Mathilde Letard
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mathilde Letard has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Dimitri Lague
Univ Rennes, CNRS, Géosciences Rennes, UMR 6118, 35000, Rennes, France
6 shared publications - Paul Leroy
Univ Rennes, CNRS, Plateforme LiDAR topo-bathymétrique, OSERen, UAR 3343, 35000, Rennes, France
6 shared publications Daniel Girardeau-Montaut
CloudCompare
5 shared publications- Laure Guerit
Univ Rennes, CNRS, Géosciences Rennes, UMR 6118, 35000, Rennes, France
5 shared publications - Olivier Payrastre
Univ Gustave Eiffel, GERS-LEE, F-44344 Bouguenais, France
5 shared publications - Philippe Steer
Géosciences Rennes, Université de Rennes, Rennes, France
5 shared publications
Latest Publications
- Open Access
- 22 July 2026
- Published by Copernicus GmbH
- p. 1-47
Reply on RC1
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- 22 July 2026
- Published by Copernicus GmbH
- p. 1-47
Open AccessReply on RC1
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- Open Access
- 22 July 2026
- Published by Copernicus GmbH
- p. 1-47
Reply on RC2
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- 22 July 2026
- Published by Copernicus GmbH
- p. 1-47
Open AccessReply on RC2
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- Open Access
- 3 July 2026
- Published by Copernicus GmbH in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XI-2-2026, 527-534
Spatiotemporal reconstruction of 4D point clouds at different time scales through implicit neural representations for topographic monitoring applications
Monitoring surface change in dynamic environments is essential to preserve the integrity of human infrastructure and livelihoods from natural hazard consequences. With the advent of 4D remote sensing, near-continuous monitoring of dynamic scenes is unlocked. However, the unordered and irregular nature of point clouds, compounded by temporally variable occlusions and diverse acquisition conditions, hinders the accurate analysis of highly information-rich 4D data. This work addresses the challenge of irregular spatiotemporal sampling in time series of 3D point clouds for the case study of a dynamic sandy beach at different time scales. We explore the use of implicit neural representations (INRs) to model 4D data as continuous spatiotemporal functions that are optimised to estimate the beach topography continuously through space and time. By comparing four model variants and assessing their performance to reconstruct spatially and temporally subsampled data, we evaluate the applicability of INRs to high-frequency topographic monitoring, especially in the context of 4D change analysis. Our results show the ability to reconstruct missing epochs from time series of 3D point clouds with centimetric to decimetric accuracy at time scales ranging from seasonal to daily observations. Our findings highlight the importance of hyperparameter tuning to enable the capture of local details in complex spatiotemporal datasets. Through this, our work lays the foundation for continuous spatiotemporal representation of dynamic scenes, supporting a potentially broad range of change analysis applications.
- 3 July 2026
- Published by Copernicus GmbH in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XI-2-2026, 527-534
Open AccessSpatiotemporal reconstruction of 4D point clouds at different time scales through implicit neural representations for topographic monitoring applications
Monitoring surface change in dynamic environments is essential to preserve the integrity of human infrastructure and livelihoods from natural hazard consequences. With the advent of 4D remote sensing, near-continuous monitoring of dynamic scenes is unlocked. However, the unordered and irregular nature of point clouds, compounded by temporally variable occlusions and diverse acquisition conditions, hinders the accurate analysis of highly information-rich 4D data. This work addresses the challenge of irregular spatiotemporal sampling in time series of 3D point clouds for the case study of a dynamic sandy beach at different time scales. We explore the use of implicit neural representations (INRs) to model 4D data as continuous spatiotemporal functions that are optimised to estimate the beach topography continuously through space and time. By comparing four model variants and assessing their performance to reconstruct spatially and temporally subsampled data, we evaluate the applicability of INRs to high-frequency topographic monitoring, especially in the context of 4D change analysis. Our results show the ability to reconstruct missing epochs from time series of 3D point clouds with centimetric to decimetric accuracy at time scales ranging from seasonal to daily observations. Our findings highlight the importance of hyperparameter tuning to enable the capture of local details in complex spatiotemporal datasets. Through this, our work lays the foundation for continuous spatiotemporal representation of dynamic scenes, supporting a potentially broad range of change analysis applications.
- Open Access
- 3 July 2026
- Published by Copernicus GmbH in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XI-2-2026, 623-630
Time-Adaptive Change Analysis through Extension of the M3C2 Algorithm using Multi-Modal Laser Scanning Data in a Salt Marsh Environment
Quantifying topographic dynamics from 3D time series is essential for a broad range of geoscientific applications. However, data acquired by laser scanning often vary between epochs, e.g. in point density and coverage. These heterogeneities present a challenge for change detection, particularly across multi-temporal and multi-modal data. We propose a new approach that enhances the analysis of spatiotemporally irregular point clouds, particularly when the sampling frequency exceeds the analysis time scale. In particular, our method extends the Multiscale Model to Model Cloud Comparison (M3C2) algorithm through adaptive temporal aggregation. Driven by a local point density requirement, our method employs both a spatial neighborhood (as in standard M3C2) and a temporal neighborhood for each core point within the point cloud time series. If the neighborhood of a core point is too sparse for surface estimation, an iterative temporal search incorporates data from adjacent epochs until either the density requirement or a maximum temporal window is reached. This adaptive process ensures sufficient local point density while preventing globally fixed temporal aggregation. We evaluate our method on a multi-modal laser scanning dataset from Mont-Saint-Michel Bay, France, comprising 38 epochs spanning a decade at daily to seasonal intervals. Results demonstrate that our method increases change detection completeness by more than 13% compared to standard M3C2 and increases accuracy by 31% compared to commonly used fixed-window averaging. Our approach thereby enhances 3D change detection in complex real-world 4D datasets, enabling higher accuracy and completeness in the analysis of surface dynamics at variable spatiotemporal scales.
- 3 July 2026
- Published by Copernicus GmbH in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XI-2-2026, 623-630
Open AccessTime-Adaptive Change Analysis through Extension of the M3C2 Algorithm using Multi-Modal Laser Scanning Data in a Salt Marsh Environment
Quantifying topographic dynamics from 3D time series is essential for a broad range of geoscientific applications. However, data acquired by laser scanning often vary between epochs, e.g. in point density and coverage. These heterogeneities present a challenge for change detection, particularly across multi-temporal and multi-modal data. We propose a new approach that enhances the analysis of spatiotemporally irregular point clouds, particularly when the sampling frequency exceeds the analysis time scale. In particular, our method extends the Multiscale Model to Model Cloud Comparison (M3C2) algorithm through adaptive temporal aggregation. Driven by a local point density requirement, our method employs both a spatial neighborhood (as in standard M3C2) and a temporal neighborhood for each core point within the point cloud time series. If the neighborhood of a core point is too sparse for surface estimation, an iterative temporal search incorporates data from adjacent epochs until either the density requirement or a maximum temporal window is reached. This adaptive process ensures sufficient local point density while preventing globally fixed temporal aggregation. We evaluate our method on a multi-modal laser scanning dataset from Mont-Saint-Michel Bay, France, comprising 38 epochs spanning a decade at daily to seasonal intervals. Results demonstrate that our method increases change detection completeness by more than 13% compared to standard M3C2 and increases accuracy by 31% compared to commonly used fixed-window averaging. Our approach thereby enhances 3D change detection in complex real-world 4D datasets, enabling higher accuracy and completeness in the analysis of surface dynamics at variable spatiotemporal scales.
- Open Access
- 3 July 2026
- Published by Copernicus GmbH in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XI-2-2026, 545-552
topo4d: Topographic 4D STAC Extension for Curating and Cataloging Multi-Source Geospatial Time Series Datasets
Spatiotemporal analysis of geospatial time series data has gained increasing attention with the emergence of 4D point clouds and automatic acquisition technologies such as permanent laser scanning (PLS), time-lapse photogrammetry, and uncrewed aerial vehicle (UAV) platforms, enabling near-continuous monitoring of Earth surface dynamics for change detection and process characterization. However, facing massive data volumes through the temporal domain, current topographic data curation practices often rely on empirically determined data processing and management, which may significantly affect reusability, interoperability, and hence processing efficiency due to the absence or heterogeneous nature of metadata. The need for standardized approaches to manage time-dependent metadata has become critical as the demands for sharing data and reproducing analysis across tools and application domains increase. We propose a topographic 4D extension (topo4d) to the SpatioTemporal Asset Catalog (STAC) framework, which provides an open and extensible specification for automatic metadata curation and FAIR data management practices. This paper demonstrates how the topo4d extension facilitates the interoperability and reusability of 4D datasets and presents the corresponding metadata curation workflows applied to two real-world environmental monitoring applications.
- 3 July 2026
- Published by Copernicus GmbH in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XI-2-2026, 545-552
Open Accesstopo4d: Topographic 4D STAC Extension for Curating and Cataloging Multi-Source Geospatial Time Series Datasets
Spatiotemporal analysis of geospatial time series data has gained increasing attention with the emergence of 4D point clouds and automatic acquisition technologies such as permanent laser scanning (PLS), time-lapse photogrammetry, and uncrewed aerial vehicle (UAV) platforms, enabling near-continuous monitoring of Earth surface dynamics for change detection and process characterization. However, facing massive data volumes through the temporal domain, current topographic data curation practices often rely on empirically determined data processing and management, which may significantly affect reusability, interoperability, and hence processing efficiency due to the absence or heterogeneous nature of metadata. The need for standardized approaches to manage time-dependent metadata has become critical as the demands for sharing data and reproducing analysis across tools and application domains increase. We propose a topographic 4D extension (topo4d) to the SpatioTemporal Asset Catalog (STAC) framework, which provides an open and extensible specification for automatic metadata curation and FAIR data management practices. This paper demonstrates how the topo4d extension facilitates the interoperability and reusability of 4D datasets and presents the corresponding metadata curation workflows applied to two real-world environmental monitoring applications.
- Open Access
- 27 April 2026
- Published by Copernicus GmbH
- Vol. 2026, 1-47
Comment on egusphere-2026-724
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- 27 April 2026
- Published by Copernicus GmbH
- Vol. 2026, 1-47
Open AccessComment on egusphere-2026-724
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- 8 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Environnemental knowledge extraction from topo-bathymetric lidar : machine learning and deep neural networds for point clouds and waveforms
Extraction de connaissances à partir de lidar topo-bathymétrique : apprentissage automatique et réseaux de neurones profonds pour les nuages de points et les formes d’onde: Les interfaces terre-eau, fortement vulnérables au changement climatique et à la pression anthropique, requièrent une surveillance accrue. Toutefois, l’observation ininterrompue des zones submergées et émergées demeure un défi en raison de la présence d’eau. La télédétection lidar topobathymétrique constitue une solution adéquate en assurant une représentation continue des zones terre-eau, matérialisée par des nuages de points 3D et des formes d’ondes 1D. Cependant, une pleine exploitation de ces données requiert des outils encore en attente de développement. Cette thèse présente plusieurs méthodes d’extraction de connaissances des données lidar topo-bathymétriques, incluant des approches de classification basées sur des nuages de points bi-spectraux et des formes d’ondes bispectrales. En outre, des réseaux de neurones profonds sont conçus pour la segmentation sémantique, la détection et la classification d’objets, ainsi que l’estimation de paramètres physiques de l’eau à partir des formes d’ondes bathymétriques. L’utilisation de modèles de transfert radiatif guide des approches visant à réduire la nécessité de données labélisées, améliorant ainsi le traitement des formes d’ondes lidar dans les eaux très peu profondes ou turbides.
- 8 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Environnemental knowledge extraction from topo-bathymetric lidar : machine learning and deep neural networds for point clouds and waveforms
Extraction de connaissances à partir de lidar topo-bathymétrique : apprentissage automatique et réseaux de neurones profonds pour les nuages de points et les formes d’onde: Les interfaces terre-eau, fortement vulnérables au changement climatique et à la pression anthropique, requièrent une surveillance accrue. Toutefois, l’observation ininterrompue des zones submergées et émergées demeure un défi en raison de la présence d’eau. La télédétection lidar topobathymétrique constitue une solution adéquate en assurant une représentation continue des zones terre-eau, matérialisée par des nuages de points 3D et des formes d’ondes 1D. Cependant, une pleine exploitation de ces données requiert des outils encore en attente de développement. Cette thèse présente plusieurs méthodes d’extraction de connaissances des données lidar topo-bathymétriques, incluant des approches de classification basées sur des nuages de points bi-spectraux et des formes d’ondes bispectrales. En outre, des réseaux de neurones profonds sont conçus pour la segmentation sémantique, la détection et la classification d’objets, ainsi que l’estimation de paramètres physiques de l’eau à partir des formes d’ondes bathymétriques. L’utilisation de modèles de transfert radiatif guide des approches visant à réduire la nécessité de données labélisées, améliorant ainsi le traitement des formes d’ondes lidar dans les eaux très peu profondes ou turbides.
