Ocean surface approximation from scattered numerical marine forecast data

Abstract

3D visualization of numerical marine forecast data still pose a challenge for oceanographer due to accuracy requirement and the highly complex behaviour of dynamic fluid. Existing ocean surface model simulate dynamics either with parametric/spectral methods or physically-based methods, which still suffer from lacking real world information. In this paper, we propose a novel coarse-to-fine ocean surface reconstruction framework to approximate ocean wave from scattered numerical marine forecast data. Under this framework, coarse curve can be approximated by employing multilevel B-Splines approximation (MBA) method, which is of C 2 continuation. Multi-layer FFT approximation algorithm is then proposed to obtain a fine surface, performing Gaussian random fluctuations in height field with high accuracy. Meanwhile mass conservation characteristic of ocean fluid was maintained by enforcing incompressibility condition. This paper details the framework and illustrates how the algorithm works to preserve the characteristic of ocean wave. Convincing experimental results are presented to validate the effectiveness of the proposed algorithm.