Articles | Volume 13, issue 1
https://doi.org/10.5194/esurf-13-167-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/esurf-13-167-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
07 Feb 2025
Research article |
| 07 Feb 2025
| 07 Feb 2025
Automatic detection of floating instream large wood in videos using deep learning
Faculty of Geosciences and Environment (FGSE), Institute of Earth Surface Dynamics (IDYST), Université de Lausanne, Quartier UNIL-Mouline – Bâtiment Géopolis, 1015 Lausanne, Switzerland
Invited contribution by Janbert Aarnink, recipient of the EGU Geomorphology Outstanding Student and PhD candidate Presentation Award 2022.
Tom Beucler
Faculty of Geosciences and Environment (FGSE), Institute of Earth Surface Dynamics (IDYST), Université de Lausanne, Quartier UNIL-Mouline – Bâtiment Géopolis, 1015 Lausanne, Switzerland
Expertise Center for Climate Extremes, Université de Lausanne, 1015 Lausanne, Switzerland
Marceline Vuaridel
Faculty of Geosciences and Environment (FGSE), Institute of Earth Surface Dynamics (IDYST), Université de Lausanne, Quartier UNIL-Mouline – Bâtiment Géopolis, 1015 Lausanne, Switzerland
Virginia Ruiz-Villanueva
Faculty of Geosciences and Environment (FGSE), Institute of Earth Surface Dynamics (IDYST), Université de Lausanne, Quartier UNIL-Mouline – Bâtiment Géopolis, 1015 Lausanne, Switzerland
Institute of Geography, University of Bern, Hallerstrasse 12, 3012 Bern, Switzerland
Related authors
No articles found.
Tabea Cache, Milton Salvador Gomez, Tom Beucler, Jovan Blagojevic, João Paulo Leitao, and Nadav Peleg
Hydrol. Earth Syst. Sci., 28, 5443–5458, https://doi.org/10.5194/hess-28-5443-2024, https://doi.org/10.5194/hess-28-5443-2024, 2024
Short summary
Short summary
We introduce a new deep-learning model that addresses the limitations of existing urban flood models in handling varied terrains and rainfall events. Our model subdivides a city into small patches and presents a novel approach to incorporate broader terrain information. It accurately predicts high-resolution flood maps across diverse rainfall events and cities (on minute and meter scales) that haven’t been seen by the model, which offers valuable insights for urban flood mitigation strategies.
Nicolas Steeb, Virginia Ruiz-Villanueva, Alexandre Badoux, Christian Rickli, Andrea Mini, Markus Stoffel, and Dieter Rickenmann
Earth Surf. Dynam., 11, 487–509, https://doi.org/10.5194/esurf-11-487-2023, https://doi.org/10.5194/esurf-11-487-2023, 2023
Short summary
Short summary
Various models have been used in science and practice to estimate how much large wood (LW) can be supplied to rivers. This contribution reviews the existing models proposed in the last 35 years and compares two of the most recent spatially explicit models by applying them to 40 catchments in Switzerland. Differences in modelling results are discussed, and results are compared to available observations coming from a unique database.
Related subject area
Physical: Geomorphology (including all aspects of fluvial, coastal, aeolian, hillslope and glacial geomorphology)
Geomorphic imprint of high-mountain floods: insights from the 2022 hydrological extreme across the upper Indus River catchment in the northwestern Himalayas
A numerical model for duricrust formation by water table fluctuations
Width evolution of channel belts as a random walk
Evidence of slow millennial cliff retreat rates using cosmogenic nuclides in coastal colluvium
Equilibrium distance from long-range dune interactions
Examination of analytical shear stress predictions for coastal dune evolution
Post-fire evolution of ravel transport regimes in the Diablo Range, CA
Landscape response to tectonic deformation and cyclic climate change since ca. 800 ka in the southern central Andes
The Aare main overdeepening on the northern margin of the European Alps: basins, riegels, and slot canyons
A simple model for faceted topographies at normal faults based on an extended stream-power law
Testing floc settling velocity models in rivers and freshwater wetlands
River suspended-sand flux computation with uncertainty estimation using water samples and high-resolution ADCP measurements
Barchan swarm dynamics from a Two-Flank Agent-Based Model
A landslide runout model for sediment transport, landscape evolution, and hazard assessment applications
Tracking slow-moving landslides with PlanetScope data: new perspectives on the satellite's perspective
Topographic metrics for unveiling fault segmentation and tectono-geomorphic evolution with insights into the impact of inherited topography, Ulsan Fault Zone, South Korea
Acceleration of coastal-retreat rates for high-Arctic rock cliffs on Brøggerhalvøya, Svalbard, over the past decade
The impact of bedrock meander cutoffs on 50 kyr scale incision rates, San Juan River, Utah
How water, temperature, and seismicity control the preconditioning of massive rock slope failure (Hochvogel)
Large structure simulation for landscape evolution models
Surficial sediment remobilization by shear between sediment and water above tsunamigenic megathrust ruptures: experimental study
Terrace formation linked to outburst floods at the Diexi palaeo-landslide dam, upper Minjiang River, eastern Tibetan Plateau
Pliocene shorelines and the epeirogenic motion of continental margins: a target dataset for dynamic topography models
Decadal-scale decay of landslide-derived fluvial suspended sediment after Typhoon Morakot
Role of the forcing sources in morphodynamic modelling of an embayed beach
A machine learning approach to the geomorphometric detection of ribbed moraines in Norway
Stream hydrology controls on ice cliff evolution and survival on debris-covered glaciers
Time-varying drainage basin development and erosion on volcanic edifices
Geomorphic risk maps for river migration using probabilistic modeling – a framework
Evolution of submarine canyons and hanging-wall fans: insights from geomorphic experiments and morphodynamic models
Riverine sediment response to deforestation in the Amazon basin
Physical modeling of ice-sheet-induced salt movements using the example of northern Germany
Geometric constraints on tributary fluvial network junction angles
A new dunetracking tool to support input parameter selection and uncertainty analyses using a Monte Carlo approach
An evaluation of flow-routing algorithms for calculating contributing area on regular grids
Downstream rounding rate of pebbles in the Himalaya
Haloturbation in the northern Atacama Desert revealed by a hidden subsurface network of calcium sulphate wedges
A physics-based model for fluvial valley width
Sub-surface processes and heat fluxes at coarse-blocky Murtèl rock glacier (Engadine, eastern Swiss Alps)
Implications for the resilience of modern coastal systems derived from mesoscale barrier dynamics at Fire Island, New York
Quantifying the migration rate of drainage divides from high-resolution topographic data
Long-term monitoring (1953–2019) of geomorphologically active sections of Little Ice Age lateral moraines in the context of changing meteorological conditions
Coevolving edge rounding and shape of glacial erratics: the case of Shap granite, UK
Dimensionless argument: a narrow grain size range near 2 mm plays a special role in river sediment transport and morphodynamics
Path length and sediment transport estimation from DEMs of difference: a signal processing approach
Influence of cohesive clay on wave–current ripple dynamics captured in a 3D phase diagram
Statistical characterization of erosion and sediment transport mechanics in shallow tidal environments – Part 1: Erosion dynamics
Statistical characterization of erosion and sediment transport mechanics in shallow tidal environments – Part 2: Suspended sediment dynamics
Geomorphological and hydrological controls on sediment export in earthquake-affected catchments in the Nepal Himalaya
Optimization of passive acoustic bedload monitoring in rivers by signal inversion
Abhishek Kashyap, Kristen L. Cook, and Mukunda Dev Behera
Earth Surf. Dynam., 13, 147–166, https://doi.org/10.5194/esurf-13-147-2025, https://doi.org/10.5194/esurf-13-147-2025, 2025
Short summary
Short summary
Short-lived, high-magnitude flood events across high mountain regions leave substantial geomorphic imprints, which are frequently triggered by excess precipitation, glacial lake outbursts, and natural dam breaches. These catastrophic floods highlight the importance of understanding the complex interaction between climatic, hydrological, and geological forces in bedrock catchments. Extreme floods can have long-term geomorphic consequences on river morphology and fluvial processes.