- Open Access
- 31 March 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-47
From point cloud to digital elevation model: Airborne topo-bathymetric LiDAR processing over the 28 km Ardèche River Gorges
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- 31 March 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-47
Open AccessFrom point cloud to digital elevation model: Airborne topo-bathymetric LiDAR processing over the 28 km Ardèche River Gorges
We present a comprehensive workflow for processing a large-scale airborne Topo-Bathymetric LiDAR (TBL) dataset acquired over the 28 km Ardèche River Gorges in France during October 2021. To address limited depth penetration, low signal-to-noise ratio, and complex topography, we integrate onboard discrete returns, a full-waveform (FWF) re-analysis, and an orthorectified full-waveform synthesis (OrthoFWF). Depth penetration, evaluated by D99 (99th percentile of retrieved water depth, D), increases from 2.88 m (discrete) to 3.70 m (FWF) and 4.48 m (OrthoFWF). Area coverage increases from 70.1 % to 79.5 % to 85.6 % of the submerged area with bathymetric data available, while the length-based coverage (the percentage of river length composed of reaches with ≥95 % bathymetric coverage) improves from 31.8 % to 54.7 % to 86.9 %. A new unsupervised classification method, using a kernel-density–derived intensity threshold, was applied to 1.3 million points, enhancing the separation of bed returns from noise within the OrthoFWF domain and improving depth extraction. The workflow includes internal flight-line geometric correction, precision benchmarking against France’s national LiDAR HD dataset, bathymetric classification with a random-forest classifier, and a targeted supplementary sonar survey to constrain the deepest reaches. To produce the final Digital Elevation Model from incomplete coverage, we compare three interpolation approaches and demonstrate that Poisson surface reconstruction yields the most morphologically realistic surfaces, particularly when constrained by sonar-derived depths. This integrated workflow substantially improves the accuracy and completeness of river bathymetry, supporting high-fidelity hydrodynamic modeling and advancing TBL applications in fluvial geomorphology.
- Open Access
- 15 March 2025
- Published by Copernicus GmbH
Extracting bathymetric information from LiDAR waveforms with 1D neural networks.
Bathymetry is a critical component in many geomorphological and ecological studies of coastal and fluvial environments. Bathymetric lidar remote sensing enables the acquisition of high-resolution, 3D data on shallow sea- and riverbeds, thus providing precise modeling of their topography. However, in certain contexts, the optical interactions between light and water make extracting bathymetry from lidar signals particularly challenging, resulting in incomplete bed coverage. This is the case in very shallow waters at the transition between land and water – where the signals of the water surface, column, and bottom become entangled – and in deep or turbid waters, where signal extinction hinders the detection of the sea- or riverbed.In this work, we explore new approaches for bathymetry extraction from lidar waveforms across diverse environments. With temporal convolutional neural networks, we show improvements in detecting the position of the sea- or riverbed using bathymetric lidar waveforms. Our experiments, conducted on simulated data representing a wide range of environmental conditions – varying turbidity, depth, and reflectance – yield promising results. They highlight the potential to extract the position of the sea- or riverbed even in simulated waveforms with low signal-to-noise ratios or highly overlapping signals – cases that have posed challenges for existing processing methods. Given the increasing popularity of bathymetric lidar, particularly with the advent of UAV-mounted sensors, enhancing waveform processing methods could help advance the surveying of submerged areas.
- 15 March 2025
- Published by Copernicus GmbH
Open AccessExtracting bathymetric information from LiDAR waveforms with 1D neural networks.
Bathymetry is a critical component in many geomorphological and ecological studies of coastal and fluvial environments. Bathymetric lidar remote sensing enables the acquisition of high-resolution, 3D data on shallow sea- and riverbeds, thus providing precise modeling of their topography. However, in certain contexts, the optical interactions between light and water make extracting bathymetry from lidar signals particularly challenging, resulting in incomplete bed coverage. This is the case in very shallow waters at the transition between land and water – where the signals of the water surface, column, and bottom become entangled – and in deep or turbid waters, where signal extinction hinders the detection of the sea- or riverbed.In this work, we explore new approaches for bathymetry extraction from lidar waveforms across diverse environments. With temporal convolutional neural networks, we show improvements in detecting the position of the sea- or riverbed using bathymetric lidar waveforms. Our experiments, conducted on simulated data representing a wide range of environmental conditions – varying turbidity, depth, and reflectance – yield promising results. They highlight the potential to extract the position of the sea- or riverbed even in simulated waveforms with low signal-to-noise ratios or highly overlapping signals – cases that have posed challenges for existing processing methods. Given the increasing popularity of bathymetric lidar, particularly with the advent of UAV-mounted sensors, enhancing waveform processing methods could help advance the surveying of submerged areas.