Search

Abstract As the need for ecosystem biodiversity assessment increases within the climate crisis framework, more and more studies using spectral variation hypothesis (SVH) are proposed to assess biodiversity at various scales. The SVH implies optical diversity (also called spectral diversity) is driven by light absorption dynamics associated with plant traits (PTs) variability (which is an indicator of functional diversity) which is, in turn, determined by biodiversity. In this study, we examined the relationship between PTs variability, optical diversity and α‐ and β‐diversity at different taxonomic ranks at the Monte Bondone grasslands, Trentino province, Italy. The results of the study showed that the PTs variability, at the α scale, was not correlated with biodiversity. On the other hand, the results observed at the community scale (β‐diversity) showed that the variation of some of the investigated biochemical and biophysical PTs was associated with the β‐diversity. We used the Mantel test to analyse the relationship between the PTs variability and species β‐diversity. The results showed a correlation coefficient of up to 0.50 between PTs variability and species β‐diversity. For higher taxonomic ranks such as family and functional groups, a slightly higher Spearman's correlation coefficient of up to 0.64 and 0.61 was observed, respectively. The SVH approach was also tested to estimate β‐diversity and we found that spectral diversity calculated by Spectral Angle Mapper showed to be a better proxy of biodiversity in the same ecosystem where the spectral diversity approach failed to estimate α‐diversity. These findings suggest that optical and PTs diversity approaches can be used to predict species diversity in the grasslands ecosystem where the species turnover is high.

Preprocessing Synthetic Aperture Radar (SAR) data is a crucial initial stage in leveraging SAR data for remote sensing applications. Terrain correction, both radiometric and geometric, and the detection of layover/shadow areas hold significant importance when SAR data are collected over mountainous regions. This study aims at investigating the impact of the Digital Elevation Model (DEM) used for terrain correction (radiometric and geometric) and for mapping layover/shadow areas on windthrow detection using COSMO SkyMed SAR images. The terrain correction was done using a radiometric and geometric terrain correction algorithm. Specifically, we evaluated five different DEMs: (i–ii) a digital terrain model and a digital surface model derived from airborne LiDAR flights; (iii) the ALOS Global Digital Surface Model; (iv) the Copernicus global DEM; and (v) the Shuttle Radar Topography Mission (SRTM) DEM. All five DEMs were resampled at 2 m and 30 m pixel spacing, obtaining a total of 10 DEMs. The terrain-corrected COSMO SkyMed SAR images were employed for windthrow detection in a forested area in the north of Italy. The findings revealed significant variations in windthrow detection across the ten corrections. The detailed LiDAR-derived terrain model (i.e., DTM at 2 m pixel spacing) emerged as the optimal choice for both pixel spacings considered.

ABSTRACTRot in commercial timber reduces the value of the wood substantially and estimating the occurrence, severity, and volume of heartwood rot would be a useful tool in decision-making to minimize economic losses. Remotely sensed data has recently been used for mapping rot on a single-tree level, and although the results have been relatively poor, some potential has been shown. This study applied area-based approaches to predict rot occurrence, rot severity, and rot volume , at an area level. Ground reference data were collected from harvester operations in 2019–2021. Predictor variables were calculated from multi-temporal remotely sensed data together with environmental variables. Response variables from the harvester data and predictor variables from remotely sensed data were aggregated to grid cells and to forest stands. Random Forest models were built for the different combinations of response variables and predictor subsets, and validated with both random- and spatial cross-validation. The results showed that it was not possible to estimate rot occurrence and rot severity with the applied modeling procedure (pR2: 0.00–0.16), without spatially close training data. The better performance of rot volume models (pR2: 0.12–0.37) was mainly due to the correlation between timber volume and rot volume.

Modeling and large-scale mapping of forest aboveground biomass (AGB) is a complicated, challenging, and expensive task. There are considerable variations in forest characteristics that create functional disparity for different models and needs comprehensive evaluation. Moreover, the human-bias involved in the process of modeling and evaluation affects the generalization of models at larger scales. In this article, we present an automated machine learning framework for modeling, evaluation, and stacking of multiple base models for AGB prediction. We incorporate a hyperparameter optimization procedure for automatic extraction of targeted features from multitemporal Sentinel-2 data that minimizes human-bias in the proposed modeling pipeline. We integrate the two independent frameworks for automatic feature extraction and automatic model ensembling and evaluation. The results suggest that the extracted target-oriented features have an excessive contribution of red-edge and short-wave infrared spectrum. The feature importance scale indicates a dominant role of summer-based features as compared to other seasons. The automated ensembling and evaluation framework produced a stacked ensemble of base models that outperformed individual base models in accurately predicting forest AGB. The stacked ensemble model delivered the best scores of R2cv = 0.71 and RMSE = 74.44 Mgha−1. The other base models delivered R2cv and RMSE ranging between 0.38–0.66 and 81.27–109.44 Mg ha−1, respectively. The model evaluation metrics indicated that the stacked ensemble model was more resistant to outliers and achieved a better generalization. Thus, the proposed study demonstrated an effective automated modeling pipeline for predicting AGB by minimizing human-bias and deployable over large and diverse forest areas.

