The phenomenon where solutions, sequences, or evolutions cease to possess the same degree of smoothness or differentiability that initial data, operators, or approximations might suggest — for example, where higher derivatives become unbounded, discontinuous, or nonexistent under evolution or limiting processes.

Principle

Regularity is propagated by the balance of smoothing and nonlinear or forcing effects; loss occurs when mechanisms that create irregularity (nonlinear steepening, singular forcing, rough coefficients) dominate the smoothing operators or when limits concentrate energy at small scales.

Demonstration

For nonlinear PDEs, shock formation in hyperbolic conservation laws shows finite-time loss of C^1-regularity even from smooth initial data; in elliptic problems, rough coefficients can produce solutions with limited Sobolev regularity, and iterative schemes can lose derivatives causing cascading regularity decay.

Misapplication

Assuming that classical a priori smoothness estimates extend across singular parameter regimes or for weak formulations without checking compatibility conditions; expecting linear superposition to preserve regularity in nonlinear settings.

Consequence

Identifying potential loss guides the choice of function spaces (weak, Sobolev, Besov), the formulation of well-posedness (weak/entropy/measure-valued solutions), and the design of numerical methods that account for limited smoothness (high-resolution shock-capturing or regularization).

Reversal

Preservation of regularity occurs when smoothing operators and compatibility conditions dominate, so initial smoothness remains or is improved under evolution or approximation.

Boundary

Refers to differentiability or Sobolev/Besov regularity of solutions to PDEs, variational problems, or sequences of functions; excludes purely topological or measure-theoretic convergence that does not address differentiability classes.

Semantic Tension

Tension with instability or loss of control: loss of regularity is a distinct technical statement about derivatives and function-space membership and not merely sensitivity of solutions to perturbations; it competes with notions of weak convergence where regularity is irrelevant.

Synthesis

Loss of regularity is the tendency for derivatives or smoothness classes to degrade under evolution, nonlinearity, rough coefficients, or singular limits, requiring weaker solution concepts and tailored analytical and numerical frameworks.