Search

The classification of ultra-high-resolution (UHR) imagery, characterized by spatial resolutions exceeding 10 cm, presents opportunities and challenges distinct from lower-resolution counterparts. Particularly, challenges are pronounced in some scenarios, such as mapping plant species in coastal environments, where similar vegetation responses and small plant sizes pose additional difficulties. The present work addressed such issues by developing a UHR vegetation cover classification model at the single plant level using data from uncrewed aerial systems (UASs) equipped with a multispectral optical sensor. The model was tested across the San Rossore Regional Park (Italy), where three pilot areas were defined as training-test-validation sites. The proposed solution consists of a hierarchical two-level-of-detail machine learning model based on object-based image analysis (OBIA) and random forest. This model considers spectral features and indices, elevation, and texture and can classify twelve plant species and two service classes (debris and sand) within the study areas. Train and test were carried out utilizing UAS flight data collected during two specific phenological periods and precise field data derived from in-situ vegetation surveys, which provided 937 herbaceous and shrub samples. The model performance was evaluated based on the error matrix and 50-fold stratified cross-validation method, obtaining an overall accuracy (OA) of 0.76 and a standard deviation of 0.08. Such assessment underscored the crucial role of texture information, in addition to radiometric and elevation. Finally, the model was tested against an unseen dataset, proving its transferability (OA equal to 0.62). Although the discussion highlights some aspects to be further improved and claims for future research, the first version of this hierarchical classification model demonstrated its potential for mapping and monitoring coastal sand dune ecosystems, providing data for understanding and, eventually, modeling ecological and biogeomorphological dynamics. [ABSTRACT FROM AUTHOR]

The proliferation of invasive plant species poses a significant ecological threat, necessitating effective mapping strategies for control and conservation efforts. Existing studies employing unmanned aerial vehicles (UAVs) and multispectral (MS) sensors in complex natural environments have predominantly relied on classical machine learning (ML) models for mapping plant species in natural environments. However, a critical gap exists in the literature regarding the use of deep learning (DL) techniques that integrate MS data and vegetation indices (VIs) with different feature extraction techniques to map invasive species in complex natural environments. This research addresses this gap by focusing on mapping the distribution of the Broad-leaved pepper (BLP) along the coastal strip in the Sunshine Coast region of Southern Queensland in Australia. The methodology employs a dual approach, utilising classical ML models including Random Forest (RF), eXtreme Gradient Boosting (XGBoost), and Support Vector Machine (SVM) in conjunction with the U-Net DL model. This comparative analysis allows for an in-depth evaluation of the performance and effectiveness of both classical ML and advanced DL techniques in mapping the distribution of BLP along the coastal strip. Results indicate that the DL U-Net model outperforms classical ML models, achieving a precision of 83%, recall of 81%, and F1–score of 82% for BLP classification during training and validation. The DL U-Net model attains a precision of 86%, recall of 76%, and F1–score of 81% for BLP classification, along with an Intersection over Union (IoU) of 68% on the separate test dataset not used for training. These findings contribute valuable insights to environmental conservation efforts, emphasising the significance of integrating MS data with DL techniques for the accurate mapping of invasive plant species. [ABSTRACT FROM AUTHOR]

This study examines the application of unsupervised classification techniques in the mapping of forest vegetation, aiming to align vegetation cover with the Braun-Blanquet classification system through remote sensing. By leveraging Landsat 8 and 9 satellite imagery and advanced clustering algorithms, specifically the Weka X-Means, this research addresses the challenge of minimizing researcher subjectivity in vegetation mapping. The methodology incorporates a two-step clustering approach to accurately classify forest communities, utilizing a comprehensive set of vegetation indices to distinguish between different types of forest ecosystems. The validation of the classification model relied on a detailed analysis of over 17,000 relevés from the "Flora" database, ensuring a high degree of accuracy in matching satellite-derived vegetation classes with field observations. The study's findings reveal the successful identification of 44 forest community types that was aggregated into seven classes of Braun-Blanquet classification system, demonstrating the efficacy of unsupervised classification in generating reliable vegetation maps. This work not only contributes to the advancement of remote sensing applications in ecological research, but also provides a valuable tool for natural resource management and conservation planning. The integration of unsupervised classification with the Braun-Blanquet system presents a novel approach to vegetation mapping, offering insights into ecological characteristics, and can be good starter point for sequestration potential of forest communities' assessment in the Republic of Tatarstan. [ABSTRACT FROM AUTHOR]

Trait-based approaches are of increasing concern in predicting vegetation changes and linking ecosystem structures to functions at large scales. However, a critical challenge for such approaches is acquiring spatially continuous plant functional trait maps. Here, six key plant functional traits were selected as they can reflect plant resource acquisition strategies and ecosystem functions, including specific leaf area (SLA), leaf dry matter content (LDMC), leaf N concentration (LNC), leaf P concentration (LPC), leaf area (LA) and wood density (WD). A total of 34 589 in situ trait measurements of 3447 seed plant species were collected from 1430 sampling sites in China and were used to generate spatial plant functional trait maps (∼1 km), together with environmental variables and vegetation indices based on two machine learning models (random forest and boosted regression trees). To obtain the optimal estimates, a weighted average algorithm was further applied to merge the predictions of the two models to derive the final spatial plant functional trait maps. The models showed good accuracy in estimating WD, LPC and SLA, with average R2 values ranging from 0.48 to 0.68. In contrast, both the models had weak performance in estimating LDMC, with average R2 values less than 0.30. Meanwhile, LA showed considerable differences between the two models in some regions. Climatic effects were more important than those of edaphic factors in predicting the spatial distributions of plant functional traits. Estimates of plant functional traits in northeastern China and the Qinghai–Tibetan Plateau had relatively high uncertainties due to sparse samplings, implying a need for more observations in these regions in the future. Our spatial trait maps could provide critical support for trait-based vegetation models and allow exploration of the relationships between vegetation characteristics and ecosystem functions at large scales. The six plant functional trait maps for China with 1 km spatial resolution are now available at 10.6084/m9.figshare.22351498 (An et al., 2023). [ABSTRACT FROM AUTHOR]

Deep learning (DL) models are gaining popularity in forest variable prediction using Earth observation (EO) images. However, in practical forest inventories, reference datasets are often represented by plot- or stand-level measurements, while high-quality representative wall-to-wall reference data for end-to-end training of DL models are rarely available. Transfer learning facilitates expansion of the use of deep learning models into areas with sub-optimal training data by allowing pretraining of the model in areas where high-quality teaching data are available. In this study, we perform a "model transfer" (or domain adaptation) of a pretrained DL model into a target area using plot-level measurements and compare performance versus other machine learning models. We use an earlier developed UNet based model (SeUNet) to demonstrate the approach on two distinct taiga sites with varying forest structure and composition. The examined SeUNet model uses multi-source EO data to predict forest height. Here, EO data are represented by a combination of Copernicus Sentinel-1 C-band SAR and Sentinel-2 multispectral images, ALOS-2 PALSAR-2 SAR mosaics and TanDEM-X bistatic interferometric radar data. The training study site is located in Finnish Lapland, while the target site is located in Southern Finland. By leveraging transfer learning, the SeUNet prediction achieved root mean squared error (RMSE) of 2.70 m and R2 of 0.882, considerably more accurate than traditional benchmark methods. We expect such forest-specific DL model transfer can be suitable also for other forest variables and other EO data sources that are sensitive to forest structure. [ABSTRACT FROM AUTHOR]

The Tibetan Plateau (TP) hosts a variety of vegetation types ranging from broadleaved and needle-leaved forests at the lower altitudes and mesic areas to alpine grassland at the higher altitudes and xeric areas. Accurate and detailed mapping of the vegetation distribution on TP is essential for an improved understanding of climate change effects on terrestrial ecosystems. Yet, existing land cover datasets of TP are either provided at a low spatial resolution or have insufficient vegetation types to characterize certain unique TP ecosystems, such as the alpine scree. Here, we produced a 10 m resolution TP land cover map with 12 vegetation classes and 3 non-vegetation classes for the year 2022 (referred as TP_LC10-2022) by leveraging state-of-the-art remote sensing approaches including the Sentinel-1 and Sentinel-2 imagery, environmental and topographic datasets, and 4 machine learning models using Google Earth Engine platform. Our dataset TP_LC10-2022 achieved an overall classification accuracy of 86.5% with a Kappa coefficient of 0.854. By comparing with 4 existing global land cover products, TP_LC10-2022 showed significant improvements in terms of reflecting local-scale vertical variations in the southeast TP region. Moreover, we found that alpine scree occupied 13.99% of the TP region which was ignored in existing land cover datasets, and that shrublands occupied 4.63% of the TP region characterized by distinct forms of deciduous shrublands and evergreen shrublands largely determined by topography and missed in existing land cover datasets. Our dataset provides a solid foundation for further analyses which need accurate delineation of these unique vegetation types in TP. The TP_LC10-2022 and the sample dataset are freely available at https://doi.org/10.5281/zenodo.8228112 (Huang et al., 2023b) and https://doi.org/10.5281/zenodo.8227942 (Huang et al., 2023a) respectively. Additionally, the classification map can be viewed through https://cold-classifier.users.earthengine.app/view/tplc10-2022. [ABSTRACT FROM AUTHOR]

