Machine learning combines inductive and automated techniques for recognizing patterns. These techniques can be used with remote sensing datasets to map aboveground biomass (AGB) with an acceptable degree of accuracy for evaluation and management of forest ecosystems. Unfortunately, statistically rigorous comparisons of machine learning algorithms are scarce. The aim of this study was to compare the performance of the 3 most common nonparametric machine learning techniques reported in the literature, vis., Support Vector Machine (SVM), k-nearest neighbor (kNN) and Random Forest (RF), with that of the parametric multiple linear regression (MLR) for estimating AGB from Landsat-5 Thematic Mapper (TM) spectral reflectance data, texture features derived from the Normalized Difference Vegetation Index (NDVI), and topographical features derived from a digital elevation model (DEM). The results obtained for 99 permanent sites (for calibration/validation of the models) established during the winter of 2011 by systematic sampling in the state of Durango (Mexico), showed that SVM performed best once the parameterization had been optimized. Otherwise, SVM could be outperformed by RF. However, the kNN yielded the best overall results in relation to the goodness-of-fit measures. The findings confirm that nonparametric machine learning algorithms are powerful tools for estimating AGB with datasets derived from sensors with medium spatial resolution. Résumé. L'apprentissage automatique combine des techniques inductives et automatisées pour la reconnaissance des formes. Ces techniques peuvent être utilisées avec des ensembles de données de télédétection pour cartographier la biomasse aérienne « aboveground biomass » (AGB) avec un degré de précision acceptable pour l'évaluation et la gestion des écosystèmes forestiers. Malheureusement, des comparaisons statistiquement rigoureuses des algorithmes d'apprentissage automatique sont rares. Le but de cette étude était de comparer les performances des 3 méthodes d'apprentissage automatique non paramétriques les plus fréquemment rapportées dans la littérature, vis., les machines à vecteurs de support « Support Vector Machine » (SVM), les k plus proches voisins « k-nearest neighbor » (kNN) et les forêts aléatoires « Random Forest » (RF), avec celle de la régression linéaire multiple paramétrique (MLR) pour l'estimation de l'AGB provenant des données de réflectance spectrale de Landsat-5 Thematic Mapper (TM), des caractéristiques de texture dérivées de l'indice de végétation par différence normalisée « Normalized Difference Vegetation Index » (NDVI) et des caractéristiques topographiques dérivées d'un modèle numérique de terrain « digital elevation model » (DEM).Les résultats obtenus pour 99 sites permanents (pour la calibration/validation des modèles) établis au cours de l'hiver 2011 par l'échantillonnage systématique dans l' État de Durango (Mexique), ont montré que les SVM montrent leurs meilleures performances une fois que le paramétrage a été optimisé. Par ailleurs, les SVM pourraient ê...

Solar radiation is affected by absorption and emission phenomena during its downward trajectory from the Sun to the Earth's surface and during the upward trajectory detected by satellite sensors. This leads to distortion of the ground radiometric properties (reflectance) recorded by satellite images, used in this study to estimate aboveground forest biomass (AGB). Atmospherically-corrected remote sensing data can be used to estimate AGB on a global scale and with moderate effort. The objective of this study was to evaluate four atmospheric correction algorithms (for surface reflectance), ATCOR2 (Atmospheric Correction for Flat Terrain), COST (Cosine of the Sun Zenith Angle), FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes) and 6S (Second Simulation of Satellite Signal in the Solar), and one radiometric correction algorithm (for reflectance at the sensor) ToA (Apparent Reflectance at the Top of Atmosphere) to estimate AGB in temperate forest in the northeast of the state of Durango, Mexico. The AGB was estimated from Landsat 5 TM imagery and ancillary information from a digital elevation model (DEM) using the non-parametric multivariate adaptive regression splines (MARS) technique. Field reference data for the model training were collected by systematic sampling of 99 permanent forest growth and soil research sites (SPIFyS) established during the winter of 2011. The following predictor variables were identified in the MARS model: Band 7, Band 5, slope (β), Wetness Index (WI), NDVI and MSAVI2. After cross-validation, 6S was found to be the optimal model for estimating AGB (R 2 = 0.71 and RMSE = 33.5 Mg¨ha´1; 37.61% of the average stand biomass). We conclude that atmospheric and radiometric correction of satellite images can be used along with non-parametric techniques to estimate AGB with acceptable accuracy.

No abstract

Understanding the linkage between accumulated fuel dryness and temporal fire occurrence risk is key for improving decision-making in forest fire management, especially under growing conditions of vegetation stress associated with climate change. This study addresses the development of models to predict the number of 10-day observed Moderate-Resolution Imaging Spectroradiometer (MODIS) active fire hotspots-expressed as a Fire Hotspot Density index (FHD)-from an Accumulated Fuel Dryness Index (AcFDI), for 17 main vegetation types and regions in Mexico, for the period 2011-2015. The AcFDI was calculated by applying vegetation-specific thresholds for fire occurrence to a satellite-based fuel dryness index (FDI), which was developed after the structure of the Fire Potential Index (FPI). Linear and non-linear models were tested for the prediction of FHD from FDI and AcFDI. Non-linear quantile regression models gave the best results for predicting FHD using AcFDI, together with auto-regression from previously observed hotspot density values. The predictions of 10-day observed FHD values were reasonably good with R 2 values of 0.5 to 0.7 suggesting the potential to be used as an operational tool for predicting the expected number of fire hotspots by vegetation type and region in Mexico. The presented modeling strategy could be replicated for any fire danger index in any region, based on information from MODIS or other remote sensors.

An accurate estimation of forests’ aboveground biomass (AGB) is required because of its relevance to the carbon cycle, and because of its economic and ecological importance. The selection of appropriate variables from satellite information and physical variables is important for precise AGB prediction mapping. Because of the complex relationships for AGB prediction, non-parametric machine-learning techniques represent potentially useful techniques for AGB estimation, but their use and comparison in forest remote-sensing applications is still relatively limited. The objective of the present study was to evaluate the performance of automatic learning techniques, support vector regression (SVR) and random forest (RF), to predict the observed AGB (from 318 permanent sampling plots) from the Landsat 8 Landsat 8 Operational Land Imager (OLI) sensor, spectral indexes, texture indexes and physical variables the Sierra Madre Occidental in Mexico. The result showed that the best SVR model explained 80% of the total variance (root mean square error (RMSE) = 8.20 Mg ha−1). The variables that best predicted AGB, in order of importance, were the bands that belong to the region of red and near and middle infrared, and the average temperature. The results show that the SVR technique has a good potential for the estimation of the AGB and that the selection of the model hyperparameters has important implications for optimizing the goodness of fit.