PICO
Understanding sediment connectivity is critical for sustainable sediment management. This research focuses on clay-dominated areas of the Northern Apennines, characterized by high landslide activity and sedimentary disequilibrium due to anthropogenic and natural disturbances. These conditions create an ideal testing ground for evaluating sediment transfer processes and restoration strategies.
The study employs a multiscale approach, integrating high-resolution digital terrain models (DTMs) at 5-meter resolution with detailed DEMs derived from drone-based LiDAR surveys (DJI Matrice 300 and L1 payload). These datasets enable detailed assessments of sediment transfer dynamics, with a focus on the influence of landslides on fluvial systems.
The SedInConnect model is used to calculate structural sediment connectivity indices, identifying pathways and barriers that influence sediment transfer and highlighting critical areas for intervention. By combining SedInConnect’s spatial analysis of connectivity with LANDPLANER’s temporal modeling of sediment fluxes, the study provides a multidimensional understanding of sediment dynamics. This integration enables the identification of vulnerable areas and the design of targeted management interventions or mitigating erosion in high-risk zones.
Drone-based LiDAR surveys represent a technological breakthrough, offering high temporal resolution and allowing frequent monitoring of topographic changes. These data are essential for detecting landslide-induced geomorphic changes and refining event-driven sediment dynamics models. By integrating field observations, remote sensing, and advanced modeling, the study delivers a robust and scalable framework for assessing sediment connectivity.
These advancements offer transformative tools for understanding and managing sediment dynamics, contributing to the development of more resilient and sustainable fluvial systems in clay-dominated landscapes. This approach is particularly valuable for designing and implementing sediment management strategies to mitigate environmental and infrastructural impacts.
How to cite: Fiorucci, F., Santangelo, M., and Rossi, M.: Multiscale Sediment Connectivity Analysis in Clay-Dominated Lithology , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12946, https://doi.org/10.5194/egusphere-egu25-12946, 2025.
Quantifying sediment transport dynamics in Alpine rivers is essential for predicting their geomorphological evolution, for managing flood risks and fluvial ecosystems, as well as for sustainable management of hydropower schemes. However, actual data on sediment transport, particularly for the bedload fraction, are often very scarce (if not absent altogether) due to the challenges inherent in collecting such information. Thus bedload transport dynamics have to be predicted at the basin scale by relying on limited (in space and time) field observations. However, models capable of simulating bedload transport at the network scale in mountain rivers are very few, and to the best of our knowledge, their validation has never been carried out.
The primary objective of this research is to evaluate the performance of the D-CASCADE model – after adapting it to work in Alpine rivers – to simulate bedload transport dynamics at a network scale in the Sulden/Solda river basin (Italian Alps). The Sulden catchment was selected due to the sediment transport monitoring station present at its outlet (130 km2) as well as for the long duration of bedload transport throughout the year due to its nivo-glacial hydrological regime. Since 2014, bedload transport has been continuously monitored in the Sulden River using geophones, which provide high-frequency data on bedload movement and capture temporal variations in bedload transport. To calibrate the geophone signals, regular bedload sampling was conducted. The data obtained from these samples provided detailed insights into the grain size distribution of the transported material at the outlet reach of the modelled network. This empirical information was crucial in fine-tuning the adapted D-CASCADE model and refining existing transport capacity formulas to characterize the connectivity properties of the Sulden network in terms of bedload flux dynamics, path lengths and velocities as well as sediment budgeting of the different reaches.
Preliminary validation of the adapted D-CASCADE model shows a promising agreement between predicted and observed bedload transport rates at the monitoring station. The model demonstrates the potential in reconstructing bedload transport patterns across the entire river network, identifying key sediment sources contributing to the overall sediment flux. Additionally, the model illustrates the spatial and temporal variability in bedload transport, highlighting the complexity of sediment dynamics in Alpine rivers.
How to cite: Pitscheider, F., Argentin, A.-L., Doolaeghe, D., Gianini, M., Repnik, L., Bizzi, S., Lane, S. N., and Comiti, F.: Network-scale modelling of bedload transport in Alpine rivers using D-CASCADE model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5869, https://doi.org/10.5194/egusphere-egu25-5869, 2025.