The delineation of management zones is one of the ways to enable the spatially differentiated management of plots using precision agriculture tools. Over the years, the spatial variability of data collected from soil and plant sampling started to be replaced by data collected by proximal and orbital sensors. As a result, the variety and volume of data have increased considerably, making it necessary to use advanced computational tools, such as machine learning, for data analysis and decision-making support. This paper presents a methodology used to establish management zones (MZ) in precision agriculture by analyzing data obtained from soil sampling, proximal sensors and orbital sensors, in experiments carried out in four plots featuring soybean–cotton and soybean–corn crops, in Mato Grosso and Paraná states, Brazil. Four procedures were evaluated, using different input data sets for the MZ delineation: (I) soil attributes, including clay content, apparent electrical conductivity or fertility, along with elevation, yield maps and vegetation indices (VIs) captured during the peak crop biomass period; (II) soil attributes in conjunction with VIs demonstrating strong correlations; (III) solely VIs exhibiting robust correlation with soil attributes and yield; (IV) VIs selected via random forests to identify the importance of the variable for estimating yield. The results showed that the VIs derived from satellite images could effectively replace other types of data. For plots where the natural spatial variability can be easily identified, all procedures favor obtaining MZ maps that allow reductions of 40% to 70% in yield variance, justifying their use. On the other hand, in plots with low natural spatial variability and that do not have reliable yield maps, different data sets used as input do not help in obtaining feasible MZ maps. For areas where anthropogenic activities with spatially differentiated treatment are already present, the exclusive use of VIs for the delineation of MZs must be carried out with reservations. [ABSTRACT FROM AUTHOR]

Wetlands are one of the major contributors of methane (CH4) emissions to the atmosphere and the intensity of emissions is driven by local environmental variables and spatial heterogeneity. Peatlands are a major wetland class and there are numerous studies that provide estimates of methane emissions at chamber or eddy covariance scales, but these are not often aggregated to the site/ecosystem scale. This study provides a robust approach to map dominant vegetation communities and to use these areas to upscale methane fluxes from chamber to site scale using a simple weighted-area approach. The proposed methodology was tested at three peatlands in Ireland over a duration of 2 years. The annual vegetation maps showed an accuracy ranging from 83 to 99% for near-natural to degraded sites respectively. The upscaled fluxes were highest (2.25 and 3.80 gC m−2 y−1) at the near-natural site and the rehabilitation (0.17 and 0.31 gC m−2 y−1), degraded (0.15 and 0.27 gC m−2 y−1) site emissions were close to net-zero throughout the study duration. Overall, the easy to implement methodology proposed in this study can be applied across various landuse types to assess the impact of peatland rehabilitation on methane emissions by mapping ecological change. [ABSTRACT FROM AUTHOR]

Global mapping of essential vegetation traits (EVTs) through data acquired by Earth-observing satellites provides a spatially explicit way to analyze the current vegetation states and dynamics of our planet. Although significant efforts have been made, there is still a lack of global and consistently derived multi-temporal trait maps that are cloud-free. Here we present the processing chain for the spatiotemporally continuous production of four EVTs at a global scale: (1) fraction of absorbed photosynthetically active radiation (FAPAR), (2) leaf area index (LAI), (3) fractional vegetation cover (FVC), and (4) leaf chlorophyll content (LCC). The proposed workflow presents a scalable processing approach to the global cloud-free mapping of the EVTs. Hybrid retrieval models, named S3-TOA-GPR-1.0-WS, were implemented into Google Earth Engine (GEE) using Sentinel-3 Ocean and Land Color Instrument (OLCI) Level-1B for the mapping of the four EVTs along with associated uncertainty estimates. We used the Whittaker smoother (WS) for the temporal reconstruction of the four EVTs, which led to continuous data streams, here applied to the year 2019. Cloud-free maps were produced at 5 km spatial resolution at 10-day time intervals. The consistency and plausibility of the EVT estimates for the resulting annual profiles were evaluated by per-pixel intra-annually correlating against corresponding vegetation products of both MODIS and Copernicus Global Land Service (CGLS). The most consistent results were obtained for LAI, which showed intra-annual correlations with an average Pearson correlation coefficient (R) of 0.57 against the CGLS LAI product. Globally, the EVT products showed consistent results, specifically obtaining higher correlation than R > 0.5 with reference products between 30 and 60° latitude in the Northern Hemisphere. Additionally, intra-annual goodness-of-fit statistics were also calculated locally against reference products over four distinct vegetated land covers. As a general trend, vegetated land covers with pronounced phenological dynamics led to high correlations between the different products. However, sparsely vegetated fields as well as areas near the equator linked to smaller seasonality led to lower correlations. We conclude that the global gap-free mapping of the four EVTs was overall consistent. Thanks to GEE, the entire OLCI L1B catalogue can be processed efficiently into the EVT products on a global scale and made cloud-free with the WS temporal reconstruction method. Additionally, GEE facilitates the workflow to be operationally applicable and easily accessible to the broader community. [ABSTRACT FROM AUTHOR]

One of the key issues in wetland monitoring is mapping vegetation. Remote sensing and machine learning are used to acquire vegetation maps, which, due to the development of sensors and data processing and analysis methods, have increasingly high accuracy. The objectives of this study were to test: (i) which of the textural information (TI) features have the highest information potential for identifying wetland communities; and (ii) whether the use of TI improves the accuracy of wetland communities mapping using hyperspectral (HS) and Airborne Laser Scanning (ALS) data. The analysis indicated that the mean and entropy features of the Gray Level Co-occurrence Matrix had the highest potential to differentiate between various wetland communities. Adding these features to the dataset resulted in a small increase (0.005) in average F1 accuracy based on HS data and 0.011 for HS and ALS scenarios in wetland communities classification, and adding TI improved the delineation of patch boundaries. A higher increase was noted for forest and scrub vegetation (by 0.019 for the HS scenario and 0.022 for the HS and ALS scenario) and rushes (only for the HS and ALS scenario 0.017). It can be concluded that it is reasonable to use textural information for mapping wetland communities, especially for areas with a high proportion of scrub and forest and rushes vegetation included in the analysis. [ABSTRACT FROM AUTHOR]

Monitoring and estimating drought impact on plant physiological processes over large regions remains a major challenge for remote sensing and land surface modeling, with important implications for understanding plant mortality mechanisms and predicting the climate change impact on terrestrial carbon and water cycles. The Orbiting Carbon Observatory 3 (OCO‐3), with its unique diurnal observing capability, offers a new opportunity to track drought stress on plant physiology. Using radiative transfer and machine learning modeling, we derive a metric of afternoon photosynthetic depression from OCO‐3 solar‐induced chlorophyll fluorescence (SIF) as an indicator of plant physiological drought stress. This unique diurnal signal enables a spatially explicit mapping of plants' physiological response to drought. Using OCO‐3 observations, we detect a widespread increasing drought stress during the 2020 southwest US drought. Although the physiological drought stress is largely related to the vapor pressure deficit (VPD), our results suggest that plants' sensitivity to VPD increases as the drought intensifies and VPD sensitivity develops differently for shrublands and grasslands. Our findings highlight the potential of using diurnal satellite SIF observations to advance the mechanistic understanding of drought impact on terrestrial ecosystems and to improve land surface modeling. [ABSTRACT FROM AUTHOR]

Feature learning of remote sensing Hyperspectral Image (HSI) using deep learning (DL) models and classification of these features using machine learning (ML) models has led to better exploitation of Earth's resources. We present a PPS framework where P, P, and S denote P rincipal Component Analysis (PCA), P KNNet (Polynomial Kernel Kervolutional Neural Network) model, and S upport Vector Classification (SVC), respectively. In this framework, we propose a novel PKNNet, a non-linear discriminative higher dimensional spatial feature extraction model. This model is capable of capturing higher order intricate pattern combinations in the image with the aid of patch-wise non-linear kernel convolution operations. The model is built in a way that achieves nonlinearity, prevents overfitting and computational expense, and lessens the effect of noise. The features retrieved using the PKNNet model are classified using SVC with Non-linear Polynomial Kernel and Linear Kernel. The experimental trials are conducted on three standard reference remote sensing HSI vegetation datasets to assess the effectiveness of the suggested framework. The experimental findings show that, when compared to current techniques, the PPS framework can produce a competitive performance by attaining 100% accuracy for all the three remote sensing vegetation datasets, when evaluated utilizing uneven and limited training samples. [ABSTRACT FROM AUTHOR]

The classification of wetland plants using unmanned aerial vehicle (UAV) and satellite synergies has received increasing attention in recent years. In this study, UAV-derived training and validation data and WorldView-3 satellite imagery are integrated in the classification of five dominant wetland plants in the Old Woman Creek (OWC) estuary, USA. Several classifiers are explored: (1) pixel-based methods: maximum likelihood (ML), support vector machine (SVM), and neural network (NN), and (2) object-based methods: Naïve Bayes (NB), support vector machine (SVM), and k-nearest neighbors (k-NN). The study evaluates the performance of the classifiers for different image feature combinations such as single bands, vegetation indices, principal components (PCs), and texture information. The results showed that all classifiers reached high overall accuracy (>85%). Pixel-based SVM and object-based NB exhibited the best performance with overall accuracies of 93.76% and 93.30%, respectively. Insignificantly lower overall accuracy was achieved with ML (92.29), followed by NN (90.95) and object-oriented SVM (90.61). The k-NN method showed the lowest (but still high) accuracy of 86.74%. All classifiers except for the pixel-based SVM required additional input features. The pixel-based SVM achieved low errors of commission and omission, and unlike the other classifiers, exhibited low variability and low sensitivity to additional image features. Our study shows the efficacy of combining very high spatial resolution UAV-derived information and the super spectral observation capabilities of WorldView-3 in machine learning for mapping wetland vegetation. [ABSTRACT FROM AUTHOR]

Accurate estimation of fuels is essential for wildland fire simulations as well as decision-making related to land management. Numerous research efforts have leveraged remote sensing and machine learning for classifying land cover and mapping forest vegetation species. In most cases that focused on surface fuel mapping, the spatial scale of interest was smaller than a few hundred square kilometers; thus, many small-scale site-specific models had to be created to cover the landscape at the national scale. The present work aims to develop a large-scale surface fuel identification model using a custom deep learning framework that can ingest multimodal data. Specifically, we use deep learning to extract information from multispectral signatures, high-resolution imagery, and biophysical climate and terrain data in a way that facilitates their end-to-end training on labeled data. A multi-layer neural network is used with spectral and biophysical data, and a convolutional neural network backbone is used to extract the visual features from high-resolution imagery. A Monte Carlo dropout mechanism was also devised to create a stochastic ensemble of models that can capture classification uncertainties while boosting the prediction performance. To train the system as a proof-of-concept, fuel pseudo-labels were created by a random geospatial sampling of existing fuel maps across California. Application results on independent test sets showed promising fuel identification performance with an overall accuracy ranging from 55% to 75%, depending on the level of granularity of the included fuel types. As expected, including the rare—and possibly less consequential—fuel types reduced the accuracy. On the other hand, the addition of high-resolution imagery improved classification performance at all levels. [ABSTRACT FROM AUTHOR]

Combinations of multi-sensor remote sensing images and machine learning have attracted much attention in recent years due to the spectral similarity of wetland plant canopy. However, the integration of hyperspectral and quad-polarization synthetic aperture radar (SAR) images for classifying marsh vegetation has still been faced with the challenges of using machine learning algorithms. To resolve this issue, this study proposed an approach to classifying marsh vegetation in the Honghe National Nature Reserve, northeast China, by combining backscattering coefficient and polarimetric decomposition parameters of C-band and L-band quad-polarization SAR data with hyperspectral images. We further developed an ensemble learning model by stacking Random Forest (RF), CatBoost and XGBoost algorithms for marsh vegetation mapping and evaluated its classification performance of marsh vegetation between combinations of hyperspectral and full-polarization SAR data and any of the lone sensor images. Finally, this paper explored the effect of different polarimetric decomposition methods and wavelengths of radar on classifying wetland vegetation. We found that a combination of ZH-1 hyperspectral images, C-band GF-3, and L-band ALOS-2 quad-polarization SAR images achieved the highest overall classification accuracy (93.13%), which was 5.58–9.01% higher than that only using C-band or L-band quad-polarization SAR images. This study confirmed that stacking ensemble learning provided better performance than a single machine learning model using multi-source images in most of the classification schemes, with the overall accuracy ranging from 77.02% to 92.27%. The CatBoost algorithm was capable of identifying forests and deep-water marsh vegetation. We further found that L-band ALOS-2 SAR images achieved higher classification accuracy when compared to C-band GF-3 polarimetric SAR data. ALOS-2 was more sensitive to deep-water marsh vegetation classification, while GF-3 was more sensitive to shallow-water marsh vegetation mapping. Finally, scattering model-based decomposition provided important polarimetric parameters from ALOS-2 SAR images for marsh vegetation classification, while eigenvector/eigenvalue-based and two-component decompositions produced a great contribution when using GF-3 SAR images. [ABSTRACT FROM AUTHOR]

The current regional-scale process-based photosynthesis models use biome-specified values of maximum carboxylation rate at 25 °C ($V_{m25}$) in simulating ecosystem gross primary productivity (GPP). These models ignore the variations in $V_{m25}$ over time and space, resulting in substantial errors in regional estimates of cropland GPP. Thus, to resolve this problem, we used the ensemble Kalman filter (EnKF) to assimilate tower-based GPP from five maize flux sites into a process-based mode to obtain the “apparent” value of $V_{m25}$ and then modeled this parameter using machine learning (ML) algorithms. The results showed that $V_{m25}$ increased during the early growing season and then decreased after reaching a peak value in the middle of the growing season. The coefficient of determination ($R^{2}$) root mean square error (RMSE) for satellite-driven coupled photosynthesis and evapotranspiration simulator (SCOPES)-Crop with EnKF-derived varied $V_{m25}$ in simulating daily GPP across all site-days increased (decreased) by 0.17 (5.63 $\mu \text {mol}\,\text {m}^{-2}\,\text {s}^{-1}$) on average compared to that for the model with fixed $V_{m25}$. We used four ML algorithms, namely artificial neural network, random forest, extreme gradient enhancement, and convolutional neural network (CNN), to model the $V_{m25}$ of maize. The CNN algorithm yielded the best results. The average of the $R^{2}$ (RMSE) values of simulated GPP using CNN-based $V_{m25}$ over the three flux sites is 0.93 (1.95 $\mu \text {mol}\,\text {m}^{-2}\,\text {s}^{-1}$), higher (smaller) than that using fixed $V_{m25}$. This study implies that representing the seasonal variations in $V_{m25}$ can facilitate improved estimates of GPP and the ML methods are useful tools for modeling the variation in $V_{m25}$. [ABSTRACT FROM AUTHOR]

Predicting wildfire spread is critical for land management and disaster preparedness. To this end, we present “Next Day Wildfire Spread,” a curated, large-scale, multivariate dataset of historical wildfires aggregating nearly a decade of remote-sensing data across the United States. In contrast to existing fire datasets based on Earth observation satellites, our dataset combines 2-D fire data with multiple explanatory variables (e.g., topography, vegetation, weather, drought index, and population density) aligned over 2-D regions, providing a feature-rich dataset for machine learning. To demonstrate the usefulness of this dataset, we implement a neural network that takes advantage of the spatial information of these data to predict wildfire spread. We compare the performance of the neural network with other machine learning models: logistic regression and random forest. This dataset can be used as a benchmark for developing wildfire propagation models based on remote-sensing data for a lead time of one day. [ABSTRACT FROM AUTHOR]

Industrial activities induce various impacts on ecosystems that influence species richness and distribution. An effective way to assess the resulting impacts on biodiversity lies in vegetation mapping. Species classification achieved through supervised machine learning algorithms at the pixel level has shown promising results using hyperspectral images and multispectral, multitemporal images. This study aims to determine whether airborne hyperspectral images with a high spatial resolution or phenological information obtained by spaceborne multispectral time series (Sentinel-2) are suitable to discriminate species and assess biodiversity in a complex impacted context. The industrial heritage of the study site has indeed induced high spatial heterogeneity in terms of stressors and species over a reduced scale. First, vegetation indices, derivative spectra, continuum removed spectra, and components provided by three feature extraction techniques, namely, Principal Component Analysis, Minimal Noise Fraction, and Independent Component Analysis, were calculated from reflectance spectra. These features were then analyzed through Sequential Floating Feature Selection. Supervised classification was finally performed using various machine learning algorithms (Random Forest, Support Vector Machines, and Regularized Logistic Regression) considering a probability-based rejection approach. Biodiversity metrics were derived from resulted maps and analyzed considering the impacts. Average Overall Accuracy (AOA) reached up to 94% using the hyperspectral image and Regularized Logistic Regression algorithm, whereas the time series of multispectral images never exceeded 72% AOA. From all tested spectral transformations, only vegetation indices applied to the time series of multispectral images increased the performance. The results obtained with the hyperspectral image degraded to the specifications of Sentinel-2 emphasize the importance of fine spatial and spectral resolutions to achieve accurate mapping in this complex context. While no significant difference was found between impacted and reference sites through biodiversity metrics, vegetation mapping highlighted some differences in species distribution. [ABSTRACT FROM AUTHOR]

