Boundary Condition Estimation of Hydrodynamic Systems Using Inverse Neuro-Numerical Modeling (INMM) Approach

摘要

Validation of a numerical hydrodynamic model, as with most modeling systems, is essential to construct a reliable management tool for coastal and estuarine systems. However, it is the most time consuming step. The uncertainty of surface elevations in the ocean boundary or the modeling domain changes is one key issue for the modeling effort. The traditional way to perform model validation is to use the closest surface elevations (or/and current station) field measurements inside the model domain to manually tune the boundary conditions until the desired criteria are met. However, this process usually requires much iteration through numerical runs, particularly for the complex boundary conditions and when there is a long distance between the boundary and measurement gages