Centre dedocumentation
scientifique

Characterizing the calibration domain of remote sensing models using convex hulls / Jean-Pierre Renaud in International journal of applied Earth observation and geoinformation, vol 112 (August 2022)  

Public [article]  inInternational journal of applied Earth observation and geoinformation > vol 112 (August 2022) . - n° 102939 Titre : Characterizing the calibration domain of remote sensing models using convex hulls Type de document : Article/Communication Auteurs : Jean-Pierre Renaud, Auteur ; Ankit Sagar , Auteur ; Pierre Barbillon, Auteur ; Olivier Bouriaud , Auteur ; Christine Deleuze, Auteur ; Cédric Vega , Auteur Année de publication : 2022 Projets : DEEPSURF / Pironon, Jacques, ARBRE / AgroParisTech Article en page(s) : n° 102939 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Statistiques [Termes IGN] données allométriques [Termes IGN] données lidar [Termes IGN] données localisées 3D [Termes IGN] échantillonnage [Termes IGN] erreur systématique [Termes IGN] étalonnage de modèle [Termes IGN] étalonnage des données [Termes IGN] extrapolation [Termes IGN] placette d'échantillonnage Résumé : (auteur) The ever-increasing availability of remote sensing data allows production of forest attributes maps, which are usually made using model-based approaches. These map products are sensitive to various bias sources, including model extrapolation. To identify, over a case study forest, the proportion of extrapolated predictions, we used a convex hull method applied to the auxiliary data space of an airborne laser scanning (ALS) flight. The impact of different sampling efforts was also evaluated. This was done by iteratively thinning a set of 487 systematic plots using nested sub-grids allowing to divide the sample by two at each level. The analysis were conducted for all alternative samples and evaluated against 56 independent validation plots. Residuals of the extrapolated validation plots were computed and examined as a function of their distance to the model calibration domain. Extrapolation was also characterized for the pixels of the area of interest (AOI) to upscale at population level. Results showed that the proportion of extrapolated pixels greatly reduced with an increasing sampling effort. It reached a plateau (ca. 20% extrapolation) with a sampling intensity of ca. 250-calibration plots. This contrasts with results on model’s root mean squared error (RMSE), which reached a plateau at a much lower sampling intensity. This result emphasizes the fact that with a low sampling effort, extrapolation risk remains high, even at a relatively low RMSE. For all attributes examined (i.e., stand density, basal area, and quadratic mean diameter) estimations were generally found to be biased for validation plots that were extrapolated. The method allows an easy identification of map pixels that are out of the calibration domain, making it an interesting tool to evaluate model transferability over an area of interest (AOI). It could also serve to compare “competing” models at a variable selection phase. From a model calibration perspective, it could serve a posteriori, to evaluate areas (in the auxiliary space) that merit further sampling efforts to improve model reliability. Numéro de notice : A2022-581 Affiliation des auteurs : LIF+Ext (2020- ) Thématique : FORET/IMAGERIE/MATHEMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.jag.2022.102939 Date de publication en ligne : 28/07/2022 En ligne : https://doi.org/10.1016/j.jag.2022.102939 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101341 [article]

Evaluation of the mixed-effects model and quantile regression approaches for predicting tree height in larch (Larix olgensis) plantations in northeastern China / Longfei Xie in Canadian Journal of Forest Research, Vol 52 n° 3 (March 2022)  

Public [article]  inCanadian Journal of Forest Research > Vol 52 n° 3 (March 2022) . - pp 309 - 319 Titre : Evaluation of the mixed-effects model and quantile regression approaches for predicting tree height in larch (Larix olgensis) plantations in northeastern China Type de document : Article/Communication Auteurs : Longfei Xie, Auteur ; Faris Rafi Almay Widagdo, Auteur ; Zheng Miao, Auteur ; Lihu Dong, Auteur ; Fengri Li, Auteur Année de publication : 2022 Article en page(s) : pp 309 - 319 Note générale : bibliographie Langues : Français (fre) Anglais (eng) Descripteur : [Vedettes matières IGN] Statistiques [Termes IGN] biométrie [Termes IGN] Chine [Termes IGN] diamètre à hauteur de poitrine [Termes IGN] hauteur des arbres [Termes IGN] Larix olgensis [Termes IGN] modèle de croissance végétale [Termes IGN] modèle de simulation [Termes IGN] régression non linéaire [Termes IGN] régression par quantile Résumé : (auteur) Tree height (H) is one of the most important tree variables and is widely used in growth and yield models, and its measurement is often time-consuming and costly. Hence, height–diameter (H–D) models have become a great alternative, providing easy-to-use and accurate tools for H prediction. In this study, H–D models were developed for Larix olgensis A. Henry in northeastern China. The Chapman–Richards function with three predictors (diameter at breast height, dominant tree height, and relative size of individual trees) performed best. Nonlinear mixed-effects (NLME) models and nonlinear quantile regressions (NQR9, nine quantiles; NQR5, five quantiles; and NQR3, three quantiles) were further used and improved the generalized H–D model, successfully providing accurate H predictions. In addition, the H predictions were calibrated using several measurements from subsamples, which were obtained from different sampling designs and sizes. The results indicated that the predictive accuracy was higher when calibrated by using any number of height measurements for the NLME model and more than three height measurements for the NQR3, NQR5, and NQR9 models. The best sampling strategy for the NLME and NQR models involved sampling medium-sized trees. Overall, the newly developed H–D models can provide highly accurate height predictions for L. olgensis. Numéro de notice : A2022-313 Affiliation des auteurs : non IGN Autre URL associée : Draft Thématique : FORET/MATHEMATIQUE Nature : Article DOI : 10.1139/cjfr-2021-0184 En ligne : https://doi.org/10.1139/cjfr-2021-0184 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100412 [article]

