In October 2023, the Rio Negro at Manaus dropped to its lowest point since Brazilian authorities began measuring it in 1903. Boats sat stranded on exposed sandbars. Entire stretches of river that normally carried cargo and connected remote communities turned to cracked mud. Then, instead of a full recovery, the 2024 dry season drove water levels down again, compounding a moisture deficit that peer-reviewed research now calls the worst in at least a century of record-keeping.
A study published in Communications Earth & Environment found that the back-to-back droughts stripped the Amazon basin of more terrestrial water storage than any comparable period in the modern observational record. The researchers traced how elevated temperatures supercharged evaporation, pulling moisture from soils at rates that far exceeded what rainfall shortfalls alone would explain. That desiccation then fed back into the drought itself: parched soils released less moisture into the atmosphere, weakening the recycling loop that normally helps generate the Amazon’s own rainfall.
Scientists say the severity cannot be chalked up to natural climate cycles. A rapid attribution analysis conducted by World Weather Attribution and led by researchers at Imperial College London concluded that human-caused warming, not the El Nino pattern active at the time, was the dominant force behind the 2023 drought. Using agricultural drought metrics such as the Standardised Precipitation-Evapotranspiration Index (SPEI), the team found that drought of this magnitude was roughly 30 times more likely in today’s climate than it would have been without anthropogenic warming. “The fingerprint of climate change on this drought is unmistakable,” said Friederike Otto, a senior lecturer in climate science at Imperial College London and a co-lead of the World Weather Attribution initiative, in a statement accompanying the analysis. In model simulations of a preindustrial world, comparable conditions were far rarer and far less intense.
A record that stretches back 120 years
The Rio Negro gauge at Manaus is one of the longest continuous hydrological records in the tropics. When researchers describe the 2023 reading as record-breaking, they are comparing it against every annual minimum logged since 1903, a dataset maintained by Brazil’s national water agency and corroborated by NOAA. Satellite imagery from NASA’s Earth Observatory captured the visible toll: sandbars where channels once ran, and vessels grounded far from navigable water.
The crisis was not confined to a single river. Gauges along the Madeira, Solimoes, and upper Amazon reported low flows and navigation constraints during the 2023 dry season, though not every site hit record levels. Taken together, the readings pointed to a basin-wide shortfall in water availability across central and southern Amazonia.
Satellite gravity instruments added depth to that picture. Data from NASA’s GRACE-FO mission and the European Space Agency’s Swarm constellation allowed researchers to track how the drought propagated from surface soils into deeper groundwater reserves. Maps of groundwater storage percentiles showed large swaths of the basin falling into extreme deficit categories during both 2023 and 2024, confirming that the damage reached well below the surface.
Deforestation weakened the basin’s defenses
The back-to-back droughts did not strike a pristine landscape. Decades of forest clearing had already degraded the Amazon’s ability to generate its own rainfall through a process scientists call moisture recycling, sometimes described as “flying rivers” of humidity that move westward across the canopy.
A study published in Nature Communications analyzed precipitation trends from 1980 to 2019 and attributed a majority of the observed rainfall decline in the southern Amazon to deforestation. Forest clearing reduces the volume of water that trees transpire back into the atmosphere, weakening the atmospheric moisture transport that sustains wet-season rains. That structural vulnerability means the basin was already operating with a thinner margin of safety when the 2023 drought arrived, and the 2024 follow-up hit a system that had not yet recovered.
Fires, fish kills, and stranded communities
On the ground, the consequences were severe. Fire activity surged across several Brazilian states as desiccated vegetation ignited. Shrinking lakes and flooded forests saw degraded water quality, and reports of fish kills emerged where water temperatures climbed and dissolved oxygen plummeted. Remote-sensing products tracking vegetation health detected patches of reduced greenness and canopy stress, particularly along deforested frontiers and in areas that had already endured repeated droughts over the prior decade.
Indigenous and riverine communities faced some of the most direct hardships: water shortages, collapsed fishing grounds, and severed access to markets and health services as rivers became impassable. Those experiences remain documented mainly through news coverage and NGO reports rather than systematic social-science surveys, leaving a gap in the formal record of who was hit hardest and which coping strategies proved effective.
What researchers still do not know
For all the data now available, several critical questions remain open. No published, peer-reviewed survey has established localized tree die-off rates across the basin, so the long-term ecological toll of the back-to-back droughts is still unclear. Field researchers have described increased leaf shedding and branch loss, but without systematic plot-based measurements, those observations are difficult to scale up.
Economic losses also lack an authoritative tally. The Brazilian government has not released an official cost assessment for either drought year, and estimates of agricultural and transportation damages vary widely. River navigation disruptions stranded cargo and isolated communities, but the full financial impact on regional supply chains has not been independently calculated.
The precise interplay between El Nino and anthropogenic warming during the 2023 event is still debated. The World Weather Attribution analysis led by Imperial College London placed human-caused warming as the primary driver, but other analyses have highlighted sea-surface temperature anomalies in the tropical Pacific and Atlantic that steered moisture away from the basin and altered atmospheric circulation. Reconciling those competing signals remains an active area of research.
Perhaps the most consequential unknown is how quickly the basin can recover. GRACE-FO data clearly show how terrestrial water storage declined, but the pace at which groundwater and deep soil moisture will recharge is uncertain. In heavily deforested or intensively farmed zones, soil compaction and land-use change may slow infiltration, prolonging hydrological deficits even if rainfall returns to normal. Whether the Amazon’s water cycle will rebound or settle into a drier baseline is a question that may take years of monitoring to answer.
Why it matters now
As of spring 2026, the Amazon basin faces another dry season with its reserves already depleted. The convergence of evidence from river gauges, satellite gravity measurements, and formal attribution studies points to a clear conclusion: the recent drought sequence cannot be explained by natural variability alone. Human-driven warming and decades of deforestation have raised the baseline of risk, making once-rare extremes more frequent and more punishing.
The remaining uncertainties, about ecosystem mortality, social impacts, and the exact balance of climate drivers, do not weaken that core finding. They do, however, underscore how much work remains before scientists, policymakers, and the communities who depend on the Amazon’s rivers can fully grasp what has already been lost and what may still be at stake.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
