Search

Abstract Governing equations are foundations for modelling, predicting, and understanding the Earth system. The Earth system is undergoing rapid change, and the conventional approaches for establishing governing equations, such as empirical generalisations, are becoming increasingly challenging to deal with the complexity and diversity of the geoscience processes we study today. In this Perspective, we explore data-driven equation discovery, a novel scientific artificial intelligence pathway, for advancing geosciences. Data-driven equation discovery identifies hidden patterns from data and transforms them into interpretable equation representations, automating and accelerating equation discovery processes. It provides a practical approach for geoscientists to model and understand complex geoscience processes based on big Earth data. The final vision is to uncover new clear, describable, and quantifiable equations in various geoscience disciplines. We summarize opportunities and highlight that challenges in this field should be addressed by interdisciplinary collaborations.

Summary: The increasing frequency and severity of droughts present a significant risk to vulnerable regions of the globe, potentially leading to substantial human displacement in extreme situations. Drought-induced displacement is a complex and multifaceted issue that can perpetuate cycles of poverty, exacerbate food and water scarcity, and reinforce socio-economic inequalities. However, our understanding of human mobility in drought scenarios is currently limited, inhibiting accurate predictions and effective policy responses. Drought-induced displacement is driven by numerous factors and identifying its key drivers, causal-effect lags, and consequential effects is often challenging, typically relying on mechanistic models and qualitative assumptions. This paper presents a novel, data-driven methodology, grounded in causal discovery, to retrieve the drivers of drought-induced displacement within Somalia from 2016 to 2023. Our model exposes the intertwined vulnerabilities and the leading times that connect drought impacts, water and food security systems along with episodes of violent conflict, emphasizing that causal mechanisms change across districts. These findings pave the way for the development of algorithms with the ability to learn from human mobility data, enhancing anticipatory action, policy formulation, and humanitarian aid.

Abstract Climate change is leading to more extreme weather hazards, forcing human populations to be displaced. We employ explainable machine learning techniques to model and understand internal displacement flows and patterns from observational data alone. For this purpose, a large, harmonized, global database of disaster-induced movements in the presence of floods, storms, and landslides during 2016–2021 is presented. We account for environmental, societal, and economic factors to predict the number of displaced persons per event in the affected regions. Here we show that displacements can be primarily attributed to the combination of poor household conditions and intense precipitation, as revealed through the interpretation of the trained models using both Shapley values and causality-based methods. We hence provide empirical evidence that differential or uneven vulnerability exists and provide a means for its quantification, which could help advance evidence-based mitigation and adaptation planning efforts.

Abstract Existing methods for fine-scale air quality assessment have significant gaps in their reliability. Purely data-driven methods lack any physically-based mechanisms to simulate the interactive process of air pollution, potentially leading to physically inconsistent or implausible results. Here, we report a hybrid multilevel graph neural network that encodes fluid physics to capture spatial and temporal dynamic characteristics of air pollutants. On a multi-air pollutant test in China, our method consistently improved extrapolation accuracy by an average of 11–22% compared to several baseline machine learning methods, and generated physically consistent spatiotemporal trends of air pollutants at fine spatial and temporal scales.

Globally, forests are net carbon sinks that partly mitigates anthropogenic climate change. However, there is evidence of increasing weather-induced tree mortality, which needs to be better understood to improve forest management under future climate conditions. Disentangling drivers of tree mortality is challenging because of their interacting behavior over multiple temporal scales. In this study, we take a data-driven approach to the problem. We generate hourly temperate weather data using a stochastic weather generator to simulate 160,000 years of beech, pine, and spruce forest dynamics with a forest gap model. These data are used to train a generative deep learning model (a modified variational autoencoder) to learn representations of three-year-long monthly weather conditions (precipitation, temperature, and solar radiation) in an unsupervised way. We then associate these weather representations with years of high biomass loss in the forests and derive weather prototypes associated with such years. The identified prototype weather conditions are associated with 5–22% higher median biomass loss compared to the median of all samples, depending on the forest type and the prototype. When prototype weather conditions co-occur, these numbers increase to 10–25%. Our research illustrates how generative deep learning can discover compounding weather patterns associated with extreme impacts.

Hybrid modeling integrates machine learning with scientific knowledge to enhance interpretability, generalization, and adherence to natural laws. Nevertheless, equifinality and regularization biases pose challenges in hybrid modeling to achieve these purposes. This paper introduces a novel approach to estimating hybrid models via a causal inference framework, specifically employing double machine learning (DML) to estimate causal effects. We showcase its use for the Earth sciences on two problems related to carbon dioxide fluxes. In the Q _10 model, we demonstrate that DML-based hybrid modeling is superior in estimating causal parameters over end-to-end deep neural network approaches, proving efficiency, robustness to bias from regularization methods, and circumventing equifinality. Our approach, applied to carbon flux partitioning, exhibits flexibility in accommodating heterogeneous causal effects. The study emphasizes the necessity of explicitly defining causal graphs and relationships, advocating for this as a general best practice. We encourage the continued exploration of causality in hybrid models for more interpretable and trustworthy results in knowledge-guided machine learning.

Abstract Land, atmosphere and climate interact constantly and at different spatial and temporal scales. In this paper we rely on causal discovery methods to infer spatial patterns of causal relations between several key variables of the carbon and water cycles: gross primary productivity, latent heat energy flux for evaporation, surface air temperature, precipitation, soil moisture and radiation. We introduce a methodology based on the convergent cross-mapping (CCM) technique. Despite its good performance in general, CCM is sensitive to (even moderate) noise levels and hyper-parameter selection. We present a robust CCM (RCCM) that relies on temporal bootstrapping decision scores and the derivation of more stringent cross-map skill scores. The RCCM method is combined with the information-geometric causal inference (IGCI) method to address the problem of strong and instantaneous variable coupling, another important and long-standing issue of CCM. The proposed methodology allows to derive spatially explicit global maps of causal relations between the involved variables and retrieve the underlying complexity of the interactions. Results are generally consistent with reported patterns and process understanding, and constitute a new way to quantify and understand carbon and water fluxes interactions.

Cirrus clouds are key modulators of Earth’s climate. Their dependencies on meteorological and aerosol conditions are among the largest uncertainties in global climate models. This work uses 3 years of satellite and reanalysis data to study the link between cirrus drivers and cloud properties. We use a gradient-boosted machine learning model and a long short-term memory network with an attention layer to predict the ice water content and ice crystal number concentration. The models show that meteorological and aerosol conditions can predict cirrus properties with R2 = 0.49. Feature attributions are calculated with SHapley Additive exPlanations to quantify the link between meteorological and aerosol conditions and cirrus properties. For instance, the minimum concentration of supermicron-sized dust particles required to cause a decrease in ice crystal number concentration predictions is 2 × 10−4 mg/m3. The last 15 hr before the observation predict all cirrus properties.

Abstract Climate change exacerbates the occurence of extreme droughts and heatwaves, increasing the frequency and intensity of large wildfires across the globe. Forecasting wildfire danger and uncovering the drivers behind fire events become central for understanding relevant climate‐land surface feedback and aiding wildfire management. In this work, we leverage Deep Learning (DL) to predict the next day's wildfire danger in a fire‐prone part of the Eastern Mediterranean and explainable Artificial Intelligence (xAI) to diagnose model attributions. We implement DL models that capture the temporal and spatio‐temporal context, generalize well for extreme wildfires, and demonstrate improved performance over the traditional Fire Weather Index. Leveraging xAI, we identify the substantial contribution of wetness‐related variables and unveil the temporal focus of the models. The variability of the contribution of the input variables across wildfire events hints into different wildfire mechanisms. The presented methodology paves the way to more robust, accurate, and trustworthy data‐driven anticipation of wildfires.

Abstract All ocean basins have been experiencing significant warming and rising sea levels in recent decades. There are, however, important regional differences, resulting from distinct processes at different timescales (temperature-driven changes being a major contributor on multi-year timescales). In view of this complexity, it deems essential to move towards more sophisticated data-driven techniques as well as diagnostic and prognostic prediction models to interpret observations of ocean warming and sea level variations at local or regional sea basins. In this context, we present a machine learning approach that exploits key ocean temperature estimates (as proxies for the regional thermosteric sea level component) to model coastal sea level variability and associated uncertainty across a range of timescales (from months to several years). Our findings also demonstrate the utility of machine learning to estimate the possible tendency of near-future regional sea levels. When compared to actual sea-level records, our models perform particularly well in the coastal areas most influenced by internal climate variability. Yet, the models are widely applicable to evaluate the patterns of rising and falling sea levels across many places around the globe. Thus, our approach is a promising tool to model and anticipate sea level changes in the coming (1–3) years, which is crucial for near-term decision making and strategic planning about coastal protection measures.

Natural disturbances imperil healthy and productive forests, but quantifying their effects at large scales is challenging. Here the authors apply machine learning to disturbance records and satellite data to quantify and map European forest vulnerability to fires, windthrows, and insect outbreaks through 1979-2018.

Detecting anomalous changes in remote sensing images is a challenging problem, where many approaches and techniques have been presented so far. We rely on the standard field of multivariate statistics of diagnostic measures, which are concerned about the characterization of distributions, detection of anomalies, extreme events, and changes. One useful tool to detect multivariate anomalies is the celebrated Cook's distance. Instead of assuming a linear relationship, we present a novel kernelized version of the Cook's distance to address anomalous change detection in remote sensing images. Due to the large computational burden involved in the direct kernelization, and the lack of out-of-sample formulas, we introduce and compare both random Fourier features and Nyström implementations to approximate the solution. We study the kernel Cook's distance for anomalous change detection in a chronochrome scheme, where the anomalousness indicator comes from evaluating the statistical leverage of the residuals of regressors between time acquisitions. We illustrate the performance of all algorithms in a representative number of multispectral and very high resolution satellite images involving changes due to droughts, urbanization, wildfires, and floods. Very good results and computational efficiency confirm the validity of the approach.

Causal discovery from observational data offers unique opportunities in many scientific disciplines: reconstructing causal drivers, testing causal hypotheses, and comparing and evaluating models for optimizing targeted interventions. Recent causal discovery methods focused on estimating the latent space of the data to get around a lack of causal sufficiency or additivity constraints. However, estimating the latent space significantly increases model complexity, compromising causal identifiability and making it hard to compare models that correspond to different causal hypotheses. We propose a kernel, non-parametric latent-space modelling approach and deal with the difficulty of comparing causal directions by measuring and controlling for the level of causal assumption fulfilment. We introduce a latent noise causal inference framework to estimate latent factors associated with the hypothesized causal direction by optimizing a loss function with kernel independence criteria. We extend the framework to work with time series using an additional time-dependent kernel regularizer. We discuss the additivity assumption and model complexity and give empirical evidence of performance in a wide range of synthetic and real causal discovery problems.

A wide variety of methods exist nowadays to address the important problem of estimating crop yields from available remote sensing and climate data. Among the different approaches, machine learning (ML) techniques are being increasingly adopted, since they allow exploiting all the information on crop progress and environmental conditions and their relations with crop yield, achieving reliable and accurate estimations. However, interpreting the relationships learned by the ML models, and hence getting insights about the problem, remains a complex and usually unexplored task. Without accountability, confidence and trust in the ML models can be compromised. Here, we develop interpretable ML approaches for crop yield estimation that allow us investigating the most informative agro-ecological drivers and influencial regions learned by the models. We conduct a set of experiments to learn the selection of agro-ecological drivers leading to best estimations of main crops grown in Europe: corn, barley and wheat. As input data, we consider a variety of multi-scale Earth observation products sensitive to canopy greenness (e.g. EVI and LAI), its water-uptake dynamics (e.g. VOD) and available water (soil moisture), as well as climatic variables from the ERA5-Land reanalysis (e.g. temperature and radiation). Our results show that the best performances (R2>0.55 for corn and R2>0.8 for both barley and wheat) are obtained when descriptors of soil, vegetation, and atmosphere status are used as input in the ML models. This combination of variables outperforms the results obtained using single variables as inputs or all variables altogether. We then further analyze the relations of input features with crop yield in the developed models by means of Gaussian Process Regression (GPR). We show how the information learned by the GPR model allows us to identify atypical or anomalous crop seasons across the study region, and investigate the underlying factors behind crop progress and failure in Europe.

Questions of causality are ubiquitous in Earth system sciences and beyond, yet correlation techniques still prevail. This Perspective provides an overview of causal inference methods, identifies promising applications and methodological challenges, and initiates a causality benchmark platform.

[This corrects the article DOI: 10.1371/journal.pone.0235885.].

Hyperspectral acquisitions have proven to be the most informative Earth observation data source for the estimation of nitrogen (N) content, which is the main limiting nutrient for plant growth and thus agricultural production. In the past, empirical algorithms have been widely employed to retrieve information on this biochemical plant component from canopy reflectance. However, these approaches do not seek for a cause-effect relationship based on physical laws. Moreover, most studies solely relied on the correlation of chlorophyll content with nitrogen, and thus neglected the fact that most N is bound in proteins. Our study presents a hybrid retrieval method using a physically-based approach combined with machine learning regression to estimate crop N content. Within the workflow, the leaf optical properties model PROSPECT-PRO including the newly calibrated specific absorption coefficients (SAC) of proteins, was coupled with the canopy reflectance model 4SAIL to PROSAIL-PRO. The latter was then employed to generate a training database to be used for advanced probabilistic machine learning methods: a standard homoscedastic Gaussian process (GP) and a heteroscedastic GP regression that accounts for signal-to-noise relations. Both GP models have the property of providing confidence intervals for the estimates, which sets them apart from other machine learners. Moreover, a GP-based sequential backward band removal algorithm was employed to analyze the band-specific information content of PROSAIL-PRO simulated spectra for the estimation of aboveground N. Data from multiple hyperspectral field campaigns, carried out in the framework of the future satellite mission Environmental Mapping and Analysis Program (EnMAP), were exploited for validation. In these campaigns, corn and winter wheat spectra were acquired to simulate spectral EnMAP data. Moreover, destructive N measurements from leaves, stalks and fruits were collected separately to enable plant-organ-specific validation. The results showed that both GP models can provide accurate aboveground N simulations, with slightly better results of the heteroscedastic GP in terms of model testing and against in situ N measurements from leaves plus stalks, with root mean square error (RMSE) of 2.1 g/m². However, the inclusion of fruit N content for validation deteriorated the results, which can be explained by the inability of the radiation to penetrate the thick tissues of stalks, corn cobs and wheat ears. GP-based band analysis identified optimal spectral settings with ten bands mainly situated in the shortwave infrared (SWIR) spectral region. Use of well-known protein absorption bands from the literature showed comparative results. Finally, the heteroscedastic GP model was successfully applied on airborne hyperspectral data for N mapping. We conclude that GP algorithms, and in particular the heteroscedastic GP, should be implemented for global agricultural monitoring of aboveground N from future imaging spectroscopy data.

Kernel methods are powerful machine learning techniques which use generic non-linear functions to solve complex tasks. They have a solid mathematical foundation and exhibit excellent performance in practice. However, kernel machines are still considered black-box models as the kernel feature mapping cannot be accessed directly thus making the kernels difficult to interpret. The aim of this work is to show that it is indeed possible to interpret the functions learned by various kernel methods as they can be intuitive despite their complexity. Specifically, we show that derivatives of these functions have a simple mathematical formulation, are easy to compute, and can be applied to various problems. The model function derivatives in kernel machines is proportional to the kernel function derivative and we provide the explicit analytic form of the first and second derivatives of the most common kernel functions with regard to the inputs as well as generic formulas to compute higher order derivatives. We use them to analyze the most used supervised and unsupervised kernel learning methods: Gaussian Processes for regression, Support Vector Machines for classification, Kernel Entropy Component Analysis for density estimation, and the Hilbert-Schmidt Independence Criterion for estimating the dependency between random variables. For all cases we expressed the derivative of the learned function as a linear combination of the kernel function derivative. Moreover we provide intuitive explanations through illustrative toy examples and show how these same kernel methods can be applied to applications in the context of spatio-temporal Earth system data cubes. This work reflects on the observation that function derivatives may play a crucial role in kernel methods analysis and understanding.

Process understanding and modeling is at the core of scientific reasoning. Principled parametric and mechanistic modeling dominated science and engineering until the recent emergence of machine learning (ML). Despite great success in many areas, ML algorithms in the Earth and climate sciences, and more broadly in physical sciences, are not explicitly designed to be physically-consistent and may, therefore, violate the most basic laws of physics. In this work, motivated by the field of algorithmic fairness, we reconcile data-driven ML with physics modeling by illustrating a nonparametric and nonlinear physics-aware regression method. By incorporating a dependence-based regularizer, the method leads to models that are consistent with domain knowledge, as reflected by either simulations from physical models or ancillary data. The idea can conversely encourage independence of model predictions with other variables that are known to be uncertain either in their representation or magnitude. The method is computationally efficient and comes with a closed-form analytic solution. Through a consistency-vs-accuracy path diagram, one can assess the consistency between data-driven models and physical models. We demonstrate in three examples on simulations and measurement data in Earth and climate studies that the proposed ML framework allows us to trade-off physical consistency and accuracy.

In this paper, we discuss the scientific outcomes of the 2015 data fusion contest organized by the Image Analysis and Data Fusion Technical Committee (IADF TC) of the IEEE Geoscience and Remote Sensing Society (IEEE GRSS). As for previous years, the IADF TC organized a data fusion contest aiming at fostering new ideas and solutions for multisource studies. The 2015 edition of the contest proposed a multiresolution and multisensorial challenge involving extremely high-resolution RGB images and a three-dimensional (3-D) LiDAR point cloud. The competition was framed in two parallel tracks, considering 2-D and 3-D products, respectively. In this paper, we discuss the scientific results obtained by the winners of the 2-D contest, which studied either the complementarity of RGB and LiDAR with deep neural networks (winning team) or provided a comprehensive benchmarking evaluation of new classification strategies for extremely high-resolution multimodal data (runner-up team). The data and the previously undisclosed ground truth will remain available for the community and can be obtained at http://www.grss-ieee.org/community/technical-committees/data-fusion/2015-ieee-grss-data-fusion-contest/. The 3-D part of the contest is discussed in the Part-B paper [1].

Forecasting crop yields is becoming increasingly important under the current context in which food security needs to be ensured despite the challenges brought by climate change, an expanding world population accompanied by rising incomes, increasing soil erosion, and decreasing water resources. Temperature, radiation, water availability and other environmental conditions influence crop growth, development, and final grain yield in a complex nonlinear manner. Machine learning (ML) techniques, and deep learning (DL) methods in particular, can account for such nonlinear relations between yield and its covariates. However, they typically lack transparency and interpretability, since the way the predictions are derived is not directly evident. Yet, in the context of yield forecasting, understanding which are the underlying factors behind both a predicted loss or gain is of great relevance. Here, we explore how to benefit from the increased predictive performance of DL methods while maintaining the ability to interpret how the models achieve their results. To do so, we applied a deep neural network to multivariate time series of vegetation and meteorological data to estimate the wheat yield in the Indian Wheat Belt. Then, we visualized and analyzed the features and yield drivers learned by the model with the use of regression activation maps. The DL model outperformed other tested models (ridge regression and random forest) and facilitated the interpretation of variables and processes that lead to yield variability. The learned features were mostly related to the length of the growing season, and temperature and light conditions during this time. For example, our results showed that high yields in 2012 were associated with low temperatures accompanied by sunny conditions during the growing period. The proposed methodology can be used for other crops and regions in order to facilitate application of DL models in agriculture.

Physically-based radiative transfer models (RTMs) help in understanding the processes occurring on the Earth’s surface and their interactions with vegetation and atmosphere. When it comes to studying vegetation properties, RTMs allows us to study light interception by plant canopies and are used in the retrieval of biophysical variables through model inversion. However, advanced RTMs can take a long computational time, which makes them unfeasible in many real applications. To overcome this problem, it has been proposed to substitute RTMs through so-called emulators. Emulators are statistical models that approximate the functioning of RTMs. Emulators are advantageous in real practice because of the computational efficiency and excellent accuracy and flexibility for extrapolation. We hereby present an “Emulator toolbox” that enables analysing multi-output machine learning regression algorithms (MO-MLRAs) on their ability to approximate an RTM. The toolbox is included in the free-access ARTMO’s MATLAB suite for parameter retrieval and model inversion and currently contains both linear and non-linear MO-MLRAs, namely partial least squares regression (PLSR), kernel ridge regression (KRR) and neural networks (NN). These MO-MLRAs have been evaluated on their precision and speed to approximate the soil vegetation atmosphere transfer model SCOPE (Soil Canopy Observation, Photochemistry and Energy balance). SCOPE generates, amongst others, sun-induced chlorophyll fluorescence as the output signal. KRR and NN were evaluated as capable of reconstructing fluorescence spectra with great precision. Relative errors fell below 0.5% when trained with 500 or more samples using cross-validation and principal component analysis to alleviate the underdetermination problem. Moreover, NN reconstructed fluorescence spectra about 50-times faster and KRR about 800-times faster than SCOPE. The Emulator toolbox is foreseen to open new opportunities in the use of advanced RTMs, in which both consistent physical assumptions and data-driven machine learning algorithms live together.

The wealth of sensory data coming from different modalities has opened numerous opportunities for data analysis. The data are of increasing volume, complexity and dimensionality, thus calling for new methodological innovations towards multimodal data processing. However, multimodal architectures must rely on models able to adapt to changes in the data distribution. Differences in the density functions can be due to changes in acquisition conditions (pose, illumination), sensors characteristics (number of channels, resolution) or different views (e.g. street level vs. aerial views of a same building). We call these different acquisition modes domains, and refer to the adaptation problem as domain adaptation. In this paper, instead of adapting the trained models themselves, we alternatively focus on finding mappings of the data sources into a common, semantically meaningful, representation domain. This field of manifold alignment extends traditional techniques in statistics such as canonical correlation analysis (CCA) to deal with nonlinear adaptation and possibly non-corresponding data pairs between the domains. We introduce a kernel method for manifold alignment (KEMA) that can match an arbitrary number of data sources without needing corresponding pairs, just few labeled examples in all domains. KEMA has interesting properties: 1) it generalizes other manifold alignment methods, 2) it can align manifolds of very different complexities, performing a discriminative alignment preserving each manifold inner structure, 3) it can define a domain-specific metric to cope with multimodal specificities, 4) it can align data spaces of different dimensionality, 5) it is robust to strong nonlinear feature deformations, and 6) it is closed-form invertible, which allows transfer across-domains and data synthesis. To authors' knowledge this is the first method addressing all these important issues at once. We also present a reduced-rank version of KEMA for computational efficiency, and discuss the generalization performance of KEMA under Rademacher principles of stability. Aligning multimodal data with KEMA reports outstanding benefits when used as a data pre-conditioner step in the standard data analysis processing chain. KEMA exhibits very good performance over competing methods in synthetic controlled examples, visual object recognition and recognition of facial expressions tasks. KEMA is especially well-suited to deal with high-dimensional problems, such as images and videos, and under complicated distortions, twists and warpings of the data manifolds. A fully functional toolbox is available at https://github.com/dtuia/KEMA.git.

