Applications of drone-based photogrammetry and multi-resolution raster data processing with fine detail for fish pass analysis

Abstract

High-resolution topographic data are crucial for accurately modeling hydraulic structures, particularly fish passes, where subtle geomorphic features govern flow heterogeneity and fish navigation. Conventional remote sensing approaches often lack the spatial resolution to resolve such microrelief. This study demonstrates the application of unmanned aerial vehicle (UAV) photogrammetry using Structure-from-Motion (SfM), which generates dense point clouds and digital elevation models (DEMs) with sub-centimeter accuracy. The UAV-SfM workflow enables rapid, non-invasive acquisition under controlled field conditions, followed by point-cloud processing in CloudCompare and rasterization into hydraulically suitable elevation models. These datasets support high-fidelity numerical modeling in 1D, 2D, and 3D, as well as the creation of physical models via 3D printing for laboratory hydraulic experiments. The study highlights the advantages of UAV-SfM in capturing complex terrain, estimating grain characteristics, and improving hydraulic model reliability, while also noting limitations such as restricted visibility areas and computational demands of large point-cloud datasets. By integrating high-resolution UAV-based topographic data, new opportunities for routine, high-detail hydraulic modeling and environmental engineering practice are opened.