Deriving a tree growth model from any existing stand growth model / Quang V. Cao in Canadian Journal of Forest Research, Vol 52 n° 2 (February 2022)  

Public [article]  inCanadian Journal of Forest Research > Vol 52 n° 2 (February 2022) . - pp 137 - 147 Titre : Deriving a tree growth model from any existing stand growth model Type de document : Article/Communication Auteurs : Quang V. Cao, Auteur Année de publication : 2022 Article en page(s) : pp 137 - 147 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Statistiques [Termes IGN] désagrégation [Termes IGN] Etats-Unis [Termes IGN] méthode des moindres carrés [Termes IGN] modèle de croissance végétale [Termes IGN] modélisation de la forêt [Termes IGN] Pinus taeda [Termes IGN] régression [Termes IGN] surface terrière Résumé : (auteur) In this study, a new method was developed to derive a tree survival and diameter growth model from any existing stand-level model, without the need for individual-tree growth data. Predictions from the derived tree model are constrained to match the number of trees and the basal area per hectare as outputted by the stand model. The tree models derived from three different stand models were evaluated against a tree model, in both unadjusted and disaggregated forms. For the same stand-level model, the derived tree model outperformed its counterpart, the disaggregated tree model. Furthermore, except for one stand model with poor performance, the tree models derived from the remaining two stand models delivered results comparable to those obtained from the unadjusted tree model. The tree model derived from one stand model even performed slightly better than the unadjusted tree model. This result is significant because the coefficients of the unadjusted and disaggregated tree models had to be estimated from tree-level growth data, whereas the derived tree model required no tree growth data at all. The methodology presented in this study should be applicable when there is no ingrowth or recruitment of new trees. Numéro de notice : A2022-311 Affiliation des auteurs : non IGN Autre URL associée : Draft Thématique : FORET/MATHEMATIQUE Nature : Article DOI : 10.1139/cjfr-2021-0106 En ligne : https://doi.org/10.1139/cjfr-2021-0106 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100408 [article]

Modeling in forestry using mixture models fitted to grouped and ungrouped data / Eric K. Zenner in Forests, vol 12 n° 9 (September 2021)  

Public [article]  inForests > vol 12 n° 9 (September 2021) . - n° 1196 Titre : Modeling in forestry using mixture models fitted to grouped and ungrouped data Type de document : Article/Communication Auteurs : Eric K. Zenner, Auteur ; Mahdi Teimouri, Auteur Année de publication : 2021 Article en page(s) : n° 1196 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Statistiques [Termes IGN] algorithme espérance-maximisation [Termes IGN] complexité [Termes IGN] diamètre à hauteur de poitrine [Termes IGN] distribution de Weibull [Termes IGN] distribution, loi de [Termes IGN] dynamique de la végétation [Termes IGN] estimation par noyau [Termes IGN] méthode du maximum de vraisemblance (estimation) [Termes IGN] modélisation de la forêt [Termes IGN] structure d'un peuplement forestier Résumé : (auteur) The creation and maintenance of complex forest structures has become an important forestry objective. Complex forest structures, often expressed in multimodal shapes of tree size/diameter (DBH) distributions, are challenging to model. Mixture probability density functions of two- or three-component gamma, log-normal, and Weibull mixture models offer a solution and can additionally provide insights into forest dynamics. Model parameters can be efficiently estimated with the maximum likelihood (ML) approach using iterative methods such as the Newton-Raphson (NR) algorithm. However, the NR algorithm is sensitive to the choice of initial values and does not always converge. As an alternative, we explored the use of the iterative expectation-maximization (EM) algorithm for estimating parameters of the aforementioned mixture models because it always converges to ML estimators. Since forestry data frequently occur both in grouped (classified) and ungrouped (raw) forms, the EM algorithm was applied to explore the goodness-of-fit of the gamma, log-normal, and Weibull mixture distributions in three sample plots that exhibited irregular, multimodal, highly skewed, and heavy-tailed DBH distributions where some size classes were empty. The EM-based goodness-of-fit was further compared against a nonparametric kernel-based density estimation (NK) model and the recently popularized gamma-shaped mixture (GSM) models using the ungrouped data. In this example application, the EM algorithm provided well-fitting two- or three-component mixture models for all three model families. The number of components of the best-fitting models differed among the three sample plots (but not among model families) and the mixture models of the log-normal and gamma families provided a better fit than the Weibull distribution for grouped and ungrouped data. For ungrouped data, both log-normal and gamma mixture distributions outperformed the GSM model and, with the exception of the multimodal diameter distribution, also the NK model. The EM algorithm appears to be a promising tool for modeling complex forest structures. Numéro de notice : A2021-721 Affiliation des auteurs : non IGN Thématique : FORET Nature : Article DOI : 10.3390/f12091196 En ligne : https://doi.org/10.3390/f12091196 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98639 [article]