Caroline Fenske, Jean Braun, François Guillocheau, and Cécile Robin
Earth Surf. Dynam., 13, 119–146, https://doi.org/10.5194/esurf-13-119-2025, https://doi.org/10.5194/esurf-13-119-2025, 2025
Short summary
Short summary
We have developed a new numerical model to represent the formation of duricrusts, which are hard mineral layers found in soils and at the surface of the Earth. We assume that the formation mechanism implies variations in the height of the water table and that the hardening rate is proportional to precipitation. The model allows us to quantify the potential feedbacks they generate on the surface topography and the thickness of the regolith/soil layer.
Jens M. Turowski, Fergus McNab, Aaron Bufe, and Stefanie Tofelde
Earth Surf. Dynam., 13, 97–117, https://doi.org/10.5194/esurf-13-97-2025, https://doi.org/10.5194/esurf-13-97-2025, 2025
Short summary
Short summary
Channel belts comprise the area affected by a river due to lateral migration and floods. As a landform, they affect water resources and flood hazard, and they often host unique ecological communities. We develop a model describing the evolution of channel-belt area over time. The model connects the behaviour of the river to the evolution of the channel belt over a timescale of centuries. A comparison to selected data from experiments and real river systems verifies the random walk approach.
Rémi Bossis, Vincent Regard, Sébastien Carretier, and Sandrine Choy
Earth Surf. Dynam., 13, 71–79, https://doi.org/10.5194/esurf-13-71-2025, https://doi.org/10.5194/esurf-13-71-2025, 2025
Short summary
Short summary
The erosion of rocky coasts occurs episodically through wave action and landslides, constituting a major natural hazard. Documenting the factors that control the coastal retreat rate over millennia is fundamental to evidencing any change in time. However, the known rates to date are essentially representative of the last few decades. Here, we present a new method using the concentration of an isotope, 10Be, in sediment eroded from the cliff to quantify its retreat rate averaged over millennia.
Jean Vérité, Clément Narteau, Olivier Rozier, Jeanne Alkalla, Laurie Barrier, and Sylvain Courrech du Pont
Earth Surf. Dynam., 13, 23–39, https://doi.org/10.5194/esurf-13-23-2025, https://doi.org/10.5194/esurf-13-23-2025, 2025
Short summary
Short summary
Using a numerical model in 2D, we study how two identical dunes interact with each other when exposed to reversing winds. Depending on the distance between the dunes, they either repel or attract each other until they reach an equilibrium distance, which is controlled by the wind strength, wind reversal frequency, and dune size. This process is controlled by the modification of wind flow over dunes of various shapes, influencing the sediment transport downstream.
Orie Cecil, Nicholas Cohn, Matthew Farthing, Sourav Dutta, and Andrew Trautz
Earth Surf. Dynam., 13, 1–22, https://doi.org/10.5194/esurf-13-1-2025, https://doi.org/10.5194/esurf-13-1-2025, 2025
Short summary
Short summary
Using computational fluid dynamics, we analyze the error trends of an analytical shear stress distribution model used to drive aeolian transport for coastal dunes, which are an important line of defense against storm-related flooding hazards. We find that compared to numerical simulations, the analytical model results in a net overprediction of the landward migration rate. Additionally, two data-driven approaches are proposed for reducing the error while maintaining computational efficiency.
Hayden L. Jacobson, Danica L. Roth, Gabriel Walton, Margaret Zimmer, and Kerri Johnson
Earth Surf. Dynam., 12, 1415–1446, https://doi.org/10.5194/esurf-12-1415-2024, https://doi.org/10.5194/esurf-12-1415-2024, 2024
Short summary
Short summary
Loose grains travel farther after a fire because no vegetation is left to stop them. This matters since loose grains at the base of a slope can turn into a debris flow if it rains. To find if grass growing back after a fire had different impacts on grains of different sizes on slopes of different steepness, we dropped thousands of natural grains and measured how far they went. Large grains went farther 7 months after the fire than 11 months after, and small grain movement didn’t change much.
Elizabeth N. Orr, Taylor F. Schildgen, Stefanie Tofelde, Hella Wittmann, and Ricardo N. Alonso
Earth Surf. Dynam., 12, 1391–1413, https://doi.org/10.5194/esurf-12-1391-2024, https://doi.org/10.5194/esurf-12-1391-2024, 2024
Short summary
Short summary
Fluvial terraces and alluvial fans in the Toro Basin, NW Argentina, record river evolution and global climate cycles over time. Landform dating reveals lower-frequency climate cycles (100 kyr) preserved downstream and higher-frequency cycles (21/40 kyr) upstream, supporting theoretical predications that longer rivers filter out higher-frequency climate signals. This finding improves our understanding of the spatial distribution of sedimentary paleoclimate records within landscapes.
Fritz Schlunegger, Edi Kissling, Dimitri Tibo Bandou, Guilhem Amin Douillet, David Mair, Urs Marti, Regina Reber, Patrick Schläfli, and Michael Alfred Schwenk
Earth Surf. Dynam., 12, 1371–1389, https://doi.org/10.5194/esurf-12-1371-2024, https://doi.org/10.5194/esurf-12-1371-2024, 2024
Short summary
Short summary
Overdeepenings are bedrock depressions filled with sediment. We combine the results of a gravity survey with drilling data to explore the morphology of such a depression beneath the city of Bern. We find that the target overdeepening comprises two basins >200 m deep. They are separated by a bedrock riegel that itself is cut by narrow canyons up to 150 m deep. We postulate that these structures formed underneath a glacier, where erosion by subglacial meltwater caused the formation of the canyons.
Stefan Hergarten
Earth Surf. Dynam., 12, 1315–1327, https://doi.org/10.5194/esurf-12-1315-2024, https://doi.org/10.5194/esurf-12-1315-2024, 2024
Short summary
Short summary
Faceted topographies are impressive footprints of active tectonics in geomorphology. This paper investigates the evolution of faceted topographies at normal faults and their interaction with a river network theoretically and numerically. As a main result beyond several relations for the geometry of facets, the horizontal displacement associated with normal faults is crucial for the dissection of initially polygonal facets into triangular facets bounded by almost parallel rivers.
Justin A. Nghiem, Gen K. Li, Joshua P. Harringmeyer, Gerard Salter, Cédric G. Fichot, Luca Cortese, and Michael P. Lamb
Earth Surf. Dynam., 12, 1267–1294, https://doi.org/10.5194/esurf-12-1267-2024, https://doi.org/10.5194/esurf-12-1267-2024, 2024
Short summary
Short summary
Fine sediment grains in freshwater can cohere into faster-settling particles called flocs, but floc settling velocity theory has not been fully validated. Combining three data sources in novel ways in the Wax Lake Delta, we verified a semi-empirical model relying on turbulence and geochemical factors. For a physics-based model, we showed that the representative grain diameter within flocs relies on floc structure and that heterogeneous flow paths inside flocs increase floc settling velocity.
Jessica Marggraf, Guillaume Dramais, Jérôme Le Coz, Blaise Calmel, Benoît Camenen, David J. Topping, William Santini, Gilles Pierrefeu, and François Lauters
Earth Surf. Dynam., 12, 1243–1266, https://doi.org/10.5194/esurf-12-1243-2024, https://doi.org/10.5194/esurf-12-1243-2024, 2024
Short summary
Short summary
Suspended-sand fluxes in rivers vary with time and space, complicating their measurement. The proposed method captures the vertical and lateral variations of suspended-sand concentration throughout a river cross-section. It merges water samples taken at various positions throughout the cross-section with high-resolution acoustic velocity measurements. This is the first method that includes a fully applicable uncertainty estimation; it can easily be applied to any other study sites.
