Assessing eco-geomorphic interactions across scales using novel UAV based remote sensing techniques and modelling

Abstract

The importance of vegetation within the fluvial domain is well established, influencing both flow and morphology, and has long been recognised as a key component of the river corridor. Despite this, adequately capturing the spatial and structural variability of vegetation for us to understand the eco-geomorphic feedbacks occurring at a range of scales remains a challenge. Currently, the focus of this research takes place at either the individual plant scale, looking into vegetation-flow interactions, or at larger scales, attempting to spatially discretise vegetation for bulk roughness metrics. Subsequently, hydrodynamic models are typically based around these bulk roughness values which exclude vegetation structure. The aim of this research is to attempt to bridge this gap and link the different scales of analysis to improve our understanding of eco-geomorphic interactions. This is achieved by: (1) Examining current remote sensing methods that may be used for fluvial research, (2) Developing a novel UAV based remote sensing system to collect plant scale data for reach scale analysis, (3) Extracting trait-based metrics for individual plants and upscaling these to reach scale extents, (4) Implementing these traits-based parameters in to a 2D hydrodynamic model. At present, the main trade offs in remote sensing centre around scale and resolution, whereby capturing larger areas reduces the detail of the phenomena being studied. Structure from Motion (SfM) photogrammetry has helped to bridge this gap yet fails to reconstruct topography in vegetated reaches and cannot resolve vegetation structure. These drawbacks have herein been overcome with the introduction of UAV based laser scanning techniques, capable of accurately capturing topography in vegetated reaches as well as resolving vegetation structure. This data can be used to extract traits-based vegetation metrics, identify individual guilds within a river corridor, and be scaled to spatially discretise vegetation structure at reach scales. Guilds are then evaluated against monitored morphological change to investigate eco-geomorphic feedbacks. These vegetation metrics and classifications are subsequently used to parameterise a 2D hydrodynamic model, showing the impact that vegetation discretisation methods have on model outputs. This research has developed methods for obtaining reach scale data on vegetation structure to better inform our understanding of eco-geomorphic feedbacks. The robustness and scalability of these methods presents future avenues of research, both within the fluvial domain and for other environmental research applications, where eco-geomorphic feedbacks have a major influence in shaping the Earth’s surface.

More information

Identifiers

ORCID for Chris Tomsett: orcid.org/0000-0002-6916-6063

ORCID for Julian Leyland: orcid.org/0000-0002-3419-9949

Catalogue record

Export record