[H-CG22] Earth surface processes related to deposition, erosion and sediment transport

[HCG22-11] Exploring a phenomenological upscaled model of lateral migration of river channels

*Rentarou Miyata¹, Noritaka Endo¹ (1.Graduate school of natural science & technology, Kanazawa University)

Keywords:fluvial landscapes evolution, lateral erosion

Rivers are recognized as complex systems consisting of phenomena of various scales of time and space, from turbulent flow (less than one minute and a few meters) to the movement of a watershed boundary (more than several thousand years and several tens kilometers). Thus, the study of fluvial landscapes is based on viewpoints of different spatio-temporal scales for different interests. LEMs (Landscape Evolution Models) have been developed for various goals and there are many variations. The key issue common to all LEM is that model-generated landscapes are (1) reaching a state of static equilibrium and (2) greatly influenced by the initial topography (Kwang et al., 2019). (1) means that the topography "consolidates" by balancing uplift and erosion at all points, which is contrary to commonly observed facts such as channel migration due to lateral erosion. (2) is incompatible with the fact that many different watersheds have similar geomorphic features. As an improvement over the traditional LEM, Langston & Tucker (2018), who implemented the lateral erosion process in a watershed-scale model, formulated correlations between (a) flow and lateral channel mobility (b) bedrock strength and lateral channel mobility (c) slope and lateral erosion. On the other hand, the results of model experiments and natural geomorphic analyses in previous studies suggest that lateral erosion is closely related to changes in sediment supply and dynamic stream width. Therefore, in order to improve model accuracy, new modeling is required in which the lateral erosion process is related to these physical quantities. This requires (I) the application of channel erosion theory with explicit relationships to sediment supply, (II) the implementation of dynamic river widths, and (III) sediment dynamics modeling on a long-term timescale. For (I), if existing models are used, information of flow velocity and water depth is required, but the computational cost (e.g., solving 2D shallow-water equations) is a significant issue. For (II), in the case of the topographic representation using a square grid, it is difficult to represent continuous changes. For (III), there are several existing models, but the timescale of the assumed physical process is much shorter than the computational time step set for a basin-scale model calculation with a light computational cost, resulting in timescale inconsistency (inefficient or unreasonable in terms of computational cost). This study examines whether it is possible to obtain a phenomenological equation for lateral channel migration of timescales suitable for light-cost watershed-scale modeling.