Answers are grounded in this paper and cite their sources.
Using the coupled Marine Environment Research and Forecasting Model and General Ocean Turbulence Model (MERF–GOTM), this study examines how K -profile parameterization (KPP) influences nonhydrostatic simulations of internal lee waves over an idealized two-dimensional supercritical sill and evaluates its performance in a realistic shallow-sea application using the Coastal and Regional Ocean COmmunity model (CROCO). The constant-mixing benchmark reproduces a downstream lee-wave packet, whereas the standard KPP configuration, with a critical bulk Richardson number (Ric) of 0.3 and an empirical turbulent-shear coefficient (Cv) of 1.6, generates excessive boundary-layer mixing, weakens crest-scale stratification, and strongly suppresses lee-wave development. Forward experiments show that, under the tested conditions, the response is controlled mainly by Ric: reducing Ric weakens excessive mixing and restores a clearer wave signal, whereas tested Cv variations produce only minor changes. Additional diagnostics indicate an inverse relationship between lee-wave amplitude and crest-adjacent upper-layer mixing, suggesting that KPP modifies the wave response by preconditioning the local density structure. Interior-only and hydrostatic experiments further indicate that spurious damping is mainly associated with the boundary-layer components. In the realistic shallow-sea case, removing boundary-layer mixing leads to insufficient mixing and larger temperature errors, whereas reducing Ric improves the thermal structure relative to the standard KPP configuration. A supplementary surrogate-based inverse physics-informed neural network (IPINN) analysis shows that the reduced-Ric diagnostic signature is recoverable from the synthetic sensitivity ensemble, supporting the parameter regime identified by the forward experiments. Overall, accurate lee-wave simulation requires limiting excessive boundary-layer mixing while retaining sufficient vertical mixing to maintain realistic stratification.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: