A5027285931- Remote Sensing and LiDAR Applications
- Forest ecology and management
- Forest Ecology and Biodiversity Studies
- Remote Sensing in Agriculture
- Forest Management and Policy
- Agriculture and Rural Development Research
- Tree Root and Stability Studies
- French Urban and Social Studies
- Forest Biomass Utilization and Management
- Soil erosion and sediment transport
- Botany and Plant Ecology Studies
- African Botany and Ecology Studies
- Aeolian processes and effects
- Soil Geostatistics and Mapping
- Horticultural and Viticultural Research
- 3D Surveying and Cultural Heritage
- Fire effects on ecosystems
- Soil Carbon and Nitrogen Dynamics
- Plant Water Relations and Carbon Dynamics
- Satellite Image Processing and Photogrammetry
- Plant Physiology and Cultivation Studies
- Archaeological and Historical Studies
- Edible Oils Quality and Analysis
- Medical Imaging and Analysis
- Species Distribution and Climate Change
Office National des Forêts
2015-2024
Centre National de la Propriété Forestière
2004-2024
Université de Lorraine
2021-2024
Institut national de l’information géographique et forestière
2021
Ministère de l'Agriculture
1989
Soil bulk density (ρ) is an important physical property, but its measurement frequently lacking in soil surveys due to the time-consuming nature of making measurement. As a result pedotransfer functions (PTFs) have been developed predict ρ from other more easily available properties. These are generally derived regression methods that aim fit single model. In this study, we use technique called Generalized Boosted Regression Modelling (GBM; Ridgeway, 2006) which combines two algorithms:...
Depuis les dernières années, de nombreuses données d’origine et nature variées deviennent aisément accessibles. Certaines, liées aux attributs forestiers, permettent d’accroître la précision des inventaires forestiers produire estimations à échelles beaucoup plus fines qu’auparavant. Sur cette base, une approche d’inventaire « multisource » a vu le jour en Finlande dans années 1990 se développe depuis France. Des gains considérables sont prévoir pour décideurs gestionnaires. Messages clés :-...
Abstract. The development of high-resolution mapping models for forest attributes based on remote sensing data combined with machine or deep learning techniques has become a prominent topic in the field observation and monitoring. This resulted availability multiple, sometimes conflicting, sources information, but, at face value, it also makes possible to learn about attribute uncertainty through joint interpretation multiple models. article seeks endorse latter by utilizing Bayesian model...
The development of high-resolution mapping models for forest attributes, driven by machine and deep learning techniques, has resulted in the widespread availability multiple information sources. While this abundance can potentially lead to confusion, it also presents an opportunity gain "extended" understanding conditions interpreting these sources collectively. This contribution focuses on latter, leveraging Bayesian Model Averaging (BMA) approach. BMA not only enables diagnosing...
As studies have underlined the sensitivity of lidar metrics to scan angles, objective this study was twofold. Firstly, we further investigated influence angle on ABA predictions stand attributes riparian (29 field plots), broadleaf (42 coniferous (31 plots) and mixed (45 forest types in France. Secondly, evaluated potential voxelisation approaches normalise effects mitigate models. To achieve these objectives, first selected a model based four with different sensitivities angle, i.e. mean...
Abstract. The development of high-resolution mapping models for forest attributes based on remote sensing data combined with machine or deep learning techniques, has become a prominent topic in the field observation and monitoring. This resulted an extensive availability multiple sources information, which can either lead to potential confusion, possibility learn both about through joint interpretation models. article seeks endorse latter, by relying Bayesian model averaging (BMA) approach,...
Multisource forest inventory methods were developed to improve the precision of national estimates. These rely on combination data and auxiliary information correlated with attributes interest. As these have been predominantly tested over coniferous forests, present study used this approach for heterogeneous complex deciduous forests in center France. The considered included a type map, Landsat 8 spectral bands derived vegetation indexes, 3D variables from photogrammetric canopy height...
Combining national forest inventory (NFI) data with auxiliary information allows downscaling and improving the precision of NFI estimates for small domains, where normally too few field plots are available to produce reliable estimates. In most situations, domains represent administrative units that could greatly vary in size forested area. poorly sampled often drop below expected standards. To tackle this issue, we introduce a algorithm generating smallest possible groups satisfying...
The French National Forest Inventory provides detailed forest information up to large national and regional scales. inventory for small areas of interest within a population is equally important decision making, such as local planning management purposes. However, sampling these with sufficient ground plots often not cost efficient. In response, area estimation has gained increasing popularity in inventory. It consists set techniques that enables predictions attributes subpopulation the help...
Forest map products are widely used and have taken benefit from progresses in the multisource forest inventory approaches, which meant to improve precision of estimates at high spatial resolution. However, estimating errors pixel-wise predictions remains difficult, reconciling statistical outcomes with is still an open important question. We address this problem using original approach relying on a model-based inference framework k-nearest neighbours (k-NN) models produce estimations related...
Abstract. Lidar scan angle can affect estimation of lidar-derived forest metrics used in area-based approaches (ABAs). As commonly first-order and various user-developed are computed the form a grid or raster, their response to angles needs be investigated similarly. The objective this study was highlight impact on 11 (9 height-based 2 other metrics) at level grid-cell. area divided into cell size 30 m. In every grid-cell, flight lines that sampled least 90% grid-cell were identified....
Many forest growth models use plot-level variables, such as stem density or basal area, predictors. Even if these variables are area-based, so that their expectations do not change with plot size when expressed per unit of a in may induce prediction bias the model is nonlinear respect to variables. In this study, we show effect on can be seen an error term affects entries and propagation theory used predict size-induced biases. For density-dependent matrix population tropical rainforest...
The ever-increasing availability of remote sensing data allows production forest attributes maps, which are usually made using model-based approaches. These map products sensitive to various bias sources, including model extrapolation. To identify, over a case study forest, the proportion extrapolated predictions, we used convex hull method applied auxiliary space an airborne laser scanning (ALS) flight. impact different sampling efforts was also evaluated. This done by iteratively thinning...
La biomecanique etudie les reactions et adaptations des etres vivants a leur environnement mecanique, par exemple aux oscillations forces exercees le vent. Au-dela theories anciennes de la securite mecanique constante, mecanobiologie recemment formalise signaux mecaniques, perception cellules vivantes reponses croissance. Ces mecanismes physiologiques font que l’arbre ne forme vraiment du bois, tissu soutien, lorsqu’il est mecaniquement stimule. croissance controlee deformations mecaniques...