Quick clays are fine-grained, highly sensitive marine deposits that are widespread across formerly glaciated regions, including Norway, Sweden, Finland, and Canada. The low remoulded strength of the quick clays makes them particularly susceptible to extensive retrogressive landslides, which pose serious challenges to society. Erosion is recognized as one of the most important pre-conditioning and triggering factor for quick clay landslide. Therefore, identification of the erosion hotspots is essential for understanding landslide initiation processes and for effective hazard mitigation in quick clay terrains. Machine learning has emerged as an effective tool for erosion hotspot mapping, allowing complex spatial patterns and nonlinear interactions among erosion-controlling factors to be identified from remote sensing–derived data. Recent studies have demonstrated that Deep Neural Networks can be effectively employed to identify erosion-prone zones in quick clay environments when sufficient labelled data are available. This study investigates whether unsupervised machine learning applied to remote sensing–derived data can effectively identify erosion hotspots in quick clay areas. A fully automated, Python-based workflow was developed for erosion hotspot mapping in quick clay areas using remote sensing–derived data. The dataset includes terrain, hydrological, environmental, and anthropogenic parameters relevant to erosion and slope instability. Initially, a total of twenty input parameters were considered. Pearson correlation coefficients were computed to assess inter-feature dependencies, and principal component analysis (PCA) was employed to evaluate feature importance. The unsupervised analysis was performed using multiple clustering techniques to capture different structural characteristics of the data where each cluster represents a distinct level of erosion susceptibility. The results suggest that the proposed unsupervised framework can effectively delineate erosion hotspots in quick clay areas and constitutes an initial step toward the development of early warning systems.
Acknowledgements
This work was supported by the Research Council of Norway through the SAFERCLAY project (Grant No. 352887). Orkun Türe was supported by the Council of Higher Education of Türkiye under the DOSAP scholarship programme and served as a visiting researcher at NGI and NTNU.

How to cite: Türe, O., Tao, R., L’Heureux, J.-S., Oguz, E. A., and Tyagi, A.: Fully Automated Unsupervised Machine Learning Framework for Mapping Erosion Hotspots in Quick Clay Areas Using Remote Sensing–Derived Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5775, https://doi.org/10.5194/egusphere-egu26-5775, 2026.

The size and mobility of landslides control their impact on both landscapes and communities. Despite their importance to understanding landslide mechanisms and associated hazards, few studies have examined the factors controlling these two characteristics, particularly at a large scale. This is especially the case for deep-seated landslides that occur across diverse geomorphological and lithological settings. Further, most research focuses on recent landslides and thus fail to consider historical processes that could be associated with environmental conditions that differ from the contemporary ones. Here, we investigate the influence of geomorphology and lithology on the size and mobility of old and recent deep-seated landslides in the North Tanganyika-Kivu Rift region in Africa, an under-researched mountainous environment located in the tropics. Based on a comprehensive inventory of ~2500 landslides, we show that mobility increases with size, especially for the old landslides. These old landslides are significantly larger than the recent ones, likely due to potential progressive landslide growth over time and  influenced by the region’s paleoseismic activity. The main controls on both the size and mobility of deep-seated landslides are lithology and, to a lesser extent, fluctuations in Lake Kivu’s level during the Holocene. Landscape rejuvenation by migrating knickpoints associated with rifting also plays a key role in determining landslide size: in rejuvenated landscapes, landslides tend to be larger than those in relict landscapes. The presence of these large landslides favours the development of smaller ones along their margins, reflecting the influence of path dependency on landslide occurrence and size. Our findings underscore the importance of considering the chronology of landslide occurrence and the long-term legacy of landscape evolution in shaping landslide characteristics.

How to cite: Mugaruka Bibentyo, T., Dille, A., Deijns, A., Nzolang, C., Dewaele, S., and Dewitte, O.: Controls on the size and mobility of deep-seated landslides in the North Tanganyika - Kivu Rift region, Africa, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5929, https://doi.org/10.5194/egusphere-egu26-5929, 2026.

Ephemeral mountain streams on the western Andean slopes remain dry most of the year, yet during intense rainfall events they generate short-lived flash floods with exceptionally high sediment transport capacity. This study investigates the hydraulic response of the upper Río Seco micro-basin (Huaycoloro catchment, Peru) under extreme rainfall scenarios, using a hydraulic–geomorphological framework that links surface hydrology with sediment mobility thresholds. Design discharges were estimated through IDF-based rainfall analysis and classical hydrological methods, while sectional hydraulic modelling using the Manning equation provided flow velocities and bed shear stresses along representative channel reaches. Results indicate mean velocities ranging from 2.4 to 3.4 m/s and shear stresses up to 215 Pa. These values exceed the critical shear stress of the coarse gravel bed by more than five times, indicating generalized sediment mobility and strong incision potential in confined steep reaches. Such conditions promote significant sediment supply from the upper basin, increasing the likelihood of downstream channel aggradation and flood hazard in peri-urban sectors of eastern Lima. To our knowledge, this is the first hydraulic–geomorphological quantification of sediment mobility thresholds in an arid Andean micro-basin under design-storm conditions. The findings provide quantitative evidence supporting the need to transition from purely water-based flood models toward sediment-inclusive risk assessments in steep ephemeral mountain catchments.

How to cite: Rosales Torres, L. and Cárdenas-Gaudry, M.: High-energy sediment dynamics in ephemeral Andean mountain streams: The case of Río Seco, Peru, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6006, https://doi.org/10.5194/egusphere-egu26-6006, 2026.

Flood-prone small mountainous catchments hosting critical infrastructure, such as bridges and transport networks, require integrated hydrologic–hydraulic analyses to ensure long-term resilience under changing climatic and land-use conditions. This study develops a coupled HEC-HMS–HEC-RAS modelling framework to quantify design discharges, inundation patterns and local hydraulic controls for the torrential stream crossing the settlement of Kato Nevrokopi in Northern Greece. Using high-resolution topographic data (DEM), GIS-based basin delineation and long-term rainfall records, design storms for multiple return periods are derived and transformed into flood hydrographs at the catchment outlet. These hydrographs force 1D steady-flow simulations in HEC-RAS, explicitly representing bridges, piers and local constrictions that act as morphodynamic bottlenecks and potential failure points under extreme flows. Model results are used to generate flood extent and water-depth maps for events up to the 1,000-year return period, identify critical cross-sections where afflux and backwater effects are most pronounced, and assess the effectiveness of alternative layout and channel-training configurations. The analysis is framed within the current EU Floods Directive 2007/60/EC and Greek legislation for stream delineation, linking quantitative hazard metrics to planning constraints and infrastructure design requirements. The work highlights how relatively simple, openly available tools, when combined with detailed geometric representation of bridges and channel morphology, can support evidence-based decisions on flood protection works, minimise over-engineering, and improve the adaptive management of critical infrastructure in steep, data-scarce basins.

How to cite: Pavlidis, K. and Valyrakis, M.: Hydrologic-hydraulic modelling and flood hazard mapping for infrastructure resilience in a small mountainous catchment on Northern Greece, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7986, https://doi.org/10.5194/egusphere-egu26-7986, 2026.

The construction of mega-canals necessitates a profound understanding of the pre-existing fluvial equilibrium to mitigate adverse geomorphic consequences, particularly in rivers with limited channel capacity. This study focuses on the intrinsic stability mechanisms of the Qin River, a typical small-to-medium-sized mountainous river in South China, prior to the implementation of the Pinglu Canal project. Field surveys and sediment analyses were conducted to characterise the natural bed state, with a focus on a morphologically representative reach. The findings indicate that the riverbed has historically maintained a strong dynamic equilibrium, supported by lateral confinement from riparian vegetation and natural armor processes unique to mountainous fluvial regimes, which are derived from tributary inputs. The analysis reveals that specific hydrodynamic thresholds and sediment connectivity are essential for maintaining this stability. Therefore, rather than hydraulic stress alone, the system's main vulnerability is determined to be the possible disruption of these established equilibrium conditions, particularly with regard to geological substrate constraints and longitudinal continuity. These results establish a scientific standard for assessing the potential disturbance risks of canalization in delicate mountainous river systems by providing a critical morphodynamic baseline.

How to cite: Zhu, S., Sun, J., Liu, C., Chen, L., and Chen, W.: Natural Riverbed Stability in a Small-to-Medium-Sized Mountainous River: A Baseline Investigation of the Qin River Prior to the Pinglu Canal Construction, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8868, https://doi.org/10.5194/egusphere-egu26-8868, 2026.

The Tel River, a major tributary of the Mahanadi River in eastern India, exhibits strong spatial and temporal variability in flow and sediment dynamics due to its monsoon-driven hydrology, heterogeneous terrain, and increasing human interventions. Soil erosion and sediment transport, although naturally driven by rainfall and surface runoff, have been significantly altered by agriculture, urbanization, and water management structures, leading to changes in soil loss, sedimentation, and degradation of water resources. Therefore, in this study, the production of soil erosion in the Tel River Basin is estimated using the Revised Universal Soil Loss Equation (RUSLE), while riverine sediment transport is simulated using ANUGA-Sed, a two-dimensional shallow-water hydrodynamic and sediment transport model based on a finite-volume scheme. The ANUGA flow and sediment modules were calibrated and validated using observed discharge and suspended sediment data from multiple gauging stations along the Tel River. Parallel simulations performed on the Param Pravega high-performance computing systems significantly reduced computation time while maintaining numerical accuracy, enabling high-resolution modelling of the entire Tel River Basin. The model was further evaluated for elasticity, computational accuracy, and optimal grid distribution per node on the HPC system, demonstrating robust scalability and efficient utilization of computational resources.

The model results show strong agreement with observations, with errors in net erosion and deposition generally below 10%. The simulations successfully reproduce the spatial patterns of sediment generation, transport, and deposition along the river network. Importantly, the model provides new insights into sediment dynamics between gauging stations where direct measurements are unavailable and captures cross-sectional channel changes associated with sediment transport processes. These results were further validated using field-based suspended sediment data collected in October 2023 at intermediate river locations using portable sampling instruments. The simulations reveal distinct zones of high erosion and deposition that are critical for understanding flood conveyance and channel stability. Overall, the results confirm that ANUGA-Sed can reliably simulate suspended sediment transport and riverbed changes in monsoon-dominated river systems.

How to cite: Dahiwale, A. V., Dutta, U., Singh, Y. K., Yendargaye, G., Prabhu, T. S. M., and Muddu, S.: FROM CATCHMENT TO CHANNEL: HIGH-PERFORMANCE PARALLEL MODELING OF SEDIMENT TRANSPORT IN THE TEL RIVER BASIN USING ANUGA Sed, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19944, https://doi.org/10.5194/egusphere-egu26-19944, 2026.

