Methodological aspects of using high-speed cameras to quantify soil splash phenomenon

Abstract Soil splash is the first step of water erosion. There are many methods for splash investigations however the use of high-speed cameras is becoming increasingly popular. This study describes the methodological aspects of using high-speed cameras and a PTV (Particle Tracking Velocimetry) system to investigate the soil splash phenomenon. The purpose of the research was to examine the impact of image processing and the choice of settings for particle tracking software on the effectiveness of detecting and identifying splashed/ejected particles (with their flight trajectories). Experiments were carried out on moisturized Haplic Luvisol soil samples and the impact of single water-drop onto the surface was recorded by three calibrated Phantom Miro M310 high-speed cameras. Registered frames were subjected to image processing in Dantec DynamicStudio software, where the selection of the optimal set of parameters/settings was validated, in the context of maximum detection of the splashed particles. The resulted images provided the input data for the Volumetric 3D PTV module (DynamicStudio) to identify/track the ejected particles and then determine their trajectories. The study presents the use of different settings (variants) of tracking modules, enabling the most efficient identification and description of the particles’ trajectories (76%). The incomplete trajectories detected (excluding starting and landing fragments), were subjected to a reconstruction process based on mathematical extrapolation. The analyze of the registered images and reconstructed trajectories of splashed particles gave the possibility to determine the following parameters describing soil splash phenomenon: a) number of splashed particles, b) ejection angle (for each particle), c) ejection velocity, d) displacement range, e) altitude (the maximal height reached by the particle).