The movement of sediment through river networks is crucial for the health and functionality of river ecosystems, flood control, and water availability. Network-scale sediment connectivity models have emerged in recent decades but lack robust validation with field measurements. Here, we perform a path length-based application of the morphological method, the Variational Mode Decomposition (VMD) method, to the Tagliamento River, a large braided river in northeastern Italy, to validate the sediment flux estimates generated by the network scale sediment connectivity model D-CASCADE.
The results indicate that D-CASCADE can generate sediment flux estimates that align with those derived from the VMD method and with values documented in literature. Furthermore, we observe that the generated path length estimates align with the expected path length based on the spacing of confluence-diffluence couplets which has been previously proposed as a proxy for path length. These results underscore the need for careful calibration of grain size distributions for specific rivers to improve model accuracy. Additionally, we identify the importance of estimating a fundamentally unknown input parameter, the active transport width (the part of the river channel where bedload is moving for a specific discharge), and its impact on the modeled sediment transport estimates. Finally, we see from the field acquisitions that even during small flood events on the Tagliamento, there is significant compensation when comparing the erosion and deposition volumes during each flood event.
These results demonstrate that the VMD method provides reasonable estimates of path length and sediment flux, thereby serving as a valuable validation tool for network-scale sediment connectivity models and increasing the robustness of the D-CASCADE model in large, complex river systems. The presented field data also help clarify when topographic changes are not a reliable representation of bedload fluxes due to high flow events or confined planform morphology, which then limits the applicability of the VMD method. Overall, the present study is a step forward in validating and refining our understanding of sediment transport processes in braided river environments and provides practical implications for the sustainable management of riverine ecosystems.
How to cite: Capito, L., Doolaeghe, D., Bozzolan, E., Surian, N., Bertoldi, W., and Bizzi, S.: Path Length and Sediment Flux Validation in Braided River Systems: Application of the VMD Method and D-CASCADE Model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15136, https://doi.org/10.5194/egusphere-egu25-15136, 2025.
The limited observational data on river bedload presents a significant challenge in understanding sediment transport processes. However, with recent advancements in computing capability, availability of remotely sensed data, and smart sensors, it is nowadays possible to model these transport processes in river networks at catchment scale. Nevertheless, the results of these models are often not robust due to inherited uncertainty and the stochastic nature of the input parameters. To manage these uncertainties and improve the robustness of model outputs, sensitivity analysis plays a crucial role. Sensitivity analysis is a method to study how changes in a numerical model's input factors contribute to variations in its output.
This project aims to apply Global Sensitivity Analysis (GSA) techniques to the D-CASCADE (Dynamic CAtchment Sediment Connectivity And Delivery) model, for the Po River network in Italy. D-CASCADE is a network-based (or graph-based) model that simulates material movement as distinct transport processes at the reach scale, or ‘cascades,’ defined by their provenance, sediment volume, and interactions downstream, at daily timestep.
To conduct the GSA, we use the SAFE toolbox, supporting both the generation of 5,000 random input factor combinations within defined ranges and distributions, as well as the quantification of the impact of each input factor's variation on the output.
In this work, we focus on the sensitivity estimation of active channel widths and riverbed slopes for every reach of the simulated network. These two input factors are key drivers of the transport capacity and the consequent sediment fluxes generated for the various sediment transport formulas implemented in D-CASCADE. The active transport width (the portion of the channel where bedload transport is active for a specific discharge) is largely unknown, even in data-rich contexts. Hydraulic slopes are also often unknown and generally replaced with topographic slopes which are largely dependent on the quality of the DEM used. Active widths and slopes are then structurally inherently uncertain although they drive the model results. Through GSA, we evaluate how simultaneous random changes in these two input factors affect the simulated sediment fluxes and budgets. Results are analyzed both at the reach scale (sensitivity to local parameters) and the network scale (sensitivity to upstream parameters).
The presented methodology allows us to obtain important information about the effects of structural uncertainties in sediment transport modelling at network scale. These findings provide a foundation for enhancing the model's accuracy and resolving uncertainty in sediment transport prediction.
How to cite: Shrestha, S., Bozzolan, E., Doolaeghe, D., Surian, N., and Bizzi, S.: Enhancing sediment transport model reliability through Sensitivity Analysis: A Case Study in the Po River, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11920, https://doi.org/10.5194/egusphere-egu25-11920, 2025.