• Developed a decomposition method for simulating layered vegetation coverage. • Simulated layered PAI and COVER continuity maps by combining SOFM and RF model. • In Jiangxi Province, more than 80% of forestland is under 20 m in height. • PAI and COVER both show an upward-then-downward trend with height, peaking at 5–10 m. The spatially continuous mapping of vegetation structure parameters in forested areas serves as a crucial foundation for research in various fields, including ecosystem ecology, climate change, hydrology, and forest management and protection. However, there is a notable scarcity of regional scale data regarding vegetation structure parameters. Therefore, our objective was to fill this data gap by providing a methodology for mapping vegetation structure parameters in forested areas at regional-scale. Global Ecosystem Dynamics Investigation (GEDI) provides precise observational data on vegetation structure parameters across the world. Using GEDI vegetation structure parameter point data, we developed a vegetation state coefficient based on the ratio of plant area index (PAI) and vegetation coverage (COVER). Subsequently, we proposed a method for decomposing vegetation coverage to simulate layered vegetation coverage. Taking Jiangxi Province as a case study, we integrated machine learning models to generate a spatially continuous map of vegetation structure parameters in forested areas, including layered coverage and layered leaf area index, in 2020. The map has a spatial resolution of 30 m and a vertical resolution of 5 m. The model exhibits excellent performance. Among the 140 sets of models, approximately 63 % of them achieved an R2 surpassing 0.6, while 89 % of the models achieved a root mean square error (RMSE) below 0.3. This study can serve as a valuable reference for the decomposition of vegetation structure parameters at the regional scale and provide a more precise depiction of the spatial characteristics of vertical vegetation structure at the regional scale. It contributes by providing data support and methodological guidance for research related to forest structure analysis, resource management, and ecological process simulation at the regional scale. [ABSTRACT FROM AUTHOR]

Improving high-resolution (meter-scale) mapping of snow-covered areas in complex and forested terrains is critical to understanding the responses of species and water systems to climate change. Commercial high-resolution imagery from Planet Labs, Inc. (Planet, San Francisco, CA, USA) can be used in environmental science, as it has both high spatial (0.7–3.0 m) and temporal (1–2 day) resolution. Deriving snow-covered areas from Planet imagery using traditional radiometric techniques have limitations due to the lack of a shortwave infrared band that is needed to fully exploit the difference in reflectance to discriminate between snow and clouds. However, recent work demonstrated that snow cover area (SCA) can be successfully mapped using only the PlanetScope 4-band (Red, Green, Blue and NIR) reflectance products and a machine learning (ML) approach based on convolutional neural networks (CNN). To evaluate how additional features improve the existing model performance, we: (1) build on previous work to augment a CNN model with additional input data including vegetation metrics (Normalized Difference Vegetation Index) and DEM-derived metrics (elevation, slope and aspect) to improve SCA mapping in forested and open terrain, (2) evaluate the model performance at two geographically diverse sites (Gunnison, Colorado, USA and Engadin, Switzerland), and (3) evaluate the model performance over different land-cover types. The best augmented model used the Normalized Difference Vegetation Index (NDVI) along with visible (red, green, and blue) and NIR bands, with an F-score of 0.89 (Gunnison) and 0.93 (Engadin) and was found to be 4% and 2% better than when using canopy height- and terrain-derived measures at Gunnison, respectively. The NDVI-based model improves not only upon the original band-only model's ability to detect snow in forests, but also across other various land-cover types (gaps and canopy edges). We examined the model's performance in forested areas using three forest canopy quantification metrics and found that augmented models can better identify snow in canopy edges and open areas but still underpredict snow cover under forest canopies. While the new features improve model performance over band-only options, the models still have challenges identifying the snow under trees in dense forests, with performance varying as a function of the geographic area. The improved high-resolution snow maps in forested environments can support studies involving climate change effects on mountain ecosystems and evaluations of hydrological impacts in snow-dominated river basins. [ABSTRACT FROM AUTHOR]

This paper presents two categories of features extraction and mapping suite, a very high-resolution suite and an ultra-resolution suite at 2 m and 0.5 m resolutions, respectively, for the differentiation and mapping of land cover and community-level vegetation types. The features extraction flow of the ultra-resolution suite involves pan-sharpening of the multispectral image, color-transformation of the pan-sharpened image, and the generation of panchromatic textural features. The performance of the ultra-resolution features extraction suite was compared with the very high-resolution features extraction suite that involves the calculation of radiometric indices and color-transformation of the multi-spectral image. This research was implemented in three mountainous ecosystems located in a cool temperate region. Three machine learning classifiers, Random Forests, XGBoost, and SoftVoting, were employed with a 10-fold cross-validation method for quantitatively evaluating the performance of the two suites. The ultra-resolution suite provided 5.3% more accuracy than the very high-resolution suite using single-date autumn images. Addition of summer images gained 12.8% accuracy for the ultra-resolution suite and 13.2% accuracy for the very high-resolution suite across all sites, while the ultra-resolution suite showed 4.9% more accuracy than the very high-resolution suite. The features extraction and mapping suites presented in this research are expected to meet the growing need for differentiating land cover and community-level vegetation types at a large scale. [ABSTRACT FROM AUTHOR]

Species distribution models (SDMs) have been widely used to create maps of expected species incidence, often using citizen science (CS) occurrence data as inputs. Environmental policy is informed by knowledge of community distributions, but there have been fewer attempts to utilise the potential of community distribution models (CDMs) to predict these. Many countries have vegetation community classification systems which include phytosociological information on individual species. Within Great Britain, the National Vegetation Classification (NVC) is the primary standard for vegetation communities, and while maps have been produced at regional scales, cost‐effective techniques are required for national scales.Published NVC occurrence records of 22 upland NVC communities in England and Wales were used as observed occurrences (presence‐only data). Predictors for the CDMs were enhanced vegetation index (EVI), elevation, slope, aspect, temperature and rainfall. Five modelling methods were investigated: generalised linear models (GLMs), support vector machines (SVMs), random forests (RFs), maximum entropy (MaxEnt) and maximum likelihood (MaxLike). Model quality was assessed via bootstrapping via area under the curve (AUC), true skill statistic (TSS) and Kappa index.There were only small differences in the accuracy of the models (median TSS model accuracy 0.742; range 0.280 to 0.873) with RF models the best overall CDM method. Across all NVC communities, summer and winter maximum temperatures and annual rainfall were the most important predictor variables. NVCs with spatially disjunct distributions in both lowlands and uplands, or that responded to localised management or environmental conditions, were poorly predicted.Synthesis and applications. Vegetation communities can be reliably predicted at large spatial scales using CDMs from extant datasets. Management practitioners can use community‐level predictions to design targeted field surveys for individual species typically associated with specific communities. Most existing CS survey schemes focus on species rather than communities. Hence future development of new CS schemes similar to the National Plant Monitoring Scheme (NPMS), that aligns with the NVC, will enable CS data to generate up‐to‐date maps of both communities and species. [ABSTRACT FROM AUTHOR]

Estimating gross primary production (GPP), the gross uptake of CO2 by vegetation, is a fundamental prerequisite for understanding and quantifying the terrestrial carbon cycle. Over the last decade, multiple approaches have been developed to derive spatiotemporal dynamics of GPP combining in situ observations and remote sensing data using machine learning techniques or semiempirical models. However, no high spatial resolution GPP product exists so far that is derived entirely from satellite-based remote sensing data. Sentinel-2 satellites are expected to open new opportunities to analyze ecosystem processes with spectral bands chosen to study vegetation between 10- and 20-m spatial resolutions with five-day revisit frequency. Of particular relevance is the availability of red-edge bands that are suitable for deriving estimates of canopy chlorophyll content that are expected to be much better than any previous global mission. Here, we analyzed whether red-edge-based and near-infrared-based vegetation indices (VIs) or machine learning techniques that consider VIs, all spectral bands, and their nonlinear interactions could predict daily GPP derived from 58 eddy covariance sites. Using linear regressions based on classic VIs, including near-infrared reflectance of vegetation (NIRv), we achieved prediction powers of $R^{2}_{\mathrm{10-fold}} = 0.51$ and an $RMSE_{\mathrm{10-fold}} = 2.95 $ [ $\mu \rm {mol \ CO_{2} m^{-2}s^{-1}}$ ] in a 10-fold cross validation. Chlorophyll index red (CIR) and the novel kernel NDVI (kNVDI) achieved significantly higher prediction powers of around $R^{2}_{\mathrm{10-fold}} \approx 0.61$ and $RMSE_{\mathrm{10-fold}} \approx 2.57$ [ $\mu \rm {mol \ CO_{2} m^{-2}s^{-1}}$ ]. Using all spectral bands and VIs jointly in a machine learning prediction framework allowed us to predict GPP with $R^{2}_{\mathrm{10-fold}} = 0.71$ and $RMSE_{\mathrm{10-fold}} = 2.68$ [ $\mu \rm {mol \ CO_{2} m^{-2}s^{-1}}$ ]. Despite the high-power prediction when machine learning techniques are used, under water-stress scenarios or heat waves, optical information alone is not enough to predict GPP properly. In general, our analyses show the potential of nonlinear combinations of spectral bands and VIs for monitoring GPP across ecosystems at a level of accuracy comparable to previous works, which, however, required additional meteorological drivers. [ABSTRACT FROM AUTHOR]

