Movement of livestock production within a country or region has implications for genetics, adaptation, well-being, nutrition, and production logistics, particularly in continental-sized countries, such as Brazil. Cattle production in Brazil from 1977 to 2011 was spatialized, and the annual midpoint of production was calculated. Changes in the relative production and acceleration of production were calculated and spatialized using ARCGIS®. Cluster and canonical discriminant analyses were performed to further highlight differences between regions in terms of cattle production. The mean production point has moved from the Center of Minas Gerais State (in the southeast region) to the North of Goiás State (in the Midwest region). This reflects changes in environmental factors, such as pasture type, temperature and humidity. Acceleration in production in the northern region of Brazil has remained strong over the years. More recently, “traditional” cattle-rearing regions, such as the south and southeast, showed a reduction in growth rates as well as a reduction in herd size or internal migration over the period studied. These maps showed that this movement tends to be gradual, with few regions showing high acceleration or deceleration rates.

Sheep production is present on all continents and has been practiced in Brazil since the colonization. In this study, the multitemporal dynamics of sheep production in Brazil is examined using official government data (Brazilian Institute for Geography and Statistics-IBGE) from 1976 to 2010. Maps of flock growth rates and growth acceleration maps by municipality were elaborated. The Southern states are seen to show a reduction in production mainly due to the wool crisis in the 1970s and 80s. The Northeast is seen to be important for meat production. More recently, centerwest and northern states have shown an increase in growth rates but this is still incipient. The maps of growth, acceleration and midpoint for sheep production showed a noticeable return to an increase in production in the South in recent years. The midpoint of production flow was in the northeast direction, which has stagnated. There was great dynamics in sheep production over the whole Brazilian territory, which affected supply chains due to the expansion of domestic and foreign markets. Areas with higher fluctuations in production are more vulnerable in terms of investment policies.

The expansion of agricultural frontiers in Brazil has caused substantial changes in land use and land cover. This research aims to analyze the space-time dynamics of soybeans and cattle production in the Brazilian territory during the period 1991–2015. The spatial analysis adopted the following procedures: (a) The change vector from the annual calculation of the midpoint of production; (b) mapping of the growth and acceleration rates of the two productions, and (c) mapping of the correlation between the time series of soybean and cattle. The results showed high rates of growth and acceleration for soy production in the South, Central-West and Matopiba regions. The growth acceleration rate identified the long-term deviations that characterized the effective soybean and cattle expansion areas. The results demonstrated the effects of Brazil’s soy moratorium contained soybean expansion into the Amazon region. However, as a side effect, the soybean production replaced cattle production in the savanna region, which in turn, migrated to the Amazon rainforest. Therefore, the present study highlights the importance of public policies that comprehensively understand the spatial-temporal dynamics of Brazilian agriculture to promote sustainable land-use practices.

The state of Amapá within the Amazon biome has a high complexity of ecosystems formed by forests, savannas, seasonally flooded vegetation, mangroves, and different land uses. The present research aimed to map the vegetation from the phenological behavior of the Sentinel-1 time series, which has the advantage of not having atmospheric interference and cloud cover. Furthermore, the study compared three different sets of images (vertical–vertical co-polarization (VV) only, vertical–horizontal cross-polarization (VH) only, and both VV and VH) and different classifiers based on deep learning (long short-term memory (LSTM), Bidirectional LSTM (Bi-LSTM), Gated Recurrent Units (GRU), Bidirectional GRU (Bi-GRU)) and machine learning (Random Forest, Extreme Gradient Boosting (XGBoost), k-Nearest Neighbors, Support Vector Machines (SVMs), and Multilayer Perceptron). The time series englobed four years (2017–2020) with a 12-day revisit, totaling 122 images for each VV and VH polarization. The methodology presented the following steps: image pre-processing, temporal filtering using the Savitsky–Golay smoothing method, collection of samples considering 17 classes, classification using different methods and polarization datasets, and accuracy analysis. The combinations of the VV and VH pooled dataset with the Bidirectional Recurrent Neuron Networks methods led to the greatest F1 scores, Bi-GRU (93.53) and Bi-LSTM (93.29), followed by the other deep learning methods, GRU (93.30) and LSTM (93.15). Among machine learning, the two methods with the highest F1-score values were SVM (92.18) and XGBoost (91.98). Therefore, phenological variations based on long Synthetic Aperture Radar (SAR) time series allow the detailed representation of land cover/land use and water dynamics.