This study investigated the influence of sampling design parameters on biomass prediction accurac... more This study investigated the influence of sampling design parameters on biomass prediction accuracy obtained from airborne lidar data. A one-factor-at-a-time and a global sensitivity analyses were applied to identify the parameters most impacting model accuracy. We focused on several lidar and field survey parameters that can be easily controlled by users. In this pine plantations study site, a decrease in pulse density (4 to 0.5 pulse/m 2) led to a small decrease in prediction accuracy (À3%). However, variability in the number of field plots, positioning accuracy, and plot size, significantly impacted model performance. To obtain a robust model, a minimum of 40 field plots, along with field plot position accuracy of 5 m or lower, and field plot radius exceeding 13 m are recommended. The minimum diameter at breast height (DBH) threshold and the choice of the allometric biomass equation were found to have lesser impacts on model accuracy. In addition, accuracies of DBH and tree height measurements were respectively shown to have a minor and negligible contribution to the prediction error. Significant field measurement costs will still be needed to ensure good-quality models for biomass mapping. However, by reducing pulse density, cost savings can be made on lidar acquisition.
Reliable measurements of the 3D distribution of Leaf Area Density (LAD) in forest canopy are cruc... more Reliable measurements of the 3D distribution of Leaf Area Density (LAD) in forest canopy are crucial for describing and modelling microclimatic and eco-physiological processes involved in forest ecosystems functioning. To overcome the obvious limitations of direct measurements, several indirect methods have been developed, including methods based on Terrestrial LiDAR scanning (TLS). This work focused on various LAD estimators used in voxel-based approaches. LAD estimates were compared to reference measurements at branch scale in laboratory, which offered the opportunity to investigate in controlled conditions the sensitivity of estimations to various factors such as voxel size, distance to scanner, leaf morphology (species), type of scanner and type of estimator. We found that all approaches to retrieve LAD estimates were highly sensitive to voxel size whatever the species or scanner and to distance to the FARO scanner. We provided evidence that these biases were caused by vegetatio...
& Key message Diachronic photogrammetric canopy height models can be used to quantify at a fine s... more & Key message Diachronic photogrammetric canopy height models can be used to quantify at a fine scale changes in dominant height and wood volume following storms. The regular renewal of aerial surveys makes this approach appealing for monitoring forest changes. & Context The increasing availability of aerial photographs and the development of dense matching algorithms open up new possibilities to assess the effects of storm events on forest canopies. & Aims The objective of this research is to assess the potential of diachronic canopy height models derived from photogrammetric point clouds (PCHM) to quantify changes in dominant height and wood volume of a broadleaved forest following a major storm. & Methods PCHMs derived from aerial photographs acquired before and after a storm event were calibrated using 25 field Handling Editors: Céline Meredieu and François de Coligny Contribution of the co-authors Jean-Pierre Renaud designed the experiment, analyzed the results, and wrote the paper. Cedric Vega, Steen Magnussen, and Sylvie Durrieu contributed to the experimental design, analysis of the results, and writing of the paper. Jonathan Lisein contributed to the production of the photogrammetric models and proofreading of the manuscript. Philippe Lejeune and Mériem Fournier proofread the manuscript. Mériem Fournier coordinated the research project. This article is part of the Topical Collection on Mensuration and modelling for forestry in a changing environment
Reçu le 4 mai; accepté le 10 novembre 1993) Résumé — La télédétection satellitaire est, d'a... more Reçu le 4 mai; accepté le 10 novembre 1993) Résumé — La télédétection satellitaire est, d'après certaines études, un outil adapté à la détection et à la cartographie des changements forestiers importants (coupes rases, incendies) sur de vastes territoires. Elle intéresse donc l'Inventaire forestier national français (IFN) dans une optique de mise à jour de sa carte des types de peuplements. Dans la présente étude nous avons cherché à détecter des changements plus subtils, tels les éclaircies. L'approche retenue consiste à normaliser des images satellitaires prises à des dates différentes et à les comparer ensuite pixel à pixel. La méthode de correction relative des effets atmosphériques proposée est basée sur les seules données image. Elle a été testée sur une zone de moyenne montagne, les monts de Lacaune (Tarn), à partir d'images Landsat TM prises à 1 et 2 ans d'intervalle (août 1988, 1990 et 1991). Les premiers résultats obtenus par comparaison des images après normalisation sont très satisfaisants en ce qui concerne la détection des coupes rases