This paper proposes a processing chain for the derivation of global Leaf Area Index (LAI), Fraction of Absorbed Photosynthetically Active Radiation (FAPAR), Fraction Vegetation Cover (FVC), and Canopy water content (CWC) maps from 15-years of MODIS data exploiting the capabilities of the Google Earth Engine (GEE) cloud platform. The retrieval chain is based on a hybrid method inverting the PROSAIL radiative transfer model (RTM) with Random forests (RF) regression. A major feature of this work is the implementation of a retrieval chain exploiting the GEE capabilities using global and climate data records (CDR) of both MODIS surface reflectance and LAI/FAPAR datasets allowing the global estimation of biophysical variables at unprecedented timeliness. We combine a massive global compilation of leaf trait measurements (TRY), which is the baseline for more realistic leaf parametrization for the considered RTM, with large amounts of remote sensing data ingested by GEE. Moreover, the proposed retrieval chain includes the estimation of both FVC and CWC, which are not operationally produced for the MODIS sensor. The derived global estimates are validated over the BELMANIP2.1 sites network by means of an inter-comparison with the MODIS LAI/FAPAR product available in GEE. Overall, the retrieval chain exhibits great consistency with the reference MODIS product (R2 = 0.87, RMSE = 0.54 m2/m2 and ME = 0.03 m2/m2 in the case of LAI, and R2 = 0.92, RMSE = 0.09 and ME = 0.05 in the case of FAPAR). The analysis of the results by land cover type shows the lowest correlations between our retrievals and the MODIS reference estimates (R2 = 0.42 and R2 = 0.41 for LAI and FAPAR, respectively) for evergreen broadleaf forests. These discrepancies could be attributed mainly to different product definitions according to the literature. The provided results proof that GEE is a suitable high performance processing tool for global biophysical variable retrieval for a wide range of applications.

The exploitation of Earth observation satellite images acquired by optical instruments requires an automatic and accurate cloud detection. Multitemporal approaches to cloud detection are usually more powerful than their single scene counterparts since the presence of clouds varies greatly from one acquisition to another whereas surface can be assumed stationary in a broad sense. However, two practical limitations usually hamper their operational use: the access to the complete satellite image archive and the required computational power. This work presents a cloud detection and removal methodology implemented in the Google Earth Engine (GEE) cloud computing platform in order to meet these requirements. The proposed methodology is tested for the Landsat-8 mission over a large collection of manually labeled cloud masks from the Biome dataset. The quantitative results show state-of-the-art performance compared with mono-temporal standard approaches, such as FMask and ACCA algorithms, yielding improvements between 4–5% in classification accuracy and 3–10% in commission errors. The algorithm implementation within the Google Earth Engine and the generated cloud masks for all test images are released for interested readers.