The non-linear feedback mechanisms and interactions between discharge-sediment supply and instream (riparian) vegetation cover generate spatio-temporal heterogeneity in braided channel forms. The present study examines such relationships among three contrasting braided rivers of India: the Brahmaputra (highly braided), the Brahmani (weakly braided) and the Netravathi (meandering-braided). Long term JRC Surface water layer, vegetation-water remote sensing indices, numerical model derived hydrological datasets and periodic field visits have been integrated to understand the vegetation–hydrology–sediment coupling across these braided river systems.  The results show that the channel forming discharges in the Brahmaputra shows a hierarchical level and extreme events dominate over the effect of sparse vegetated landforms. In weakly braided reaches, channel-in-channel form oscillates between two extreme nodes depending upon the intensity of disturbing events. For rivers with meandering-braided transition form, channels are relatively stable and riparian vegetation cover generate a stable geometry and absence of floodplain sediment storage.   

How to cite: Pradhan, C.: Integrating Google Earth Engine Cloud Computing and Fluvial Surveys to Quantify Vegetation–Hydrology–Sediment Coupling in Contrasting Braided River Systems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21498, https://doi.org/10.5194/egusphere-egu26-21498, 2026.

Coastal protection infrastructures such as rubble-mound breakwaters (RMBs) demand frequent geometric inspection to quantify armor-layer dynamics and support reproducible structural monitoring. While UAV-based photogrammetry and LiDAR are established reference techniques for rapid 3D mapping, high revisit rates remain operationally constrained by wind sensitivity, sensor payload limits, and regulatory flight restrictions. Videogrammetry complements these approaches by increasing inter-frame overlap and mitigating missed-trigger acquisitions, especially useful in complex coastal scenes (e.g., those affected by occlusions between armor units and block interstices). As in conventional photogrammetry, videogrammetry relies on image redundancy and self-calibration rather than highly sophisticated instrumentation. Despite this potential, consumer-grade action cameras remain scarcely validated for multi-epoch 3D monitoring in coastal engineering, mainly due to wide-angle lens distortion and coarse onboard GNSS geotag precision.

This study assesses pole-mounted GoPro videogrammetry for multi-temporal 3D relative change detection in the emerged portions of a detached rubble-mound breakwater at Cabedelo do Douro (PT). Two survey epochs were acquired in July 2024 and November 2024 to characterize the above-water zone, inspecting the seaward slope, the landward armor-toe transition, and the horizontal crest platform segment at one of the heads of the RMB. Frames were extracted at 1 Hz and processed in Metashape using an SfM-MVS (Structure-from-Motion Multi-View Stereo) self-calibrating camera model. Multi-epoch point clouds were coregistered in CloudCompare with ICP (Iterative Closest Point) refinement over stable crest and toe areas, and 3D changes were quantified using M3C2 (Multiscale Model-to-Model Cloud Comparison), generating signed distance maps and detection histograms. A concurrent UAV-RTK survey, supported by additional GNSS-measured ground control points (GCPs), served as a geometric benchmark.

Mean ActionCam-to-UAV sensor offsets were +0.06 m, confirming that, despite potentially unstable absolute georeferencing in GoPro-derived reconstructions, the resulting point clouds preserve sufficient geometric and scale consistency to support relative multi-temporal 3D change detection and the identification of concrete armor-unit displacements. Results confirm that pole-mounted videogrammetry supports rapid, repeatable, low-cost SHM (Structural Health Monitoring) observations, providing defensible detection thresholds and reproducible change-detection limits for engineering interpretation and maintenance support.

How to cite: Martínez Olmedo, V., Bento, A. M., Arza-García, M., and Gonçalves, J. A.: Feasibility of Action Camera-Based Videogrammetry for Multi-Temporal 3D Monitoring of Rubble-Mound Breakwaters, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2250, https://doi.org/10.5194/egusphere-egu26-2250, 2026.

River deltas play a crucial role in the transport of sediments and nutrients between river catchments and the sea. Scientific studies have demonstrated that Arctic deltas have a significant potential for sediment retention. Ongoing climate change is accelerating the thawing of permafrost, which largely constitutes the substrate of Arctic deltas, thereby affecting the morphological and hydrological evolution of these low-lying tundra systems.

The aim of this study is to estimate changes in the surface area and flood storage capacity of deltaic lakes using remote sensing methods. Optical and radar satellite data from Sentinel-2 and RADARSAT-2 were used, obtained under a grant from the Canadian Space Agency (application no. RCM CSA-RC-FORM-0003), together with advanced tools for spatial and radar data analysis. The selected study area is an eastern part of the Mackenzie River Delta (Canada, Northwest Territories), namely Big Lake, located near the city of Inuvik, approximately 130 km from the Beaufort Sea. The Big Lake is a through-flow lake with an area of about 800 ha. It is part of a system of approximately 2,000 lakes that maintain year-round connectivity with the East Channel, one of the main distributary channels conveying water within the delta.