Dominic T. Robson and Andreas C. W. Baas
Earth Surf. Dynam., 12, 1205–1226, https://doi.org/10.5194/esurf-12-1205-2024, https://doi.org/10.5194/esurf-12-1205-2024, 2024
Short summary
Short summary
Barchans are fast-moving sand dunes which form large populations (swarms) on Earth and Mars. We show that a small range of model parameters produces swarms in which dune size does not vary downwind – something that is observed in nature but not when using earlier models. We also show how the shape of dunes and the spatial patterns they form are affected by wind direction. This work furthers our understanding of the interplay between environmental drivers, dune interactions, and swarm properties.
Jeffrey Keck, Erkan Istanbulluoglu, Benjamin Campforts, Gregory Tucker, and Alexander Horner-Devine
Earth Surf. Dynam., 12, 1165–1191, https://doi.org/10.5194/esurf-12-1165-2024, https://doi.org/10.5194/esurf-12-1165-2024, 2024
Short summary
Short summary
MassWastingRunout (MWR) is a new landslide runout model designed for sediment transport, landscape evolution, and hazard assessment applications. MWR is written in Python and includes a calibration utility that automatically determines best-fit parameters for a site and empirical probability density functions of each parameter for probabilistic model implementation. MWR and Jupyter Notebook tutorials are available as part of the Landlab package at https://github.com/landlab/landlab.
Ariane Mueting and Bodo Bookhagen
Earth Surf. Dynam., 12, 1121–1143, https://doi.org/10.5194/esurf-12-1121-2024, https://doi.org/10.5194/esurf-12-1121-2024, 2024
Short summary
Short summary
This study investigates the use of optical PlanetScope data for offset tracking of the Earth's surface movement. We found that co-registration accuracy is locally degraded when outdated elevation models are used for orthorectification. To mitigate this bias, we propose to only correlate scenes acquired from common perspectives or base orthorectification on more up-to-date elevation models generated from PlanetScope data alone. This enables a more detailed analysis of landslide dynamics.
Cho-Hee Lee, Yeong Bae Seong, John Weber, Sangmin Ha, Dong-Eun Kim, and Byung Yong Yu
Earth Surf. Dynam., 12, 1091–1120, https://doi.org/10.5194/esurf-12-1091-2024, https://doi.org/10.5194/esurf-12-1091-2024, 2024
Short summary
Short summary
Topographic metrics were used to understand changes due to tectonic activity. We evaluated the relative tectonic activity along the Ulsan Fault Zone (UFZ), one of the most active fault zones in South Korea. We divided the UFZ into five segments, based on the spatial variation in activity. We modeled the landscape evolution of the study area and interpreted tectono-geomorphic history during which the northern part of the UFZ experienced asymmetric uplift, while the southern part did not.
Juditha Aga, Livia Piermattei, Luc Girod, Kristoffer Aalstad, Trond Eiken, Andreas Kääb, and Sebastian Westermann
Earth Surf. Dynam., 12, 1049–1070, https://doi.org/10.5194/esurf-12-1049-2024, https://doi.org/10.5194/esurf-12-1049-2024, 2024
Short summary
Short summary
Coastal rock cliffs on Svalbard are considered to be fairly stable; however, long-term trends in coastal-retreat rates remain unknown. This study examines changes in the coastline position along Brøggerhalvøya, Svalbard, using aerial images from 1970, 1990, 2010, and 2021. Our analysis shows that coastal-retreat rates accelerate during the period 2010–2021, which coincides with increasing storminess and retreating sea ice.
Aaron T. Steelquist, Gustav B. Seixas, Mary L. Gillam, Sourav Saha, Seulgi Moon, and George E. Hilley
Earth Surf. Dynam., 12, 1071–1089, https://doi.org/10.5194/esurf-12-1071-2024, https://doi.org/10.5194/esurf-12-1071-2024, 2024
Short summary
Short summary
The rates at which rivers erode their bed can be used to interpret the geologic history of a region. However, these rates depend significantly on the time window over which you measure. We use multiple dating methods to determine an incision rate for the San Juan River and compare it to regional rates with longer timescales. We demonstrate how specific geologic events, such as cutoffs of bedrock meander bends, are likely to preserve material we can date but also bias the rates we measure.
Johannes Leinauer, Michael Dietze, Sibylle Knapp, Riccardo Scandroglio, Maximilian Jokel, and Michael Krautblatter
Earth Surf. Dynam., 12, 1027–1048, https://doi.org/10.5194/esurf-12-1027-2024, https://doi.org/10.5194/esurf-12-1027-2024, 2024
Short summary
Short summary
Massive rock slope failures are a significant alpine hazard and change the Earth's surface. Therefore, we must understand what controls the preparation of such events. By correlating 4 years of slope displacements with meteorological and seismic data, we found that water from rain and snowmelt is the most important driver. Our approach is applicable to similar sites and indicates where future climatic changes, e.g. in rain intensity and frequency, may alter the preparation of slope failure.
Julien Coatléven and Benoit Chauveau
Earth Surf. Dynam., 12, 995–1026, https://doi.org/10.5194/esurf-12-995-2024, https://doi.org/10.5194/esurf-12-995-2024, 2024
Short summary
Short summary
The aim of this paper is to explain how to incorporate classical water flow routines into landscape evolution models while keeping numerical errors under control. The key idea is to adapt filtering strategies to eliminate anomalous numerical errors and mesh dependencies, as confirmed by convergence tests with analytic solutions. The emergence of complex geomorphic structures is now driven exclusively by nonlinear heterogeneous physical processes rather than by random numerical artifacts.
Chloé Seibert, Cecilia McHugh, Chris Paola, Leonardo Seeber, and James Tucker
EGUsphere, https://doi.org/10.5194/egusphere-2024-2011, https://doi.org/10.5194/egusphere-2024-2011, 2024
Short summary
Short summary
We propose a new mechanism of widespread surficial co-seismic sediment entrainment by seismic motions in subduction earthquakes. Our physical experiments show that shear from sediment-water relative velocities from long-period earthquake motions can mobilize synthetic fine marine sediment. High frequency vertical shaking can enhance this mobilization. According to our results, the largest tsunamigenic earthquakes that rupture to the trench may be distinguishable in the sedimentary record.
Jingjuan Li, John D. Jansen, Xuanmei Fan, Zhiyong Ding, Shugang Kang, and Marco Lovati
Earth Surf. Dynam., 12, 953–971, https://doi.org/10.5194/esurf-12-953-2024, https://doi.org/10.5194/esurf-12-953-2024, 2024
Short summary
Short summary
In this study, we investigated the geomorphology, sedimentology, and chronology of Tuanjie (seven terraces) and Taiping (three terraces) terraces in Diexi, eastern Tibetan Plateau. Results highlight that two damming and three outburst events occurred in the area during the late Pleistocene, and the outburst floods have been a major factor in the formation of tectonically active mountainous river terraces. Tectonic activity and climatic changes play a minor role.
Andrew Hollyday, Maureen E. Raymo, Jacqueline Austermann, Fred Richards, Mark Hoggard, and Alessio Rovere
Earth Surf. Dynam., 12, 883–905, https://doi.org/10.5194/esurf-12-883-2024, https://doi.org/10.5194/esurf-12-883-2024, 2024
Short summary
Short summary
Sea level was significantly higher during the Pliocene epoch, around 3 million years ago. The present-day elevations of shorelines that formed in the past provide a data constraint on the extent of ice sheet melt and the global sea level response under warm Pliocene conditions. In this study, we identify 10 escarpments that formed from wave-cut erosion during Pliocene times and compare their elevations with model predictions of solid Earth deformation processes to estimate past sea level.
