Science Serving Maryland's Coasts


Continuous Mapping of Channel Features for Monitoring Aquatic Habitat and Sediment Flux in Coastal Systems

Principal Investigator: 

Matthew Baker

Start/End Year: 

2017 to 2019


University of Maryland, Baltimore County

Co-Principal investigator: 

Alex Rittle, University of Maryland Baltimore County

Strategic focus area: 

Resilient ecosystem processes and responses


Traditional mapping of riverbed features is performed through cross-sectional surveys at limited locations within a river segment. Such methodology leads to an incomplete picture of riverine processes, including aquatic habitat variability and sediment fluxes. Recently work has shown that cameras mounted on an unmanned aerial vehicle (UAV) can be used to continuously map fluvial features at a relatively low cost. UAVs can significantly increase efficiency of data collection, while also providing continuous data, which is not feasible with cross-sectional methods. The purpose of this study is to use UAV aerial imagery for mapping bathymetry, sediment distribution, and channel hydraulics over a continuous reach of the Patapsco River. There are three objectives to this study: Objective1) use a UAV to develop high-accuracy orthophotos of a continuous section of the Patapsco River; Objective 2) Inger channel depth and sediment facies from the developed orthophotos; and Objective 3) Derive first- and second-order hydraulic and geomorphic variables to aid in river characterization. For Objective 1, a Phantom 4 Professional with an integrated 4K, polarized camera will be flown over 3 different segments of the Patapsco. Images extracted from flight videos will be incorporated into a Structure-from-Motion (SfM)workflow to produce geometrically and geographically accurate orthophotos. For Objective 2, the orthophotos produced from Objective 1 will be used to measure channel bathymetry (water depth) and map sediment facies. A digital elevation model (DEM) developed from the SfM workflow will be exported into a GIS environment (ArcGIS v10.3) to plot channel longitudinal profiles. Supervised classification and object oriented classification schemes will be used to automate mapping of sediment facies. For objective 3, using the first order variables derived in Objective 2, second-order hydraulic and geomorphic variables will be calculated including velocity, channel and water surface slope, Froude number, stream power, and stream stress. The hydraulic parameters calculated from the orthophotos will be compared to the same parameters calculated from surveyed, cross-sectional data in order to determine how continuous datasets can alter understanding of fundamental fluvial hydraulics and channel processes. This research will allow practitioners to use less expensive, more efficient methods for mapping channel features. By increasing accuracy of channel measurements, this project will improve understanding of real channel hydraulics and sediment transport, including channel bathymetry. This study would improve spatial characterizations of fish and macroinvertebrate habitat, including anadromous and catadromous fish extant in coastal, riverine systems. This study will also improve analysis on the ongoing Bloede Dam removal project by replacing cross-section data with a continuous map, which will be used in pre-and post-removal analyses