et des éclaircies. D'autres changements tels que dégagements dans des reboisements, travaux sur pistes, incendies sont aussi détectés. Ces résultats sont encourageants et il faut maintenant se préoccuper de la façon dont ils peuvent être cartographiés et présentés auprès des utilisateurs des données IFN. cartographie forestière / télédétection satellitaire / détection des changements / prétraitements / correction radiométrique / éclaircies Summary — A method of comparing satellite images to detect forest changes in an area of rugged terrain (southwestern France). According to the literature, satellite remote sensing is a tool adapted to the detection and mapping of major forest changes (clearcuts, burned areas) on vast territories. The French forest survey service, Inventaire Forestier National (IFN), is therefore interested in its use for updating its map of stand types. The present study attempts to detect small changes, mainly thinnings. The approach consists in normalizing satellite images taken at different dates and comparing them pixel by pixel. The proposed method of relative correction of atmospheric effects is based on image data only. It has been tested in a relief area, Monts de Lacaune (Tarn, southwestern France), and on Landsat TM images taken in August 1988, 1990 and 1991. The initial results, obtained by image comparison after normalization, concerning clearcuts and thinnings are good. Other changes, such as cleaning, forest road works, and burned areas are also detected. These results are encouraging and the work now consists in actual mapping of the changes in a way which is suitable for users. forest mapping / satellite remote sensing / change detection / preprocessing / radiometric correction / thinning
We explored the potential of airborne laser scanner (ALS) data to improve Bayesian models linking... more We explored the potential of airborne laser scanner (ALS) data to improve Bayesian models linking biodiversity indicators of the understory vegetation to environmental factors. Biodiversity was studied at plot level and models were built to investigate species abundance for the most abundant plants found on each study site, and for ecological group richness based on light preference. The usual abiotic explanatory factors related to climate, topography and soil properties were used in the models. ALS data, available for two contrasting study sites, were used to provide biotic factors related to forest structure, which was assumed to be a key driver of understory biodiversity. Several ALS variables were found to have significant effects on biodiversity indicators. However, the responses of biodiversity indicators to forest structure variables, as revealed by the Bayesian model outputs, were shown to be dependent on the abiotic environmental conditions characterizing the study areas. Lower responses were observed on the lowland site than on the mountainous site. In the latter, shade-tolerant and heliophilous species richness was impacted by vegetation structure indicators linked to light penetration through the canopy. However, to reveal the full effects of forest structure on biodiversity indicators, forest structure would need to be measured over much wider areas than the plot we assessed. It seems obvious that the forest structure surrounding the field plots can impact biodiversity indicators measured at plot level. Various scales were found to be relevant depending on: the biodiversity indicators that were modelled, and the ALS variable. Finally, our results underline the utility of lidar data in abundance and richness models to characterize forest structure with variables that are difficult to measure in the field, either due to their nature or to the size of the area they relate to.
International Journal of Applied Earth Observation and Geoinformation, 2011
This paper presents a method for individual tree crown extraction and characterisation from a Can... more This paper presents a method for individual tree crown extraction and characterisation from a Canopy Surface Model (CSM). The method is based on a conventional algorithm used for localising LM on a smoothed version of the CSM and subsequently for modelling the tree crowns around each maximum at the plot level. The novelty of the approach lies in the introduction of controls on both the degree of CSM filtering and the shape of elliptic crowns, in addition to a multi-filtering level crown fusion approach to balance omission and commission errors. The algorithm derives the total tree height and the mean crown diameter from the elliptic tree crowns generated. The method was tested and validated on a mountainous forested area mainly covered by mature and even-aged black pine (Pinus nigra ssp. nigra [Arn.]) stands. Mean stem detection per plot, using this method, was 73.97 %. Algorithm performance was affected slightly by both stand density and heterogeneity (i.e. tree diameter classes' distribution). The total tree height and the mean crown diameter were estimated with root mean squared error values of 1.83 m and 1.48 m respectively. Tree heights were slightly underestimated in flat areas and overestimated on slopes. The average crown diameter was underestimated by 17.46 % on average.