Gregory A. Ruetenik, Ken L. Ferrier, and Odin Marc
Earth Surf. Dynam., 12, 863–881, https://doi.org/10.5194/esurf-12-863-2024, https://doi.org/10.5194/esurf-12-863-2024, 2024
Short summary
Short summary
Fluvial sediment fluxes increased dramatically in Taiwan during Typhoon Morakot in 2009, which produced some of the heaviest landsliding on record. We analyzed fluvial discharge and suspended sediment concentration data at 87 gauging stations across Taiwan to quantify fluvial sediment responses since Morakot. In basins heavily impacted by landsliding, rating curve coefficients sharply increased during Morakot and then declined exponentially with a characteristic decay time of <10 years.
Nil Carrion-Bertran, Albert Falqués, Francesca Ribas, Daniel Calvete, Rinse de Swart, Ruth Durán, Candela Marco-Peretó, Marta Marcos, Angel Amores, Tim Toomey, Àngels Fernández-Mora, and Jorge Guillén
Earth Surf. Dynam., 12, 819–839, https://doi.org/10.5194/esurf-12-819-2024, https://doi.org/10.5194/esurf-12-819-2024, 2024
Short summary
Short summary
The sensitivity to the wave and sea-level forcing sources in predicting a 6-month embayed beach evolution is assessed using two different morphodynamic models. After a successful model calibration using in situ data, other sources are applied. The wave source choice is critical: hindcast data provide wrong results due to an angle bias, whilst the correct dynamics are recovered with the wave conditions from an offshore buoy. The use of different sea-level sources gives no significant differences.
Thomas J. Barnes, Thomas V. Schuler, Simon Filhol, and Karianne S. Lilleøren
Earth Surf. Dynam., 12, 801–818, https://doi.org/10.5194/esurf-12-801-2024, https://doi.org/10.5194/esurf-12-801-2024, 2024
Short summary
Short summary
In this paper, we use machine learning to automatically outline landforms based on their characteristics. We test several methods to identify the most accurate and then proceed to develop the most accurate to improve its accuracy further. We manage to outline landforms with 65 %–75 % accuracy, at a resolution of 10 m, thanks to high-quality/high-resolution elevation data. We find that it is possible to run this method at a country scale to quickly produce landform inventories for future studies.
Eric Petersen, Regine Hock, and Michael G. Loso
Earth Surf. Dynam., 12, 727–745, https://doi.org/10.5194/esurf-12-727-2024, https://doi.org/10.5194/esurf-12-727-2024, 2024
Short summary
Short summary
Ice cliffs are melt hot spots that increase melt rates on debris-covered glaciers which otherwise see a reduction in melt rates. In this study, we show how surface runoff streams contribute to the generation, evolution, and survival of ice cliffs by carving into the glacier and transporting rocky debris. On Kennicott Glacier, Alaska, 33 % of ice cliffs are actively influenced by streams, while nearly half are within 10 m of streams.
Daniel O'Hara, Liran Goren, Roos M. J. van Wees, Benjamin Campforts, Pablo Grosse, Pierre Lahitte, Gabor Kereszturi, and Matthieu Kervyn
Earth Surf. Dynam., 12, 709–726, https://doi.org/10.5194/esurf-12-709-2024, https://doi.org/10.5194/esurf-12-709-2024, 2024
Short summary
Short summary
Understanding how volcanic edifices develop drainage basins remains unexplored in landscape evolution. Using digital evolution models of volcanoes with varying ages, we quantify the geometries of their edifices and associated drainage basins through time. We find that these metrics correlate with edifice age and are thus useful indicators of a volcano’s history. We then develop a generalized model for how volcano basins develop and compare our results to basin evolution in other settings.
Brayden Noh, Omar Wani, Kieran B. J. Dunne, and Michael P. Lamb
Earth Surf. Dynam., 12, 691–708, https://doi.org/10.5194/esurf-12-691-2024, https://doi.org/10.5194/esurf-12-691-2024, 2024
Short summary
Short summary
In this paper, we propose a framework for generating risk maps that provide the probabilities of erosion due to river migration. This framework uses concepts from probability theory to learn the river migration model's parameter values from satellite data while taking into account parameter uncertainty. Our analysis shows that such geomorphic risk estimation is more reliable than models that do not explicitly consider various sources of variability and uncertainty.
Steven Y. J. Lai, David Amblas, Aaron Micallef, and Hervé Capart
Earth Surf. Dynam., 12, 621–640, https://doi.org/10.5194/esurf-12-621-2024, https://doi.org/10.5194/esurf-12-621-2024, 2024
Short summary
Short summary
This study explores the creation of submarine canyons and hanging-wall fans on active faults, which can be defined by gravity-dominated breaching and underflow-dominated diffusion processes. The study reveals the self-similarity in canyon–fan long profiles, uncovers Hack’s scaling relationship and proposes a formula to estimate fan volume using canyon length. This is validated by global data from source-to-sink systems, providing insights into deep-water sedimentary processes.
Anuska Narayanan, Sagy Cohen, and John R. Gardner
Earth Surf. Dynam., 12, 581–599, https://doi.org/10.5194/esurf-12-581-2024, https://doi.org/10.5194/esurf-12-581-2024, 2024
Short summary
Short summary
This study investigates the profound impact of deforestation in the Amazon on sediment dynamics. Novel remote sensing data and statistical analyses reveal significant changes, especially in heavily deforested regions, with rapid effects within a year. In less disturbed areas, a 1- to 2-year lag occurs, influenced by natural sediment shifts and human activities. These findings highlight the need to understand the consequences of human activity for our planet's future.
Jacob Hardt, Tim P. Dooley, and Michael R. Hudec
Earth Surf. Dynam., 12, 559–579, https://doi.org/10.5194/esurf-12-559-2024, https://doi.org/10.5194/esurf-12-559-2024, 2024
Short summary
Short summary
We investigate the reaction of salt structures on ice sheet transgressions. We used a series of sandbox models that enabled us to experiment with scaled-down versions of salt bodies from northern Germany. The strongest reactions occurred when large salt pillows were partly covered by the ice load. Subsurface salt structures may play an important role in the energy transition, e.g., as energy storage. Thus, it is important to understand all processes that affect their stability.
Jon D. Pelletier, Robert G. Hayes, Olivia Hoch, Brendan Fenerty, and Luke A. McGuire
EGUsphere, https://doi.org/10.5194/egusphere-2024-1153, https://doi.org/10.5194/egusphere-2024-1153, 2024
Short summary
Short summary
On the gently sloping landscapes next to mountain fronts, junction angles tend to be lower (more acute), while in bedrock landscapes where the initial landscape or tectonic forcing is likely more spatially variable, junction angles tend to be larger (more obtuse). We demonstrate this using an analysis of ~20 million junction angles for the U.S.A., augmented by analyses of the Loess Plateau, China, and synthetic landscapes.
Julius Reich and Axel Winterscheid
EGUsphere, https://doi.org/10.5194/egusphere-2024-579, https://doi.org/10.5194/egusphere-2024-579, 2024
Short summary
Short summary
Analysing the geometry and the dynamics of riverine bedforms (so-called dunetracking) is important for various fields of application and contributes to a sound and efficient river and sediment management. We developed a new tool, which enables a robust estimation of bedform characteristics and with which comprehensive sensitivity analyses can be carried out. Using a test dataset, we show that the selection of input parameters of dunetracking tools can have a significant impact on the results.
