The Amazon basin, a vast and vital ecosystem, is experiencing increasingly severe droughts exacerbated by climate change. This poses significant challenges, particularly for the millions of people living in the region, many of whom depend on river systems for transportation, livelihoods, and access to essential resources. Rivers serve as the primary means of medium- to long-distance transportation in the Amazon, connecting communities and facilitating trade. However, the intensification of the hydrological cycle, characterized by more frequent and intense droughts and floods, directly threatens this reliance on river transport. The unprecedented 2023 drought, resulting in fish kills, river dolphin mortality, disrupted navigation, and widespread wildfires, serves as a stark warning of the potential consequences. Hydrological droughts, defined as sustained periods of below-normal water levels, severely impact inland navigation by reducing vessel speeds, limiting access to ports, and requiring reduced vessel loads. In extreme cases, navigation can be completely halted, leading to the complete isolation of villages for extended periods. This isolation results in shortages of essential goods, such as food, fuel, and medicine, and limits access to healthcare and education. Previous research has highlighted various impacts of droughts on Amazonian communities, but a systematic, spatially explicit analysis at the scale of the entire basin has been lacking. This study addresses this gap by examining the impacts of droughts from 2000 to 2020 on the Brazilian Amazon, focusing on the consequences of reduced accessibility and mobility due to disrupted river navigation. The study leverages multiple data sources, including hydrological data from river gauges and information gleaned from media reports, to provide a comprehensive understanding of these impacts.

Existing literature has documented various impacts of droughts on Amazonian riverine communities (“ribeirinhos”). Studies have described the consequences of droughts on these communities, including impacts on their access to resources and services. However, these studies often focused on specific case studies or lacked the broad spatial scope needed to understand the basin-wide implications. Other studies have concentrated on the hydrological impacts of droughts, emphasizing the crucial role of inland water transport for local populations. While some work has described the various impacts of droughts on riverine communities, a systematic, spatially explicit analysis of these impacts at the scale of the Amazon basin has been absent. This study seeks to fill this gap, building on existing research by providing a more comprehensive spatial and temporal assessment of the impacts of hydrological droughts on communities within the Brazilian Amazon.

This study integrates three primary approaches to assess the impact of extreme droughts (2000-2020) on communities in the Brazilian Amazon. **1. Spatial Analysis of Settlement Proximity:** The study first analyzed the proximity of both Indigenous and non-Indigenous settlements to major water bodies and roads. This involved calculating the shortest straight-line distances from settlements to water bodies (during high and low water seasons) and roads using GIS techniques and data from various sources such as the Brazilian Institute of Geography and Statistics (IBGE) and the National Foundation for Indigenous Peoples (FUNAI). The analysis differentiated between high and low water seasons to account for seasonal variations in river accessibility. The analysis identified settlements that are highly reliant on river transport, especially during low-water periods. **2. Analysis of Digital News Media:** To gather information on the impacts of droughts, the researchers collected and analyzed news articles from digital media outlets. This involved using web scraping tools and manual screening of articles focusing on droughts, navigation difficulties, and community isolation in the Brazilian Amazon. A total of 70 news articles yielded 142 statements detailing drought impacts. These statements were categorized based on the types of activities and services affected, and the spatial and temporal distribution of these impacts were mapped. The inclusion criteria for the articles ensured that only relevant information was retained, addressing the inherent limitations of this type of data source. **3. Hydrological Data Analysis:** The study also analyzed hydrological data from 90 river gauges across the Brazilian Amazon. The researchers focused on the period from 2000 to 2021, selecting gauges located in areas mentioned in the news articles. Time series of water levels were analyzed to determine the duration of low-water periods in each hydrological year. The 80th percentile (P80) of the long-term water level duration curve was used as a threshold to define low-water levels. The analysis then compared the duration of low-water periods during drought years to the median duration across the entire historical time series. Data preprocessing involved filtering out gauges significantly influenced by dams and handling data gaps in the time series. The methodology carefully considered data limitations, such as the accuracy of inundation maps and the presence of data gaps in hydrological records, employing conservative approaches to minimize biases in the analysis.

