Abstract

During the past decades, the systematic investigation of the morphodynamics of meandering rivers mostly involved the theoretical-analytical methodology. The development of analytical models enabled the definition of equilibrium conditions, stability and evolution of river meanders and to investigate the interaction between planform and bedform processes and mechanisms. In recent years the new branch of remote sensing applied to river morphodynamics has been constantly developing simultaneously to the rapid increase of computational and satellite resources. The remote sensing analysis is nowadays employed in a wide range fields in geophysics; for this reason, the past years have seen the prolific development of numerous algorithms for remote sensing analysis. However, remote sensing of meandering river morphodynamics has not been consistently integrated with morphodynamic modelling so far. There is a lack of sophisticated algorithms for the extraction of extensive morphodynamic information from the available remotely sensed data; this gap prevented researchers from seeking systematic validation of analytical models to define their range of applicability, and to exploit their potential for improved insight on observations in real world meandering rivers. The evolutionary dynamics of the channel width, at local and bend scale, as well as the dynamics of bars in meandering rivers represent two major unsettled issues in our present understanding of river meandering dynamics. In this thesis I first provide a systematic methodology for the automated extraction of meandering river morphodynamic information from multitemporal, multispectral remotely sensed data, coded in the PyRIS software. Moreover, I develop an analytical model to investigate the long-term planform evolution of periodic sequences of meander bends incorporating spatio-temporal variations of channel curvature, width and slope. A first model component predicts the temporal evolution of the channel width and slope based on a novel treatment of the sediment continuity at the reach scale. A second model component is a fully analytical, evolutionary model of periodic meanders with spatially and temporally oscillating width accounting for nonlinear feedbacks in flow and sediment transport by means of a two-parameters perturbation approach. Application of the PyRIS software to several long reaches of free-flowing meandering rivers allows me to develop a consistent set of observations on the temporal and spatial evolution of channel width and curvature with unprecedented level of detail. Furthermore, model outcomes indicate that meander-averaged width and slope invariably decrease during meander development, and that the temporal adjustment of the hydraulic geometry is controlled by the ratio between the evolutionary timescales of planform and riverbed, quantified from the analyzed meandering rivers dataset. The nonlinear perturbation model indicates that width and curvature co-evolve according to a hysteretic behavior in time and predicts that the meander belt width dramatically decreases when the meander resonance threshold is crossed. The modelling approach predicts wider-at-bend meanders when the bank pull is dominant with respect to bar push, which in turn promotes meander bends that are wider at inflections. Analytical modeling and remote sensing analysis are mostly integrated through a statistical approach; bend-scale evolutionary analysis of planform descriptors such as channel width, width oscillations and curvature in large pristine meandering rivers exhibit good agreement with the outcomes of the proposed analytical models. Finally, the integration between analytical modeling and remote sensing analysis allows me to identify the key processes controlling the interaction between migrating sediment bars and planform-driven steady point bars. The conditions for the formation of migrating bars in meandering rivers are mostly related to the production of sediment supply by the basin, contrarily to the widespread idea that meandering rivers exhibiting migrating bars typically display lower values of the channel curvature.