Catastrophic sediment release in fluvial systems is largely driven by landsliding that occurs naturally in mountain belts during extreme events, such as earthquakes or storms. Sediments are routed through the river system until they are stored either permanently in alluvial fans and lakes or temporarily in floodplains. The river response to such catastrophic sediment release has already been studied with 2D numerical models using a single effective grain size. Yet, in natural systems, the sediment grain size distribution can span several orders of magnitude and evolves during transport.
The role played by the grain size distribution on morphodynamics depends on transport modes and on grain size interactions. On one hand, fine sediments that tend to be transported in suspension and thus higher in the water column than coarse sediments contribute to floodplain formation and maintenance. On the other hand, coarse sediments that tend to be immobile or transported as bed load contribute to armouring of the channel bed surface that prevents its degradation and in turn leads to channel widening.
Assuming a single effective grain size may limit accurate forecasting of morphodynamic and sedimentological changes in rivers systems during landslide-induced sediment cascades. Modelling the response of a river reach in 3D, meaning that both morphodynamics (2D) and stratigraphy (1D) are resolved may be challenging due to computational time and computer memory. To cope with these limitations, we propose a 2.5D numerical model as a simplified approach. It incorporates: i) a multi-grain size sediment transport model with the ability to capture the transport of suspended and bedload material as well as the dispersion rate and sediment sorting patterns of various grain sizes such as armouring and downstream fining (threshold of motion and explicit grain-size specific entrainment and deposition rates), ii) an explicit transfer of sediment from the river channel to adjacent floodplains (based on the vertical distribution in the water column), iii) freely evolving channel width and slope, and iv) an algorithm to handle channel and floodplain sedimentary records (stratigraphic layers).
We conducted numerical simulations on a constricted river reach that consists of a straight channel with a floodplain on both sides. Numerical simulations reveal: i) how the grain-size specific signals propagate in a river reach and are preserved in the channel and floodplain stratigraphy in response to a catastrophic sediment release, and ii) how the channel width adjusts with stochastic flow conditions and sediment supply.
These preliminary results were obtained in the context of the SCALEES (Signature of sediment CAscades following Landslides triggered by Extreme Events in the Stratigraphy) project funded by the European Union. The combination of empirical data with numerical simulations will allow us to predict for the first time the full signal (all grain sizes) of sediment cascades preserved in stratigraphy in response to an extreme event at the scale of a catchment. It will also pave the way for inverting the stratigraphic record of landslide induced sediment cascades for quantitative insights into their response amplitudes and relaxation times.
How to cite: Le Minor, M., Lague, D., Howarth, J., and Davy, P.: Coupling a channel width evolution model and a multi-grain size sediment transport model: a simplified approach to predict the response of a river reach to a catastrophic sediment release, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14139, https://doi.org/10.5194/egusphere-egu25-14139, 2025.
Understanding sediment transport routes is crucial for predicting geomorphic changes driven by natural and anthropogenic processes in coastal and fluvial systems. Sediment tracing methods are vital to gain such understanding, but common sediment tracers are usually limited to following sediment released from a specific point. In our TRacking Ameland Inlet Living lab Sediment (TRAILS) project, we explore the use of natural luminescence signals of minerals to trace nourished sediment grains on an ebb-tidal delta. Towards this, we obtained and analyzed sediment samples from the Dutch Wadden Sea, where a mega-nourishment in the Ameland inlet ebb-tidal delta aims to address the sediment demand of the nearby coast and basin.
Insufficient luminescence signal resetting, e.g. poor bleaching, due to limited light exposure, can serve as a tool for sediment tracing by examining variations in the degree of bleach characteristics of luminescence signals with differing bleaching sensitivities within a single grain. This can inform us about light-exposure of that grain and therefore about sediment transport history, as explored by Reimann et al (2015) for a beach nourishment project at the Dutch coastline. Firstly, we hypothesize that slow-to-bleach signals reveal information about the end-member type: native grains in our ebb-tidal delta will be well-bleached in comparison to nourished grains. Secondly, we hypothesize that fast-to-bleach signals give insight into the transport history of grains: native grains will be more or less fully reset within the dynamic tidal reworking system of the Wadden Sea whilst nourished grains will still inherit part of their original signal. Combining information derived from slow- and fast-to-bleach signals thus provides a promising novel approach for tracing sediment grains in dynamic subaqueous environments, and thereby reveals sediment transport pathways of nourished sand grains.
