Fourier Spectral Analysis of Free-surface Behavior Based on Imagery Shot in the Field

Although there are some researches focusing on free-surface behavior in laboratory scale flows, there are few studies focusing on free-surface behavior in the actual river field, which might be based on the difficulty of measuring surface profiles in the filed. In the past laboratory studies, free-surface fluctuations are simply described as a superposition of turbulence-generated fluctuations and gravity waves. The dominant behavior of free-surface has been considered to change depending on by the Froude number and the wall roughness; i.e. energy occupied by gravity waves grows with increase of the Froude number and the roughness height relative to the still water depth. However, the dependency of free-surface behavior on these parameters are not fully understood. Additionally, the effect of the Reynolds number has been neglected though it is one of the most important parameters characterizing open-channel turbulent flows.
In this study, river surface imagery shot from a riverbank was used to extract characteristics of the free-surface behavior in river flows. Instead of physical measurements of water surface profile, image intensities were utilized that represent water surface fluctuations by introducing a simple assumption; bright/dark portion of images corresponds to high/low water level qualitatively. Fourier spectral analysis was applied to the post-processed images to estimate a wavenumber-frequency spectrum of free-surface fluctuations. As the result, the dominant behavior of free-surface in the river flow was found to be described as a superposition of turbulence-generated fluctuations, and gravity waves including stationary waves generated as the results of the interaction between the gravity waves and the flow. Additionally, the most dominant behavior was found to vary according to the dimensionless parameter relating to the wavenumber of stationary waves. This parameter is directly connected to the Froude number and the still water depth by
employing the deep-water assumption. These results mostly agree well with the past studies on the laboratory scale flows except for excepting that the cases in which turbulence occupies more energy than gravity waves. This difference is enough