Hugo Do Nascimento Bendini

Perennial crops are a key component of sustainable agriculture due to their higher carbon storage compared to annual systems. Improving their mapping is essential to support effective public policies. This study investigates the use of field observations, waveform Light Detection and Ranging (LiDAR) data from the Global Ecosystem Dynamics Investigation (GEDI), and hyperspectral imagery from EnMAP to identify spectral and structural features and assess their separability. Field data were used to develop an interpretation key for the analyzed classes. GEDI Level 1B waveforms were normalized to extract metrics such as the number of peaks, while structural attributes, including canopy height (RH95), were derived from GEDI Level 2A. From EnMAP imagery, spectral reflectance was obtained and processed using mean values, first- and second-order derivatives, and continuum removal. Differences among classes were evaluated using Tukey’s test and principal component analysis (PCA). Results show that waveform structure and canopy height significantly distinguish secondary vegetation from perennial crops. Additionally, spectral regions in the visible and shortwave infrared present significant differences between these classes. These findings demonstrate that the structural and spectral variables could improve the discrimination of perennial crops and help overcome challenges in separating them from secondary vegetation.

Perennial crops are a key component of sustainable agriculture due to their higher carbon storage compared to annual systems. Improving their mapping is essential to support effective public policies. This study proposes an integrated approach combining field observations, waveform LiDAR data from Global Ecosystem Dynamics Investigation (GEDI), and hyperspectral imagery from EnMAP to enhance perennial crop classification.Field data were used to develop an interpretation key for the analyzed classes. GEDI Level 1B waveforms were normalized to extract metrics such as the number of peaks, while structural attributes, including canopy height (RH95), were derived from GEDI Level 2A. From EnMAP imagery, spectral reflectance was obtained and processed using mean values, first- and second-order derivatives, and continuum removal. Differences among classes were evaluated using Tukey’s test and principal component analysis (PCA). Results show that waveform structure and canopy height significantly distinguish secondary vegetation from perennial crops. Additionally, spectral regions in the visible and shortwave infrared present significant differences between these classes. These findings demonstrate that integrating structural and spectral variables improves the discrimination of perennial crops and helps overcome challenges in separating them from secondary vegetation.

Brazil, a leading rice producer globally, faces challenges in systematically mapping its diverse rice fields due to varying cropping systems, climates, and planting calendars. Existing rice mapping methods often rely on complex techniques like deep learning or microwave imagery, posing limitations for large-scale mapping. This study proposes a novel approach utilizing Sentinel-2 spectral–temporal metrics (STMs) in conjunction with a random forest classifier for rice paddy mapping. By extracting diverse STMs and training both regional and global classifiers, we validated the method across independent areas. While regional models tended to overestimate rice areas, the global model effectively reduced discrepancies between our data and the reference maps, achieving an overall classifier accuracy exceeding 80%. Despite the need for further refinement to address confusion with other crops, STM exhibits promise for national-scale rice paddy mapping in Brazil.

Cover crops play a pivotal role in mitigating climate change by bolstering carbon sequestration through biomass production and soil integration. However, current methods for quantifying cover crop biomass lack spatial precision and objectivity. Thus, our research aimed to devise a remote-sensing-based approach to estimate cover crop biomass across various species and mixtures during fallow periods in France. Leveraging Sentinel-2 optical data and machine learning algorithms, we modeled biomass across 50 fields representative of France’s diverse cropping practices and climate types. Initial tests using traditional empirical relationships between vegetation indices/spectral bands and dry biomass revealed challenges in accurately estimating biomass for mixed cover crop categories due to spectral interference from grasses and weeds, underscoring the complexity of modeling diverse agricultural conditions. To address this challenge, we compared several machine learning algorithms (Support Vector Machine, Random Forest, and eXtreme Gradient Boosting) using spectral bands and vegetation indices from the latest available image before sampling as input. Additionally, we developed an approach that incorporates dense optical time series of Sentinel-2 data, generated using a Radial Basis Function for interpolation. Our findings demonstrated that a Random Forest model trained with dense time series data during the cover crop development period yielded promising results, with an average R-squared (r²) value of 0.75 and root mean square error (RMSE) of 0.73 t·ha⁻¹, surpassing results obtained from methods using single-image snapshots (r² of 0.55). Moreover, our approach exhibited robustness in accounting for factors such as crop species diversity, varied climatic conditions, and the presence of weed vegetation—essential for approximating real-world conditions. Importantly, its applicability extends beyond France, holding potential for global scalability. The availability of data for model calibration across diverse regions and timeframes could facilitate broader application.