Fourier analysis has a great amount of applications. For example, it appears to be a very suited ... more Fourier analysis has a great amount of applications. For example, it appears to be a very suited tool for vineyard detection in aerial images, due to the periodic patterns induced by this culture. Vine-plot mapping and monitoring are crucial issues in land management, particularly for areas where vineyards are dominant, like in some French regions. In this context, the availability
Revue Francaise de Photogrammetrie et de Teledetection
Reliable and very height spatial resolution data on forest structure are necessary to elaborate s... more Reliable and very height spatial resolution data on forest structure are necessary to elaborate silvicultural guide and to optimize silvicultural strategies. Airborne laser scanning data allows to extract 3D information from individual trees growing within forest stands. Such dense information information, including tree positions, height of the apices and crown dimensions are required to drive growth or forest dynamic models. The paper introduce a method for crown segmentation and characterisation based on a lidar digital canopy model. The whole method can be summarize by the following 3 steps. First, potential tree apices are estimated from local maxima. Second, tree crowns are delineated for each potential tree apex using a region growing algorithm. Finally, tree apices are extracted filtering out the potential tree apices falling within the crown associated with an apex of higher height. The method has been tested and validated for black pine dominated stands growing in a mounta...
The aim of this study was to develop a methodology to delineate accurately tree crowns using digi... more The aim of this study was to develop a methodology to delineate accurately tree crowns using digital elevation models independently of source (laser scanned data or stereoscopic pairs) and to identify the complementary information that can be obtained from high resolution imagery. This methodology uses the Watershed algorithm as the baseline. However the direct application of this algorithm generates over-segmentation for complex tree-crowns and non-wooded areas. To solve these problems a preprocessing stage consisting of a mixture modeling adaptive thresholding technique was developed to remove non-wooded areas. Results obtained on different plots of land are discussed putting forward the influential factors.
Les avancées récentes dans le domaine de la micro-informatique ont mis à la portée d'un large... more Les avancées récentes dans le domaine de la micro-informatique ont mis à la portée d'un large public des techniques sophistiquées de traitement d'image. Les utilisateurs de photographies aériennes disposent ainsi d'outils performants de traitement et d'utilisation de ces images leur permettant notamment la réalisation d'orthophotographies, que l'on peut aisément intégrer à un logiciel de traitement de l'information géographique (SIG). Ce dossier fait le point sur le principe de l'orthorectification, ses contraintes en matériel, logiciels et données et aborde son application concrète au travers d'une étude hydrologique. / The fast progress of microcomputers has made available new mass storage capacities and new techniques of image processing. Users of aerial photographs have now access to friendly softwares, allowing the integration of these images in a digital environment. These tools allow to orthorectify aerial photographs to integrate them in G...
Assessing forest aboveground biomass at global scale is crucial to address the challenge of susta... more Assessing forest aboveground biomass at global scale is crucial to address the challenge of sustainable management of forest resources and to strengthen forest-based climate change mitigation. To achieve this goal relying on spaceborne lidar missions is acknowledged to be a highly relevant solution. However, if this is taken as a given from the measurement point of view, the premise that spaceborne observation is the most suitable solution to provide information for sustainable management of forest resources is worth discussing. In this paper we suggest to take a fresh look at measurement processes designed to support the monitoring of Earth resources. We discuss the sustainability of Earth observation from space considering (1) issues that call into question the assumption that Earth-orbiting platform will always be available to the civilian remote sensing community and (2) issues concerning environmental impacts of space activity on the Earth. This leads us to suggest some actions...
