Riparian vegetation provides crucial ecosystem services but the reliable estimation of its impacts on the hydrodynamic and transport processes is limited by the oversimplified description of plants in numerical models. The aim of this paper is to improve the representation of the interactions between flow and flexible woody vegetation in the 2-dimensional hydrodynamic model D-Flow FM. We generalized the existing total resistance approach, implementing a two-way flow-vegetation interaction process allowing to correctly describe effective bed shear stress. The novelties compared to other numerical models are the realistic representation of woody plants by considering the leaf and stem area indices, the full reconfiguration process including the streamlining of leaves and bending of the stems under flow forcing, and the flexibility-generated modifications specifically for submerged conditions. 1D simulations representative of floodplain shrubs and young trees demonstrated the 3-7 times higher accuracy of the new model compared to the original model describing plants as rigid cylinders. The consideration of the foliation process supports reliable simulations from leafless to densely foliated conditions. As one of the first validation studies for submerged vegetation, we found that the proposed von K & aacute;rm & aacute;n scaling factor improved the predictions. Further, we preliminarily validated the new bed shear stress computation method for the rarely measured case of complex reconfiguring vegetation. We support the mainstreaming of these developments by integrating the described numerical scheme into the open-source Delft3D FM software, making it accessible to end-users. Finally, the developed model decreased the uncertainty in flow hydraulics, indicating an improved description of the physical system. Thus, the advances serve for improving understanding of the impacts of varying vegetation conditions on broader hydrological, transport and water quality processes in rivers and floodplains. As the developed model has constant vegetative parameter values under variable conditions, it is expected to improve predictions particularly under non-calibrated conditions.
The active-layer model used to account for mixed-size sediment morphodynamic processes may be ill-posed under certain circumstances. Well-posedness guarantees the existence of a unique solution continuously depending on the problem data. When a model becomes ill-posed, infinitesimal perturbations to a solution grow infinitely fast. Apart from the fact that this behaviour cannot represent a physical process, numerical simulations of an ill-posed model continue to change as the grid is refined. For this reason, ill-posed models cannot be used as predictive tools. There exists a regularisation strategy based on a preconditioning method that guarantees that the one-dimensional active-layer model is well-posed. Here, we show that the extension of this strategy to two dimensions does not regularise the model and we propose a different regularisation strategy based on diffusion that guarantees that both the one-dimensional and two-dimensional active-layer models are well-posed. We implement the strategy in Delft3D Flexible Mesh and show an application.
Field surveys for monitoring river morphology are resource-intensive, especially in large river systems. Earth observation (EO) satellites provide valuable alternatives, yet the integrated use of multiple missions remains underexplored. This study develops a framework that combines Landsat, Sentinel-1, Sentinel-2 and Ice, Cloud and Land Elevation Satellite-2 (ICESat-2) data to estimate periodically inundated bed topography and morphological change in the large alluvial Jamuna River. The method achieved an accuracy of similar to 1.5 m of RMSE in topography estimation, with a tendency to overestimate elevations with a mean bias of 1.24 m while preserving slope. By relating water occurrence probability to elevation, morphological changes were quantified for elevation bands, enabling the estimation of area and volume dynamics while accounting for interannual hydrological variability. Compared with the earlier EO-based morphological monitoring tool-which tends to overestimate erosion and accretion areas-the proposed framework reduced false detections (similar to 65%) by accounting for hydrological variability and refining change-classification thresholds. Estimated volumes were comparable in magnitude to EO-based estimates reported in the literature. However, validation with field data revealed that erosion and accretion volumes were underestimated primarily due to the inability to detect changes in aperiodic bed and an inconsistent temporal scale of field and satellite data. Despite these limitations, the method successfully captured dominant reach-scale morphological processes and is suitable for deriving first-order estimates of morphological changes in large rivers with seasonally fluctuating water levels. While further research is needed to better quantify errors/uncertainties and validate the SWOT-based water surface elevation profile correction, this work lays a foundation for using multi-satellite EO data for large-scale river morphological assessments.
The contribution of tidal trapping to salt dispersion has been well described for well-mixed estuaries, in terms of barotropic filling and emptying of the traps. How traps contribute to salt dispersion in deeper, partially stratified systems remains underexplored. We investigate the dispersive effect of temporary storage of saltwater in harbors adjacent to a partially stratified estuary using field observations and numerical modeling. Our results show that instantaneous channel-harbor salt exchange is dominated by density-driven exchange flows arising from baroclinic pressure gradients between the channel and the harbors. This pressure gradient, and consequently the exchange flow, reverses during the tide due to tidal variations in main-channel salinity. Quantification of the trapping-induced additional salt transport from individual basins reveals substantial differences in contributions of individual basins. These differences are linked to a region in the main channel where the tidal salinity range has a minimum, thus limiting the set-up of baroclinic pressure gradients, reducing exchange flow strength and tidal trapping. Analysis of the density-driven exchange reveals that it scales with the tidal salinity range raised to the power 3/2. Using this relationship, we derive an expression for the dispersion coefficient associated with density-driven tidal trapping. This formulation indicates that the resulting dispersion is governed by the main-channel tidal excursion length and the propagation speed of the density current within the trap, and that the dispersion coefficient scales with the square root of the along-channel salinity gradient, in contrast to tidal trapping driven by basin filling and emptying, which is independent of this gradient.
Fragility functions are often used to determine earthquake related damage to houses and infrastructures. To be reliable, such fragility functions need to be specific for the structure, the subsoil and the earthquake characteristics. This study derives fragility curves for embankments located on liquefiable soils subjected to induced seismicity in the Groningen region, The Netherlands. The results are based on more than 3000 nonlinear fully coupled dynamic analyses of embankments on liquefiable soils using PLAXIS2D with the PM4Sand model. Both the liquefaction criterion and the input motions used are Groningen-specific and capture the response of embankments under increasing levels of shaking intensity. The fragility curves are developed to evaluate the seismic risk of embankments on liquefiable soil, with the damage states (DSs) described in terms of embankment settlement. The effectiveness of the criteria, including sufficiency, efficiency, practicality, and proficiency, is assessed based on the relationship between intensity measures (IMs) and engineering demand parameters (EDPs). The paper also presents a novel general model that allows for the derivation of fragility curves beyond the specific predefined configurations, providing a more comprehensive analysis. This approach allows for a more holistic understanding of the system’s response and provides valuable insights into the contributions of various intermediate factors.