In the last decades, several methodologies for estimating crop phenology based on remote sensing data have been developed and used to create different algorithms. Although many studies have been conducted to evaluate the different methodologies, a comprehensive understanding of the potential of the different current algorithms to detect changes in the growing season is still lacking, especially in large regions and with more than one crop per season. Therefore, this work aimed to evaluate different phenological metrics extraction methodologies. Using data from over 1500 fields distributed across Brazil’s central area, six algorithms, including CropPhenology, Digital Earth Australia tools package (DEA), greenbrown, phenex, phenofit, and TIMESAT, to extract soybean crop phenology were applied. To understand how robust the algorithms are to different input sources, the NDVI and EVI2 time series derived from MODIS products (MOD13Q1 and MOD09Q1) and from Sentinel-2 satellites were used to estimate the sowing date (SD) and harvest date (HD) in each field. The algorithms produced significantly different phenological date estimates, with Spearman’s R ranging between 0.26 and 0.82 when comparing sowing and harvesting dates. The best estimates were obtained using TIMESAT and phenex for SD and HD, respectively, with R greater than 0.7 and RMSE of 16–17 days. The DEA tools and greenbrown packages showed higher sensitivity when using different data sources. Double cropping is an added challenge, with no method adequately identifying it.

The increasing deforestation and fires since 2019 raises concerns about the irreversible destruction of the Brazilian Amazon. Our goal was to better understand these changes in south-west Pará across different land-tenure and farm systems and between the terms of President Rousseff, Temer, and Bolsonaro. We reconstructed deforestation and fire history using all Landsat and Sentinel-2 observations from 2014 to 2020 and assessed, using quasi-experimental methods, the average treatment effects of each presidency on deforestation and fires across land-tenure and farm types. Deforestation nearly quadrupled to 1,201 km², particularly during Bolsonaro in undesignated areas and conservation units and on medium-sized farms (p < 0.001). Burning increased to 4,805 km² and in all tenure types (p < 0.001). The increase was strongest in agrarian settlements and conservation units and on medium and large farms. Our observations show the importance of clarifying land-tenure and re-strengthening disincentives of environmental infractions, which have been weakened specifically under President Bolsonaro.

In Brazil, irrigated agriculture is responsible for 46% of withdrawals of water bodies and 67% of use concerning the total water abstracted volume, representing the most significant consumptive use in the country. Understanding how different crops use water over time is essential for planning and managing water allocation, water rights, and farming production. In this work, we propose a methodology to estimate water used in agriculture irrigated by center pivots in the municipality of Itobi, São Paulo, in the Brazilian Savanna (known as Cerrado), which has strong potential for agricultural and livestock production. The methodology proposed for the water use estimate is based on mapping crops irrigated by center pivots for the 2015/2016 crop year and actual evapotranspiration (ETa). ETa is derived from the Operational Simplified Surface Energy Balance model (SSEBop) and parameterized for edaphoclimatic conditions in Brazil (SSEBop-Br). Three meteorological data sources (INMET, GLDAS, CFSv2) were tested for estimating ETa. The water use was estimated for each meteorological data source, relating the average irrigation balance and the total area for each crop identified in the map. We evaluated the models for each crop present in the center pivots through global accuracy and f1-score metrics, and f1-score was more significant than 0.9 for all crops. The potato was the crop that consumed the most water in irrigation, followed by soy crops, beans, carrots, and onions, considering the three meteorological data sources. The total water volume consumed by center pivots in the municipality of Itobi in the 2015/2016 agricultural year for each meteorological data source was 3.2 million m³ (INMET), 2.5 million m³; (GLDAS), and 1.8 million m³ (CFSv2).

The Brazilian Savanna is the second largest biogeographical region in Brazil and present different vegetation types, consisting mostly of tropical savannas, grasslands, and forests. The forest types have different tree cover and floristic composition, which is associated to leaf deciduousness. Considering the importance of Cerrado to biodiversity conservation and the maintaining of environmental services, the development of methods to map the different forest types in Cerrado is important for conservation programmes, subsidize restauration plains, and to allow estimations of carbon sink and stock. Mapping heterogeneous tropical areas, such as the Brazilian Savanna, is very complex due to the natural factors and peculiarities of the vegetation types, and it's still particularly challenging to separate between different forest formations. In this study we tested machine learning approaches based on the use of dense image time series, in order to evaluate the separability Dry Tropical Forests and Savanna woodlands. We considered the Brazilian State of Tocantins as the study area, which is located in the Northern region of the country. RF classification of Landsat dense time series showed an overall accuracy of 0.85005, while the LSTM approach presented an overall accuracy of 0.88601, with the highest f1-score for the savanna woodlands class, suggesting the capability of the recurrent neural networks on handling complex long-term dependencies such as the EVI dense time series data. This study showed the potential for the development of a semi-automatic method for discriminating the different types of forest formations in the Brazilian Savanna, based on remote sensing.

Classifying dense satellite image time series has become a necessity, especially with the recent efforts to create analysis ready data cubes. Approaches developed to perform this task are usually pixel-based. Even though these approaches can achieve good results, they do not take advantage of the intrinsic spatial correlation of geographic data nor do they consider spatial heterogeneity along with the time series. Region-based classification is a suitable solution to incorporate contextual information for dense satellite image time series classification. In this article, we introduce a new segmentation method based on a superpixel approach. This method creates multitemporal superpixels, which are meaningful regions in space and time. To evaluate the performance of the proposed method, tests were performed on two data sets using a total of 23 ground-truth references. Experimental results showed that the method performed well, achieving a good boundary agreement and obtaining high scores on the three metrics used for evaluation.