This study investigated the influence of sampling design parameters on biomass prediction accurac... more This study investigated the influence of sampling design parameters on biomass prediction accuracy obtained from airborne lidar data. A one-factor-at-a-time and a global sensitivity analyses were applied to identify the parameters most impacting model accuracy. We focused on several lidar and field survey parameters that can be easily controlled by users. In this pine plantations study site, a decrease in pulse density (4 to 0.5 pulse/m 2) led to a small decrease in prediction accuracy (À3%). However, variability in the number of field plots, positioning accuracy, and plot size, significantly impacted model performance. To obtain a robust model, a minimum of 40 field plots, along with field plot position accuracy of 5 m or lower, and field plot radius exceeding 13 m are recommended. The minimum diameter at breast height (DBH) threshold and the choice of the allometric biomass equation were found to have lesser impacts on model accuracy. In addition, accuracies of DBH and tree height measurements were respectively shown to have a minor and negligible contribution to the prediction error. Significant field measurement costs will still be needed to ensure good-quality models for biomass mapping. However, by reducing pulse density, cost savings can be made on lidar acquisition.
Reliable measurements of the 3D distribution of Leaf Area Density (LAD) in forest canopy are cruc... more Reliable measurements of the 3D distribution of Leaf Area Density (LAD) in forest canopy are crucial for describing and modelling microclimatic and eco-physiological processes involved in forest ecosystems functioning. To overcome the obvious limitations of direct measurements, several indirect methods have been developed, including methods based on Terrestrial LiDAR scanning (TLS). This work focused on various LAD estimators used in voxel-based approaches. LAD estimates were compared to reference measurements at branch scale in laboratory, which offered the opportunity to investigate in controlled conditions the sensitivity of estimations to various factors such as voxel size, distance to scanner, leaf morphology (species), type of scanner and type of estimator. We found that all approaches to retrieve LAD estimates were highly sensitive to voxel size whatever the species or scanner and to distance to the FARO scanner. We provided evidence that these biases were caused by vegetatio...
& Key message Diachronic photogrammetric canopy height models can be used to quantify at a fine s... more & Key message Diachronic photogrammetric canopy height models can be used to quantify at a fine scale changes in dominant height and wood volume following storms. The regular renewal of aerial surveys makes this approach appealing for monitoring forest changes. & Context The increasing availability of aerial photographs and the development of dense matching algorithms open up new possibilities to assess the effects of storm events on forest canopies. & Aims The objective of this research is to assess the potential of diachronic canopy height models derived from photogrammetric point clouds (PCHM) to quantify changes in dominant height and wood volume of a broadleaved forest following a major storm. & Methods PCHMs derived from aerial photographs acquired before and after a storm event were calibrated using 25 field Handling Editors: Céline Meredieu and François de Coligny Contribution of the co-authors Jean-Pierre Renaud designed the experiment, analyzed the results, and wrote the paper. Cedric Vega, Steen Magnussen, and Sylvie Durrieu contributed to the experimental design, analysis of the results, and writing of the paper. Jonathan Lisein contributed to the production of the photogrammetric models and proofreading of the manuscript. Philippe Lejeune and Mériem Fournier proofread the manuscript. Mériem Fournier coordinated the research project. This article is part of the Topical Collection on Mensuration and modelling for forestry in a changing environment
Reçu le 4 mai; accepté le 10 novembre 1993) Résumé — La télédétection satellitaire est, d'a... more Reçu le 4 mai; accepté le 10 novembre 1993) Résumé — La télédétection satellitaire est, d'après certaines études, un outil adapté à la détection et à la cartographie des changements forestiers importants (coupes rases, incendies) sur de vastes territoires. Elle intéresse donc l'Inventaire forestier national français (IFN) dans une optique de mise à jour de sa carte des types de peuplements. Dans la présente étude nous avons cherché à détecter des changements plus subtils, tels les éclaircies. L'approche retenue consiste à normaliser des images satellitaires prises à des dates différentes et à les comparer ensuite pixel à pixel. La méthode de correction relative des effets atmosphériques proposée est basée sur les seules données image. Elle a été testée sur une zone de moyenne montagne, les monts de Lacaune (Tarn), à partir d'images Landsat TM prises à 1 et 2 ans d'intervalle (août 1988, 1990 et 1991). Les premiers résultats obtenus par comparaison des images après normalisation sont très satisfaisants en ce qui concerne la détection des coupes rases