filtering.

Consequence

Critical slowing down provides early-warning signals for transitions: increased autocorrelation, increased variance, and longer decay times, which can be quantified to anticipate bifurcations or phase changes.

Reversal

Critical speeding up: in some systems the dominant timescale shortens near a parameter change when stabilizing feedbacks strengthen, producing faster recovery rather than slowing.

Boundary

Pertains to continuous bifurcations or second-order phase transitions and their vicinity; does not apply to abrupt first-order transitions without a nearby vanishing linear stability exponent or to systems dominated by nonlocal memory.

Semantic Tension

Slowing due to criticality vs slowing due to external time-varying forcing: both produce longer correlations, but only the former reflects an intrinsic eigenvalue approaching zero and hence a nearby bifurcation.

Synthesis

Critical slowing down is the intrinsic lengthening of relaxation times as a system approaches a critical point, caused by a vanishing linear stability rate and manifested in elevated correlation times and variance.