Management intensity deeply influences meadow structure and functioning, therefore affecting grassland ecosystem services. Conservation and management measures, including European Common Agricultural Policy subsidies, should therefore be based on updated and publicly available data about management intensity. The mowing frequency is a crucial trait to describe meadows management intensity, but the potential of using vegetation indices from Sentinel-2 imagery for its retrieval has not been fully exploited. In this work we developed on the Google Earth Engine platform a four-phases algorithm to identify mowing frequency, including i) vegetation index time-series computing, ii) smoothing and resampling, iii) mowing detection, and iv) majority analysis. Mowing frequency during 2020 of 240 ha of grassland fields in the Italian Alps was used for algorithm optimization and evaluation. Six vegetation indexes (EVI, GVMI, MTCI, NDII, NDVI, RENDVI783.740) were tested as input to the proposed algorithm. The Normalized Difference Infrared Index (NDII) showed the best performance, resulting in mean absolute error of 0.07 and 93% overall accuracy on average at the four sites used for optimization, at pixel resolution. A slightly lower accuracy (mean absolute error = 0.10, overall accuracy = 90%) was obtained aggregating the maps to management parcels. The algorithm showed a good generalization ability, with a similar performance between global and local optimization and an average mean absolute error of 0.12 and an overall accuracy of 89% on average on the sites not used for parameters optimization. The lowest accuracies occurred in intensively managed grasslands surveyed by one satellite orbit only. This study demonstrates the suitability of the proposed algorithm to monitor very fragmented grasslands in complex mountain ecosystems. Google Earth Engine was used to develop the model and will enable researchers, agencies and practitioners to easily and quickly apply the code to map grassland mowing frequency for extensive grasslands protection and conservation, for mowing event verification, or for forage system characterization.

Bark beetles cause severe damage to European forests leading to impacts on many sectors, from the environmental to the economical. Timely mapping of the different stages of an attack is very important. Remote sensing has been widely used to map bark beetle damage using both airborne and satellite data. Newly available satellite multispectral data with a daily revisit time and high spatial resolution has the potential to monitor an attack in all its phases. This study explores the spectral separability of bark beetle infestation stages using the Planet imagery at individual tree level. Multi-temporal spectral analysis of 78 trees in different stages of a spruce bark beetle attack was carried out. Bands and vegetation indexes derived from 42 multispectral images were compared to eleven field surveys over a time span of approximately four months. The spectral separability analysis was done considering three criteria exploring: 1) the significance of the differences, 2) the magnitude of the differences and 3) the separability in a supervised classification context. The field surveys reported different effects depending on the season of the bark beetle attack - spring vs. summer. Spectral bands and indexes extracted from trees in the healthy and red-stage were significantly different. Trees in the green-attack stage at the end of the summer showed a statistically significant difference from healthy trees. The separability measured with a supervised classifier showed that it is possible to separate healthy, green-attack and red-stage trees with high accuracy values (kappa accuracy above 0.9).

Aboveground biomass (AGB) is an important forest attribute directly linked to the forest carbon pool. The use of satellite remote sensing (RS) data has increased for AGB prediction due to their large footprint and low-cost availability. However, they have been limited due to saturation effect that leads to low prediction precision. In this article, we propose an innovative and dynamic architecture based on generative neural network that extracts target oriented generative features for predicting forest AGB using satellite RS data. These features are more resilient to mixed forest types and geographical conditions as compared to the traditional features and models. The effectiveness of the proposed features was assessed by experiments performed using multispectral, synthetic aperture radar, and combined dual-source datasets. The proposed model achieved best performance in terms of precision, model agreement, and overfitting as compared to the other conventional models for all analyzed datasets. The t-distributed stochastic neighbor embedding scatterplots of the generative features clearly show one dimension of the feature space associated with the target AGB. Feature importance analysis indicated that the produced generative features were more significant than the conventional analytical features. Also, the model provided a robust framework for homogeneous fusion of multisensor features from satellite RS data for predicting AGB.

Making oil palm agriculture as efficient as possible is essential to ensuring that this economically important crop can be grown sustainably. To determine how oil palm growth rates vary across tropical landscapes, we used repeat airborne LiDAR data to map the height growth of >500,000 oil palms in Malaysian Borneo over a two-year period coinciding with the 2015–16 El Niño drought. Despite uncharacteristically dry and hot weather conditions, oils palms grew an average of 1.6 m yr−1 in height over this period. However, oil palm growth rates varied across the landscape and in relation to plant age, tending to be fastest for younger individuals growing closer to forest edges, further from rivers and at higher elevations. Our results highlight the ability of oil palms to grow even during periods of drought and showcase how cutting-edge remote sensing technologies can help improve the efficiency and sustainability of oil palm agriculture.

