Building a High-Precision 2D Hydrodynamic Flood Model Using UAV Photogrammetry and Sensor Network Monitoring

• Published in: Water (Basel) 2017-11, Vol.9 (11), p.861
• Main Authors: Langhammer, Jakub, Bernsteinová, Jana, Miřijovský, Jakub
• Format: Article
• Language: English
• Subjects: Automation; Bathymetry; Buildings; Catchments; Czech Republic; Data processing; Data sources; Drone aircraft; Flooding; Floods; Gauges; Geospatial data; Hydrologic data; Hydrologic models; Hydrology; Imagery; Measuring instruments; Modelling; Mountains; Navigation; Navigation satellites; Network management systems; Photogrammetry; Reconstruction; Remote monitoring; Runoff; Sensors; Temporal resolution; Two dimensional models; Unmanned aerial vehicles; Water levels; Workflow
• ISSN: 2073-4441; 2073-4441
• DOI: 10.3390/w9110861

This paper explores the potential of the joint application of unmanned aerial vehicle (UAV)-based photogrammetry and an automated sensor network for building a hydrodynamic flood model of a montane stream. UAV-based imagery was used for three-dimensional (3D) photogrammetric reconstruction of the stream channel, achieving a resolution of 1.5 cm/pixel. Automated ultrasonic water level gauges, operating with a 10 min interval, were used as a source of hydrological data for the model calibration, and the MIKE 21 hydrodynamic model was used for building the flood model. Three different horizontal schematizations of the channel—an orthogonal grid, curvilinear grid, and flexible mesh—were used to evaluate the effect of spatial discretization on the results. The research was performed on Javori Brook, a montane stream in the Sumava (Bohemian Forest) Mountains, Czech Republic, Central Europe, featuring a fast runoff response to precipitation events and that is located in a core zone of frequent flooding. The studied catchments have been, since 2007, equipped with automated water level gauges and, since 2013, under repeated UAV monitoring. The study revealed the high potential of these data sources for applications in hydrodynamic modeling. In addition to the ultra-high levels of spatial and temporal resolution, the major contribution is in the method’s high operability, enabling the building of highly detailed flood models even in remote areas lacking conventional monitoring. The testing of the data sources and model setup indicated the limitations of the UAV reconstruction of the stream bathymetry, which was completed by the geodetic-grade global navigation satellite system (GNSS) measurements. The testing of the different model domain schematizations did not indicate the substantial differences that are typical for conventional low-resolution data, proving the high reliability of the tested modeling workflow.