Alexander B. Prescott, Jon D. Pelletier, Satya Chataut, and Sriram Ananthanarayan
EGUsphere, https://doi.org/10.5194/egusphere-2024-1138, https://doi.org/10.5194/egusphere-2024-1138, 2024
Short summary
Short summary
Many Earth surface processes are controlled by the spatial pattern of surface water flow. We review commonly used methods for predicting such spatial patterns in digital landform models and document the pros and cons of commonly used methods. We propose a new method that is designed to minimize those limitations and show that it works well in a variety of test cases.
Prakash Pokhrel, Mikael Attal, Hugh D. Sinclair, Simon M. Mudd, and Mark Naylor
Earth Surf. Dynam., 12, 515–536, https://doi.org/10.5194/esurf-12-515-2024, https://doi.org/10.5194/esurf-12-515-2024, 2024
Short summary
Short summary
Pebbles become increasingly rounded during downstream transport in rivers due to abrasion. This study quantifies pebble roundness along the length of two Himalayan rivers. We demonstrate that roundness increases with downstream distance and that the rates are dependent on rock type. We apply this to reconstructing travel distances and hence the size of ancient Himalaya. Results show that the ancient river network was larger than the modern one, indicating that there has been river capture.
Aline Zinelabedin, Joel Mohren, Maria Wierzbicka-Wieczorek, Tibor Janos Dunai, Stefan Heinze, and Benedikt Ritter
EGUsphere, https://doi.org/10.5194/egusphere-2024-592, https://doi.org/10.5194/egusphere-2024-592, 2024
Short summary
Short summary
In order to interpret the formation processes of subsurface salt wedges and polygonal patterned grounds from the northern Atacama Desert, we present a multi-methodological approach. Due to the high salt content of the wedges, we suggest that their formation is dominated by subsurface salt dynamics requiring moisture. We assume that the climatic conditions during the wedge growth were slightly wetter than today, offering the potential to use the wedges as palaeoclimate archives.
Jens Martin Turowski, Aaron Bufe, and Stefanie Tofelde
Earth Surf. Dynam., 12, 493–514, https://doi.org/10.5194/esurf-12-493-2024, https://doi.org/10.5194/esurf-12-493-2024, 2024
Short summary
Short summary
Fluvial valleys are ubiquitous landforms, and understanding their formation and evolution affects a wide range of disciplines from archaeology and geology to fish biology. Here, we develop a model to predict the width of fluvial valleys for a wide range of geographic conditions. In the model, fluvial valley width is controlled by the two competing factors of lateral channel mobility and uplift. The model complies with available data and yields a broad range of quantitative predictions.
Dominik Amschwand, Jonas Wicky, Martin Scherler, Martin Hoelzle, Bernhard Krummenacher, Anna Haberkorn, Christian Kienholz, and Hansueli Gubler
EGUsphere, https://doi.org/10.5194/egusphere-2024-172, https://doi.org/10.5194/egusphere-2024-172, 2024
Short summary
Short summary
Rock glaciers are comparatively climate-resilient coarse-debris permafrost landforms. We estimate the energy budget of the seasonally thawing active layer (AL) of rock glacier Murtèl (Swiss Alps) based on a novel sub-surface sensor array. In the coarse-blocky AL during the thaw season, heat is transferred by thermal radiation and air convection. The ground heat flux is largely used to melt ground ice in the AL that protects to some degree the permafrost body beneath.
Daniel J. Ciarletta, Jennifer L. Miselis, Julie C. Bernier, and Arnell S. Forde
Earth Surf. Dynam., 12, 449–475, https://doi.org/10.5194/esurf-12-449-2024, https://doi.org/10.5194/esurf-12-449-2024, 2024
Short summary
Short summary
We reconstructed the evolution of Fire Island, a barrier island in New York, USA, to identify drivers of landscape change. Results reveal Fire Island was once divided into multiple inlet-separated islands with distinct features. Later, inlets closed, and Fire Island’s landscape became more uniform as human activities intensified. The island is now less mobile and less likely to resist and recover from storm impacts and sea level rise. This vulnerability may exist for other stabilized barriers.
Chao Zhou, Xibin Tan, Yiduo Liu, and Feng Shi
Earth Surf. Dynam., 12, 433–448, https://doi.org/10.5194/esurf-12-433-2024, https://doi.org/10.5194/esurf-12-433-2024, 2024
Short summary
Short summary
The drainage-divide stability provides new insights into both the river network evolution and the tectonic and/or climatic changes. Several methods have been proposed to determine the direction of drainage-divide migration. However, how to quantify the migration rate of drainage divides remains challenging. In this paper, we propose a new method to calculate the migration rate of drainage divides from high-resolution topographic data.
Moritz Altmann, Madlene Pfeiffer, Florian Haas, Jakob Rom, Fabian Fleischer, Tobias Heckmann, Livia Piermattei, Michael Wimmer, Lukas Braun, Manuel Stark, Sarah Betz-Nutz, and Michael Becht
Earth Surf. Dynam., 12, 399–431, https://doi.org/10.5194/esurf-12-399-2024, https://doi.org/10.5194/esurf-12-399-2024, 2024
Short summary
Short summary
We show a long-term erosion monitoring of several sections on Little Ice Age lateral moraines with derived sediment yield from historical and current digital elevation modelling (DEM)-based differences. The first study period shows a clearly higher range of variability of sediment yield within the sites than the later periods. In most cases, a decreasing trend of geomorphic activity was observed.
Paul A. Carling
Earth Surf. Dynam., 12, 381–397, https://doi.org/10.5194/esurf-12-381-2024, https://doi.org/10.5194/esurf-12-381-2024, 2024
Short summary
Short summary
Edge rounding in Shap granite glacial erratics is an irregular function of distance from the source outcrop in northern England, UK. Block shape is conservative, evolving according to block fracture mechanics – stochastic and silver ratio models – towards either of two attractor states. Progressive reduction in size occurs for blocks transported at the sole of the ice mass where the blocks are subject to compressive and tensile forces of the ice acting against a bedrock or till surface.
Gary Parker, Chenge An, Michael P. Lamb, Marcelo H. Garcia, Elizabeth H. Dingle, and Jeremy G. Venditti
Earth Surf. Dynam., 12, 367–380, https://doi.org/10.5194/esurf-12-367-2024, https://doi.org/10.5194/esurf-12-367-2024, 2024
Short summary
Short summary
River morphology has traditionally been divided by the size 2 mm. We use dimensionless arguments to show that particles in the 1–5 mm range (i) are the finest range not easily suspended by alluvial flood flows, (ii) are transported preferentially over coarser gravel, and (iii), within limits, are also transported preferentially over sand. We show how fluid viscosity mediates the special status of sediment in this range.
Lindsay Marie Capito, Enrico Pandrin, Walter Bertoldi, Nicola Surian, and Simone Bizzi
Earth Surf. Dynam., 12, 321–345, https://doi.org/10.5194/esurf-12-321-2024, https://doi.org/10.5194/esurf-12-321-2024, 2024
Short summary
Short summary
We propose that the pattern of erosion and deposition from repeat topographic surveys can be a proxy for path length in gravel-bed rivers. With laboratory and field data, we applied tools from signal processing to quantify this periodicity and used these path length estimates to calculate sediment transport using the morphological method. Our results highlight the potential to expand the use of the morphological method using only remotely sensed data as well as its limitations.
Xuxu Wu, Jonathan Malarkey, Roberto Fernández, Jaco H. Baas, Ellen Pollard, and Daniel R. Parsons
Earth Surf. Dynam., 12, 231–247, https://doi.org/10.5194/esurf-12-231-2024, https://doi.org/10.5194/esurf-12-231-2024, 2024
Short summary
Short summary
The seabed changes from flat to rippled in response to the frictional influence of waves and currents. This experimental study has shown that the speed of this change, the size of ripples that result and even whether ripples appear also depend on the amount of sticky mud present. This new classification on the basis of initial mud content should lead to improvements in models of seabed change in present environments by engineers and the interpretation of past environments by geologists.