Grassland fractional vegetation cover (FVC) accurate mapping on a large scale is crucial, since degraded grasslands contribute less to provisioning services, carbon storage, water purification, erosion control and biodiversity conservation. The spatial and temporal resolution of Sentinel-2 (S2) and PlanetScope (PS) data has never been explored for grassland FVC estimation so far and will enable researchers and agencies to quantify and map timelier and more precisely grassland processes. In this paper we compare FVC estimation models developed from Landsat-8 (L8), S2 and PS imagery. The reference grassland FVC dataset was obtained on the Paganella ski runs (46.15°N, 11.01°E, Italy) applying unsupervised classification to nadir grassland RGB photographs taken from 1.35 m above the soil. Fractional Response Models between reference FVC and 18 vegetation indices (VIs) extracted from satellite imagery were fitted and analysed. Then, leave-one-out cross validation and spatiotemporal change analysis were also performed. Our study confirms the robustness of the commonly used VIs based on the difference between NIR and the red wavelength region (R2 = 0.91 for EVI using S2 imagery) and indicate that VIs based on the red-edge spectral region are the best performing for PS imagery (R2 = 0.89 for RECI). Only medium to high spatial resolution imagery (S2 and PS) precisely mapped spatial patterns at the study site, since grasslands FVC varies at a fine scale. Previously available imagery at medium to low spatial and temporal resolution (e.g., L8) may still be interesting for analysis requiring long time-series of data.

This article develops a robust feature fusion approach to enhance the classification performance of very high resolution (VHR) remote sensing images. Specifically, a novel two-stage multiple feature fusion (TsF) approach is proposed, which includes an intragroup and an intergroup feature fusion stages. In the first fusion stage, multiple features are grouped by clustering, where redundant information between different types of features is eliminated within each group. Then, features are pairwisely fused in an intergroup fusion model based on the guided filtering method. Finally, the fused feature set is imported into a classifier to generate the classification map. In this work, the original VHR spectral bands and their attribute profiles are taken as examples as input spectral and spatial features, respectively, in order to test the performance of the proposed TsF approach. Experimental results obtained on two QuickBird datasets covering complex urban scenarios demonstrate the effectiveness of the proposed approach in terms of generation of more discriminative fusion features and enhancing classification performance. More importantly, the fused feature dimensionality is limited at a certain level; thus, the computational cost will not be significantly increased even if multiple features are considered.

High-resolution (HR) satellite images, due to the technical constraints on spectral and spatial resolutions, usually contain only several broad spectral bands but with a very high spatial resolution. This provides rich spatial details of the objects on the Earth surface, while their spectral discrimination is relatively low. Recently, the increase of the satellite revisit times made it possible to acquire more frequent data coverage for finer classification. In this article, we proposed a novel multitemporal deep fusion network (MDFN) for short-term multitemporal HR images classification. Specifically, a two-branch structure of MDFN is designed, which includes a long short-term memory (LSTM) and a convolutional neural network (CNN). The LSTM branch is mainly used to learn the joint expression of different temporal-spectral features. For the CNN branch, the three-dimensional (3-D) convolution is firstly applied along the temporal and spectral dimensions to jointly learn the temporal-spatial and spectral-spatial information, respectively, and then the 2-D convolution is performed along the spatial dimension to further extract the spatial context information. Finally, features generated from the two different branches are fused to obtain the discriminative high-level semantic information for classification. Experimental results carried on two real multitemporal HR remote sensing datasets demonstrate that the proposed MDFN provides better classification performance over the state-of-the-art methods, and it also shows the potentiality to use short-term multitemporal HR images for more accurate land use/land cover mapping.

Change detection from multi-temporal remote sensing images is an effective way to identify the burned areas after forest fires. However, the complex image scenario and the similar spectral signatures in multispectral bands may lead to many false positive errors, which make it difficult to exact the burned areas accurately. In this paper, a novel-burned area change detection approach is proposed. It is designed based on a new Normalized Burn Ratio-SWIR (NBRSWIR) index and an automatic thresholding algorithm. The effectiveness of the proposed approach is validated on three Landsat-8 data sets presenting various fire disaster events worldwide. Compared to eight index-based detection methods that developed in the literature, the proposed approach has the best performance in terms of class separability (2.49, 1.74 and 2.06) and accuracy (98.93%, 98.57% and 99.51%) in detecting the burned areas. Simultaneously, it can also better suppress the complex irrelevant changes in the background.

