Author(s): Melissa Latella; Tommaso Raimondo; Carlo Camporeale
Linked Author(s): Melissa Latella
Keywords: LiDAR; Point clouds; Riparian vegetation; Field measurements; River modelling
Abstract: In non-arid environments, the morphodynamic evolution of rivers and their floodplains is strictly related to the dynamics of riparian vegetation and the continuous feedback it exchanges with water and sediments. The proper quantification of vegetation geometry and biomass along -- and within -- river channels is needed to carry on realistic fluvial modeling activities, which account for the additional flow resistance due to the presence of plants and the related pattern of topographic erosion and accretion. Vegetation measurements also underlie the activities addressing the management of riparian ecosystems in the frame of river restoration. In recent years, the use of remote sensing techniques to support river sciences has increasingly widened since several branches of remote sensing have started to deal with the characterization of vegetation and water resources at various spatial scales. At the scale of fluvial bars up to the whole riparian corridor of a river segment, Light Detection And Ranging (LiDAR) has particularly drawn attention so far. LiDAR sensors can be mounted on terrestrial or airborne supports for this kind of application and yield georeferenced point clouds. Point clouds can provide a large number of features about the ground and vegetation if properly processed, but the limitations linked to their acquisition mode and data processing must be carefully taken into account during applications not to undermine the results’ reliability. In the present work, we describe the workflow to identify individual trees and shrubs along rivers and to estimate their biomass through the integration of local in-situ measurements and LiDAR acquisitions. After defining a study area along the Orco River in Italy, we process and analyze the data obtained by airborne laser scanners to individuate the strengths and weaknesses of the available techniques. We also discuss how these techniques can be combined with field measurements to increase their reliability. Furthermore, we explain how literature and new algorithms can be blended to estimate tree and shrub features and biomass from the LiDAR point clouds for both individual plants and stands on the river bars and the floodplain. The final aim of the work is to provide indications about how to acquire, process, and apply LiDAR data to address different issues of the river sciences, such as the setting of hydrodynamic models and ecomorphodynamic studies.