Andrea D'Alpaos, Davide Tognin, Laura Tommasini, Luigi D'Alpaos, Andrea Rinaldo, and Luca Carniello
Earth Surf. Dynam., 12, 181–199, https://doi.org/10.5194/esurf-12-181-2024, https://doi.org/10.5194/esurf-12-181-2024, 2024
Short summary
Short summary
Sediment erosion induced by wind waves is one of the main drivers of the morphological evolution of shallow tidal environments. However, a reliable description of erosion events for the long-term morphodynamic modelling of tidal systems is still lacking. By statistically characterizing sediment erosion dynamics in the Venice Lagoon over the last 4 centuries, we set up a novel framework for a synthetic, yet reliable, description of erosion events in tidal systems.
Davide Tognin, Andrea D'Alpaos, Luigi D'Alpaos, Andrea Rinaldo, and Luca Carniello
Earth Surf. Dynam., 12, 201–218, https://doi.org/10.5194/esurf-12-201-2024, https://doi.org/10.5194/esurf-12-201-2024, 2024
Short summary
Short summary
Reliable quantification of sediment transport processes is necessary to understand the fate of shallow tidal environments. Here we present a framework for the description of suspended sediment dynamics to quantify deposition in the long-term modelling of shallow tidal systems. This characterization, together with that of erosion events, allows one to set up synthetic, yet reliable, models for the long-term evolution of tidal landscapes.
Emma L. S. Graf, Hugh D. Sinclair, Mikaël Attal, Boris Gailleton, Basanta Raj Adhikari, and Bishnu Raj Baral
Earth Surf. Dynam., 12, 135–161, https://doi.org/10.5194/esurf-12-135-2024, https://doi.org/10.5194/esurf-12-135-2024, 2024
Short summary
Short summary
Using satellite images, we show that, unlike other examples of earthquake-affected rivers, the rivers of central Nepal experienced little increase in sedimentation following the 2015 Gorkha earthquake. Instead, a catastrophic flood occurred in 2021 that buried towns and agricultural land under up to 10 m of sediment. We show that intense storms remobilised glacial sediment from high elevations causing much a greater impact than flushing of earthquake-induced landslides.
Mohamad Nasr, Adele Johannot, Thomas Geay, Sebastien Zanker, Jules Le Guern, and Alain Recking
Earth Surf. Dynam., 12, 117–134, https://doi.org/10.5194/esurf-12-117-2024, https://doi.org/10.5194/esurf-12-117-2024, 2024
Short summary
Short summary
Hydrophones are used to monitor sediment transport in the river by listening to the acoustic noise generated by particle impacts on the riverbed. However, this acoustic noise is modified by the river flow and can cause misleading information about sediment transport. This article proposes a model that corrects the measured acoustic signal. Testing the model showed that the corrected signal is better correlated with bedload flux in the river.
Cited articles
Aarnink, J. and Beucler, T.: Codebase for Automatic Detection of Instream Large Wood in Videos Using Deep Learning, GitHub [code], https://github.com/janbertoo/Instream_Wood_Detection (last access: 1 March 2024), 2024. a
Aarnink, J., Vuaridel, M., and Ruiz-Villanueva, V.: Database for Automatic Detection of Instream Large Wood in Videos Using Deep Learning, Zenodo [data set], https://doi.org/10.5281/zenodo.10822254, 2024. a
Àlex Solé Gómez, Scandolo, L., and Eisemann, E.: A learning approach for river debris detection, Int. J. Appl. Earth Obs., 107, 102682, https://doi.org/10.1016/j.jag.2022.102682, 2022. a
Andreoli, A., Comiti, F., and Lenzi, M. A.: Characteristics, distribution and geomorphic role of large woody debris in a mountain stream of the Chilean Andes, Earth Surf. Proc. Land., 32, 1675–1692, https://doi.org/10.1002/esp.1593, 2007. a
Benda, L. E. and Sias, J. C.: A quantitative framework for evaluating the mass balance of in-stream organic debris, Foreset Ecol. Manag., 172, 1–16, https://doi.org/10.1016/S0378-1127(01)00576-X, 2003. a
Bengio, Y., Courville, A., and Vincent, P.: Representation learning: A review and new perspectives, IEEE T. Pattern Anal., 35, 1798–1828, https://doi.org/10.1109/TPAMI.2013.50, 2013. a
Bochkovskiy, A., Wang, C.-Y., and Liao, H.-Y. M.: YOLOv4: Optimal Speed and Accuracy of Object Detection, arXiv [preprint], https://doi.org/10.48550/ARXIV.2004.10934, 2020. a
Casado-García, A., Heras, J., Milella, A., and Marani, R.: Semi-supervised deep learning and low-cost cameras for the semantic segmentation of natural images in viticulture, Precis. Agric., 23, 2001–2026, https://doi.org/10.1007/s11119-022-09929-9, 2022. a
Collins, B. D., Montgomery, D. R., Fetherston, K. L., and Abbe, T. B.: The floodplain large-wood cycle hypothesis: A mechanism for the physical and biotic structuring of temperate forested alluvial valleys in the North Pacific coastal ecoregion, Geomorphology, 139-140, 460–470, https://doi.org/10.1016/j.geomorph.2011.11.011, 2012. a
Curran, J. H. and Wohl, E.: Large woody debris and flow resistance in step-pool channels, Cascade Range, Washington, Geomorphology, 51, 141–157, https://doi.org/10.1016/S0169-555X(02)00333-1, 2003. a
Dalal, N. and Triggs, B.: Histograms of oriented gradients for human detection, in: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), vol. 1, 886–893, https://doi.org/10.1109/CVPR.2005.177, 2005. a
De Cicco, P. N., Paris, E., Ruiz-Villanueva, V., Solari, L., and Stoffel, M.: In-channel wood-related hazards at bridges: A review, River Res. Appl., 34, 617–628, https://doi.org/10.1002/rra.3300, 2018. a
Dibike, Y. and Solomatine, D.: River flow forecasting using artificial neural networks, Phys. Chem. Earth Pt. B, 26, 1–7, https://doi.org/10.1016/S1464-1909(01)85005-X, 2001. a
Diehl, T.: Potential Drift Accumulation at Bridges, Elsevier, https://doi.org/10.1016/S1464-1909(01)85005-X, 1997. a
Duan, K., Bai, S., Xie, L., Qi, H., Huang, Q., and Tian, Q.: CenterNet: Keypoint triplets for object detection, Proceedings of the IEEE International Conference on Computer Vision, October 2019, 6568–6577, Seoul, South Korea, https://doi.org/10.1109/ICCV.2019.00667, 2019. a, b
Galia, T., Ruiz-Villanueva, V., Tichavský, R., Šilhán, K., Horáček, M., and Stoffel, M.: Characteristics and abundance of large and small instream wood in a Carpathian mixed-forest headwater basin, Forest Ecol. Manag., 424, 468–482, https://doi.org/10.1016/j.foreco.2018.05.031, 2018. a
Ghaffarian, H., Piégay, H., Lopez, D., Rivière, N., MacVicar, B., Antonio, A., and Mignot, E.: Video-monitoring of wood discharge: first inter-basin comparison and recommendations to install video cameras, Earth Surf. Proc. Land., 45, 2219–2234, https://doi.org/10.1002/esp.4875, 2020. a, b