Luminescence tracing methods rely on quantitative information about the potential and efficiency of subaqueous signal resetting. In a one-day experiment we quantified bleaching potential, that is, the light intensity and spectrum as a function of time, depth and tidal stage, and bleaching efficiency, that is the degree of bleaching of slow- and fast-to-bleach luminescence signals (de Boer et al., 2024a). Strongest subaqueous light attenuation took place during low tide when sediment concentrations are the highest, we also observed stronger attenuation of the ultraviolet part of the light spectrum. Light-sensitive luminescence signals, such as low-temperature feldspar IRSL, bleached more rapidly than less light-sensitive signals, such as high-temperature feldspar post-IR IRSL. None of the investigated signals were fully reset after 13.5 hours of light exposure, even for subaerially exposed samples. We then collected and analyzed over 100 sediment samples from the Ameland ebb-tidal delta. Using an EMCCD camera (de Boer et al., 2024b), we imaged a multitude of single-grain luminescence signals to explore the native or nourished origin of these sand grains. Ultimately, we aim to integrate these findings with Lagrangian sediment transport models to better understand spatial and temporal coastal sediment dynamics and inform coastal nourishment strategies (Pearson et al., 2022).
How to cite: de Boer, A.-M., Pearson, S. G., Pannozzo, N., Kooistra, T. J., van Prooijen, B., and Wallinga, J.: Luminescence imaging of single grains of sand reveals their sediment transport history, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6632, https://doi.org/10.5194/egusphere-egu25-6632, 2025.
Predicting the flow conditions required to entrain sediment in gravel-bed rivers is essential for many issues, such as river ecology, flood risk assessment, river restoration and sustainable river management, among many others. In this regard, the critical Shields parameter is the most commonly used metric to characterise particle entrainment in bedload transport studies. Consequently, the determination of critical Shields is fundamental to the prediction of bedload transport in gravel-bed rivers. Conventional field studies have focused on estimating a reach-averaged Shields stress, despite the large spatial variability that this parameter exhibits at the reach scale. This is largely due to the lack of standardised field approaches for characterising Shields stress in a spatially distributed manner. In this work, we propose a field-based procedure for estimating the frequency distribution of critical Shields at the patch scale in a gravel-bed river, based on the measurement of resistance to movement of individual clasts and a number of variables related to the position and orientation of grains. Following this procedure, we have approximated the patch-scale variability of particle entrainment conditions in a gravel bar of the Upper Cinca River, located in the southern watershed of the Pyrenees. The results (mean Shields ~0.03) are consistent with previous estimates of critical Shields in this river and with established theory of particle entrainment in gravel-bed rivers. We believe that this method has great potential to provide valuable field information on particle entrainment.
Ackowledgements: This research benefitted from the methods and outcomes of the MorphHab research project (PID2019-104979RB-I00/AEI/10.13039/501100011033, Ministry of Science, Innovation and Universities (MICINN), Government of Spain). The work by the first author is also part of the 2023–2026 grant signed between the Spanish Directorate General for Water (DGA-MITERD; Government of Spain) and the Spanish National Research Council (CSIC-Ministry of Science, Innovation and Universities), which includes action “Sedimentary Morphodynamics” (20233TE012: IGME-CSIC; Tarquín 2 Project).
How to cite: Vázquez-Tarrío, D., Carrero-Carralero, E., López, R., Ville, F., Vericat, D., and Batalla, R. J.: A field-based protocol to approximate variability in critical Shields coarse-bed rivers at the patch scale, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11090, https://doi.org/10.5194/egusphere-egu25-11090, 2025.
