The transfer of parameter values, constitutive coefficients, or calibrated closures from one model, scale, or context to another to ensure consistent parametrization across representations or to initialize dependent components.

Principle

Map parameters between models or scales using physically motivated scaling laws, matched moments, calibration protocols, or statistical inference while accounting for scale dependence and contextual adjustments.

Demonstration

Passing eddy-viscosity closure coefficients estimated from high-resolution simulations to a coarse Reynolds-averaged model after rescaling by cell size and time-stepping, or supplying soil hydraulic parameters measured at laboratory scale to a field-scale hydrological model with uncertainty inflation.

Misapplication

Blindly copying parameters without rescaling or bias correction across scales or regimes, producing systematic model errors, inconsistency between coupled components, or unstable numerical behavior.

Consequence

Promotes coherence across model hierarchies, speeds calibration by reusing informed values, and enables consistent initialization of coupled components when scaling and uncertainty are properly handled.

Reversal

Reversal is parameter inference or learning: inferring effective coarse parameters from fine-scale data or from observations; passing parameters and inferring them are complementary in model development.

Boundary

Valid only when parameters have a meaningful mapping between contexts; excludes parameters that are emergent, context-specific, or deliberately model-architecture-dependent without a defined transfer rule.

Semantic Tension

Tension between using fixed transferred parameters for simplicity and developing scale-aware parameterizations that adapt parameters as functions of resolution, regime, or state.

Synthesis

A disciplined protocol for transporting constitutive information between models that combines scaling rules, calibration, and uncertainty accounting so that transferred parameters maintain coherence and validity in the target context.