Timely assessment of crop maturity contributes to optimized harvesting schedules while limiting food loss/waste at the farm level. Maturity assessments are typically performed via costly and time-consuming in situ methods. This study aimed to evaluate pod size crop maturity using imaging spectroscopy via unmanned aerial systems (UASs), as well as identifying discriminating wavelengths, using snap bean as a proxy crop. The research utilized a UAS-mounted hyperspectral imager in the visible-to-near-infrared region. Two years’ worth of data were collected at two different geographical locations for six different snap bean cultivars. Our approach consisted of calibration to reflectance, vegetation detection, noise reduction, creating classification bins, and feature selection. We used our previously published feature selection library, Jostar, and utilized ant colony optimization and simulated annealing to detect five spectral features and Plus-L Minus-R to identify one to ten features. We utilized decision trees and random forest classifiers for the classification task. Our findings show that, given the proper wavelengths, accurate pod maturity assessment is feasible for large-sieve cultivars (F1 score = 0.79–0.91), separating sieve sizes between ready-to-harvest and not ready-to-harvest pods. These spectral features were in the ~450, ~530, ~660, 700–720, ~740, and ~760 nm regions. This bodes well for the potential extension of results to an operational, multispectral sensor, tuned with the identified bands, thereby negating the need for a costly hyperspectral system. However, this proposition mandates further investigation, including data acquisition from geographical locations with variable climates, and quantifying noise for the hyperspectral imager to compare results with noisier datasets. [ABSTRACT FROM AUTHOR]

The solutions of vegetation height and the mapping of terrain in forest have always been concerning and are still challenging problems. The purpose of this article is to provide a new technical approach for the digital terrain model (DTM) and forest topographic survey by aerial photogrammetry, in consideration of the forest vegetation height modeling problem. Based on an aerial digital orthophoto map and a digital surface model (DSM), the spectral features and geometric features that are related to forest vegetation height are analyzed and extracted. The nonlinear correlation maximal information coefficient, maximum asymmetry score, and Pearson linear correlation coefficient between feature factors and vegetation height are listed, and the correlations are evaluated as the basis for factors selection. Two kinds of support vector regression algorithms were adopted to establish the machine learning for forest vegetation height model (VHM). Therefore, the DSM can be corrected to DTM. The experimental results show that the accuracy of the forest VHM is better than 1 m. Thus, the proposed method is proved to be feasible and practical. It provides a low-cost and high-efficiency method for the VHM and DTM in forest areas by photogrammetry. [ABSTRACT FROM AUTHOR]

In modern industries, early fault detection is crucial for maintaining process safety and product quality. Process data contains information on the entire plant acting as a map for visualization of relationships between various plant units, making data-driven process monitoring a key technology for efficiency enhancement. This article focuses on review of process monitoring techniques reported for metal etching process, which is a batch operation carried out in semiconductor manufacturing industry. Various machine learning (and deep learning) techniques applied to date for fault detection and diagnosis of metal etching process are surveyed. Detailed survey of research work on different techniques and the reported results are presented in graphical (pie chart and bar chart) and tabular format. The review article further presents the pros and cons, gaps and future directions in the techniques applied in metal etching process. [ABSTRACT FROM AUTHOR]

Questions: Understanding spatial variation in floristic composition is crucial to quantify the extent, patchiness and connectivity of distinct habitats and their spatial relationships. Broad‐scale variation in floristic composition and the degree of uniqueness of different regions remains poorly mapped and understood in several areas across the globe. We here aim to map vegetation heterogeneity in Amazonia. Location: Middle Juruá river region, Amazonas State, Brazil. Methods: We mapped four plant groups by applying machine learning to scale up locally observed community composition and using environmental and remotely sensed variables as predictors, which were obtained as GIS layers. To quantify how reliable our predictions were, we made an assessment of model transferability and spatial applicability. We also compared our floristic composition map to the official Brazilian national‐level vegetation classification. Results: The overall performance of our floristic models was high for all four plant groups, especially ferns, and the predictions were found to be spatially congruent and highly transferable in space. For some areas, the models were assessed not to be applicable, as the field sampling did not cover the spectral or environmental characteristics of those regions. Our maps show extensive habitat heterogeneity across the region. When compared to the Brazilian vegetation classification, floristic composition was relatively homogeneous within dense forests, while floristic heterogeneity in rainforests classified as open was high. Conclusion: Our maps provide geoecological characterization of the regions and can be used to test biogeographical hypotheses, develop species distribution models and, ultimately, aid science‐based conservation and land‐use planning. [ABSTRACT FROM AUTHOR]

Spectral measurements are commonly applied for the nondestructive estimation of leaf parameters, such as the concentrations of chlorophyll a and b, carotenoid, anthocyanin, brown pigment, leaf water content, and leaf mass per area for the quantification of vegetation physiology. The most popular way to estimate these parameters is by using spectral vegetation indices. The use of biochemical models allows us to use the full wavelength range (400–2500 nm) and to physically interpret the result. However, their performance is usually lower than that of supervised machine learning regression techniques. Machine learning regression techniques, on the other hand, have the disadvantage that the relationship between estimated parameters and the reflectance/transmission spectra is unclear. In this article, a hybrid between a supervised learning method and physical modeling for the estimation of leaf parameters is proposed. In this method, a machine learning regression technique is applied to learn a mapping from the true hyperspectral data set to a data set that follows the PROSPECT model. The PROSPECT model then reveals the actual leaf parameters. Two mapping methods, based on Gaussian processes (GPs) and kernel ridge regression (KRR) are proposed. As an alternative, mapping onto the leaf absorption spectra is proposed as well. The proposed methodology not only estimates the leaf parameters with a lower error but also solves the interpretation problem of the parameters estimated by the advanced machine learning regression techniques. This method is validated on the ANGERS and LOPEX data set. [ABSTRACT FROM AUTHOR]

Ecological resources are an important material foundation for the survival, development, and self-realization of human beings. In-depth and comprehensive research and understanding of ecological resources are beneficial for the sustainable development of human society. Advances in observation technology have improved the ability to acquire long-term, cross-scale, massive, heterogeneous, and multi-source data. Ecological resource research is entering a new era driven by big data. Traditional statistical learning and machine learning algorithms have problems with saturation in dealing with big data. Deep learning is a method for automatically extracting complex high-dimensional nonlinear features, which is increasingly used for scientific and industrial data processing because of its ability to avoid saturation with big data. To promote the application of deep learning in the field of ecological resource research, here, we first introduce the relationship between deep learning theory and research on ecological resources, common tools, and datasets. Second, applications of deep learning in classification and recognition, detection and localization, semantic segmentation, instance segmentation, and graph neural network in typical spatial discrete data are presented through three cases: species classification, crop breeding, and vegetation mapping. Finally, challenges and opportunities for the application of deep learning in ecological resource research in the era of big data are summarized by considering the characteristics of ecological resource data and the development status of deep learning. It is anticipated that the cooperation and training of cross-disciplinary talents may promote the standardization and sharing of ecological resource data, improve the universality and interpretability of algorithms, and enrich applications with the development of hardware. [ABSTRACT FROM AUTHOR]

The severe accident at the Chornobyl Nuclear Power Plant (ChNPP) in 1986 resulted in extraordinary contamination of the surrounding territory, which necessitated the creation of the Chornobyl Exclusion Zone (ChEZ). During the accident, liquidation materials contaminated by radioactive fallout (e.g., contaminated soil and trees) were buried in so-called Radioactive Waste Temporary Storage Places (RWTSPs). The exact locations of these burials were not always sufficiently documented. However, for safety management, including eventual remediation works, it is crucial to know their locations and rely on precise hazard maps. Over the past 34 years, most of these so-called trenches and clamps have been exposed to natural processes. In addition to settlement and erosion, they have been overgrown with dense vegetation. To date, more than 700 burials have been thoroughly investigated, but a large number of burial sites (approximately 300) are still unknown. In the past, numerous burials were identified based on settlement or elevation in the decimeter range, and vegetation anomalies that tend to appear in the immediate vicinity. Nevertheless, conventional detection methods are time-, effort- and radiation dose-intensive. Airborne gamma spectrometry and visual ground inspection of morphology and vegetation can provide useful complementary information, but it is insufficient for precisely localizing unknown burial sites in many cases. Therefore, sensor technologies, such as UAV-based lidar and multispectral imagery, have been identified as potential alternative solutions. This paper presents a novel method to detect radioactive waste sites based on a set of prominent features generated from high-resolution remote sensing data in combination with a random forest (RF) classifier. Initially, we generate a digital terrain model (DTM) and 3D vegetation map from the data and derive tree-based features, including tree density, tree height, and tree species. Feature subsets compiled from normalized DTM height, fast point feature histograms (FPFH), and lidar metrics are then incorporated. Next, an RF classifier is trained on reference areas defined by visual interpretation of the DTM grid. A backward feature selection strategy reduces the feature space significantly and avoids overfitting. Feature relevance assessment clearly demonstrates that the members of all feature subsets represent a final list of the most prominent features. For three representative study areas, the mean overall accuracy (OA) is 98.2% when using area-wide test data. Cohens' kappa coefficient κ ranges from 0.609 to 0.758. Additionally, we demonstrate the transferability of a trained classifier to an adjacent study area (OA = 93.6%, κ = 0.452). As expected, when utilizing the classifier on geometrically incorrect and incomplete reference data, which were generated from old maps and orthophotos based on visual inspection, the OA decreases significantly to 65.1% (κ = 0.481). Finally, detection is verified through 38 borings that successfully confirm the existence of previously unknown buried nuclear materials in classified areas. These results demonstrate that the proposed methodology is applicable to detecting area-wide unknown radioactive biomass burials in the ChEZ. [ABSTRACT FROM AUTHOR]

