A functional inequality that relates the L2 norm of a function to its L1 norm and its Dirichlet energy (L2 norm of the gradient), typically taking the form ||f||_2^{2+4/n} ≤ C ||∇f||_2^2 ||f||_1^{4/n} on R^n, and used to derive decay and smoothing estimates for the heat equation and semigroups.

Principle

It expresses a trade-off between mass concentration (L1), energy dissipation (Dirichlet form), and L2 amplitude: higher Dirichlet energy forces L2-norm decay relative to the conserved L1 mass according to a dimension-dependent exponent.

Demonstration

On R^n, applying the inequality to a nonnegative initial datum for the heat equation yields decay rates for the L2 norm of the solution and can be combined with semigroup interpolation to obtain upper bounds on the heat kernel; illustrative scenario: showing algebraic time-decay of L2 norms for diffusive evolution.

Misapplication

Using the standard Euclidean Nash form on bounded domains or manifolds without adjusting constants or accounting for spectral gaps, or applying it to functions that lack integrable gradient, leads to misleading conclusions about decay rates.

Consequence

Correct application produces explicit decay estimates for parabolic flows, controls on semigroup norms, and a pathway to heat-kernel upper bounds that reflect the space dimension and diffusion strength.

Reversal

Interpreting Nash as a lower bound on energy given L1 and L2 sizes is generally false; reversing direction would assert coercivity not provided by the inequality unless supplemented by spectral or geometric hypotheses.

Boundary

Valid under assumptions of sufficient regularity and integrability of f (typically f∈L1∩H1) and with constants depending on the ambient space and its volume growth; modifications are required on bounded domains, manifolds, or discrete spaces.

Semantic Tension

Closely allied to Sobolev and Gagliardo–Nirenberg inequalities and to logarithmic Sobolev inequalities; tension arises in whether one frames the result as an Lp interpolation, a heat-kernel tool, or a form of functional isoperimetry.

Synthesis

The Nash inequality links mass, energy, and L2 amplitude through a dimensionally scaled interpolation that yields decay and smoothing for diffusion processes: it is most effective where gradient integrability holds and geometry does not introduce spectral gaps, and must be adapted when domain or topology affects dissipation.