The presented results are based on satellite and hydrological analyses conducted at the beginning of the ice-free water period, occurring at the turn of May and June. The study includes a comparison of satellite observations with gauge data. To determine the extent and volume of floodwaters, the Normalized Difference Water Index (NDWI), advanced radar data analyses, and statistical analyses of hydrological data from Water Survey of Canada (WSC) were applied. Satellite imagery acquired during open-water seasons made it possible to delineate shoreline extents and the associated water surface elevations. Selected years from the period 2011–2024 were analysed; for example, it was estimated that at the turn of May and June 2024 the lake stored approximately 8.2 million m³ of water over a period of 49 days.

Considering sediment transport, the Mackenzie River is the largest supplier to the Arctic Ocean, delivers more than 100 million tonnes of sediment annually. Previous studies characterise these sediments as predominantly fine-grained fractions that are easily transported. The presence of an organic-rich catchment combined with the magnitude of fluvial sediment transport highlights the importance of understanding the mechanisms governing sediment distribution, quantities, and areas of deposition within the delta system.

This research is being conducted with the permission of the Government of Canada – North West Territories (NWT) – research licence number 17694 which was issued under application number 6131 and financed by the Polish Ministry of Education and Science - National Research Agency, title: Evaluation of the settling velocity and trapping capacity of sediments in lakes in the Great Arctic River deltas, grant no. 2023/50/O/ST10/00597.

How to cite: Ciepłowski, D. and Habel, M.: Remote sensing analysis of water dynamics within floodplain lakes in the eastern part of the Mackenzie River delta, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20785, https://doi.org/10.5194/egusphere-egu26-20785, 2026.

We conducted a two-month-long cryoseismic monitoring study in the Schirmacher Oasis, East Antarctica, to investigate icequake activity caused by the movement and melting of ice sheets. For this purpose, we deployed a Raspberry Shake seismometer on the Antarctic land and ice sheet for a month. Through a comparative analysis of the recorded seismic data, we gained insights into ice dynamics and diurnal icequake patterns. The Raspberry Shake instrumentation, powered by solar energy, offers a cost-effective approach for establishing a dense seismic network. During installation, the seismometer, solar controller, and Li-ion battery were housed in a wooden box lined with nitrile foam for insulation. The analysis suggests that icequake detections follow a distinct diurnal pattern, with more events occurring during the daytime. Furthermore, we also observe interdependence between icequake detections and high wind speeds.We use a multi STA/LTA approach for event detection on a continuous 11-day period while the seismometer was on ice. We detect 2249 icequake events, which are further manually classified into three categories. More than half of icequakes (67%) belong to a shallow origin and some are indicative of deep icequakes (9%).These findings highlight the need for a denser seismic network and more detailed investigations to further understand the impact of climate change on melting ice sheets.

How to cite: Sattasi, N. S., Silwal, V., T. M., M., Ahamad, A., Suthar, A., and Negi, S. S.: Cryoseismic monitoring in the Schirmacher Oasis, East Antarctica, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19976, https://doi.org/10.5194/egusphere-egu26-19976, 2026.

Bathymetry plays a critical role in determining the occurrence and stability of landfast sea ice, although its seasonal impact on sub-Arctic ice-covered shelves has yet to be thoroughly quantified and understood. Our study explores the ways in which nearshore bathymetry and coastal topography influence the spatial distribution, seasonal persistence, and variability of landfast sea ice, with an emphasis on shallow embayments of James Bay. Our hypothesis suggests that factors like coastal orientation and bathymetry provide extent and stability to the landfast sea ice in the James Bay region, rather than being exclusively governed by marine and atmospheric factors. Using satellite-derived observations of landfast sea-ice delineations, regional bathymetric datasets, and information on coastal geomorphological configuration, this analysis will quantify statistical relationships among the landfast-ice edge extent and persistence metrics with the bathymetric thresholds and coastal orientations. Initial findings indicate that recurrent landfast ice extents are larger and their persistence is higher when there is a shallow water column, undulating bathymetry with mounds, and/or offshore features. Our observations support the hypothesis that bathymetry plays a crucial role in determining the presence and stability of landfast sea ice. By explicitly correlating bathymetry and geomorphology with landfast ice phenology and stability indicators, our research aims to advance both conceptual and quantitative understandings within coastal ice modelling frameworks and refine projections concerning the response of landfast sea ice to ongoing Arctic amplification and climate change.

How to cite: Banerjee, D., Gupta, K., and Mukhopadhyay, A.: Geomorphological controls on the persistence and extent of Landfast Sea Ice in James Bay Region, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15851, https://doi.org/10.5194/egusphere-egu26-15851, 2026.

Understanding transformations of the climate system in the geological past is essential for predicting and mitigating the effects of global climate change in the next future. The geological record provides a unique archive that documents long-term fluctuations of environmental variables, including seasonality. Seasonality appears to have played a crucial role in extreme climate transitions, highlighting the importance of constraining its variability in the past. Increased seasonality is often associated with colder conditions and the development of ice accumulations, making it a key parameter for understanding and forecasting climate change.

Species of the brachiopod Gigantoproductus are giants within the Palaeozoic sedentary benthos, characterised by exceptional size and thick shells, reaching over 30 cm in width and more than 1 cm in shell thickness. These features make them unparalleled bioarchives for palaeoecological and palaeoclimatic reconstructions, enabling the investigation of long-term changes during key intervals of past climate change.

In this study, specimens of Gigantoproductus semiglobosus from upper Visean (Mississippian, Carboniferous) successions of western Ireland (Aran Islands and the Burren) were subjected to detailed diagenetic screening and subsequently analysed using a sclerochemical approach (δ¹⁸O, δ¹³C). These analyses were used to reconstruct seasonal variability and to provide additional evidence for the timing of Mississippian phases of the Late Palaeozoic Ice Age (LPIA).