Vegetation of Cameroon includes a variety of landscape types with high biodiversity. Ecologicalmonitoring of Yaoundé requires visualization of vegetation types in context of climate change. Vegetation Indices (VIs) derived from Sentinel-2 multispectral satellite image were analyzed in SAGA GIS to separate wetland biomes, as well as savannah and tropical rainforests. The methodology includes computing 6 VIs: NDVI, DVI, SAVI, RVI, TTVI, CTVI. The VIs shown correlation of data with vegetation distribution rising from wetlands, grassland, savanna, and shrub land towards tropical rainforests, increasing values alongwith canopy greenness, while also being inversely proportional to soils, urban spaces and Sanaga River. The study contributed to the environmental studies of Cameroon and demonstration of the satellite image processing. [ABSTRACT FROM AUTHOR]

Oil palm has become well known for its oil palm yields that can be used to produce food, biodiesel and biogas. The rapid expansion of oil palm plantations over large areas has changed the land use and land cover of surroundings. Changes in land covers can be mapped and later used for further analysis. However, obtaining and classifying large coverages require massive amounts of data and computing resources and the skills and time of analysts. The Remote Ecosystem Monitoring Assessment Pipeline (REMAP) provides a cloud computing platform that hosts an open-source stacked Landsat data that allows land cover classification to be implemented using a built-in random forest supervised machine learning algorithm. Classifications were performed with the aid of predictor layers to discriminate the following land covers in Peninsular Malaysia: oil palm, built-up, bare soil, water, forest, other vegetation and paddy. The classification performed on period 1 (1999–2003) and period 2 (2014–2017) data produced an overall accuracy of 80.34% and 79.53% respectively. The analysis of the changes in oil palm distributions from period 1 to period 2 indicated an increment of 23.59%. Further analysis revealed that oil palm expansion in Peninsular Malaysia only minimally affected forested area and is mostly resulted from the conversion of less productive crops to oil palm. Results prove the land cover mapping and change detection capabilities of REMAP as a cloud computing platform for large areas. Despite its limitations, REMAP has the potential to achieve fast-paced mapping over large areas and monitor land changes in oil palm distributions. [ABSTRACT FROM AUTHOR]

Great efforts have been devoted to improving the performance of scene classification. However, it is still a challenging task because of the complex background and diverse objects in scene images. To address this issue, multiple resolution block feature (MRBF) is proposed for remote-sensing scene classification. It is a unified and effective scene representation, consisting of completed double cross pattern (CDCP) combined with fisher vectors (FV). Specifically, in order to capture more robust and richer scene information, multiple resolution block descriptor is devised based on CDCP. After that, it is combined with FV to construct unified MRBF, which can fully exploit discriminative information from the block descriptor. Finally, the scene classification is achieved by kernel extreme learning machine. Extensive evaluations on four benchmark scene data-sets demonstrate the effectiveness and superiority of the proposed MRBF method for scene classification. [ABSTRACT FROM AUTHOR]

An unprecedented spectroscopic data stream will soon become available with forthcoming Earth-observing satellite missions equipped with imaging spectroradiometers. This data stream will open up a vast array of opportunities to quantify a diversity of biochemical and structural vegetation properties. The processing requirements for such large data streams require reliable retrieval techniques enabling the spatiotemporally explicit quantification of biophysical variables. With the aim of preparing for this new era of Earth observation, this review summarizes the state-of-the-art retrieval methods that have been applied in experimental imaging spectroscopy studies inferring all kinds of vegetation biophysical variables. Identified retrieval methods are categorized into: (1) parametric regression, including vegetation indices, shape indices and spectral transformations; (2) nonparametric regression, including linear and nonlinear machine learning regression algorithms; (3) physically based, including inversion of radiative transfer models (RTMs) using numerical optimization and look-up table approaches; and (4) hybrid regression methods, which combine RTM simulations with machine learning regression methods. For each of these categories, an overview of widely applied methods with application to mapping vegetation properties is given. In view of processing imaging spectroscopy data, a critical aspect involves the challenge of dealing with spectral multicollinearity. The ability to provide robust estimates, retrieval uncertainties and acceptable retrieval processing speed are other important aspects in view of operational processing. Recommendations towards new-generation spectroscopy-based processing chains for operational production of biophysical variables are given. [ABSTRACT FROM AUTHOR]

Abstract Crop type mapping at the field level is necessary for a variety of applications in agricultural monitoring and food security. As remote sensing imagery continues to increase in spatial and temporal resolution, it is becoming an increasingly powerful raw input from which to create crop type maps. Still, automated crop type mapping remains constrained by a lack of field-level crop labels for training supervised classification models. In this study, we explore the use of random forests transferred across geographic distance and time and unsupervised methods in conjunction with aggregate crop statistics for crop type mapping in the US Midwest, where we simulated the label-poor setting by depriving the models of labels in various states and years. We validated our methodology using available 30 m spatial resolution crop type labels from the US Department of Agriculture's Cropland Data Layer (CDL). Using Google Earth Engine, we computed Fourier transforms (or harmonic regressions) on the time series of Landsat Surface Reflectance and derived vegetation indices, and extracted the coefficients as features for machine learning models. We found that random forests trained on regions and years similar in growing degree days (GDD) transfer to the target region with accuracies consistently exceeding 80%. Accuracies decrease as differences in GDD expand. Unsupervised Gaussian mixture models (GMM) with class labels derived using county-level crop statistics classify crops less consistently but require no field-level labels for training. GMM achieves over 85% accuracy in states with low crop diversity (Illinois, Iowa, Indiana, Nebraska), but performs sometimes no better than random when high crop diversity interferes with clustering (North Dakota, South Dakota, Wisconsin, Michigan). Under the appropriate conditions, these methods offer options for field-resolution crop type mapping in regions around the world with few or no ground labels. Highlights • Crop type maps of >80% accuracy created without in-season field data in some regions. • Fourier transform coefficients of Landsat time series distinguish crop types. • Random forest classifiers transfer well across regions of similar growing degree days. • Gaussian mixture models cluster crop types accurately in low crop diversity regions. • Aggregate crop statistics allow automatic assignment of crop types to clusters. [ABSTRACT FROM AUTHOR]

Abstract The Landsat archive presents a unique source for mapping and monitoring of shrublands. Still, efficient and accurate mapping approaches are needed that provide shrub cover fraction estimates over space and time. The spectral signal of shrubs is composed of green vegetation and non-photosynthetic vegetation as well as varying fractions of soil, grass, and shadow, which makes a direct mapping of fraction cover challenging. In this study, we mapped 32 years of shrub cover fractions from annual Landsat best observation composites covering a rockrose (genus Cistus) ecosystem in the Baixo Alentejo region of southern Portugal. Fraction mapping was based on a multi-year support vector regression model trained with synthetically mixed data from a multi-annual image spectral library. Resulting fractions maps were validated using reference information derived from high-resolution satellite imagery available for 10 out of 32 years. Fraction maps reproduced the spatial-temporal patterns of shrub cover in the study region very well, with an average mean absolute error over all validation years of 13.7%. For individual validation years, mean absolute errors ranged from 7.7% to 17.1%. Our consistent modeling framework led to a reliable annual shrub cover fraction time-series, which allowed identifying areas of stable shrub cover and areas with different types and intensities of change. Such long-term shrub cover fraction monitoring is of great value for land use assessments and contributes to a thorough understanding and management of Earth's arid and semi-arid ecosystems. Highlights • A 32-year time-series of shrub fractions was retrieved for Baixo Alentejo. • A multi-year SVR model applied to annual Landsat composites produced accurate maps. • Fraction time-series supported the assessment of shrub cover change processes. [ABSTRACT FROM AUTHOR]

Forthcoming spaceborne imaging spectrometers will provide novel opportunities for mapping urban composition globally. To move from case studies for single cities towards comparative and more operational analyses, generalized models that may be transferred throughout space are desired. In this study, we investigated how single regression models can be spatially generalized for vegetation-impervious-soil (VIS) mapping across multiple cities. The combination of support vector regression (SVR) with synthetically mixed training data generated from spectral libraries was used for fraction mapping. We developed three local models based on separate spectral libraries from Berlin (Germany), Brussels (Belgium), and Santa Barbara (U.S.), and a generalized model based on a combined multi-site spectral library. To examine the performance and transferability of the generalized model compared to local models, we first applied all model variants to simulated Environmental Mapping and Analysis Program (EnMAP) data from the three cities that were represented in the models, i.e., known sites. Next, we transferred the models to two unknown sites not represented in the models, San Francisco Bay Area (U.S.) and Munich (Germany). In the first mapping constellation, results demonstrated that the generalized model was capable of accurately mapping VIS fractions across all three known sites. Average mean absolute errors (AV-MAEs) were 8.5, 12.2, and 11.0% for Berlin, Brussels, and Santa Barbara. The performance of the generalized model was very similar to the local models, with ∆AV-MAEs falling within a range of ±0.7%. A detailed assessment of fraction maps and class-wise accuracies confirmed that modeling errors related to remaining limitations of urban mapping based on optical remote sensing data rather than to the choice between a local or generalized model. For the second mapping constellation, the generalized model proved to be useful for mapping vegetation and impervious fractions in the unknown sites. MAEs for both cover types were 5.4 and 10.9% for the San Francisco Bay Area, and 6.3 and 15.4% for Munich. In contrast, the three local models were only found to have similar accuracies as the generalized model for one of the two sites or for individual VIS categories. Despite the enhanced transferability of the generalized model to the unknown sites, deficiencies remained for accurate soil mapping. MAEs were 22.4 and 12.3%, and high over - and underestimations were observed at the low and high end of the fraction range. These shortcomings indicated possible limitations of the spectral libraries to account for the spectral characteristics of soils in the unknown sites. Overall, we conclude that the combination of SVR and synthetically mixed training data generated from multi-site libraries constitutes a flexible modeling approach for generalized urban mapping across multiple cities. [ABSTRACT FROM AUTHOR]

