The CHOVE-CHUVA project is a Space for Climate Observatory initiative aiming at developing operational tools to monitor socio-environmental dynamics in the Brazilian state of Mato Grosso, in the Southern Amazon. This project focuses on the dissemination of remote sensing-based spatial information to monitor the evolution of climate variables and land use dynamics, especially regarding agriculture, natural vegetation and hydrological resources. The platform is enhanced by the collection of collaborative data about the adoption of specific land use types (e.g. forest restoration and crop-livestock-forest integrated systems) encouraged by the Brazilian program for a low-carbon agriculture (ABC plan).

In recent decades, oil palm plantations have expanded significantly in Indonesia. Palm oil production currently relies on two plantation types with different economic, social and environmental impacts: (1) industrial plantations and (2) smallholder plantations. While efforts have been made to characterize these plantation types, this objective remains challenging for the remote sensing community. Consequently, this study assesses the potential of Sentinel-1 textural metrics in distinguishing between industrial and smallholder plantations, offering potential solutions to this persistent challenge. We used machine learning algorithms (Random Forest vs. eXtreme Gradient Boosting Tree) with textural features calculated by the Grey Level Co-occurrence Matrix. The results confirmed the potential of Sentinel-1 textural metrics to discriminate OP plantation types. The XGBTree model achieved a higher Kappa (0.71) than the Random Forest model (0.63). Moreover, the contrast, dissimilarity, GLCM-Mean, and GLCMVariance metrics were the most explanatory for discriminating industrial and smallholder plantations. This study possesses limitations, especially concerning its applicability on a broader scale. However, leveraging cloud computing tools like Google Earth Engine could aid in scaling up the methodology.

The Pantanal, a large wetland in the centre of the South American continent, is a heterogeneous ecosystem and has complex relationships with climatic and geomorphic factors. Here, we investigated the relationships between woodland phenology, rainfall, and flooding patterns along the north-south gradient of the Pantanal. The average values of monthly Normalized Difference Vegetation Index (NDVI) from Harmonized Landsat Sentinel-2 (HLS) time series, rainfall from the Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS), and water level from gauge stations for the 2015-2022 period were used to monitor phenology, rainfall, and flooding cycles. The relationships were assessed using Pearson’s correlation. Rainfall has steady patterns across the gradient, and phenology positively responds to it, notably southward, where deciduous trees dominate. Flooding is regionally asynchronous. Although the correlation between phenology and flooding is predominantly negative in the south, patches with positive responses indicate that flood seasonalities are important drivers of local vegetation variability.

As one of the most important application for remote sensing monitoring, land cover mapping has witnessed notable advancements in data acquisition, algorithmic diversity, and classification accuracy. Despite the instrumental role data-driven algorithms have played in the development of global land cover products, their inherent limitations as “black box” methods often fall short of meeting end-users’ specific requirements. In this study, built upon the foundation of the earlier land cover monitoring platform [FROM-GLC plus(FGP)], a data and knowledge dual-driven framework (FGP 2.0) was developed as a user-adaptive framework for intelligent remote sensing land cover mapping. By incorporating ontology-based semantic descriptions with advanced data-driven algorithms, FGP 2.0 provides the capacity for both traditional annual mapping and emerging dynamic mapping. Our results illustrate that FGP 2.0 significantly improves the overall accuracy of annual maps by ~5%, and dynamic maps by ~20% compared to FGP. Moreover, an operational dynamic mapping tool has been developed on the Google Earth engine (GEE), enabling the generation of near-real-time land cover maps for any given place. With an extensible and flexible mapping framework, FGP 2.0 demonstrates the potential of customized land cover monitoring results to suit different application scenarios. This innovative approach not only meets the current demand for reliable annual and dynamic land cover maps but also sets a new benchmark for the integration of geoscientific expertise with machine learning techniques in remote sensing monitoring.

The rapid expansion of a very active agricultural frontier in the Southern Brazilian Amazon induces high deforestation rates that influence local to global water cycles. In this study, we analysed landscape indicators derived from Brazilian MapBiomas land use maps in order to assess deforestation patterns in watersheds dominated by crop or pasture lands. Our results indicate that the proportion of forest is on average higher in watersheds where pasture prevails (53,7%) compared to watersheds where soybean prevails (42,4%). On the contrary, we also found that riparian buffers in soybean dominant watersheds are better preserved (75,7% covered by forest) than in pasture dominant watersheds (60,0% of forested area). Finally, we emphasize the benefit of monitoring land use impacts on stream reach scale, as new remote sensing technologies are under development.

Monitoring forest degradation at fine scale over large area is critical from an environmental point of view since it provides crucial information for many ecological applications. We introduce an automatic method based on optical Landsat time series (2000–2017) to detect and quantify forest disturbances and to identify the causes of forest degradation. The method is based on i) an automatic spectral unmixing to detect forest's disturbances and on ii) landscape metrics and temporal indicators to detect the causes of forest degradation. We applied the approach in the Brazilian Amazon municipality of Paragominas to map forested areas affected by reduced impact logging, conventional logging or illegal logging and fires.