et des éclaircies. D'autres changements tels que dégagements dans des reboisements, travaux sur pistes, incendies sont aussi détectés. Ces résultats sont encourageants et il faut maintenant se préoccuper de la façon dont ils peuvent être cartographiés et présentés auprès des utilisateurs des données IFN. cartographie forestière / télédétection satellitaire / détection des changements / prétraitements / correction radiométrique / éclaircies Summary — A method of comparing satellite images to detect forest changes in an area of rugged terrain (southwestern France). According to the literature, satellite remote sensing is a tool adapted to the detection and mapping of major forest changes (clearcuts, burned areas) on vast territories. The French forest survey service, Inventaire Forestier National (IFN), is therefore interested in its use for updating its map of stand types. The present study attempts to detect small changes, mainly thinnings. The approach consists in normalizing satellite images taken at different dates and comparing them pixel by pixel. The proposed method of relative correction of atmospheric effects is based on image data only. It has been tested in a relief area, Monts de Lacaune (Tarn, southwestern France), and on Landsat TM images taken in August 1988, 1990 and 1991. The initial results, obtained by image comparison after normalization, concerning clearcuts and thinnings are good. Other changes, such as cleaning, forest road works, and burned areas are also detected. These results are encouraging and the work now consists in actual mapping of the changes in a way which is suitable for users. forest mapping / satellite remote sensing / change detection / preprocessing / radiometric correction / thinning
We explored the potential of airborne laser scanner (ALS) data to improve Bayesian models linking... more We explored the potential of airborne laser scanner (ALS) data to improve Bayesian models linking biodiversity indicators of the understory vegetation to environmental factors. Biodiversity was studied at plot level and models were built to investigate species abundance for the most abundant plants found on each study site, and for ecological group richness based on light preference. The usual abiotic explanatory factors related to climate, topography and soil properties were used in the models. ALS data, available for two contrasting study sites, were used to provide biotic factors related to forest structure, which was assumed to be a key driver of understory biodiversity. Several ALS variables were found to have significant effects on biodiversity indicators. However, the responses of biodiversity indicators to forest structure variables, as revealed by the Bayesian model outputs, were shown to be dependent on the abiotic environmental conditions characterizing the study areas. Lower responses were observed on the lowland site than on the mountainous site. In the latter, shade-tolerant and heliophilous species richness was impacted by vegetation structure indicators linked to light penetration through the canopy. However, to reveal the full effects of forest structure on biodiversity indicators, forest structure would need to be measured over much wider areas than the plot we assessed. It seems obvious that the forest structure surrounding the field plots can impact biodiversity indicators measured at plot level. Various scales were found to be relevant depending on: the biodiversity indicators that were modelled, and the ALS variable. Finally, our results underline the utility of lidar data in abundance and richness models to characterize forest structure with variables that are difficult to measure in the field, either due to their nature or to the size of the area they relate to.
International Journal of Applied Earth Observation and Geoinformation, 2011
This paper presents a method for individual tree crown extraction and characterisation from a Can... more This paper presents a method for individual tree crown extraction and characterisation from a Canopy Surface Model (CSM). The method is based on a conventional algorithm used for localising LM on a smoothed version of the CSM and subsequently for modelling the tree crowns around each maximum at the plot level. The novelty of the approach lies in the introduction of controls on both the degree of CSM filtering and the shape of elliptic crowns, in addition to a multi-filtering level crown fusion approach to balance omission and commission errors. The algorithm derives the total tree height and the mean crown diameter from the elliptic tree crowns generated. The method was tested and validated on a mountainous forested area mainly covered by mature and even-aged black pine (Pinus nigra ssp. nigra [Arn.]) stands. Mean stem detection per plot, using this method, was 73.97 %. Algorithm performance was affected slightly by both stand density and heterogeneity (i.e. tree diameter classes' distribution). The total tree height and the mean crown diameter were estimated with root mean squared error values of 1.83 m and 1.48 m respectively. Tree heights were slightly underestimated in flat areas and overestimated on slopes. The average crown diameter was underestimated by 17.46 % on average.