Ghaffarian, H., Lemaire, P., Zhi, Z., Tougne, L., MacVicar, B., and Piégay, H.: Automated quantification of floating wood pieces in rivers from video monitoring: a new software tool and validation, Earth Surf. Dynam., 9, 519–537, https://doi.org/10.5194/esurf-9-519-2021, 2021. a, b
Haschenburger, J. K. and Rice, S. P.: Changes in woody debris and bed material texture in a gravel-bed channel, Geomorphology, 60, 241–267, https://doi.org/10.1016/j.geomorph.2003.08.003, 2004. a
Hassan, M. A., Hogan, D. L., Bird, S. A., May, C. L., Gomi, T., and Campbell, D.: Spatial and temporal dynamics of wood in headwater streams of the pacific northwest, J. Am. Water Resour. As., 41, 899–919, https://doi.org/10.1111/j.1752-1688.2005.tb04469.x, 2005. a
Hortobágyi, B., Vaudor, L., Ghaffarian, H., and Piégay, H.: Inter-basin comparison of wood flux using random forest modelling and repeated wood extractions in unmonitored catchments, Hydrol. Process., 38, 1–19, https://doi.org/10.1002/hyp.15176, 2024. a
Innocenti, L., Bladé, E., Sanz-Ramos, M., Ruiz-Villanueva, V., Solari, L., and Aberle, J.: Two-Dimensional Numerical Modeling of Large Wood Transport in Bended Channels Considering Secondary Current Effects, Water Resour. Res., 59, 1–16, https://doi.org/10.1029/2022WR034363, 2023. a
Jodas, D. S., Brazolin, S., Yojo, T., de Lima, R. A., Velasco, G. D. N., Machado, A. R., and Papa, J. P.: A Deep Learning-based Approach for Tree Trunk Segmentation, in: 2021 34th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), 370–377, https://doi.org/10.1109/SIBGRAPI54419.2021.00057, 2021. a
Kalsotra, R. and Arora, S.: Background subtraction for moving object detection: explorations of recent developments and challenges, The Visual Computer, 38, 4151–4178, https://doi.org/10.1007/s00371-021-02286-0, 2021. a
Kaur, R. and Singh, S.: A comprehensive review of object detection with deep learning, Digital Signal Processing, 132, 2023. a
Keller, E. A., MacDonald, A., Tally, T., and Merrit, N. J.: Effects of large organic debris on channel morphology and sediment storage in selected tributaries of Redwood Creek, northwestern California, US Geological Survey Professional Paper, 1454, 1–29, 1995. a
Lassettre, N. S. and Kondolf, G. M.: Large woody debris in urban stream channels: Redefining the problem, River Res. Appl., 28, 1477–1487, https://doi.org/10.1002/rra.1538, 2012. a
Lassettre, N. S., Piegay, H., Dufour, S., and Rollet, A.: Decadal changes in distribution and frequency of wood in a free meandering river, the Ain River, France, Earth Surf. Proc. Land., 33, 1098–1112, https://doi.org/10.1002/esp.1605, 2008. a
Le Coz, J., Patalano, A., Collins, D., Guillén, N. F., García, C. M., Smart, G. M., Bind, J., Chiaverini, A., Le Boursicaud, R., Dramais, G., and Braud, I.: Crowdsourced data for flood hydrology: Feedback from recent citizen science projects in Argentina, France and New Zealand, J. Hydrol., 541, 766–777, https://doi.org/10.1016/j.jhydrol.2016.07.036, 2016. a
Lecun, Y., Bengio, Y., and Hinton, G.: Deep learning, Nature, 521, 436–444, https://doi.org/10.1038/nature14539, 2015. a, b
Lemaire, P., Piégay, H., MacVicar, B., Vaudor, L., Mouquet-Noppe, C., and Tougne, L.: An automatic video monitoring system for the visual quantification of driftwood in large rivers, presented at the 3rd International Conference on Wood in World Rivers (WWR3-2015), Padova, Italy, 6–10 July 2015. a
Li, K., Wan, G., Cheng, G., Meng, L., and Han, J.: Object detection in optical remote sensing images: A survey and a new benchmark, ISPRS J. Photogramm., 159, 296–307, https://doi.org/10.1016/j.isprsjprs.2019.11.023, 2020. a
Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C. Y., and Berg, A. C.: SSD: Single shot multibox detector, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 9905 LNCS, 21–37, Springer, https://doi.org/10.1007/978-3-319-46448-0_2, 2016. a
Lucía, A., Comiti, F., Borga, M., Cavalli, M., and Marchi, L.: Dynamics of large wood during a flash flood in two mountain catchments, Nat. Hazards Earth Syst. Sci., 15, 1741–1755, https://doi.org/10.5194/nhess-15-1741-2015, 2015. a
Lyn, D., Cooper, T., and Yi, Y.-K.: Debris Accumulation at Bridge Crossings: Laboratory and Field Studies, Publication FHWA/IN/JTRP-2003/10, Joint Transportation Research Program, Indiana Department of Transportation and Purdue University, West Lafayette, Indiana, https://doi.org/10.1016/j.ejrh.2023.101348, 2003. a
MacVicar, B., Piegay, H., Henderson, A., Comiti, F., Oberlin, C., and Pecorari, E.: Quantifying the temporal dynamics of wood in large rivers: field trials of wood surveying, dating, tracking, and monitoring techniques, Earth Surf. Proc. Land., 34, 2031–2046, https://doi.org/10.1002/esp.1888, 2009. a
Maxwell, A. E., Warner, T. A., and Fang, F.: Implementation of machine-learning classification in remote sensing: an applied review, Int. J. Remote Sens., 39, 2784–2817, https://doi.org/10.1080/01431161.2018.1433343, 2018. a
Panici, D.: An Experimental and Numerical Approach to Modeling Large Wood Displacement in Rivers, Water Resour. Res., 57, 1–18, https://doi.org/10.1029/2021WR029860, 2021. a
Platts, W. S., Armour, C., Booth, G. B., Bryant, M., Bufford, J. L., Cuplin, P., Jensen, S., Lienkaemper, G. W., Wayne Minshall, G., Monsen, S. B., Nelson, R. L., Sedell, J. R., and Tuhy, J. S.: Methods for evaluating riparian habitats with applications to management., General Technical Report – US Department of Agriculture, Forest Service, 1987. a
Pucci, A., Eickmeier, D., Sousa, H. S., Giresini, L., Matos, J. C., and Holst, R.: Fragility Analysis Based on Damaged Bridges during the 2021 Flood in Germany, Appl. Sci., 13, 1–21, https://doi.org/10.3390/app131810454, 2023. a
Ravazzolo, D., Mao, L., Picco, L., and Lenzi, M.: Tracking log displacement during floods in the Tagliamento River using RFID and GPS tracker devices, Geomorphology, 228, 226–233, https://doi.org/10.1016/j.geomorph.2014.09.012, 2015. a
Redmon, J., Divvala, S., Girshick, R., and Farhadi, A.: You Only Look Once: Unified, Real-Time Object Detection, in: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 779–788, 27–30 June 2016, Las Vegas, Nevada, USA, https://doi.org/10.1109/CVPR.2016.91, 2016. a, b
Ren, S., He, K., Girshick, R., and Sun, J.: Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks, IEEE T. Pattern. Anal., 39, 1137–1149, https://doi.org/10.1109/TPAMI.2016.2577031, 2017. a
Ribeiro, M. T., Singh, S., and Guestrin, C.: Why should I trust you? Explaining the predictions of any classifier, in: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1135–1144, ACM, https://doi.org/10.1145/2939672.2939778, 2016. a, b