Pathogenic wood decay fungi such as species of are some of the most serious forest pathogens in Europe, causing rot of tree boles and loss of growth, with estimated economic losses of eight hundred million euros per year. In conifers with low resinous heartwood such as species of and , these fungi are commonly confined to heartwood and thus external infection signs on the bark or foliage of trees are normally absent. Consequently, determining the extent of disease presence in a forest stand with field surveys is not practical for guiding forest management decisions such as optimal rotation time. Remote sensing technologies such as airborne laser scanning and aerial imagery are already used to reduce the reliance on fieldwork in forest inventories. This study aimed to use remote sensing to detect rot in spruce ( L. Karst.) forests in Norway. An airborne hyperspectral imager provided information for classifying the presence or absence of rot in a single-tree-based framework. Ground reference data showing the presence of rot were collected by harvest machine operators during the harvest of forest stands. Random forest and support vector machine algorithms were used to classify the presence and absence of rot. Results indicate a 64% overall classification accuracy for presence-absence classification of rot, although additional work remains to make the classifications usable for practical forest management.HeterobasidionPiceaAbiesPicea abies

Numerous species of pathogenic wood decay fungi, including members of the genera Heterobasidion and Armillaria, exist in forests in the northern hemisphere. Detection of these fungi through field surveys is often difficult due to a lack of visual symptoms and is cost-prohibitive for most applications. Remotely sensed data can offer a lower-cost alternative for collecting information about vegetation health. This study used hyperspectral imagery collected from unmanned aerial vehicles (UAVs) to detect the presence of wood decay in Norway spruce (Picea abies L. Karst) at two sites in Norway. UAV-based sensors were tested as they offer flexibility and potential cost advantages for small landowners. Ground reference data regarding pathogenic wood decay were collected by harvest machine operators and field crews after harvest. Support vector machines were used to classify the presence of root, butt, and stem rot infection. Classification accuracies as high as 76% with a kappa value of 0.24 were obtained with 490-band hyperspectral imagery, while 29-band imagery provided a lower classification accuracy (~60%, kappa = 0.13).

Tree species composition is an essential attribute in stand-level forest management inventories and remotely sensed data might be useful for its estimation. Previous studies on this topic have had several operational drawbacks, e.g., performance studied at a small scale and at a single tree-level with large fieldwork costs. The current study presents the results from a large-area inventory providing species composition following an operational area-based approach. The study utilizes a combination of airborne laser scanning and hyperspectral data and 97 field sample plots of 250 m collected over 350 km of productive forest in Norway. The results show that, with the availability of hyperspectral data, species-specific volume proportions can be provided in operational forest management inventories with acceptable results in 90% of the cases at the plot level. Dominant species were classified with an overall accuracy of 91% and a kappa-value of 0.73. Species-specific volumes were estimated with relative root mean square differences of 34%, 87%, and 102% for Norway spruce ( (L.) Karst.), Scots pine ( L.), and deciduous species, respectively. A novel tree-based approach for selecting pixels improved the results compared to a traditional approach based on the normalized difference vegetation index.22Picea abiesPinus sylvestris

Insect outbreaks affect forests, causing the deaths of trees and high economic loss. In this study, we explored the detection of European spruce bark beetle (Ips typographus, L.) outbreaks at the individual tree crown level using multispectral satellite images. Moreover, we explored the possibility of tracking the progression of the outbreak over time using multitemporal data. Sentinel-2 data acquired during the summer of 2020 over a bark beetle–infested area in the Italian Alps were used for the mapping and tracking over time, while airborne lidar data were used to automatically detect the individual tree crowns and to classify tree species. Mapping and tracking of the outbreak were carried out using a support vector machine classifier with input vegetation indices extracted from the multispectral data. The results showed that it was possible to detect two stages of the outbreak (i.e., early, and late) with an overall accuracy of 83.4%. Moreover, we showed how it is technically possible to track the evolution of the outbreak in an almost bi-weekly period at the level of the individual tree crowns. The outcomes of this paper are useful from both a management and ecological perspective: it allows forest managers to map a bark beetle outbreak at different stages with a high spatial accuracy, and the maps describing the evolution of the outbreak could be used in further studies related to the behavior of bark beetles.

Wood decay caused by pathogenic fungi in Norway spruce forests causes severe economic losses in the forestry sector, and currently no efficient methods exist to detect infected trees. The detection of wood decay could potentially lead to improvements in forest management and could help in reducing economic losses. In this study, airborne hyperspectral data were used to detect the presence of wood decay in the trees in two forest areas located in Etnedal (dataset I) and Gran (dataset II) municipalities, in southern Norway. The hyperspectral data used consisted of images acquired by two sensors operating in the VNIR and SWIR parts of the spectrum. Corresponding ground reference data were collected in Etnedal using a cut-to-length harvester while in Gran, field measurements were collected manually. Airborne laser scanning (ALS) data were used to detect the individual tree crowns (ITCs) in both sites. Different approaches to deal with pixels inside each ITC were considered: in particular, pixels were either aggregated to a unique value per ITC (i.e., mean, weighted mean, median, centermost pixel) or analyzed in an unaggregated way. Multiple classification methods were explored to predict rot presence: logistic regression, feed forward neural networks, and convolutional neural networks. The results showed that wood decay could be detected, even if with accuracy varying among the two datasets. The best results on the Etnedal dataset were obtained using a convolution neural network with the first five components of a principal component analysis as input (OA = 65.5%), while on the Gran dataset, the best result was obtained using LASSO with logistic regression and data aggregated using the weighted mean (OA = 61.4%). In general, the differences among aggregated and unaggregated data were small.