Understanding the variabilities of bedload mobility is fundamental in predicting the likelihood of erosion and deposition patterns in gravel bedded rivers, which subsequently assist towards modelling geomorphic adjustment and flood risk change over reach to catchment scales. In most applications, the shear stress required to initiate sediment transport (τ*c) is typically assumed from relations with channel slope or the median bed material grain size, and is generally assumed temporally constant. However, flume investigations identify important relations between grain arrangement (for example, grain protrusion and imbrication) and sediment flux, which vary in response to flood history. Given the complexities of river systems, grain-scale linkages between water-working history, bedload characteristics, and grain mobility remain largely unexplored in the field.
We use a combination of gravel bed microtopography data, collected via structure-from-motion photogrammetry, and in-situ grain resistance tests to resolve a grain force balance model for 45 upland gravel surfaces across England and Wales. Grain resistance forces (FR), and subsequent estimates of τ*c, are used to explore grain scale drivers of particle mobility, as well as their spatial and temporal variabilities. We interpret flow histories of sampled surfaces using typical water-working indicators (including bed surface roughness, imbrication extent and grain size sorting). Water-working metrics are compared against resistance force distributions, to address the hypothesis that conditioned surfaces exhibit systematically higher mobility thresholds. We also consider the relative role of grain shape on bed topography and stability trends. In practical application, our findings can offer more targeted, process-based, estimates of τ*c for a given channel reach, even when grain surface characteristics are only known qualitatively. Such improvements in τ*c estimates are critical in furthering our ability to predict sediment fluxes and geomorphic change in gravel dominated channels, particularly in response to climate change, where the temporal sensitivity of τ*c is likely to be important.
How to cite: Baynes, E., Whitfield, D., Rice, S., Jeffries, R., and Mathers, K.: Exploring Variabilities in Gravel Mobility Using Force-Balance Models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11412, https://doi.org/10.5194/egusphere-egu25-11412, 2025.
The stability or, conversely, the mobilisation of riverbeds varies in time and space, making it a complex phenomenon to study. The size and intensity of mobile grains can lead to disturbance of the substrate that supports physical habitats and can therefore control the presence of aquatic organisms. The degree of mobility of a given grain size fraction (GSF) can be approximated by assessing the proportion of its grains that are mobile. On the other hand, the relative degree of mobility can be expressed as the ratio between the proportion pi of this fraction among all mobilised grains compared to its initial proportion at the surface of the bed Fi. A condition of full mobility is reached when the fraction's proportion among the mobilised grains is equal to or greater than its initial proportion at the surface (pi /Fi≥ 1). An underrepresented fraction is said to be partially mobile (pi /Fi<1).
We present PhotoMOB, a GIS-based tool to characterise (i) grain shape (i.e. axis size, orientation, roundness, compactness, elongation), (ii) patch organisation (i.e. proportion of fine material cover, proportion of overlapping grains) and (iii) mobility magnitude of gravel river beds from repeated digital photographs taken before and after targeted hydrological events. It is based on the detection and the comparison of the shape of grains identified at the same coordinates (location). PhotoMOB allows identification of coincident grains (immobile) and new grains (mobile). Several variables can be extracted from this categorisation, such as: the overall proportion of mobile or immobile grains (in number or surface area), the maximum mobile or immobile diameters, the proportion per individual GSF of grains that remain immobile (stable) and newly identified grains. In addition, changes in fine material cover, grain overlap can be assessed and the percentiles of the surface grain shape distribution before and after a targeted hydrological event, as well as the distribution of exclusively immobile and/or mobilised grains, can be calculated.
Automatic classification applied to perfect (manual) digitisation of grains gives mean absolute errors for fractional mobility estimation of less than 3%, while automatic classification applied to automated digitisation with 10 minutes of manual grain boundary revision gives errors of around 8%. This approach has been developed, tested and applied in gravel-bed mountain rivers affected by hydropeaking, which induces partial mobility.
Ackowledgements: This work is carried out in the background of the projects MorphHab PID2019-104979RB-I00 / AEI / 10.13039/501100011033) and Undammed TED2021-130815B-C31 / MCIN/AEI/10.13039/501100011033, funded by the Spanish Ministry of Science, Innovation and Universities and the EU “NextGenerationEU”/PRTR. All authors are part of the Fluvial Dynamics Research Group –RIUS, a consolidated group recognized by the Generalitat de Catalunya (2021 SGR 01114).
