Disclose river morphodynamics processes through experiments

Rivers’ behavior is increasingly of interest to wide engineering and scientific communities. In fact, fluvial dynamics have an impact on infrastructure and anthropic environments as well as on natural habitats. In addition, the economic values of river as routes of commerce is great, as is the importance of precious minerals deriving from fluvial sedimentary structures. One of the more fascinating river’s peculiarity is the wide range of space and time scales that impact on the overall fluvial dynamics: from the small scales typical of turbulence that are responsible for the motion of single bed’s grains to wider scales typical of fluvial catchment and climate changes. Such a variety of impacting scales firstly accounting for the difficulties of fluvial investigations. Moreover, a second source of difficulty comes from the wide interactions between the fluvial scales: typical hydrodynamics scales influences and being influenced by the scales of sediment transport and bed deformation. It follows that non-linear and non-trivial behavior characterizes the river dynamic. Notwithstanding, great improvements in the fluvial knowledges have been done up to now. Nowadays, fluvial engineer and researchers can take advantage of field studies, laboratory experiments and hydrodynamic models to improve and deepen the study of the various fluvial phenomena. A rough chronology of fluvial advances dates back to 50s the development of field studies and laboratory experiments, and to 80s the growth of mathematical theories and numerical models. All these investigation methods are nowadays being improved and each result to be fundamental to the others. Field study represents the only way to study the real fluvial system, but imply two disadvantages: firstly, they are difficult to be performed (mainly during flood periods) and expensive, and secondly are site-sensitive (i.e., it is difficult to drawn general laws and lessons from a specific fluvial environment). Laboratory experiments are useful to operate in a reductionist manner, since they allow to isolate and separate physical issues that in nature are merged and somehow hidden. Being a strong simplification of complex natural phenomena, experiments must be carefully designed and run in order to be a realistic reproduction of what happens in nature. Mathematical theories and physically-based numerical models are a strongly quantitative way to face to fluvial behavior. Nowadays, we can see an increasingly number of theoretical models and this is due to the increase of computing power of computers and to advances in mathematical modeling. At the same time, the 4 a) b) d) e) f) g) c) Figure 1.1: Examples of various river channel patterns. From a) to g): Brahmaputra River, India (10 km wide braid plain), Rakaia River, New Zealand (1.7 km widebraid plain), Allier River, France (0.8 km wide meander belt), Koyukuk River, Alaska (10 km wide meander belt), Columbia river, Canada (2.1 km wide fluvial valley), Escalante River, Utah (60 m wide channel) and Nanedi Valles, Mars (2 km wide channel) (after Kleinhans (2010)). Introduction 5 theoretical advances risk to provide models that, even though mathematically refined, are not useful to resolve practical fluvial problems. It follows that a correct mix of field observation, experiments, and theories can be the only way to face the fascinating and still not completely discovered fluvial world. The consequence of all the physical processes that characterize rivers is the wide and fascinating range of planimetric patterns that a river can exhibit. On the basis of the channel form on the horizontal plane, rivers are traditionally classified as straight, braiding or meandering (e.g., Leopold and Wolman, 1957). Each of these planforms present different mobility on the floodplain and different degree of pattern stability. Moreover, every river planform has its peculiar natural habitat, in terms of different riparian vegetation, geochemical characteristics, and fauna. Figure 1.1 shows a good example of the different channel patterns that are present in nature. The planimetric patterns shown from panels 1.1a to 1.1g are qualitatively characterized by a decreasingly stream power versus bank strength ratio. Another peculiar feature of rivers, common to all the planimetric patterns, is the intrinsic bed instability. Once the motion threshold is exceeded the water flow over a granular bed leads the bed to presents transversely oriented wave-like features, for example ripples and dunes. This bedforms travel beneath the current, take part to sediment transport, and increase the hydraulic resistances. Dunes are of the order of magnitude of water depth, presenting in nature typical wavelength of 100-102m. Figure 1.2 shows an example of a train of dunes obtained in laboratory. Figure 1.2: Dunes in an our laboratory run. Flow was from right to left. For scaling, channel width was 50 cm. Major sizes are typical of another ubiquitous bedforms, called bars. River bars are 6 longitudinal sedimentary accumulation, submerged and moved only during high flows. Bars can assume a classical alternate configuration with respect to channel axis (alternate bar) or be present on the inner side of bends (forced bars). Bars play different morphodynamics role on different river planform: e.g., they can trigger and enhance bend evolution in meandering river and separate the single channels of braiding network. Figure 1.3 reports two alternate bar in our experiments on the pseudo-meandering pattern. Figure 1.3: Alternate bars in an experimental channel. Flow was from left to right. For scaling, the distance between the triangular markers was about 3 m. In our work we study both issues about fluvial planforms and bedforms, investigating some connections between planimetric and bed deformation. In particular we face with a planimetric configuration called "pseudo-meandering". The pseudomeandering pattern exhibits several features of both meandering rivers (alternate bars, migrating bends and asymmetrical cross-sections) and braiding rivers (flow diversion and tendency to create secondary channels due to the development of a chute channel between the inner side of the bar and the bank) which coexist in the same reach. Thanks to an experimental approach and some field observation we demonstrate how such pattern is strictly influenced and determined by the water discharge variability. Fluvial planforms were also focused in experiments reproducing some pattern changes (from braiding to single-thread) that are induced by strong sediment supply decline, that mainly happens caused by anthropic activities and infrastructures. Bedforms issue are instead presented by coupling experiments and a mathematical models, with the aim to investigate and clarify the initial stages of alternate bar formation. In particular we present results showing how our model is able to predict the wavelength selection typical of alternate bar. A great part of the experimental runs presented in this work has taken advantage Geomorphological background 7 of a new instrument that is able to profile the flume bed during the run and in a non– invasive way. We underwent this newly-developed device to several trials to test its accuracy. The maximum errors in the bed’s elevation measurement resulted to be less than 1 mm in hydraulic conditions that are typical of morphodynamics runs. The present thesis is organized as follows. Chapter 2 presents a general introduction about fluvial geomorphology, introducing the various river planforms that are present in nature and the bedforms typical of river’s beds. In chapter 3, the novel instrument used to scan in a non–invasive way the flume bed is described. The following chapters represent the core of the experimental researches: experiments regarding the influence of a varying discharge on a pseudomeadering channel are reported in chapter 4, chapter 5 is devoted to elucidate the transition from multi to single-thread fluvial patterns, and in chapter 6 a new theory and its experimental verification is developed to explain the wavelength selection typical of alternate bars.