Our results show that δ¹⁸O profiles from well-preserved shells record high seasonal variations (Δδ¹⁸O = 0.9 to 1.9 ‰ corresponding to a ΔT = 4 to 11 °C) for palaeoequatorial settings, as also observed in coeval species of Gigantoproductus from the UK (Angiolini et al., 2019). This seasonal variation is much higher than that recorded in comparable shallow water, low latitude environments both nowadays and in the distant past. The pronounced seasonality recorded by several species of Gigantoproductus from western Ireland and the UK at low palaeolatitudes supports the onset of a sustained Gondwanan glaciation in the late Visean. Also, the palaeogeographic distribution of the species of Gigantoproductus and the geochemical composition of their shells indicate that low-latitude Mississippian ocean waters did not experience a temperature decrease at the onset of the Gondwanan glaciation, but rather a marked increase in seasonal variability.

Overall, this study highlights the importance of resolving long-term changes in seasonality, using fossil carbonate shells as palaeoclimatic archives during different intervals of climate change, in both the recent and distant past, to better understand and predict long-term transformations of the climate system.

References

Angiolini et al. (2019). The giants of the phylum Brachiopoda: a matter of diet? Palaeontology, Vol. 62, Part 6, pp. 889–917

How to cite: Crippa, G., Angiolini, L., Azmy, K., Cannaò, E., Doyle, E., Della Porta, G., Murray, J., O’Connell, M., Viaretti, M., and Harper, D. A. T.: Seasonal variability at the onset of the Late Palaeozoic Ice Age: insights from Gigantoproductus shells, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12124, https://doi.org/10.5194/egusphere-egu26-12124, 2026.

The Cretaceous–Paleogene (K/Pg) mass extinction represents one of the most severe crises in Earth history, with marked regional variations in the tempo of pre- and post-extinction environmental stress and ecological recovery. The Um Sohryngkew River (USR) section of Meghalaya (NE India) provides a unique perspective on stress and recovery dynamics in a marine setting proximal to the Deccan Traps. This study integrates planktonic foraminiferal assemblage data with sedimentological observations and bulk-carbonate δ¹³C measurements to reconstruct the nature and duration of marine stress and to constrain the timing of ecological and carbon-cycle recovery in the eastern Tethyan realm. This integrated, high-resolution multi-proxy approach was previously lacking for this Deccan-proximal archive, and provides a critical constraint on how volcanogenic forcing modulated K/Pg stress and recovery at regional to global scales.

The late Maastrichtian record at USR indicates highly stressed surface-ocean conditions. Planktonic assemblages are dominated by small opportunistic taxa, particularly Guembelitria cretacea (>80%), with strong dwarfing, dominance of thin-walled morphotypes, poor preservation, and a near absence of heavily calcified taxa (e.g., Pseudotextularia spp., Globotruncana spp.). These assemblage and preservation features point to sustained calcification stress and unfavourable conditions for carbonate production in surface waters, consistent with enhanced nutrient input and surface-water acidification under intensified continental weathering/runoff and volcanogenic CO₂ emissions. Following the K/Pg boundary, planktonic foraminiferal abundance (4 tests/g) and diversity remained markedly suppressed through the early Danian. The post-boundary interval is similarly characterised by persistent dominance of small opportunistic taxa (>30%; e.g., Guembelitria spp. and Chiloguembelina spp.) and continued dwarfing, indicating sustained calcification stress and hindered ecosystem rebuilding. Bulk-carbonate δ¹³C indicates delayed carbon-cycle recovery, beginning only after ~750 kyr at USR compared to ~200–300 kyr at many distal sites. Ecological recovery lagged further, with low diversity and small test sizes persisting for ~2 Myr until biozone P1c, indicating decoupling between carbon-cycle recovery and biological reorganization under continued environmental forcing.

The first robust evidence for ecological improvement appears in planktonic foraminiferal biozone P1c, where assemblages become more diverse and better preserved, test sizes increase, and morphogroup proportions stabilise. These changes suggest improved conditions for calcification, progressive strengthening of the pelagic carbonate system, and a more efficient biological pump. By biozones P1c–P2, community structure indicates that ecological balance was largely restored, and carbonate production increased steadily towards a better-developed carbonate-factory environment. Comparison with global K/Pg records suggests that recovery mechanisms in the USR section broadly mirror global ecological and biogeochemical feedbacks, but their timing is substantially delayed relative to distal sections. Importantly, similar evidence for prolonged stress and delayed recovery has also been documented from the Krishna–Godavari Basin of southern India, supporting a coherent regional pattern in marine environments proximal to the Deccan Traps. Together, these Deccan-proximal records highlight strong spatial heterogeneity in post-K/Pg recovery trajectories, including a delayed return to stable carbon cycling, carbonate production, and ecosystem structure.

How to cite: Patra, S., Punekar, J., Srivastava, P., Rawat, S., Bhadran, A., and Girishbai, D.: Delayed carbon-cycle stabilization and ecological recovery across the K/Pg boundary: evidence from the Um Sohryngkew River section, Meghalaya (India), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10713, https://doi.org/10.5194/egusphere-egu26-10713, 2026.

