Conserving tropical forests is crucial for the environment and future of our climate. Tropical rainforests worldwide, including the Brazilian Legal Amazon (BLA), offer exceptional ecosystem services. However, the disturbances that have been occurring more frequently within them are endangering their key role in tackling climate change. An alternative approach for preserving the intact forests that remain in the BLA is the delimitation of Indigenous Lands (ILs), which can, additionally, ensure the well-being of the traditional peoples inhabiting there. An increase in deforestation rates of the BLA in recent years, due to the weakening of the Brazilian environmental policy, is not confined to unprotected areas but is also occurring within ILs. Under this scenario, mining, not allowed in ILs, is a growing threat in these protected areas. Thus, using the freely available MapBiomas dataset, we have quantified for the first time the total mining area within ILs of the BLA from 1985 to 2020. Such activity jumped from 7.45 km² in 1985 to 102.16 km² in 2020, an alarming increase of 1271%. Three ILs (Kayapó, Mundurukú, and Yanomami) concentrated 95% of the mining activity within ILs in 2020 and, therefore, they require closer monitoring. Most of the mining in ILs in 2020 (99.5%) was related to gold extraction. A total of 25 of the 31 ILs of the BLA where mining activity was detected in at least one of 36 years analyzed (~81% of them) had a statistically significant increasing trend according to the Mann–Kendall test at 5%. The datasets used or cited in this study (MapBiomas, PRODES, and DETER) enable the monitoring of the current status of ILs, and the identification of emerging trends related to illegal activities. Therefore, they are critical tools for legal authorities.

Flash floods are critical events for emergency management, yet their modeling remains highly challenging, even in smart cities approaches. Physically based hydrological models are often unsuitable at small spatiotemporal scales due to their computational complexity and dependence on detailed local parameters, which are rarely available during flash floods. With the growing availability of hydrological data, machine learning (ML) has emerged as a promising alternative. This work performs a Systematic Literature Review (SLR) to improve our understanding of the research landscape on ML applications for flash flood forecasting, a significant subset of flash flood modeling. From more than 1,200 papers published until January 2024 in Web of Science, SCOPUS/Elsevier, Springer/Nature, and Wiley, 50 were selected following PRISMA guidelines. The inclusion and exclusion criteria removed reviews, retractions, papers focused on post-flood damage assessment (not forecasting), and those with time resolutions of 6 hours or more, retaining only studies with fine-scale temporal data (<6 hours). For each paper, we extracted information on forecasting horizon, study area size, input data, ML techniques, and outcomes (regression or classification). Results show a sharp rise in ML-based flash flood research, with China leading (38%). Nearly all studies rely on rainfall, discharge, and water level data - often in combination. Long short-term memory (LSTM) networks dominate (60%). Unfortunately, only 10% of the selected studies provide access to their datasets. This lack of transparency poses a major barrier to reproducibility, inhibits fair comparative evaluation of models, and ultimately slows methodological progress in flash flood forecasting. Furthermore, our review highlights that no method consistently outperforms others. This variability in performance is likely influenced by factors such as regional hydrological characteristics (e.g., differences between arid and tropical basins), variations in input data quality, and the length of the forecast horizon (e.g., 1- vs. 6-hour prediction). Lastly, we recommend advancing this field through integration with early warning systems, creation of benchmarks, open data practices, and stronger multidisciplinary collaboration.

Over the decades, hydropower complexes have been built in several hydrographic basins of Brazil including the Amazon region. Therefore, it is important to understand the effects of these constructions on the environment and local communities. This work presents a land use and land cover change temporal analysis considering a 33-year period (1985–2018) in the direct influence zone of the Braço Norte Hydropower Complex, Brazilian Amazonia, using the Collection 4.1 level 3 of the freely available MapBiomas dataset. Additionally, we have assessed the Brazilian Amazon large-scale deforestation process acting as a land use and land cover change driver in the study area. Our findings show that the most impacted land cover was forest formation (from 414 km² to 287 km², a reduction of 69%), which primarily shifted into pasturelands (increase of 664%, from 40 km² to 299 km²). The construction of the hydropower complex also triggered indirect impacts such as the presence of urban areas in 2018 and the consequent increased local demand for crops. Together with the ongoing large-scale Amazonian deforestation process, the construction of the complex has intensified changes in the study area as 56.42% of the pixels were changed between 1985 and 2018. This indicates the importance of accurate economic and environmental impact studies for assessing social and environmental consequences of future construction in this unique region. Our results reveal the need for adopting special policies to minimize the impact of these constructions, for example, the creation of Protected Areas and the definition of locally-adjusted parameters for the ecological-economic zoning considering environmental and social circumstances derived from the local actors that depend on the natural environment to subsist such as indigenous peoples, riverine population, and artisanal fishermen.

Current forecasts estimate that almost 68% of the global population will be living in urban centers by 2050. As a result, the increase in impermeable surface area can result in severe hydrological impacts, such as the increase in surface runoff and the frequency of floods and their magnitude. Thus, this work analyzes the performance of the hydrodynamic model HEC-RAS for assessing the extent of flood-prone areas, using two digital terrain models (DTM) with different spatial resolutions (5 and 0.50 m). Four different computing intervals (1, 15, 30, and 60 s) were adopted aiming to evaluate the simulations outputs performance. Additionally, reported data by the civil defense are used for calibration and validation. In general, the model showed to be a powerful tool in the identification of susceptible areas to urban flooding. The simulated results in this work provide crucial geographic information when identifying spots with the highest risk of flooding, which should receive priority attention during such events. The simulations with a spatial resolution of 5 m showed the flood maps with the largest coverage of the flooded points (278 points out of 286—97.20%), within the shortest computation times. We highlight that the more refined DTM derived from spatial images did not produce the best flood simulation compared to the DTM with a spatial resolution of 5 m derived from orthoimages.

The great majority of the world population resides nowadays in urban areas. Understanding their physical and social structure, and especially their urban land use pattern dynamics throughout time, becomes crucial for successful, effective management of such areas. This study is committed to simulate and predict urban land use change in a pilot city belonging to the São Paulo Metropolitan Region, southeast of Brazil, by means of a cellular automata model associated with the Markov chain. This model is driven by data derived from orbital and airborne remotely sensed images and is parameterized by the Bayesian weights of evidence method. Several layers related to infrastructure and biophysical aspects of the pilot city, São Caetano do Sul, were used as evidence in the simulation process. Alternative non-stationary scenarios were generated for the short-run, and the results obtained from past simulations were statistically validated using a multiresolution “goodness-of-fit” metric relying on fuzzy logic. The best simulations reached fuzzy similarity indices around 0.25–0.58 for small neighborhood windows when an exponential decay approach was employed for the analysis, and approximately 0.65–0.95 when a constant decay and larger windows were considered. The adopted Bayesian inference method proved to be a good parameterization approach for simulating processes of urban land use change involving multiple classes and transitions.