How to cite: Ville, F., Vericat, D., Rennie, C., and Batalla, R. J.: PhotoMOB: a GIS tool to monitor spatial and temporal bed mobility at the patch scale, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16407, https://doi.org/10.5194/egusphere-egu25-16407, 2025.
Predicting when riverbed sediment is mobile is essential for managing the morphology and ecology of gravel-bed river channels. However, our ability to predict critical shear stress (τc) is still such that predictions are only accurate to an order of magnitude at best. One aspect which is often overlooked when predicting grain entrainment, and which likely contributes to our poor predictions of τc, is the role of any cohesive material surrounding the gravel grains. This material could be clay, as is commonly found in gravel-bed rivers draining agricultural catchments, and/or biological, such as produced by caddisfly larvae, mussels and biofilms. To assess the potential impact of non-biological cohesion we parameterise a force-balance grain entrainment model to demonstrate that adding plausible values of cohesion can produce an order of magnitude increase in τc. We compare our results to two sets of field measurements of grain entrainment forces. The first set are from Bury Green Brook, UK, where there is local variation in the amount of clay matrix in the gravel bed and we assess differences in entrainment forces between individual grains. The second set comprises data from multiple sites with varying amounts of fines in the bed and we compare average entrainment forces. Our field data are consistent with the model results, demonstrating the potential importance of accounting for cohesion when predicting τc. Finally, we demonstrate that cohesive forces from clay are also sensitive to water content, and so may be most important in ephemeral channels.
How to cite: Hodge, R., Yager, E., Voepel, H., Leyland, J., Sear, D., and Sitorus, D.: The impact of cohesive material on gravel entrainment, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12444, https://doi.org/10.5194/egusphere-egu25-12444, 2025.
Sediment transport plays a crucial role in shaping our natural and engineered landscapes, affecting everything from riverbanks to coastal regions and ecological habitats. It is key to effective disaster management, helping predict and mitigate the effects of floods, landslides, and coastal erosion. However, modeling how sediment moves through water and landscapes remains a complex challenge. This complexity is due to the unpredictable nature of turbulent flows and sediment movement, compounded by issues such as natural variability, lack of sufficient data for accurate model testing, and high computational demand.
This research introduces an innovative approach by integrating Lagrangian turbulent velocity theory into sediment transport models. By developing a new model that utilizes differentiable stochastic processes, this study aims to enhance our ability to predict and understand how particles behave in turbulent flows. This advanced modeling technique addresses key challenges like the unpredictability, intermittent behavior, and memory effects associated with particle movement in turbulent conditions. Ultimately, this research seeks to refine our understanding of sediment dynamics, pushing the boundaries of existing models and providing more reliable tools for environmental management.
How to cite: Tsai, C.: Lagrangian Stochastic Sediment Dynamics in Turbulent Flows, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19574, https://doi.org/10.5194/egusphere-egu25-19574, 2025.
Soil erosion by water from upland areas results in on-site and offsite problems in the catchment and aquatic ecosystems such as lakes and rivers. As such, up-to-date information on erosion rates and suspended sediment fluxes in rivers is indispensable to manage the impact of soil erosion, sediment transport, and sedimentation. However, detailed spatial and temporal information on erosion and sediment dynamics is rare, as it is labor-intensive and costly to obtain, particularly in developing countries. The advent of satellite remote sensing applications has provided the opportunity to monitor sediment fluxes by assessing the suspended sediment concentration (SSC) of rivers and lakes. This approach may provide a cost-effective alternative to ground-based sampling schemes. However, satellite-based approaches to monitor sediment fluxes require calibration and validation as the relation between SSC and optical properties of the water recorded by satellite sensors may vary with changing sediment properties.