The traditional workflow in palynology begins with the removal of rock minerals through acid digestion and heavy liquid separation, followed by mounting the organic residue on a glass slide, and analysing it under a transmitted light microscope. Using the microscope, palynologists manually identify and assign the observed particles to predefined categories within a designated counting area on each slide. Counting typically continues until a target number of particles has been reached (often between 200 to 300).

Beyond the commonly analysed palynomorphs such as pollen, spores, and dinoflagellate cysts, palynological slides may also contain a diverse range of acid resistant organic sedimentary particles, including freshwater algae, phytoclasts, amorphous organic matter, and many others. Examining the full spectrum of these particles is known as palynofacies analysis. It is one of the most powerful methods for reconstructing depositional environments in sedimentary rocks, as it relies on the distribution and relative abundances of these particles.

However, traditional counting methods for palynological and palynofacies analysis present several limitations. The counting area is rarely defined with precision, making it difficult to reproduce analyses. As a result, if any annotations need to be corrected, the entire counting workflow must be repeated. A particularly challenging aspect is the objective estimation of particles such as amorphous organic matter or phytoclasts, which are always fragmented and do not exist as discrete entities. Moreover, identification accuracy can vary substantially between analysts depending on experience, introducing challenges for reproducibility, comparability, and integration across datasets.

Digitizing palynological slides offers a promising opportunity to reduce subjectivity and personal bias by enabling particle annotation directly on high resolution digital images. This approach also supports iterative analysis, allowing annotations to be updated or refined without repeating the microscopy workflow. Through the ArtPOP project, we aim to develop objective, widely applicable annotation tool that enhance the robustness of paleoenvironmental reconstructions and facilitate integration across diverse palynological datasets. In this presentation, we provide an overview of challenges and advantages associated with digitizing the palynological workflow. We also present our preliminary results of the AI-augmented annotation of selected sedimentary particles.

How to cite: Śliwińska, K. K. and Andrianov, N.: ArtPOP - Automated RecogniTion of Palynomorphs and Organic sedimentary Particles , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10337, https://doi.org/10.5194/egusphere-egu26-10337, 2026.

The hydrological response of a basin is fundamentally controlled by geomorphic processes, structures, and physiographic characteristics. Horton’s geomorphological laws, basin topology, and kinematic properties have long been employed to derive flood response in ungauged basins through various Geomorphological Instantaneous Unit Hydrograph (GIUH) frameworks. This study investigates ten ungauged tributary sub-basins of the Shilabati River in eastern India to analyse how basin morphometry and topology regulate travel-time distribution of water particles and flash-flood potential. The Width Function Instantaneous Unit Hydrograph (WFIUH), a GIUH variant, is applied to derive the geomorphological control on peak flow and time to peak, while the morphometric analysis is performed to investigate the effect of basin characteristics on these hydrologic response parameters. The WFIUH is obtained using the flow length extracted from the SRTM DEM, together with spatially variable and fixed hillslope velocities estimated from land use-land cover and slope using the Soil Conservation Services (SCS), uniform-flow, and Manning’s velocity formulae. Due to the absence of observed streamflow, WFIUH results are evaluated against the Geomorpho-climatic Instantaneous Unit Hydrograph (GcIUH) derived from climate-dependent channel velocity and drainage network topology, as well as observed flood events. 
Results show that all variable-velocity WFIUHs have longer time bases and a lower peak flow than fixed-velocity WFIUHs, because the highest velocity cells are associated with the smallest drainage contributing areas. The SCS-based variable velocity WFIUH aligns with the GcIUH, reproducing both the peak flow and time to peak of the IUH more accurately compared to the other methods. Small, circular, and comparatively steeper sub-basins exhibit shorter times to peak (8.5-10.5 hours), indicating a high flash-flood potential, mainly in sub-basins 3-6. On the contrary, elongated and well-bifurcated sub-basins reveal slightly delayed peaks (10.5-15.5 h) but remain capable of producing moderate-to-high floods due to their larger drainage areas, as confirmed by the flash flood event in 2025 in sub-basins 1, 8-10. Correlation analysis reveals that circularity ratio, relief ratio, and hypsometric integral are positively associated with peak flow, suggesting enhanced flow synchronization in compact and steep sub-basins. In contrast, time to peak shows moderate to strong negative correlations with these parameters and positive correlations with stream length and bifurcation ratios, indicating delayed response in elongated and highly branched drainage networks due to dispersed flow paths.
Therefore, basin morphometry and drainage network topology effectively govern hydrologic responses of the sub-basins. The spatially variable SCS velocity-based WFIUH provides a more realistic depiction of hydrologic response in ungauged sub-basins. Hence, this method is well-suited for event-based lumped hydrological modelling as well as for sub-basin prioritization in flash flood risk assessment.

How to cite: Das, T. and Das, S.: Geomorphic controls on flood response using the Width Function Instantaneous Unit Hydrograph framework , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-570, https://doi.org/10.5194/egusphere-egu26-570, 2026.