The study's key findings reveal a strong correlation between severe droughts and negative impacts on Amazonian communities. **Spatial Distribution of Vulnerable Communities:** The analysis of settlement proximity reveals that a significant portion (89% of non-Indigenous localities and 92% of Indigenous villages) are within 5 km of major water bodies during high-water season, highlighting their strong dependence on river transport. However, this reliance becomes a vulnerability during low-water seasons, especially for settlements further than 5 km from roads, where accessibility is significantly reduced. The study pinpoints regions in the states of Amazonas, Acre, Pará, and Roraima as particularly vulnerable to isolation during droughts. **Temporal Distribution of Drought Impacts:** The analysis of news articles shows that the droughts of 2005, 2010, and 2016 were the most severe in terms of reported impacts on local populations, corroborating findings from previous hydrological studies. These years are associated with anomalously low rainfall and the longest low-water periods, significantly disrupting inland water transport. October, September, and August were identified as months with the most reported impacts. **Cascading Impacts of Droughts:** The study demonstrates the cascading effects of droughts on various aspects of life for riverine communities. Disruptions to inland water transport directly impacted the supply of essential goods (food, fuel, medicine), access to services (healthcare, education), and economic activities (fishing, tourism). The analysis of news articles reveals a wide range of impacts across sectors, with the most frequent mentions being transportation problems, food shortages, and community isolation. **Hydrological Data Corroboration:** The hydrological data analysis confirmed the severity of the droughts of 2005, 2010, and 2016, revealing that low-water periods were extended by an average of 36.8 days compared to the median across the historical record, reaching more than 100 days in some cases. This prolonged period of low water levels directly contributes to the observed impacts on communities. **Specific Impacts:** The study detailed impacts of the droughts on various aspects of life and livelihoods of the riverine communities. The most frequent negative consequences included difficulties accessing healthcare, disruptions to education, shortages of food and essential goods, increasing transportation costs, and prolonged community isolation. The droughts also impacted economic activities such as fishing, with some reports indicating both increased opportunity and reduced accessibility due to changes in fish distribution and water quality.

This study provides crucial insights into the vulnerability of Amazonian communities to severe droughts, filling a significant gap in understanding the impacts of these events. The findings highlight the critical dependence of many communities on river navigation, which is severely disrupted during periods of prolonged low water. The study’s multi-methodological approach, combining spatial analysis, media analysis, and hydrological data, strengthens the conclusions by providing a robust assessment of both the spatial and temporal distribution of impacts. The study confirms the severity of past droughts and underscores their cascading effects on various sectors, from the supply of essential goods to access to services and economic opportunities. The spatially explicit analysis is especially valuable in pinpointing the most vulnerable communities and regions, providing crucial information for targeted interventions. The strong link between prolonged low-water periods and the observed societal impacts emphasizes the urgent need for adaptive strategies. The integration of data from diverse sources increases confidence in the findings and highlights the importance of employing multi-faceted methodologies in studying the impacts of climate change on vulnerable populations.

This research demonstrates the significant and widespread impacts of severe droughts on riverine communities in the Brazilian Amazon, particularly regarding disruptions to river navigability and resulting community isolation. The findings highlight the urgent need for proactive and comprehensive adaptation strategies that address the cascading effects of droughts across multiple sectors. Future research should focus on more detailed local case studies, expanding the study to encompass the entire Amazon basin, and incorporating future climate projections to assess the potential magnitude of future impacts. Collaborative, transdisciplinary approaches that integrate local knowledge with scientific expertise are essential for developing effective and equitable solutions.

This study has several limitations. The accuracy of the spatial analysis is constrained by the resolution of inundation maps and the completeness of road datasets, particularly concerning less formal road networks. The analysis of news articles relies on media coverage, which might not capture the full extent of impacts in all areas. Data gaps in hydrological time series limit the analysis of water level duration curves for some gauges. The study focuses solely on the Brazilian Amazon and may not fully reflect the diverse experiences of communities across the entire Amazon basin. Finally, using web scraping has its shortcomings because it limits the amount of articles searched and may not identify all relevant articles. Future studies should strive to overcome these limitations.