Wind disturbances represent the main source of damage in European forests, affecting them directly (windthrows) or indirectly due to secondary damages (insect outbreaks and forest fires). The assessment of windthrows damages is very important to establish adequate management plans and remote sensing can be very useful for this purpose. Many types of optical remote sensing data are available with different spectral, spatial and temporal resolutions, and many options are possible for data acquisition, i.e. immediately after the event or after a certain time. The objective of this study is to compare the windthrows mapping capabilities of two multispectral satellite constellations (i.e. Sentinel-2 and PlanetScope) characterized by very different spectral, spatial and temporal resolutions, and to evaluate the impact of the acquisition conditions on the mapping results. The analysed area, with an extent of 732 km2, is located in the Trentino-South Tyrol region (Italy) which was affected by the Vaia storm on the 27th-30th of October 2018, causing serious forest damages. The change vector analysis technique was used to detect the windthrows. For each data source, two pairs of images were considered: 1) pre- and post- event images acquired as close as possible to the event; 2) pre- and post- event images acquired at optimal conditions, i.e. at similar phenological state and similar illumination conditions. The results obtained with the two satellite constellations are very similar despite their different resolutions. Data acquired in optimal conditions allowed having the best detection rate (accuracy above 80 %), while data acquired just after the event showed many limitations. Improved spatial resolution (PlanetScope data) allows for a better delineation of the borders of the windthrow areas and of the detection of smaller windthrow patches.

Abstract Background Forests are a key component of the global carbon cycle, and research is needed into the effects of human-driven and natural processes on their carbon pools. Airborne laser scanning (ALS) produces detailed 3D maps of forest canopy structure from which aboveground carbon density can be estimated. Working with a ALS dataset collected over the 8049-km2 Wellington Region of New Zealand we create maps of indigenous forest carbon and evaluate the influence of wind by examining how carbon storage varies with aspect. Storms flowing from the west are a common cause of disturbance in this region, and we hypothesised that west-facing forests exposed to these winds would be shorter than those in sheltered east-facing sites. Methods The aboveground carbon density of 31 forest inventory plots located within the ALS survey region were used to develop estimation models relating carbon density to ALS information. Power-law models using rasters of top-of-the-canopy height were compared with models using tree-level information extracted from the ALS dataset. A forest carbon map with spatial resolution of 25 m was generated from ALS maps of forest height and the estimation models. The map was used to evaluate the influences of wind on forests. Results Power-law models were slightly less accurate than tree-centric models (RMSE 35% vs 32%) but were selected for map generation for computational efficiency. The carbon map comprised 4.5 million natural forest pixels within which canopy height had been measured by ALS, providing an unprecedented dataset with which to examine drivers of carbon density. Forests facing in the direction of westerly storms stored less carbon, as hypothesised. They had much greater above-ground carbon density for a given height than any of 14 tropical forests previously analysed by the same approach, and had exceptionally high basal areas for their height. We speculate that strong winds have kept forests short without impeding basal area growth. Conclusion Simple estimation models based on top-of-the canopy height are almost as accurate as state-of-the-art tree-centric approaches, which require more computing power. High-resolution carbon maps produced by ALS provide powerful datasets for evaluating the environmental drivers of forest structure, such as wind.

Plant biodiversity is an important feature of grassland ecosystems, as it is related to the provision of many ecosystem services crucial for the human economy and well-being. Given the importance of grasslands, research has been carried out in recent years on the potential to monitor them with novel remote sensing techniques. In this study, the optical diversity (also called spectral diversity) approach was adopted to check the potential of using high-resolution hyperspectral images to estimate α-diversity in grassland ecosystems. In 2018 and 2019, grassland species composition was surveyed and canopy hyperspectral data were acquired at two grassland sites: Monte Bondone (IT-MBo; species-rich semi-natural grasslands) and an experimental farm of the University of Padova, Legnaro, Padua, Italy (IT-PD; artificially established grassland plots with a species-poor mixture). The relationship between biodiversity (species richness, Shannon’s, species evenness, and Simpson’s indices) and optical diversity metrics (coefficient of variation-CV and standard deviation-SD) was not consistent across the investigated grassland plant communities. Species richness could be estimated by optical diversity metrics with an R = 0.87 at the IT-PD species-poor site. In the more complex and species-rich grasslands at IT-MBo, the estimation of biodiversity indices was more difficult and the optical diversity metrics failed to estimate biodiversity as accurately as in IT-PD probably due to the higher number of species and the strong canopy spatial heterogeneity. Therefore, the results of the study confirmed the ability of spectral proxies to detect grassland α-diversity in man-made grassland ecosystems but highlighted the limitations of the spectral diversity approach to estimate biodiversity when natural grasslands are observed. Nevertheless, at IT-MBo, the optical diversity metric SD calculated from post-processed hyperspectral images and transformed spectra showed, in the red part of the spectrum, a significant correlation (up to R = 0.56, p = 0.004) with biodiversity indices. Spatial resampling highlighted that for the IT-PD sward the optimal optical pixel size was 1 cm, while for the IT-MBo natural grassland it was 1 mm. The random pixel extraction did not improve the performance of the optical diversity metrics at both study sites. Further research is needed to fully understand the links between α-diversity and spectral and biochemical heterogeneity in complex heterogeneous ecosystems, and to assess whether the optical diversity approach can be adopted at the spatial scale to detect β-diversity. Such insights will provide more robust information on the mechanisms linking grassland diversity and optical heterogeneity.