Progress in advanced radiative transfer models (RTMs) led to an improved understanding of reflectance (R) and sun-induced chlorophyll fluorescence (SIF) emission throughout the leaf and canopy. Among advanced canopy RTMs that have been recently modified to deliver SIF spectral outputs are the energy balance model SCOPE and the 3D models DART and FLIGHT. The downside of these RTMs is that they are computationally expensive, which makes them impractical in routine processing, such as scene generation and retrieval applications. To bypass their computational burden, a computationally effective technique has been proposed by only using a limited number of model runs, called emulation. The idea of emulation is approximating the original RTM by a surrogate machine learning model with low computation time. However, a concern is whether the emulator reaches sufficient accuracy. To this end, we analyzed key aspects of emulator development that may impact the precision of emulating SCOPE-like R and SIF spectra, being: (1) type of machine learning, (2) type of dimensionality reduction (DR) method, and (3) number of components and lookup table (LUT) size. The machine learning family of Gaussian processes regression and neural networks were found best suited to function as emulators. The classical principal component analysis (PCA) remains a robust DR method, but the number of components needs to be optimized depending on the complexity of the spectral data. Based on a small Latin hypercube sampling LUT of 500 samples (70% used for training) covering a selection of SCOPE input variables, the best-performing emulators can reconstruct any combination for the selected SCOPE input variables with relative errors along the spectral range below 2% for R and 4% for SIF. That is sufficient for a precise reconstruction for the large majority of possible combinations, and errors can be further reduced when increasing LUT size for training. As a proof of concept, we imported the best-performing emulators into a newly developed Automated Scene Generator Module (A-SGM) to generate a R and SIF synthetic scene of a vegetated surface. Using emulators as alternative of SCOPE reduced the processing time from the order of days to the order of minutes while preserving sufficient accuracy.

This paper presents and evaluates multitemporal LAI estimates derived from Sentinel-2A data on rice cultivated area identified using time series of Sentinel-1A images over the main European rice districts for the 2016 crop season [...]

Physically-based radiative transfer models (RTMs) help understand the interactions of radiation with vegetation and atmosphere. However, advanced RTMs can be computationally burdensome, which makes them impractical in many real applications, especially when many state conditions and model couplings need to be studied. To overcome this problem, it is proposed to substitute RTMs through surrogate meta-models also named emulators. Emulators approximate the functioning of RTMs through statistical learning regression methods, and can open many new applications because of their computational efficiency and outstanding accuracy. Emulators allow fast global sensitivity analysis (GSA) studies on advanced, computationally expensive RTMs. As a proof-of-concept, three machine learning regression algorithms (MLRAs) were tested to function as emulators for the leaf RTM PROSPECT-4, the canopy RTM PROSAIL, and the computationally expensive atmospheric RTM MODTRAN5. Selected MLRAs were: kernel ridge regression (KRR), neural networks (NN) and Gaussian processes regression (GPR). For each RTM, 500 simulations were generated for training and validation. The majority of MLRAs were excellently validated to function as emulators with relative errors well below 0.2%. The emulators were then put into a GSA scheme and compared against GSA results as generated by original PROSPECT-4 and PROSAIL runs. NN and GPR emulators delivered identical GSA results, while processing speed compared to the original RTMs doubled for PROSPECT-4 and tripled for PROSAIL. Having the emulator-GSA concept successfully tested, for six MODTRAN5 atmospheric transfer functions (outputs), i.e., direct and diffuse at-surface solar irradiance ( E d i f , E d i r ), direct and diffuse upward transmittance ( T d i r , T d i f ), spherical albedo (S) and path radiance ( L 0 ), the most accurate MLRA’s were subsequently applied as emulator into the GSA scheme. The sensitivity analysis along the 400–2500 nm spectral range took no more than a few minutes on a contemporary computer—in comparison, the same analysis in the original MODTRAN5 would have taken over a month. Key atmospheric drivers were identified, which are on the one hand aerosol optical properties, i.e., aerosol optical thickness (AOT), Angstrom coefficient (AMS) and scattering asymmetry variable (G), mostly driving diffuse atmospheric components, E d i f and T d i f ; and those affected by atmospheric scattering, L 0 and S. On the other hand, as expected, AOT, AMS and columnar water vapor (CWV) in the absorption regions mostly drive E d i r and T d i r atmospheric functions. The presented emulation schemes showed very promising results in replacing costly RTMs, and we think they can contribute to the adoption of machine learning techniques in remote sensing and environmental applications.