Fourier analysis has a great amount of applications. For example, it appears to be a very suited ... more Fourier analysis has a great amount of applications. For example, it appears to be a very suited tool for vineyard detection in aerial images, due to the periodic patterns induced by this culture. Vine-plot mapping and monitoring are crucial issues in land management, particularly for areas where vineyards are dominant, like in some French regions. In this context, the availability
Revue Francaise de Photogrammetrie et de Teledetection
Reliable and very height spatial resolution data on forest structure are necessary to elaborate s... more Reliable and very height spatial resolution data on forest structure are necessary to elaborate silvicultural guide and to optimize silvicultural strategies. Airborne laser scanning data allows to extract 3D information from individual trees growing within forest stands. Such dense information information, including tree positions, height of the apices and crown dimensions are required to drive growth or forest dynamic models. The paper introduce a method for crown segmentation and characterisation based on a lidar digital canopy model. The whole method can be summarize by the following 3 steps. First, potential tree apices are estimated from local maxima. Second, tree crowns are delineated for each potential tree apex using a region growing algorithm. Finally, tree apices are extracted filtering out the potential tree apices falling within the crown associated with an apex of higher height. The method has been tested and validated for black pine dominated stands growing in a mounta...
The aim of this study was to develop a methodology to delineate accurately tree crowns using digi... more The aim of this study was to develop a methodology to delineate accurately tree crowns using digital elevation models independently of source (laser scanned data or stereoscopic pairs) and to identify the complementary information that can be obtained from high resolution imagery. This methodology uses the Watershed algorithm as the baseline. However the direct application of this algorithm generates over-segmentation for complex tree-crowns and non-wooded areas. To solve these problems a preprocessing stage consisting of a mixture modeling adaptive thresholding technique was developed to remove non-wooded areas. Results obtained on different plots of land are discussed putting forward the influential factors.
Les avancées récentes dans le domaine de la micro-informatique ont mis à la portée d'un large... more Les avancées récentes dans le domaine de la micro-informatique ont mis à la portée d'un large public des techniques sophistiquées de traitement d'image. Les utilisateurs de photographies aériennes disposent ainsi d'outils performants de traitement et d'utilisation de ces images leur permettant notamment la réalisation d'orthophotographies, que l'on peut aisément intégrer à un logiciel de traitement de l'information géographique (SIG). Ce dossier fait le point sur le principe de l'orthorectification, ses contraintes en matériel, logiciels et données et aborde son application concrète au travers d'une étude hydrologique. / The fast progress of microcomputers has made available new mass storage capacities and new techniques of image processing. Users of aerial photographs have now access to friendly softwares, allowing the integration of these images in a digital environment. These tools allow to orthorectify aerial photographs to integrate them in G...
Assessing forest aboveground biomass at global scale is crucial to address the challenge of susta... more Assessing forest aboveground biomass at global scale is crucial to address the challenge of sustainable management of forest resources and to strengthen forest-based climate change mitigation. To achieve this goal relying on spaceborne lidar missions is acknowledged to be a highly relevant solution. However, if this is taken as a given from the measurement point of view, the premise that spaceborne observation is the most suitable solution to provide information for sustainable management of forest resources is worth discussing. In this paper we suggest to take a fresh look at measurement processes designed to support the monitoring of Earth resources. We discuss the sustainability of Earth observation from space considering (1) issues that call into question the assumption that Earth-orbiting platform will always be available to the civilian remote sensing community and (2) issues concerning environmental impacts of space activity on the Earth. This leads us to suggest some actions...
Uploads
Papers by S. Durrieu