An international data science challenge, called National Ecological Observatory Network—National Institute of Standards and Technology data science evaluation, was set up in autumn 2017 with the goal to improve the use of remote sensing data in ecological applications. The competition was divided into three tasks: (1) individual tree crown (ITC) delineation, for identifying the location and size of individual trees; (2) alignment between field surveyed trees and ITCs delineated on remote sensing data; and (3) tree species classification. In this paper, the methods and results of team Fondazione Edmund Mach (FEM) are presented. The ITC delineation (Task 1 of the challenge) was done using a region growing method applied to a near-infrared band of the hyperspectral images. The optimization of the parameters of the delineation algorithm was done in a supervised way on the basis of the Jaccard score using the training set provided by the organizers. The alignment (Task 2) between the delineated ITCs and the field surveyed trees was done using the Euclidean distance among the position, the height, and the crown radius of the ITCs and the field surveyed trees. The classification (Task 3) was performed using a support vector machine classifier applied to a selection of the hyperspectral bands and the canopy height model. The selection of the bands was done using the sequential forward floating selection method and the Jeffries Matusita distance. The results of the three tasks were very promising: team FEM ranked first in the data science competition in Task 1 and 2, and second in Task 3. The Jaccard score of the delineated crowns was 0.3402, and the results showed that the proposed approach delineated both small and large crowns. The alignment was correctly done for all the test samples. The classification results were good (overall accuracy of 88.1%, kappa accuracy of 75.7%, and mean class accuracy of 61.5%), although the accuracy was biased toward the most represented species.

Forest aboveground biomass (AGB) is a prime forest parameter that requires global level estimates to study the global carbon cycle. Light detection and ranging (LiDAR) is the state-of-the-art technology for AGB prediction but it is expensive, and its coverage is restricted to small areas. On the contrary, spaceborne Earth observation data are effective and economical information sources to estimate and monitor AGB at a large scale. In this paper, we present a study on the use of different spaceborne multispectral remote sensing data for the prediction of forest AGB. The objective is to evaluate the effects of temporal, spectral, and spatial capacities of multispectral satellite data for AGB prediction. The study was performed on multispectral data acquired by Sentinel-2, RapidEye, and Dove satellites which are characterized by different spatial resolutions, temporal availability, and number of spectral bands. A systematic process of least absolute shrinkage and selection operator (lasso) variable selection generalized linear modeling, leave-one-out cross-validation, and analysis was accomplished on each satellite dataset for AGB prediction. Results point out that the multitemporal data based AGB models were more effective in prediction than the single-time models. In addition, red-edge and short wave infrared (SWIR) channel dependent variables showed significant improvement in the modeling results and contributed to more than 50% of the selected variables. Results also suggest that high spatial resolution plays a smaller role than spectral and temporal information in the prediction of AGB. The overall analysis emphasizes a good potential of spaceborne multispectral data for developing sophisticated methods for AGB prediction especially with specific spectral channels and temporal information.

Red-edge (RE) spectral vegetation indices (SVIs)—combining bands on the sharp change region between near infrared (NIR) and visible (VIS) bands—alongside with SVIs solely based on NIR-shoulder bands (wavelengths 750–900 nm) have been shown to perform well in estimating leaf area index (LAI) from proximal and remote sensors. In this work, we used RE and NIR-shoulder SVIs to assess the full potential of bands provided by Sentinel-2 (S-2) and Sentinel-3 (S-3) sensors at both temporal and spatial scales for grassland LAI estimations. Ground temporal and spatial observations of hyperspectral reflectance and LAI were carried out at two grassland sites (Monte Bondone, Italy, and Neustift, Austria). A strong correlation (R2 > 0.8) was observed between grassland LAI and both RE and NIR-shoulder SVIs on a temporal basis, but not on a spatial basis. Using the PROSAIL Radiative Transfer Model (RTM), we demonstrated that grassland structural heterogeneity strongly affects the ability to retrieve LAI, with high uncertainties due to structural and biochemical PTs co-variation. The RENDVI783.740 SVI was the least affected by traits co-variation, and more studies are needed to confirm its potential for heterogeneous grasslands LAI monitoring using S-2, S-3, or Gaofen-5 (GF-5) and PRISMA bands.