Reliable fire severity mapping is a vital resource for fire scientists and land management agencies globally. Satellite derived pre- and post-fire differenced severity indices (∆FSI), such as the differenced Normalised Burn Ratio (∆NBR), are widely used to map the severity of large wildfires. Fire severity classification is commonly undertaken through the identification of severity class thresholds in ∆FSI. However, several shortcomings have been identified with severity classifications using ∆FSI, including poor delineation of low fire severity classes, and unsatisfactory performance when ∆FSI classification thresholds are applied to new landscapes. Our study assesses the performance of the Random Forest classifier (RF) for improving the accuracy of satellite based wildfire severity mapping across heterogeneous landscapes using Landsat imagery. We collected point based fire severity training data (n = 10,855) from sixteen large wildfires occurring across south-eastern Australia between 2006 and 2016. The predictive accuracy of fire severity classification using ∆NBR and the RF incorporating numerous spectral indices, was assessed using bootstrapping and cross validation. Image acquisition and index calculation for each fire was undertaken in Google Earth Engine (GEE). Results of the bootstrapping validation show that the RF classifier had very high classification accuracy (>95%) for unburnt (UB), crown scorch (CS) and crown consumption (CC) severity classes, and high classification accuracy (>74%) for low severity classes (crown unburnt, CU; partial crown scorch, PCS). The RF classification outperformed the ∆NBR classification for all severity classes, increasing classification accuracy by between 6%–21%. Cross validation using independent fires produced similar median classification accuracy as the bootstrapping validation, though the RF classification of CU was substantially reduced to ~55%. ∆NBR, ∆NDWI and ∆NDVI were the three most important variables in the RF model. The Landsat satellite platform used to derive the indices had little effect on classification accuracy. Maps produced using the RF classifier in GEE had similar spatial patterns in fire severity classes as maps produced using time-consuming hand digitisation of aerial images. GEE was found to be a highly efficient platform for image acquisition, processing and production of severity maps. Our study shows that fire severity mapping using RF classifiers provides a robust method for broad scale mapping of fire severity across heterogeneous landscapes. Furthermore, the GEE-based classification framework supports the operational application of this approach in a land management agency context for the rapid production of fire severity maps. [ABSTRACT FROM AUTHOR]

Monitoring biodiversity and how anthropogenic pressures impact this is critical, especially as anthropogenically driven climate change continues to affect all ecosystems. Intertidal areas are exposed to particularly high levels of pressures owing to increased population density in coastal areas. Traditional methods of monitoring intertidal areas do not provide datasets with full coverage in a cost-effective or timely manner, and so the use of remote sensing to monitor these areas is becoming more common. Monitoring of ecologically important monospecific habitats, such as seagrass beds, using remote sensing techniques is well documented. However, the ability for multispectral data to distinguish efficiently and accurately between classes of vegetation with similar pigment composition, such as seagrass and green algae, has proved difficult, often requiring hyperspectral data. A machine learning approach was used to differentiate between soft-bottom intertidal vegetation classes when exposed at low tide, comparing 6 different multi- and hyperspectral remote and in situ sensors. For the library of 366 spectra, collected across Northern Europe, high accuracy (>80%) was found across all spectral resolutions. While a higher spectral resolution resulted in higher accuracy, there was no discernible increase in accuracy above 10 spectral bands (95%: Sentinel-2 MSI sensor with a spatial resolution of 20 m). This work highlights the ability of multispectral sensors to discriminate intertidal vegetation types, while also showing the most important wavelengths for this discrimination (∼530 and ∼ 730 nm), giving recommendations for spectral ranges of future satellite missions. The ability for multispectral sensors to aid in accurate and rapid intertidal vegetation classification at the taxonomic resolution of classes, could be a significant contribution for future sustainable and effective ecosystem management. • Multispectral data can distinguish green macrophytes with high accuracy. • This accuracy is achievable at satellite spectral resolutions currently available. • This opens new avenues for biodiversity studies of intertidal ecosystems. • Minimal gain in accuracy is achieved by increased spectral resolution. [ABSTRACT FROM AUTHOR]

An unsupervised machine-learning workflow is proposed for estimating fractional landscape soils and vegetation components from remotely sensed hyperspectral imagery. The workflow is applied to EO-1 Hyperion satellite imagery collected near Ibirací, Minas Gerais, Brazil. The proposed workflow includes subset feature selection, learning, and estimation algorithms. Network training with landscape feature class realizations provide a hypersurface from which to estimate mixtures of soil (e.g. 0.5 exceedance for pixels: 75% clay-rich Nitisols, 15% iron-rich Latosols, and 1% quartz-rich Arenosols) and vegetation (e.g. 0.5 exceedance for pixels: 4% Aspen-like trees, 7% Blackberry-like trees, 0% live grass, and 2% dead grass). The process correctly maps forests and iron-rich Latosols as being coincident with existing drainages, and correctly classifies the clay-rich Nitisols and grasses on the intervening hills. These classifications are independently corroborated visually (Google Earth) and quantitatively (random soil samples and crossplots of field spectra). Some mapping challenges are the underestimation of forest fractions and overestimation of soil fractions where steep valley shadows exist, and the under representation of classified grass in some dry areas of the Hyperion image. These preliminary results provide impetus for future hyperspectral studies involving airborne and satellite sensors with higher signal-to-noise and smaller footprints. [ABSTRACT FROM AUTHOR]

Abstract: Vegetation maps are models of the real vegetation patterns and are considered important tools in conservation and management planning. Maps created through traditional methods can be expensive and time‐consuming, thus, new more efficient approaches are needed. The prediction of vegetation patterns using machine learning shows promise, but many factors may impact on its performance. One important factor is the nature of the vegetation–environment relationship assessed and ecological redundancy. We used two datasets with known ecological redundancy levels (strength of the vegetation–environment relationship) to evaluate the performance of four machine learning (ML) classifiers (classification trees, random forests, support vector machines, and nearest neighbor). These models used climatic and soil variables as environmental predictors with pretreatment of the datasets (principal component analysis and feature selection) and involved three spatial scales. We show that the ML classifiers produced more reliable results in regions where the vegetation–environment relationship is stronger as opposed to regions characterized by redundant vegetation patterns. The pretreatment of datasets and reduction in prediction scale had a substantial influence on the predictive performance of the classifiers. The use of ML classifiers to create potential vegetation maps shows promise as a more efficient way of vegetation modeling. The difference in performance between areas with poorly versus well‐structured vegetation–environment relationships shows that some level of understanding of the ecology of the target region is required prior to their application. Even in areas with poorly structured vegetation–environment relationships, it is possible to improve classifier performance by either pretreating the dataset or reducing the spatial scale of the predictions. [ABSTRACT FROM AUTHOR]

This work presents a methodology based on Remotely Piloted Aircraft Systems (RPAS) RGB imagery for the analysis and characterization of the structure and condition of complex habitat mosaics of high conservation value in mountain wetlands. Structure from Motion (SfM) image reconstruction techniques were applied on a collection of RGB photographs to derive ultra-high resolution (2,5 cm) digital surface models and ortho-mosaics. Geographical object-based image analysis (GEOBIA) was used for the automatic discrimination of vegetation types for habitat condition assessment via multi-scale object-oriented classifications integrating machine-learning classification techniques with other decision rules. In particular, four vegetation classes were assessed, namely woody vegetation (scrub plants higher than 10 cm), bog herbaceous vegetation, non-bog herbaceous vegetation (as part of wet heathland vegetation mosaics) and areas with scarce or no vegetation (rocky habitats and areas of bare ground). The outputs of the classification were validated against field data of detailed vegetation coverage survey. Results allowed us the automatic and accurate discrimination of habitat types with different management demands, e.g. wet heaths against bogs, as well as the diagnosis of structural characteristics critical for their conservation, such as the ratio of cover herbaceous/woody species or the presence of erosion features. [ABSTRACT FROM AUTHOR]

