Inaccuracy introduced by LiDAR-generated cross sections and its impact on 1D hydrodynamic simulations

摘要

High-quality topographic data is an essential prerequisite for any hydrologic simulation. In recent years, LiDAR measurements have been used as a reliable method for obtaining such data even though it has its own flaws. This article focuses on analysis of error introduced by LiDAR-based elevation scanning due to the fact that the near-IR laser beam (1,064 nm) cannot penetrate water masses. Such error may affect derived structures (cross sections) and it influences hydrodynamic simulations as well. Motivation for our research was mainly the insufficient amount of measured cross sections on tributary channels in our study area (Moravian-Silesian Region) and evaluation/correction of error dragged into hydrodynamic modeling when LiDAR-based cross sections are used instead. In our article, we compared two data sets of cross sections: real cross sections created from direct measurements of the river channel and cross sections of the same river channel extracted from LiDAR (9 cross sections, 0.2 m LiDAR DEM resolution, gauge stations and flood-plains data were used). As it was expected, the inaccuracies between these two cross-section data sets radically affected the results of the HEC-RAS simulation in the study area (the confluence of rivers Olse and Stonavka). Some known methods for depth estimation and flow area estimation were used for correction of these errors and as a conclusion of the study a recommendation is presented: inaccuracies introduces by LiDAR's inability to penetrate water masses must be taken into account and must be corrected according to the catchment-specific conditions. We also showed that double-linear depth estimation method works well when dealing with low-flow conditions in upland catchment study area.