Water surface and velocity measurement-river and flume
2014-09-01T13:13:21Z (GMT) by
Understanding the flow of water in natural watercourses has become increasingly important as climate change increases the incidence of extreme rainfall events which cause flooding. Vegetation in rivers and streams reduce water conveyance and natural vegetation plays a critical role in flood events which needs to be understood more fully. A funded project at Loughborough University is therefore examining the influence of vegetation upon water flow, requiring measurement of both the 3-D water surface and flow velocities. Experimental work therefore requires the measurement of water surface morphology and velocity (i.e. speed and direction) in a controlled laboratory environment using a flume but also needs to be adaptable to work in a real river. Measuring the 3D topographic characteristics and velocity field of a flowing water surface is difficult and the purpose of this paper is to describe recent experimental work to achieve this. After reviewing past work in this area, the use of close range digital photogrammetry for capturing both the 3D water surface and surface velocity is described. The selected approach uses either two or three synchronised digital SLR cameras in combination with PhotoModeler for data processing, a commercial close range photogrammetric package. One critical aspect is the selection and distribution of appropriate floating marker points, which are critical if automated and appropriate measurement methods are to be used. Two distinct targeting approaches are available: either large and distinct specific floating markers or some fine material capable of providing appropriate texture. Initial work described in this paper uses specific marker points, which also provide the potential measuring surface velocity. The paper demonstrates that a high degree of measurement and marking automation is possible in a flume environment, where lighting influences can be highly controlled. When applied to a real river it is apparent that only lower degrees of automation are practicable. The study has demonstrated that although some automation is possible for point measurement, point matching needs to be manually guided in a natural environment where lighting cannot be controlled.