Evaluating the use of stereo-photogrammetry for gravel-bed roughness analysis

Abstract

In this work, three different measurement technologies are employed to study the hydraulic roughness in a laboratory flume with a gravel-bed. The gravel-bed topography is measured with (i) an acoustic bed-profiler; (ii) a hand-held laser-scanner; and (iii) stereo-photogrammetry. Digital Elevation Models (DEMs) are obtained with each technique. Statistical roughness analyses of the DEMs are undertaken by estimating the vertical roughness length, represented by the standard deviation of bed elevations. In addition, Probability Distribution Functions (PDFs) and two-dimensional second-order structure functions of bed elevations are used for the analyses. All three measurement techniques were found precise enough to reach the grain-scale of laboratory gravel-beds. However, important disparities were found between the DEMs, which were not always visible statistically. The acoustic bed-profiler and the laser-scanner were found suitable for gravel-bed roughness characterisation. With stereo-photogrammetry, the complex operational processes must be improved to obtain realistic high-resolution DEMs and to derive faithful statistics.