Regularized regression: A new tool for investigating and predicting tree growth / Stuart I. Graham in Forests, vol 12 n° 9 (September 2021)  

Public [article]  inForests > vol 12 n° 9 (September 2021) . - n° 1283 Titre : Regularized regression: A new tool for investigating and predicting tree growth Type de document : Article/Communication Auteurs : Stuart I. Graham, Auteur ; Ariel Rokem, Auteur ; Claire Fortunel, Auteur ; Nathan J.B. Kraft, Auteur ; Janneke Hille Ris Lambers, Auteur Année de publication : 2021 Article en page(s) : n° 1283 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Statistiques [Termes IGN] croissance des arbres [Termes IGN] inférence statistique [Termes IGN] interpolation [Termes IGN] modèle de simulation [Termes IGN] modélisation de la forêt [Termes IGN] placette d'échantillonnage [Termes IGN] régression Résumé : (auteur) Neighborhood models have allowed us to test many hypotheses regarding the drivers of variation in tree growth, but require considerable computation due to the many empirically supported non-linear relationships they include. Regularized regression represents a far more efficient neighborhood modeling method, but it is unclear whether such an ecologically unrealistic model can provide accurate insights on tree growth. Rapid computation is becoming increasingly important as ecological datasets grow in size, and may be essential when using neighborhood models to predict tree growth beyond sample plots or into the future. We built a novel regularized regression model of tree growth and investigated whether it reached the same conclusions as a commonly used neighborhood model, regarding hypotheses of how tree growth is influenced by the species identity of neighboring trees. We also evaluated the ability of both models to interpolate the growth of trees not included in the model fitting dataset. Our regularized regression model replicated most of the classical model’s inferences in a fraction of the time without using high-performance computing resources. We found that both methods could interpolate out-of-sample tree growth, but the method making the most accurate predictions varied among focal species. Regularized regression is particularly efficient for comparing hypotheses because it automates the process of model selection and can handle correlated explanatory variables. This feature means that regularized regression could also be used to select among potential explanatory variables (e.g., climate variables) and thereby streamline the development of a classical neighborhood model. Both regularized regression and classical methods can interpolate out-of-sample tree growth, but future research must determine whether predictions can be extrapolated to trees experiencing novel conditions. Overall, we conclude that regularized regression methods can complement classical methods in the investigation of tree growth drivers and represent a valuable tool for advancing this field toward prediction. Numéro de notice : A2021-720 Affiliation des auteurs : non IGN Thématique : FORET Nature : Article DOI : 10.3390/f12091283 En ligne : https://doi.org/10.3390/f12091283 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98636 [article]

JUST: MATLAB and python software for change detection and time series analysis / Ebrahim Ghaderpour in GPS solutions, vol 25 n° 3 (July 2021)  

Permalink

Stability problems for stochastic models: Theory and applications / Alexander Zeifman (2021)  

Permalink

Self-tuning robust adjustment within multivariate regression time series models with vector-autoregressive random errors / Boris Kargoll in Journal of geodesy, vol 94 n° 5 (May 2020)  

Permalink

A breakpoint detection in the mean model with heterogeneous variance on fixed time-intervals / Olivier Bock in Statistics and Computing, vol 29 n° 1 (February 2020)  

Permalink

Volcano-seismic transfer learning and uncertainty quantification with bayesian neural networks / Angel Bueno in IEEE Transactions on geoscience and remote sensing, vol 58 n° 2 (February 2020)  

Permalink

Applied and computational statistics / Sorana D. Bolboacă (2020)  

Permalink

Bayesian inference on complicated data / Niansheng Tang (2020)  

Permalink

Development of new homogenisation methods for GNSS atmospheric data. Application to the analysis of climate trends and variability / Annarosa Quarello (2020)  

Permalink

Planar maps, random walks and circle packing / Asaf Nachmias (2020)  

Permalink

Réponses de la productivité des forêts aux fluctuations météorologiques : biais et surestimations des estimations de terrain / Olivier Bouriaud (2020)  

Permalink