ANALYZING ANALOG RIVER TABLE TOPOGRAPHY WITH TIME-LAPSE AND STRUCTURE FROM MOTION

The use of analog river models like the Little River EM4 is becoming more common in academia to demonstrate fluvial processes; however, these systems lack a robust, quantitative means of measurement. As of now observations of change are primarily collected by repeat surveying cross sections on the model, consequently limiting the use of the model to classroom demonstrations rather than research and data collection. This system is not adequate for quantifying the 3D evolution of rivers and other geomorphic processes. We improve upon this starting point by describing a means of two- and three-dimensional surveying of model topography.

The two dimensional data collection is achieved using a stationary camera above the river table for time lapse photography, allowing for uninterrupted collection of data of the changing river system for an extended period of time. This imagery can be used in Matlab or GIS to measure lateral movements of the channel as well as channel narrowing, and widening.

Three-dimensional measurements are achieved using close-range photogrammetry with imagery collected from a system of equally distributed DSLR camera stations around the table. These images are fed into Structure from Motion software (Pix4D) to create a three-dimensional model of the river table at a point in time. By taking multiple sets of photos during the course of a simulation, a series of models are created, which can be differenced against each other to measure things like incision rate, sedimentation rate, and changes in channel dimensions. This method of data collection permits the researcher to collect data without stopping or disturbing the flow of the river table, and can be performed by one individual instead of requiring multiple people to move equipment. This combination of data collection sources will allow river tables to be more effectively used as research tools for understanding fluvial geomorphology processes.