Estimating changes of riverine landscapes and riverbeds by using airborne LiDAR data and river cross-sections

Today, Airborne Laser Scanning (ALS), also referred to as airborne LiDAR, derived Digital Terrain Models (DTMs) and Digital Surface Models (DSMs) are used in different scientific disciplines, such as hydrology, geomorphology, forestry, archaeology and others. In geomorphology, ALS data are used for studies on landslides, soil erosion, mass movements, glacial geomorphology, river geomorphology, and many others. In the field of river geomorphology, ALS data sets provide information on riverine vegetation, the water level and water-land-boundaries, the elevation of the riparian foreland and their roughness. Small-footprint ALS systems used for topographic data acquisition operate mainly in the near-infrared wavelength. Thus, topographic ALS is not able to penetrate water but to provide a highly detailed representation of the dry land. Therefore, a method to derive Digital Bathymetric Models (DBMs) by using river cross-sections acquired by terrestrial field surveys is presented in this paper. The DBM, which is combined with the ALS-DTM to a DTM of the watercourse, is the basis for calculating changes of the riverbed and the riverine landscape between two ALS data epochs. The first step of the DBM delineation method is to separate water from land in the ALS data. A raster-based approach to derive the Water-Extent-Polygon (WEP) is presented which incorporates the signal strength of the ALS backscatter (referred to as intensity or amplitude), terrain slope and height differences between the DTM and DSM (i.e. the so-called normalized DSM, nDSM). In the second step, the river centerline is extracted by applying a shrinking algorithm to the WEP. Subsequently, a dense array of 2D-transects, perpendicular to the centerline, is defined. For these 2D-transects the heights are interpolated linearly from the measured river cross-sections. From the obtained 3D point cloud representing the riverbed a raster model can be calculated by applying a suitable interpolation technique. In the final step, the DTM and the DBM are combined to a DTM of the watercourse. For two available ALS-DTM data sets (years 2003 and 2006) the respective watercourse DTMs are calculated based on terrestrial measured river cross-section. data sets. By computing difference-models changes in the water level between the two ALS-DTMs are calculated. To estimate the accumulation and erosion potential of the riverbed between the two periods, the difference-model of watercourse DTMs is used. The results show the potential of using ALS in combination with river cross-section data as input for DBM modeling, watercourse DTM generation, riverine landscape and riverbed change detection. The main objectives of the paper are on presenting an accurate WEP delineation approach and a workflow to model a watercourse DTM.