Ruiz-Villanueva, V., Wyżga, B., Mikuś, P., Hajdukiewicz, M., and Stoffel, M.: Large wood clogging during floods in a gravel-bed river: the Długopole bridge in the Czarny Dunajec River, Poland, Earth Surf. Proc. Land., 42, 516–530, https://doi.org/10.1002/esp.4091, 2017. a
Ruiz-Villanueva, V., Mazzorana, B., Bladé, E., Bürkli, L., Iribarren-Anacona, P., Mao, L., Nakamura, F., Ravazzolo, D., Rickenmann, D., Sanz-Ramos, M., Stoffel, M., and Wohl, E.: Characterization of wood-laden flows in rivers, Earth Surf. Proc. Land., 44, 1694–1709, https://doi.org/10.1002/esp.4603, 2019. a
Ruiz-Villanueva, V., Aarnink, J., Gibaja del Hoyo, J., Finch, B., and Vuaridel, M.: Integrating flow-, sediment- and instream wood-regimes during e-flows in the Spöl River (Swiss Alps), IAHR, https://doi.org/10.3850/IAHR-39WC2521711920221000, 2022. a
Sanhueza, D., Picco, L., Ruiz-Villanueva, V., Iroumé, A., Ulloa, H., and Barrientos, G.: Quantification of fluvial wood using UAVs and structure from motion, Geomorphology, 345, 106837, https://doi.org/10.1016/j.geomorph.2019.106837, 2019. a
Schenk, E. R., Mouline, B., Hupp, C. R., and Richter, J. M.: Large wood budget and transport dynamics on a large river using radio telemetry, Earth Surf. Proc. Land., 39, 487–498, https://doi.org/10.1002/esp.3463, 2013. a
Schwindt, S., Meisinger, L., Negreiros, B., Schneider, T., and Nowak, W.: Transfer learning achieves high recall for object classification in fluvial environments with limited data, Geomorphology, 455, 109185, https://doi.org/10.1016/j.geomorph.2024.109185, 2024. a
Sejr, J. H., Schneider-Kamp, P., and Ayoub, N.: Surrogate Object Detection Explainer (SODEx) with YOLOv4 and LIME, Machine Learning and Knowledge Extraction, 3, 662–671, https://doi.org/10.3390/make3030033, 2021. a
Shorten, C. and Khoshgoftaar, T.: A survey on Image Data Augmentation for Deep Learning, Journal of Big Data, 6, 1–48, https://doi.org/10.1186/s40537-019-0197-0, 2019. a
Sokolova, M., Cordova, M., Nap, H., van Helmond, A., Mans, M., Vroegop, A., Mencarelli, A., and Kootstra, G.: An integrated end-to-end deep neural network for automated detection of discarded fish species and their weight estimation, ICES J. Mar. Sci., 80, 1911–1922, https://doi.org/10.1093/icesjms/fsad118, 2023. a
Sun, R., Lei, T., Chen, Q., Wang, Z., Du, X., Zhao, W., and Nandi, A. K.: Survey of Image Edge Detection, Frontiers in Signal Processing, 2, 1–13, https://doi.org/10.3389/frsip.2022.826967, 2022. a
Swaroop, P. and Sharma, N.: An Overview of Various Template Matching Methodologies in Image Processing, International Journal of Computer Applications, 153, 8–14, https://doi.org/10.5120/ijca2016912165, 2016. a
Taskesen, E.: Python package clustimage is for unsupervised clustering of images, https://erdogant.github.io/clustimage (last access: 1 March 2024), 2021. a
Taylor, L. S., Quincey, D. J., and Smith, M. W.: Evaluation of low-cost Raspberry Pi sensors for structure-from-motion reconstructions of glacier calving fronts, Nat. Hazards Earth Syst. Sci., 23, 329–341, https://doi.org/10.5194/nhess-23-329-2023, 2023. a
Tian, J., Jin, Q., Wang, Y., Yang, J., Zhang, S., and Sun, D.: Performance analysis of deep learning-based object detection algorithms on COCO benchmark: a comparative study, J. Eng. Appl. Sci., 1–18, https://doi.org/10.1186/s44147-024-00411-z, 2024. a
Van Der Maaten, L. and Hinton, G.: Visualizing data using t-SNE, J. Mach. Learn. Res., 9, 2579–2625, 2008. a
van Lieshout, C., van Oeveren, K., van Emmerik, T., and Postma, E.: Automated River Plastic Monitoring Using Deep Learning and Cameras, Earth and Space Science, 7, 1–14, https://doi.org/10.1029/2019EA000960, 2020. a, b
Viso.ai: Viso Suite: The One No Code Computer Vision Platform, https://viso.ai/ (last access: 1 March 2024), 2022. a
Wang, C.-Y., Bochkovskiy, A., and Liao, H.-Y. M.: YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors, arXiv [preprint], https://doi.org/10.48550/arXiv.2207.02696, 2022. a, b, c
Wohl, E.: A legacy of absence: Wood removal in US rivers, Prog. Phys. Geog., 38, 637–663, https://doi.org/10.1177/0309133314548091, 2014. a
Wohl, E., Lininger, K. B., Fox, M., Baillie, B. R., and Erskine, W. D.: Instream large wood loads across bioclimatic regions, Forest Ecol. Manag., 404, 370–380, https://doi.org/10.1016/j.foreco.2017.09.013, 2017. a
Wohl, E., Scott, D. N., and Lininger, K. B.: Spatial Distribution of Channel and Floodplain Large Wood in Forested River Corridors of the Northern Rockies, Water Resour. Res., 54, 7879–7892, https://doi.org/10.1029/2018WR022750, 2018. a
Wohl, E., Kramer, N., Ruiz-Villanueva, V., Scott, D. N., Comiti, F., Gurnell, A. M., Piegay, H., Lininger, K. B., Jaeger, K. L., Walters, D. M., and Fausch, K. D.: The natural wood regime in rivers, BioScience, 69, 259–273, https://doi.org/10.1093/biosci/biz013, 2019. a
Xu, Y. and Goodacre, R.: On Splitting Training and Validation Set: A Comparative Study of Cross-Validation, Bootstrap and Systematic Sampling for Estimating the Generalization Performance of Supervised Learning, Journal of Analysis and Testing, 2, 249–262, 2018. a
Zhang, Z., Ghaffarian, H., Macvicar, B., Vaudor, L., Antonio, A., Michel, K., and Piégay, H.: Video monitoring of in-channel wood: From flux characterization and prediction to recommendations to equip stations, Earth Surf. Proc. Land., 46, 822–836, https://doi.org/10.1002/esp.5068, 2021. a, b, c
Zheng, Z., Wang, P., Liu, W., Li, J., Ye, R., and Ren, D.: Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression, AAAI, 34, https://doi.org/10.1609/aaai.v34i07.6999, 2020. a, b
Zoph, B., Cubuk, E. D., Ghiasi, G., Lin, T.-Y., Shlens, J., and Le, Q. V.: Learning Data Augmentation Strategies for Object Detection, in: Computer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, 23–28 August 2020, Proceedings, Part XXVII, 566–583, Springer-Verlag, Berlin, Heidelberg, https://doi.org/10.1007/978-3-030-58583-9_34, 2020. a
Zou, X.: A Review of Object Detection Techniques, in: 2019 International Conference on Smart Grid and Electrical Automation (ICSGEA), 251–254, 10–11 August 2019, Xiangtan, China, https://doi.org/10.1109/ICSGEA.2019.00065, 2019. a
Short summary
This study presents a novel convolutional-neural-network approach for detecting instream large wood in rivers, addressing the need for flexible monitoring methods across diverse data sources. Using a database of 15 228 fully labelled images, the model achieved a weighted mean average precision of 67 %. Fine-tuning parameters and sampling techniques can improve performance by over 10 % in some cases, offering valuable insights into ecosystem management.
This study presents a novel convolutional-neural-network approach for detecting instream large...