Incipient sediment motion in turbulent flows remains difficult to characterize and predict because the underlying hydrodynamic forces are highly intermittent and events are sparse in time, even in well-controlled experiments. This study investigates whether temporal deep-learning architectures can detect the onset of particle motion directly from high-frequency velocity time series measured by an instrumented “smart sphere” [1, 2], without explicit force or torque measurements. The workflow includes detrending and cleaning of raw signals, physics-informed signal transforms (e.g. smoothed velocity, acceleration, jerk, and kinematic impulse proxies), segmentation with sliding windows, and supervised training of temporal deep-learning architectures, including recurrent, convolutional, and attention-based models, using class-imbalance mitigation such as focal loss, class weighting, and data augmentation.
Hyperparameter optimization is performed automatically with Optuna, and model performance is assessed using ROC and precision–recall curves, confusion matrices and time-resolved prediction performance. Results show that all tested architectures can learn consistent kinematic signatures preceding incipient motion from single-axis velocity time series, with models incorporating attention mechanisms achieving the highest recall on rare motion-onset events, consistent with their ability to focus on intermittent, high-magnitude kinematic bursts preceding entrainment. These findings demonstrate that deep learning applied to smart-particle sensor data can provide an efficient, non-intrusive tool for particle-scale sediment transport monitoring and real-time–capable event detection. The approach is directly relevant to the session’s focus on particle-scale transport mechanics and data-driven upscaling, and opens avenues for integrating deep-learning-based event detection into multi-scale sediment transport models in geophysical and engineered flows.

References
[1] Al-Obaidi, K., Xu, Y., & Valyrakis, M. (2020). The design and calibration of instrumented particles for assessing water infrastructure hazards. Journal of Sensor and Actuator Networks, 9(3), 36.
[2] AlObaidi, K., & Valyrakis, M. (2021). Linking the explicit probability of entrainment of instrumented particles to flow hydrodynamics. Earth Surface Processes and Landforms, 46(12), 2448-2465.

How to cite: Mavris, I. and Valyrakis, M.: Rare-event detection of incipient sediment motion from smart-particle time series using deep learning, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11678, https://doi.org/10.5194/egusphere-egu26-11678, 2026.

Study of flow-vegetation interactions in river channels is necessary to comprehend its importance in sediment transport and morphological changes. Numerous laboratory experiments, numerical modelling, and field data have been collected and analyzed by researchers throughout decades. Previous laboratory experiments simulating vegetation majorly studied the impacts from vegetation shoot width and density. However, studies showed that along with the shape and size of vegetation, root-soil binding capacity also plays an important role in the morphological changes in the channel. To test this theory, we conducted experiments using a flume of 15 m in length and 1.8 m in width at the University of British Columbia. The main channel and floodplain width considered is 60 cm each. Two sets of experiments with and without vegetation roots in the floodplains were conducted. 3D printer was used to model the floodplain vegetation (see Figure). In the case of vegetation with roots, we considered it as a taproot system with a spiral structure attached to the simple root-shoot system as seen in the figure. Preliminary tests showed vegetation with roots was able to sustain the force of flow in different discharges in a better way without getting uprooted compared to vegetation without roots. Furthermore, there is also a difference in the morphology of the channels between the with and without roots experiments. The initial study showed that incorporating vegetation roots in the laboratory provides a more effective means of understanding flow-vegetation interactions and channel evolution. Furthermore, this study will also be helpful for the advancement of nature-based solutions like soil bioengineering techniques.

                           Simple root-shoot system                                                                               Taproot-shoot system

How to cite: Barman, J. and Hassan, M.: Can vegetation root simulation in the laboratory lead to better understanding of flow-vegetation interactions?, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5549, https://doi.org/10.5194/egusphere-egu26-5549, 2026.

How to cite: Gailleton, B., Steer, P., Cordonnier, G., and Clubb, F.: Efficient Hydrodynamic Modeling at the Landscape Scale: Quantifying River Width and Shear Stress Variability to Decode Tectonic Signals, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11152, https://doi.org/10.5194/egusphere-egu26-11152, 2026.

Calcareous valves of various ostracod species from the Miocene (Burdigalian) Quilon Formation, Kerala Basin, southwest India, were separated and identified up to the species level. The 15 most abundant species were selected to determine the carbon and oxygen isotope composition, with 2 to 5 replicates to assess the variation among individual valves within each species. The δ¹³C ratios range from 0.56 to -4.65‰ VPDB with a standard deviation range between 0.08 to 0.53‰. The δ¹⁸O ratios varied between -2.57 to -4.25‰ VPDB with a standard deviation between 0.12‰ and 0.46‰. The seawater δ¹⁸O values were calculated using the empirical equation by Kim and Neil (1997), and they range between -3.08‰ to -0.01‰ (VSMOW), with an average of -1.85‰ (VSMOW). This study also tries to categorise the species into distinct habitat groups, namely the open ocean, mixed estuarine and shallow-marine environment with significant coastal upwelling influence, based on their isotopic composition. The results were compared with the habitats of their extant relatives at the family and genus levels, as well as information derived from valve ornamentations. Ostracods, namely Phlyctenophora meridionalis, Paranesidea cf. gajensis, Bairdoppilata sp., and Krithe autochthona inhabited a range of settings from shallow to deeper marine environments. The species Aurila singhi, Paractinocythereis gujaratensis, Stigmatocythere sp., Actinocythereis sp., Trachyleberis sp., Neocyprideis murudensis, Pokornyella chaasraensis, and Tenedocythere keralaensis are identified to inhabit an estuarine or shallow-marine environment influenced by freshwater influx. Whereas Paijenborchellina prona, Cytherelloidea sp., and Loxoconcha confinis show an indication of a shallow-marine environment with significant coastal upwelling influence.