Vegetation and land-cover information is critical for sustainable environmental management in urban areas. Remote sensing has increasingly been used to derive such information, yet it has been challenged by the spectral and spatial complexity in the urban environment. In this study, we developed a multiple classifier system (MCS) to help improve remote-sensing-based vegetation and land-cover mapping in a large metropolitan area. MCSs, although considered as an emerging hot topic and a promising trend in pattern recognition, have not received the attention it deserves in the remote-sensing community. Our work consisted of several components. First, we identified a group of commonly used pattern recognizers from different families of statistical learning algorithms as base classifiers. Then, we implemented them to derive land-cover information from a satellite image covering the study site. Last, we adopted a weighting and combination method to generate the final map. Results indicate that there is statistically significant difference in the classification accuracy between the MCS developed and each base classifier considered. Comparing with the base classifiers, the MCS produced not only about 5-8% higher overall classification accuracy but also the most stable categorical accuracies. Moreover, the MCS generated a larger accuracy improvement for spectrally complex classes than for relatively homogenous ones, suggesting its comparative advantage in reducing classification errors caused by class ambiguity. The novelties of our work are with the demonstration of how MCSs can be operationally used to improve image classification from large remote sensor data sets with complex patterns and with the insight into the behaviour of MCSs in relation to the complexity of individual classes. These findings can help promote the use of MCSs as an emerging premier approach for image classification by the remote-sensing community. [ABSTRACT FROM AUTHOR]

This paper presents spiking neural networks (SNNs) for remote sensing spatiotemporal analysis of image time series, which make use of the highly parallel and low-power-consuming neuromorphic hardware platforms possible. This paper illustrates this concept with the introduction of the first SNN computational model for crop yield estimation from normalized difference vegetation index image time series. It presents the development and testing of a methodological framework which utilizes the spatial accumulation of time series of Moderate Resolution Imaging Spectroradiometer 250-m resolution data and historical crop yield data to train an SNN to make timely prediction of crop yield. The research work also includes an analysis on the optimum number of features needed to optimize the results from our experimental data set. The proposed approach was applied to estimate the winter wheat (Triticum aestivum L.) yield in Shandong province, one of the main winter-wheat-growing regions of China. Our method was able to predict the yield around six weeks before harvest with a very high accuracy. Our methodology provided an average accuracy of 95.64%, with an average error of prediction of 0.236 t/ha and correlation coefficient of 0.801 based on a nine-feature model. [ABSTRACT FROM PUBLISHER]

Smallholder farms dominate in many parts of the world, particularly Sub-Saharan Africa. These systems are characterized by small, heterogeneous, and often indistinct field patterns, requiring a specialized methodology to map agricultural land cover. Using a variety of sites in South Africa, we present a new approach to mapping agricultural fields, based on efficient extraction of a vast set of simple, highly correlated, and interdependent features, followed by a random forest classifier. We achieved similar high performance across agricultural types, including the spectrally indistinct smallholder fields as well as the more easily distinguishable commercial fields, and demonstrated the ability to generalize performance across large geographic areas. In sensitivity analyses, we determined multi-temporal information provided greater gains in performance than the addition of multi-spectral bands available in DigitalGlobe Worldview-2 imagery. [ABSTRACT FROM AUTHOR]

The new generation of Earth observation sensors with high spatial resolution can provide detailed information for change detection. The widely used methods for high-resolution image change detection rely on textural/structural features. However, these spatial features always produce high-dimensional data space since they are related to a series of parameters, e.g., window sizes and directions. Machine learning methods are also commonly employed, but their performances are subject to the quantity and quality of the training samples, and hence, much effort should be made to collect the high-quality samples. To address these problems, in this study, a novel multiindex automatic change detection method is proposed for the high-resolution imagery. The notable advantages of the proposed model include the following: 1) Complicated urban scenes are represented by a set of low dimensional but semantic information indexes, replacing the high-dimensional but low-level features (e.g., textural and structural features), and 2) the change detection model is carried out automatically without using training samples since the information indexes can directly indicate the primitive urban classes. The multiindex representation refers to the enhanced vegetation index, the water index, and the recently developed morphological building index. Experiments were conducted on the multitemporal WorldView-2 images over Shenzhen City (south of China) and Kuala Lumpur (the capital of Malaysia), where promising results were achieved by the proposed method. Moreover, the traditional methods based on the state-of-the-art textural/morphological features were also implemented for the purpose of comparison, which further validates the advantages of our proposed model. [ABSTRACT FROM AUTHOR]

High spatial resolution mapping of natural resources is much needed for monitoring and management of species, habitats and landscapes. Generally, detailed surveillance has been conducted as fieldwork, numerical analysis of satellite images or manual interpretation of aerial images, but methods of object-based image analysis (OBIA) and machine learning have recently produced promising examples of automated classifications of aerial imagery. The spatial application potential of such models is however still questionable since the transferability has rarely been evaluated. We investigated the potential of mosaic aerial orthophoto red, green and blue (RGB)/near infrared (NIR) imagery and digital elevation model (DEM) data for mapping very fine-scale vegetation structure in semi-natural terrestrial coastal areas in Denmark. The Random Forest (RF) algorithm, with a wide range of object-derived image and DEM variables, was applied for classification of vegetation structure types using two hierarchical levels of complexity. Models were constructed and validated by cross-validation using three scenarios: (1) training and validation data without spatial separation, (2) training and validation data spatially separated within sites, and (3) training and validation data spatially separated between different sites. Without spatial separation of training and validation data, high classification accuracies of coastal structures of 92.1% and 91.8% were achieved on coarse and fine thematic levels, respectively. When models were applied to spatially separated observations within sites classification accuracies dropped to 85.8% accuracy at the coarse thematic level, and 81.9% at the fine thematic level. When the models were applied to observations from other sites than those trained upon the ability to discriminate vegetation structures was low, with 69.0% and 54.2% accuracy at the coarse and fine thematic levels, respectively. Evaluating classification models with different degrees of spatial correlation between training and validation data was shown to give highly different prediction accuracies, thereby highlighting model transferability and application potential. Aerial image and DEM-based RF models had low transferability to new areas due to lack of representation of aerial image, landscape and vegetation variation in training data. They do, however, show promise at local scale for supporting conservation and management with vegetation mappings of high spatial and thematic detail based on low-cost image data. [ABSTRACT FROM AUTHOR]

• Vegetation communities in the largest karst wetland of China achieved fine classifications. • Fusion of JL101K and UAV multispectral images improved classification accuracy (1.9%–4.0%) • Classification performance of LightGBM algorithm outperformed XGBoost (0.6%–2.5%) and RF algorithms (1.6%–3.5%). • UAV multispectral images produced the highest overall accuracy (87.8%) in three data sources. • Red-edge band and DSM provide great contributions to identify vegetation communities. Fine classification of wetland vegetation communities using machine learning algorithm and high spatial resolution images have attracted increased attention. However, there exist several challenges in image fusion, data dimension reduction and algorithm tuning. To resolve these issues, this paper attempts to fuse Unmanned Aerial Vehicle (UAV) images with spaceborne Jilin-1 (JL101K) multispectral images for classifying vegetation communities of karst wetland using the optimized Random Forest (RF), Extreme gradient boosting (XGBoost) and Light Gradient Boosting (LightGBM) algorithms. This study also quantitatively evaluates image fusion quality from spatial detail and spectral fidelity, and explores the effects of different image feature combinations and classifiers on mapping vegetation communities by variable selection and dimensionality reduction. Finally, this paper further evaluates and quantifies the importance and contribution rate of feature variables for typical vegetation communities using Recursive feature elimination (RFE) algorithm. The results showed that: (1) the Gram-Schmidt (GS)algorithm produced the high-quality fusion image of JL101K and UAV, and the fusion image achieved higher overall accuracy (82.8%) than the original JL101K multispectral image; (2) UAV multispectral image and its derivatives (scheme 3) achieved the highest overall accuracy (87.8%) in all classification schemes; (3) The optimized object-based LightGBM algorithm outperformed XGBoost and RF algorithm, which provided an improvement of 0.6%∼3.5% in overall accuracy (OA). McNemar's test indicated that there existed significant differences in vegetation communities' classification between the three algorithms. (4) The average accuracy (AA) of vegetation communities in karst wetlands was mainly ranged from 60% to 90%. The water hyacinth and herbaceous vegetation were sensitive to the Mean Digital Surface Model (DSM) and Standard RedEdge band. [ABSTRACT FROM AUTHOR]

Temporal generalization allows a trained classification algorithm to be applied to multiple images across time to derive reliable classification map products. It is a challenging remote-sensing research topic since the results are dependent on the selection of atmospheric correction methods, classification algorithms, validation processes, and their varying combinations. This study examined the temporal generalization of sub-pixel vegetation mapping using multiple Landsat images (1990, 1996, 2004, and 2010). All Landsat images were processed with two atmospheric correction methods: simple dark object subtraction (DOS) and the Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS) algorithm. For the sub-pixel vegetation mapping of the 2004 Landsat image, we used high-resolution OrbView-3 images as a training/validation data set and compared three machine learning algorithms (neural networks, random forests, and classification and regression trees) for their classification performance. The trained classifiers were then applied to other Landsat images (1990, 1996, and 2010) to derive sub-pixel vegetation map products. For the 2004 Landsat image classification, cross-validation shows similar classification results for neural networks (root mean square error (RMSE) = 0.099) and random forests (RMSE = 0.100) algorithms, and both are better than classification and regression trees (RMSE = 0.123). Pseudo-invariant pixels between 2004 and 2010 were used as validation points to evaluate the temporal generalizability of classification algorithms. Simple DOS and LEDAPS atmospheric correction resulted in similar accuracy statistics. The neural-network-based classifier performed best in generating reliable sub-pixel vegetation map products across time. [ABSTRACT FROM PUBLISHER]