Protection and recovery of natural resource and biodiversity requires accurate monitoring at multiple scales. Airborne Laser Scanning (ALS) provides high-resolution imagery that is valuable for monitoring structural changes to vegetation, providing a reliable reference for ecological analyses and comparison purposes, especially if used in conjunction with other remote-sensing and field products. However, the potential of ALS data has not been fully exploited, due to limits in data availability and validation. To bridge this gap, the global network for airborne laser scanner data (GlobALS) has been established as a worldwide network of ALS data providers that aims at linking those interested in research and applications related to natural resources and biodiversity monitoring. The network does not collect data itself but collects metadata and facilitates networking and collaborative research amongst the end-users and data providers. This letter describes this facility, with the aim of broadening participation in GlobALS.

Competitive interactions are important predictors of tree growth. Spatial and temporal changes in resource availability, and variation in species and spatial patterning of trees alter competitive interactions, thus affecting tree growth and, hence, biomass. Competition indices are used to quantify the level of competition among trees. As these indices are normally computed only over small areas, where field measurements are done, it would be useful to have a tool to predict them over large areas. On this regard, remote sensing, and in particular light detection and ranging (lidar) data, could be the perfect tool. The objective of this study was to use lidar metrics to predict competition (on the basis of distance-dependent competition indices) of individual trees and to relate them with tree aboveground biomass (AGB). The selected study area was a mountain forest area located in the Italian Alps. The analyses focused on the two dominant species of the area: Silver fir (Abies alba Mill.) and Norway spruce (Picea abies (L.) H. Karst). The results showed that lidar metrics could be used to predict competition indices of individual trees (R2 above 0.66). Moreover, AGB decreased as competition increased, suggesting that variations in the availability of resources in the soil, and the ability of plants to withstand competition for light may influence the partitioning of biomass.

The use of remote sensing data for tree species classification in tropical forests is still a challenging task, due to their high floristic and spectral diversity. In this sense, novel sensors on board of unmanned aerial vehicle (UAV) platforms are a rapidly evolving technology that provides new possibilities for tropical tree species mapping. Besides the acquisition of high spatial and spectral resolution images, UAV-hyperspectral cameras operating in frame format enable to produce 3D hyperspectral point clouds. This study investigated the use of UAV-acquired hyperspectral images and UAV-photogrammetric point cloud (PPC) for classification of 12 major tree species in a subtropical forest fragment in Southern Brazil. Different datasets containing hyperspectral visible/near-infrared (VNIR) bands, PPC features, canopy height model (CHM), and other features extracted from hyperspectral data (i.e., texture, vegetation indices-VIs, and minimum noise fraction-MNF) were tested using a support vector machine (SVM) classifier. The results showed that the use of VNIR hyperspectral bands alone reached an overall accuracy (OA) of 57% (Kappa index of 0.53). Adding PPC features to the VNIR hyperspectral bands increased the OA by 11%. The best result was achieved combining VNIR bands, PPC features, CHM, and VIs (OA of 72.4% and Kappa index of 0.70). When only the CHM was added to VNIR bands, the OA increased by 4.2%. Among the hyperspectral features, besides all the VNIR bands and the two VIs (NDVI and PSSR), the first four MNF features and the textural mean of 565 and 679 nm spectral bands were pointed out as more important to discriminate the tree species according to Jeffries−Matusita (JM) distance. The SVM method proved to be a good classifier for the tree species recognition task, even in the presence of a high number of classes and a small dataset.

Light detection and ranging (lidar) data are nowadays a standard data source in studies related to forest ecology and environmental mapping. Medium/high point density lidar data allow to automatically detect individual tree crowns (ITCs), and they provide useful information to predict stem diameter and aboveground biomass of each tree represented by a detected ITC. However, acquisition of field data is necessary for the construction of prediction models that relate field data to lidar data and for validation of such models. When working at ITC level, field data collection is often expensive and time-consuming as accurate tree positions are needed. Active learning (AL) can be very useful in this context as it helps to select the optimal field trees to be measured, reducing the field data collection cost. In this study, we propose a new method of AL for regression based on the minimization of the field data collection cost in terms of distance to navigate between field sample trees, and accuracy in terms of root mean square error of the predictions. The developed method is applied to the prediction of diameter at breast heights (DBH) and aboveground biomass (AGB) of individual trees by using their height and crown diameter as independent variables and support vector regression. The proposed method was tested on two boreal forest datasets, and the obtained results show the effectiveness of the proposed selecting strategy to provide substantial improvements over the different iterations compared to a random selection. The obtained RMSE of DBH/AGB for the first dataset was 5.09 cm/95.5 kg with a cost equal to 8256/6173 m by using the proposed multi-objective method of selection. However, by using a random selection, the RMSE was 5.20 cm/102.1 kg with a cost equal to 28,391/30,086 m. The proposed approach can be efficient in order to get more accurate predictions with smaller costs, especially when a large forest area with no previous field data is subject to inventory and analysis.