How to cite: m s, A., kannan, P., and V Kapur, V.: Ecological and hydrological reconstruction of the western Indian coastal ocean during the Early Miocene (Burdigalian) based on the oxygen and carbon isotopes of multiple ostracod species., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-455, https://doi.org/10.5194/egusphere-egu26-455, 2026.

Methane stored in shallow marine sediments significantly affects seafloor stability, and influence ocean-atmosphere interactions. Since methane is a potent greenhouse gas, its release influences regional biogeochemical cycles and benthic ecosystems. Along continental margins, favourable conditions promote biogenic methanogenesis and gas hydrate formation. Understanding how methane migrates beneath the base of hydrate stability is therefore essential, particularly because hydrate dissociation near the feather edge of continental slope releases methane to the seabed. Pockmarks form when gas escapes from shallow overpressure zones. Overpressure may develop through hydrate dissociation or through the accumulation of free gas below low-permeability layers. Once pressure exceeds the sealing capacity of the overlying sediments, gas can migrate upward and eventually vent at the seabed.

In the offshore Taranaki Basin, west of New Zealand’s North Island, high-resolution 3D seismic data reveal ~300 pockmarks between 300-700 m water depth. Beneath many of these pockmarks, the seismic data show tiers of near-vertically stacked shallow-gas bright spots, indicating focused migration pathways in the shallow subsurface across the foresets of a prograding clinoform system.

The theoretical stability limit for pure methane hydrates locally aligns with the shallowest bright anomalies. However, most anomalies lie within the free-gas zone landward of the methane-hydrate outcrop and beneath large parts of the pockmark field. Over the past ~16 kyr, bottom-water temperatures along the slope have warmed by ~2.25 °C, shifting the hydrate-stability feather edge downslope by ~1.7 km. This warming-driven retreat  can account for only ~20% of the observed pockmarks. While the presence of gas hydrates can deflect gas updip, there is no clear seismic evidence for a bottom-simulating reflection. Instead, gas appears to ascend upslope through a range of stratigraphic heterogeneities, such as cyclic steps that climb obliquely, scour rims, channel cuts, and levee deposits, which collectively provide localized pathways for migration.

In gently dipping (2-3°) slope, free gas beneath the hydrate stability zone would preferentially migrate updip along permeable strata toward the shelf edge. However, 3D seismic data show bright spots concentrated within scour rims, channel levees, and the crests of cyclic steps that act as effective traps updip of the upper limit of hydrate stability at the clinoform foresets. Gas is accumulated within levee deposits of vertically aggrading and laterally shifting channel-levee systems, where repeated cut-and-fill cycles build stacked fining-upward units. The climbing geometry of cyclic steps redirects gas vertically upslope along their crests, enhancing upward migration, while fine-grained scour infill inhibit lateral migration.3D visualization shows that such traps form multiple tiers of shallow-gas pockets linked by focused gas-flow. Together, these relationships demonstrate that fluid migration is strongly controlled by sedimentary architecture shaped by turbidity current-controlled depositional processes at the foresets of the prograding clinoforms. The clustering of numerous pockmarks above these vertically stacked gas zones strongly indicates that stratigraphic focusing, rather than along-slope migration at the base of the hydrate stability zone, controls gas ascent.

How to cite: Bhattacharya, I. and Sarkar, S.: Stratigraphic Controls on Gas Migration and Pockmark Formation at the foreset of a Prograding Clinoform System west of North Island, New Zealand, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-724, https://doi.org/10.5194/egusphere-egu26-724, 2026.

Benthic ostracods serve as effective bioindicators of sediment quality and metal enrichment in coastal systems, but quantitative tools linking their community structure to multi-metal contamination are limited. This study develops and validates two ostracod-based biotic indices, the Ostracoda Assemblage Pollution Index (OAPI) and its reduced form, Mini-OAPI, to evaluate benthic ecological responses to metal contamination on the Vedaranyam shelf, Bay of Bengal. Twenty-eight surface sediment samples were analysed for Fe, Mn, Cr, Cu, Ni, Pb, and Zn concentrations along with ostracod assemblage data. The indices integrate species diversity, functional guild composition, and normalized pollution load to produce a tolerance-weighted ecological deviation measure. The OAPI includes diversity, evenness, guild shift, and pollution load, performs best in data-rich settings, while Mini-OAPI shows stable diagnostic behaviour under data-limited conditions and consistently captures ecological responses along contamination gradients. Normalization to a 0-1 scale and the use of standardized disturbance classes (Low = 0.00-0.33; Moderate = 0.34-0.66; High = 0.67-1.00) ensure comparability across marine and estuarine systems. A unimodal diversity-pollution pattern consistent with the Intermediate Disturbance Hypothesis and weak Cu-organic associations indicate complex metal-biota interactions. These indices provide transferable, tolerance-weighted tools for ecological assessment and understanding ecosystem responses to environmental change.

Keywords: Ostracoda Assemblage Pollution Index (OAPI); Mini-OAPI; benthic bioindicators; non-linear ecological modelling; Intermediate Disturbance Hypothesis (IDH); copper paradox; marine pollution assessment.

How to cite: Palanichamy, P., Vimal Kanth, S., Chellamuthu, S. N., Subramani, R., and Hussain, S. M.: Non-linear ecological responses of Ostracod communities to multi-metal pollution based on tolerance-weighted indices from the Vedaranyam shelf, Bay of Bengal, India., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12898, https://doi.org/10.5194/egusphere-egu26-12898, 2026.