The Lençóis Maranhenses National Park (LMNP) constitutes the largest coastal dune field in South America. It is a remarkable reservoir of biodiversity facing important preservation challenges due to the rapid development of anthropogenic activities, including tourism. The objective of this study is to introduce preliminary results on the understanding of seasonal sand dune dynamics. For this purpose, we used Sentinel 2 time series from 29/07/17 to 17/09/18 in order to monitor the intra-annual migration of interdunal ponds. The method relied on three steps: 1) automatic Spectral Mixture Analysis to estimate the proportion of vegetation, mineral and water in each pixel, 2) rule-based classification and 3) extraction of interdunal areas to assess their dynamics. The average offset distance was 21.6 meters and the angle was 72.6°, corresponding to the main west-southwest wind direction. Additional studies to assess the long term expansion of the dune field are necessary.

Mapping vegetation in the tropics is of primary importance to assess its contribution to important ecosystem services. This implies to implement methods to capture the vegetation gradient that characterizes land cover in these regions. Linear Mixture Models have long been used to monitor this gradient. In the present study, we automatically unmixed six Landsat 8 images of a study area in the Republic of Congo. We then computed the weighted average fraction of mineral, vegetation and water/shadow classes for each pixel in order to produce an annual (nearly) cloud-free unmixed image. Finally this product is pre-classified into six semantic classes ranging from “very dark” to “very bright” classes to discriminate the vegetation gradient based on its visual appearance. Results indicate the ability of the approach to classify fine land cover classes while still keeping textural information of the raw image.

Mapping croplands is a challenging problem in a context of climate change and evolving agricultural calendars. Classification based on MODIS vegetation index time series is performed in order to map crop types in the Brazilian state of Mato Grosso. We used the recently developed Dense Bag-of-Temporal-SIFT-Words algorithm, which is able to capture temporal locality of the data. It allows the accurate detection of around 70% of the agricultural areas. It leads to better classification rates than a baseline algorithm, discriminating more accurately classes with similar profiles.

Increases in the frequency of extreme events, such as the occurrence of high temperatures, are prone to produce severe effects on summer crop yields especially soybeans and maize. Under a climate change scenario, the physical parameters of the Earth's surface, such as temperature, water availability and evapotranspiration, are expected to change over the next decades. We investigated the variability of soybean yields associated with crop canopy temperatures during key development that are sensitive to the occurrence of high temperatures in Mato Grosso State, Brazil. In the present paper, we propose that the temperature fluctuations around the optimum level in the crop canopy can cause favorable effects on soybean yields in MT State/Brazil. In order to evaluate the above mentioned hypothesis, we investigated the effects of canopy temperature on soybean yield during flowering to the grain filling periods using Aqua and Terra/MODIS (Moderate Resolution Imaging Spectroradiometer) satellite data, between 2003 and 2010. Comparison of spatially interpolated maps show that yield variations are positively related to canopy-LST during of flowering period, with R2=0.60 and RMSD=6.2%. Overall results show that increases in canopy-LST temperature in Mato Grosso State, during flowering/grain filling periods, are related to higher soybean yield averages.

Indigenous lands represent an efficient way to protect indigenous communities and environment in Brazil. However, these lands are also highly affected y the land use changes occuring in its surroundings. We quantified the land use changes in the Xingu basin based on MODIS EVI data between 2000 and 2006. We estimated the deforested area inside and outside the indigenous lands, the crop expansion and intensification around the protected areas. Our results indicate that, even if indigenous lands are efficient to limit deforestation (97.5% of deforestation is outside the indigenous lands), crop expansion and intensification (double crop systems) are increasing rapidly, what may imply pollution of headwaters of the Xingu river which crosses the protected area.

Time series of MODIS vegetation indices are widely used to map vegetation. However, some noise can affect the temporal profiles. Thus, many techniques have been developed to smooth them. Four algorithms are applied on crop pixels in the Brazilian Amazonian State of Mato Grosso. Comparisons led to the selection of the Weighted Least Squares (WLS) algorithm and the Savitzky-Golay (SG) filter. Those techniques were computed on MODIS data in order to detect six crop classes. Tests of separability show that the smoothed data improved the potential of separability at each MODIS sub-period. Moreover, supervised classifications were then realized. The WLS data refined efficiently the classification result when using C4.5 decision tree. When using the Maximum Likelihood and Spectral Angle Mapper classifiers, the smoothed data did not improve the classification results as compared with those obtained through original MODIS data. However, it required fewer input MODIS images to reach good results. The SG filter led to better results than the WLS algorithm when using those classifiers.