In forest management, site index information is essential for planning silvicultural operations and forecasting forest development. Site index is most commonly expressed as the average height of the dominant trees at a certain index age, and can be determined either by photo interpretation, field measurements, or projection of age combined with height estimates from remote sensing. However, recently it has been shown that site index can be accurately predicted from bi-temporal airborne laser scanner (ALS) data. Furthermore, single-time hyperspectral data have also been shown to be correlated to site index. The aim of the current study was to compare the accuracy of modelling site index using (1) data from bi-temporal ALS; (2) single-time hyperspectral data with different types of preprocessing; and (3) combined bi-temporal ALS and single-time hyperspectral data. The period between the ALS acquisitions was 11 years. The preprocessing of the hyperspectral data included an atmospheric correction and/or a normalization of the reflectance. Furthermore, a selection of pixels was carried out based on NDVI and compared to using all pixels. The results showed that bi-temporal ALS data explained about 70% (R2) of the variation in the site index, and the RMSE values from a cross-validation were 3.0 m and 2.2 m for spruce- and pine-dominated plots, respectively. Corresponding values for the different single-time hyperspectral datasets were 54%, 3.9 m, and 2.5 m. With bi-temporal ALS data and hyperspectral data used in combination, the results indicated that the contribution from the hyperspectral data was marginal compared to just using bi-temporal ALS. We also found that models constructed with normalized hyperspectral data produced lower RMSE values compared to those constructed with atmospherically corrected data, and that a selection of pixels based on NDVI did not improve the results compared to using all pixels.

The linearity and scale-dependency of ecosystem biodiversity and productivity relationships (BPRs) have been under intense debate. In a changing climate, monitoring BPRs within and across different ecosystem types is crucial, and novel remote sensing tools such as the Sentinel-2 (S2) may be adopted to retrieve ecosystem diversity information and to investigate optical diversity and productivity patterns. But are the S2 spectral and spatial resolutions suitable to detect relationships between optical diversity and productivity? In this study, we implemented an integrated analysis of spatial patterns of grassland productivity and optical diversity using optical remote sensing and Eddy Covariance data. Across-scale optical diversity and ecosystem productivity patterns were analyzed for different grassland associations with a wide range of productivity. Using airborne optical data to simulate S2, we provided empirical evidence that the best optical proxies of ecosystem productivity were linearly correlated with optical diversity. Correlation analysis at increasing pixel sizes proved an evident scale-dependency of the relationships between optical diversity and productivity. The results indicate the strong potential of S2 for future large-scale assessment of across-ecosystem dynamics at upper levels of observation.

In this study, the potential of multispectral airborne laser scanner (ALS) data to model and predict some forest characteristics was explored. Four complementary characteristics were considered, namely, aboveground biomass per hectare, Gini coefficient of the diameters at breast height, Shannon diversity index of the tree species, and the number of trees per hectare. Multispectral ALS data were acquired with an Optech Titan sensor, which consists of three scanners, called channels, working in three wavelengths (532 nm, 1064 nm, and 1550 nm). Standard ALS data acquired with a Leica ALS70 system were used as a reference. The study area is located in Southern Norway, in a forest composed of Scots pine, Norway spruce, and broadleaf species. ALS metrics were extracted for each plot from both elevation and intensity values of the ALS points acquired with both sensors, and for all three channels of the ALS multispectral sensor. Regression models were constructed using different combinations of metrics. The results showed that all four characteristics can be accurately predicted with both sensors (the best R2 being greater than 0.8), but the models based on the multispectral ALS data provide more accurate results. There were differences regarding the contribution of the three channels of the multispectral ALS. The models based on the data of the 532 nm channel seemed to be the least accurate.

Fusion of ALS and hyperspectral data can offer a powerful basis for the discrimination of tree species and enables an accurate prediction of species-specific attributes. In this study, the fused airborne laser scanning (ALS) data and hyperspectral images were used to model and predict the total and species-specific volumes based on three forest inventory approaches, namely the individual tree crown (ITC) approach, the semi-ITC approach, and the area-based approach (ABA). The performances of these inventory approaches were analyzed and compared at the plot level in a complex Alpine forest in Italy. For the ITC and semi-ITC approaches, an ITC delineation algorithm was applied. With the ITC approach, the species-specific volumes were predicted with allometric models for each crown segment and aggregated to the total volume. For the semi-ITC and ABA, a multivariate k-most similar neighbor method was applied to simultaneously predict the total and species-specific volumes using leave-one-out cross-validation at the plot level. In both methods, the ALS and hyperspectral variables were important for volume modeling. The total volume of the ITC, semi-ITC, and ABA resulted in relative root mean square errors (RMSEs) of 25.31%, 17.41%, 30.95% of the mean and systematic errors (mean differences) of 21.59%, −0.27%, and −2.69% of the mean, respectively. The ITC approach achieved high accuracies but large systematic errors for minority species. For majority species, the semi-ITC performed slightly better compared to the ABA, resulting in higher accuracies and smaller systematic errors. The results indicated that the semi-ITC outperformed the two other inventory approaches. To conclude, we suggest that the semi-ITC method is further tested and assessed with attention to its potential in operational forestry applications, especially in cases for which accurate species-specific forest biophysical attributes are needed.