Here, we used empirical models to estimate SSC values from optical sentinel 2 data for Lake Tana in Ethiopia using in situ collected water samples. Moreover, in situ reflectance data, which were measured using an ASD Field Spec 4 spectroradiometer instrument, from water samples collected at Lake Tana are used as well. SSC and in situ reflectance measurements were conducted for 546 water samples collected from the lake, particularly from the river plumes of the two most important rivers draining to Lake Tana, i.e. Gumara and Giligel Abay. The sample SSC values ranged from 1.50 mg/L to 4,146 mg/L. The samples were classified into two categories: low SSC (≤ 250 mg/L) and high SSC (> 250 mg/L), as the optical properties of water are significantly influenced by its constituents. The individual bands in the NIR and visible spectrum exhibited a good correlation (R2 = 0.73, RMSE = 30.69 mg/L) for low SSC-values over Lake Tana. Moreover, the multilinear regression (MLR) analysis using both the visible and NIR bands of low SSC-conditions improved results compared to using individual bands (R2 = 0.84, RMSE = 23.37 mg/L). In contrast, high SSC water samples from Lake Tana did not correlate well with individual bands. However, combining the NIR and red bands generally improved the estimation of SSC for high SSC values (R2 = 0.9, RMSE = 0.26 mg/L).
The established relations between optical properties and field-based SSC values will be applied to long-term timeseries of optical data to assess the temporal variations in sediment concentration in rivers draining to Lake Tana, and in Lake Tana itself. These timeseries will be compared to optical data on vegetation changes in the catchment to identify hot spots in both space and time that are responsible for elevated fluxes of sediment to Lake Tana.
How to cite: Ashagrie Simegn, A., Worku Awoke, G., Annys, S., Frankl, A., and Verstraeten, G.: Estimation of suspended sediment concentration using satellite remote sensing data in Lake Tana, Ethiopia, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15661, https://doi.org/10.5194/egusphere-egu25-15661, 2025.
The suspended sediment concentration (SSC) in rivers is commonly indirectly estimated from optical turbidity measurements in one single point. Calibration based on regular sampling and laboratory analysis allows converting the optical turbidity into SSC. The regular sampling and laboratory analysis are time-consuming and expensive.
A hydro-acoustic multi-frequency approach has advantages as an alternative to optical turbidity measurements. Backscatter data collected with hydro-acoustic echosounders contains information on the suspended particles along an entire profile. The conversion of backscatter into SSCs from an acoustic single-frequency system, like most standard ADCPs are, requires knowledge on the characteristics of the suspended particles, in particular on their average size and the grain size distribution. These characteristics can be estimated by analysing water samples in the laboratory.
The present contribution reports measurements of the SSC along an entire profile with a multi-frequency system. The multi-frequency approach allows estimating the particle characteristics from the backscatter data. Hence, the conversion of backscatter into SSC does not require water samples and laboratory analysis anymore. The potential of the hydro-acoustic multi-frequency approach is illustrated with in-situ river measurements and laboratory experiments that cover a broad range of sediment concentrations and sediment characteristics.
How to cite: Höllrigl, J., Blanckaert, K., Hurther, D., Fromant, G., and Storck, F. R.: Suspended sediment measurements by hydro-acoustic multi-frequency echosounders, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18605, https://doi.org/10.5194/egusphere-egu25-18605, 2025.
Suspended sediment concentration (SSC) significantly impacts water quality, aquatic ecosystems, and reservoir capacity, making accurate prediction vital for effective watershed management. Traditional empirical and physically based models often struggle to handle the complexities and non-linear dynamics of sediment transport. Machine learning (ML) techniques, with their ability to model non-linear relationships and process large datasets, offer a promising alternative. This study explores the application of ML models, including extra trees (ET), random forest (RF), categorical boosting (CatBoost), and extreme gradient boosting (XGBoost) and their combination with genetic programming (GP), to predict SSC. Key environmental variables such as precipitation, streamflow, and seasonality are used as inputs, and the models are trained and validated using historical hydrological data. The SHapley Additive exPlanations (SHAP) framework is employed to interpret the models, offering insights into the influence of each input variable on SSC predictions. Results demonstrate that ML models outperform traditional approaches in accuracy and robustness, particularly in capturing peak sediment events. The findings underline the potential of ML in improving SSC prediction and guiding sustainable watershed management practices.
Keywords: Suspended Sediment Concentration (SSC), Machine Learning (ML), SHAP Values, Hydrological Modeling, Sediment Transport, Watershed Management.
How to cite: Lamane, H., Mouhir, L., Moussadek, R., Baghdad, B., and El Bilali, A.: Leveraging Machine Learning for accurate and interpretable suspended sediment concentration predictions, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-629, https://doi.org/10.5194/egusphere-egu25-629, 2025.
