Workflow for hydrologic modelling with sUAS-acquired aerial imagery

Abstract

Operating with a conceptual workflow for the appropriate processing of high-spatial resolution small unmanned aircraft system (sUAS) data for hydrologic modelling of floodplains during flood events, this research investigated the effects of input data fidelity on hydrologic model generation. A digital surface model (DSM) and co-registered orthophoto mosaic of a stretch of the Drau River in southern Austria was generated. A digital terrain model (DTM) was then approximated from the generated DSM to within a vertical root-mean-square error (RMSE) of 4.65 cm. Horizontal and random metrics of roughness were calculated based on the DSM and then used to determine spatially-varying values of Manning’s n coefficient across an observed floodplain. A distributed two-dimensional hydrologic model of a river channel and simulated floodplain was performed using open source hydraulic modelling software. The effectiveness of each roughness metric at multiple pixel resolutions to model significant variations in reported velocity, water surface elevation and depth during a flood event was tested in a controlled sensitivity analysis. Results indicated that accuracy in hydrologic modelling is impacted by image spatial resolution. Research results conclude by emphasizing the importance of context in hydrologic modelling scenarios to effectively meet the needs of end users.

Keywords:

• hydrologic modelling
• sUAS
• spatial resolution
• digital terrain models
• roughness

Acknowledgments

The authors thank the reviewers for their constructive and insightful comments.

Disclosure statement

The authors declare that no conflict of interest exists in the development and dissemination of the research presented in this paper.

Additional information

Funding

This research was conducted with generous funding from the Austrian Marshall Plan Foundation at the Carinthia University of Applied Sciences.