A Comparative Assessment of Multi-Source Generation of Digital Elevation Models for Fluvial Landscapes Characterization and Monitoring

Abstract

Imaging and measuring the Earth’s relief with sensors mounted upon unmanned aerial vehicles is an increasingly frequently used and promising method of remote sensing. In the context of fluvial geomorphology and its applications, e.g., landform mapping or flood modelling, the reliable representation of the land surface on digital elevation models is crucial. The main objective of the study was to assess and compare the accuracy of state-of-the-art remote sensing technologies in generating DEMs for riverscape characterization and fluvial monitoring applications. In particular, we were interested in DAP and LiDAR techniques comparison, and UAV applicability. We carried out field surveys, i.e., GNSS-RTK measurements, UAV and aircraft flights, on islands and sandbars within a nature reserve on a braided section of the Vistula River downstream from the city of Warsaw, Poland. We then processed the data into DSMs and DTMs based on four sources: ULS (laser scanning from UAV), UAV-DAP (digital aerial photogrammetry), ALS (airborne laser scanning), and satellite Pléiades imagery processed with DAP. The magnitudes of errors are represented by the cross-reference of values generated on DEMs with GNSS-RTK measurements. Results are presented for exposed sediment bars, riverine islands covered by low vegetation and shrubs, or covered by riparian forest. While the average absolute height error of the laser scanning DTMs oscillates around 8–11 cm for most surfaces, photogrammetric DTMs from UAV and satellite data gave errors averaging more than 30 cm. Airborne and UAV LiDAR measurements brought almost the perfect match. We showed that the UAV-based LiDAR sensors prove to be useful for geomorphological mapping, especially for geomorphic analysis of the river channel at a large scale, because they reach similar accuracies to ALS and better than DAP-based image processing.

Decadal evolution of fluvial islands and its controlling factors along the lower Yangtze River

“…Numerous researchers have studied the formation and sedimentary processes associated with transient fluvial islands (Zanoni et al, 2008;Wang and Xu, 2018;Zheng et al, 2019;Leli et al, 2020;Lou et al, 2022;do Amaral et al, 2023). Previous studies have examined various processes of fluvial islands, such as their migration and erosion, and found that these processes have an impact on banks, island bodies, and river channels (Promma et al, 2007;Wintenberger et al, 2015;Sudra et al, 2023). For example, Zanoni et al (2008) examined the long-term evolution of island in a braided river and analyzed the effects of river flow regime and sediment delivery.…”

An Object-Based Ground Filtering of Airborne LiDAR Data for Large-Area DTM Generation

“…For creating a DTM, light detection and ranging (LiDAR) [9][10][11], radar [12], and photogrammetric methods [13][14][15] are generally used for collecting the three-dimensional coordinates of the earth. Then, the DTM generation method separates the bare earth from all other ground-standing objects on the earth's surface.…”