Deep inside the Amazon Basin, invisible rivers move through the sky. Vast columns of moisture — evaporated from the forest canopy and exhaled through billions of leaves — travel thousands of kilometres westward, watering croplands and cities across the South American continent. Scientists call them flying rivers. A landmark new study published in Nature Communications Earth and Environment has finally put a price tag on them: $20 billion every year.
The research analysed the rainfall generated by tropical forests using high-resolution atmospheric modelling and estimates that each square metre of Amazon forest contributes between 240 and 300 litres of rainfall annually to surrounding regions. Across the entire rainforest — an area roughly the size of the contiguous United States — that adds up to a freshwater subsidy of staggering proportions, delivered entirely free of charge to farms, hydropower reservoirs, and municipal water systems that have never been asked to account for it.
"We have been drawing on this service for centuries without paying for it," the study's authors note. "The question now is whether we can use this knowledge to protect what remains before it is gone."
The Science of Flying Rivers
The mechanism behind this service is both elegant and precarious. Through a process called evapotranspiration, trees pull groundwater upward through their root systems and release it as water vapour through their leaves. In the Amazon, this process operates at such a massive scale that the forest essentially creates its own weather — generating rainfall not just locally but hundreds of kilometres downwind. Brazil's agricultural heartland in the Mato Grosso, the farms of Bolivia and Peru, and the snowpack feeding Andean glaciers all depend partly on this moisture-recycling engine.
The flying rivers concept is not new — Brazilian atmospheric scientist Eneas Salati first identified the phenomenon in the 1970s — but the Nature Communications study represents one of the most comprehensive economic valuations of the service to date. By combining rainfall data with agricultural and hydrological models, the researchers translated a physical process into the language that policy makers and finance ministries actually respond to: money.
What makes the new valuation politically explosive is its timing. The Amazon has lost roughly 17 percent of its original forest cover, with deforestation rates fluctuating with each change of government in Brasília. Climate scientists warn the forest is approaching a tipping point — estimated at around 20 to 25 percent loss — beyond which the flying rivers could weaken irreversibly, triggering a feedback loop of drying that accelerates further forest die-back. The $20 billion figure is not a fixed asset. It is a declining one.
The Policy Gap: Pricing What Markets Ignore
For climate negotiators and forest economists, the study arrives as a critical input into one of the most contested debates at international climate talks: payments for ecosystem services. The concept — that nations or companies benefiting from forest-derived rainfall, carbon sequestration, or biodiversity should pay the countries that maintain those forests — has circulated through COP negotiations for decades. It has produced frameworks including REDD+ (Reducing Emissions from Deforestation and Forest Degradation), which channels climate finance toward forest preservation.
Yet the sums channelled through such mechanisms remain a fraction of what the science now suggests the Amazon's services are worth. Brazil received approximately $1 billion through REDD+ mechanisms across the last decade — against a rainfall service the new study values at $20 billion annually. The arithmetic of underpayment is stark.
Paulo Moutinho, a senior researcher at the Amazon Environmental Research Institute (IPAM), has long argued that the international community fundamentally undervalues what standing forests provide. "When we talk about deforestation, we tend to frame it as a biodiversity loss or a carbon loss," he has noted. "But the hydrological service is equally critical — and it crosses borders in ways that make it a genuinely global public good."
The transboundary nature of flying rivers creates both the strongest argument for international payment mechanisms and the greatest governance challenge. Argentina, Paraguay, and Uruguay benefit from Amazon-derived rainfall but bear none of the costs of forest protection. Agricultural exporters in São Paulo state depend on the moisture recycled from forests in Pará or Amazonas that their own industry's expansion may be helping to destroy.
The Monetisation Dilemma
Not all climate advocates are comfortable with the direction this analysis points. Assigning a dollar value to the forest's rainfall service is intellectually satisfying and politically useful — but it raises a question that the study's discussion section directly acknowledges: does monetising nature protect it, or does it merely make it tradeable?
The history of carbon markets offers a cautionary tale. Emissions trading systems designed to put a price on carbon have in some cases enabled continued pollution by allowing companies to buy offsets rather than reduce emissions directly. Critics worry that a payments-for-rainfall framework could follow the same trajectory — allowing wealthy nations to purchase credit for the Amazon's services without addressing the underlying economic pressures driving deforestation: cattle ranching, soy agriculture, illegal logging, and inadequate land rights for Indigenous communities.
Indigenous peoples' organisations, which represent communities that have protected the largest remaining forest areas at significant personal risk, have consistently argued that any payment system must centre their land rights, not peripheral financial flows to distant governments. Studies consistently show that Indigenous-managed territories in the Amazon have substantially lower deforestation rates than equivalent areas under other governance regimes — a fact the new study's policy recommendations implicitly acknowledge.
Urgency at the Negotiating Table
The $20 billion figure lands in a climate finance landscape already under acute strain. Developed nations have repeatedly failed to deliver on the $100 billion annual climate finance pledge made at Copenhagen in 2009. Against that backdrop, advocates hope the new valuation shifts the terms of negotiation by making the cost of inaction legible in economic language that finance ministries and trade negotiators understand.
"The forest has always been providing this service," said one climate economist familiar with the study. "What we now have is a receipt."
In climate policy, as across environmental challenges, urgency must meet solutions — science demands action, but despair achieves nothing. The Amazon's flying rivers have sustained a continent for millennia without acknowledgment or compensation. The question before negotiators, finance ministers, and heads of government is whether the world will act on what the science has now made undeniable — before the rivers stop flowing.
