Analysis of river bed surface roughnesses using 2D wavelet transform-based methods

When sedimentation occurs in a river channel, the topography of the bed surface changes, which in turn effects the flow characteristics. It is therefore important to be able to characterize the bedsurface topography during and after the sedimentation process. This paper describes results from a preliminary study conducted to analyze river bed sediment surface data using the two-dimensional discrete wavelet transform. We show how the global bed surface topography may be characterized in terms of a novel parameter, the form size distribution, and also that pertinent local bedform features may be highlighted using wavelet thresholding-based techniques. In addition, we test the stationary wavelet transform as an enhanced method of thresholding for feature extraction and we show the difference in behavior between the bed surface geometry and the geometry of a synthetic fractal surface.