A slowdown in a major Atlantic Ocean circulation could make California’s atmospheric rivers wetter by the end of this century, even though the storms form over the Pacific. That is the result of a climate-model experiment led by researchers at the University of California, Riverside.
In a 2026 paper in Nature Communications, lead author Mohima Sultana Mimi and colleagues compared a future in which the Atlantic Meridional Overturning Circulation, or AMOC, weakens with a counterfactual simulation in which it is held near its present strength. Under the high-emissions scenario they tested, AMOC weakening increased the frequency of atmospheric rivers reaching the west coast of North America and added substantially to the rain they delivered in California.
This is a simulation result. It describes what one climate model does under a particular emissions pathway, not a certain forecast of California’s weather.
This is also one study, not settled consensus. Its more interesting contribution is the mechanism it identifies: cooling in the North Atlantic can rearrange high-altitude atmospheric circulation far enough downstream to strengthen Pacific winds that steer moisture towards California.
The study compared two versions of the same future
The research team used the Community Earth System Model version 2, known as CESM2. In one six-member set of simulations, greenhouse gas concentrations followed the high-emissions SSP5-8.5 pathway and the AMOC was allowed to weaken naturally. The researchers called this the free-AMOC case.
They then ran another six-member ensemble with the same external climate forcing but manipulated freshwater in the North Atlantic to keep the AMOC close to its strength around the year 2000. Comparing the two groups gave the team a way to isolate the circulation’s influence from the broader effect of a warmer atmosphere.
The paper compares conditions near the end of the century, from 2076 to 2100, with a historical reference period of 1990 to 2014.
That experimental design is the paper’s strength. Climate change affects atmospheric rivers in several overlapping ways, including by allowing warmer air to hold more water. Holding the AMOC steady in one version of the model lets the researchers ask a narrower question: what changes because this particular ocean circulation slows?
The fivefold figure needs careful reading
For California as a whole, atmospheric-river precipitation increased by 0.07 meters, or 7 centimeters, per year in the free-AMOC simulation. In the fixed-AMOC case, the increase was 0.01 meters, or 1 centimeter. The authors describe this as the weakened circulation amplifying the projected increase by more than fivefold.
That does not mean each atmospheric river becomes five times stronger.
The comparison concerns the change in annual precipitation attributed to atmospheric rivers across the state, averaged over a 25-year future period. It combines changes in how often these systems occur with how much rain they produce. The model found that AMOC weakening raised winter atmospheric-river frequency along the west coast of North America by as much as 5.21 percent.
In wintertime California, the modeled AMOC contribution to atmospheric-river precipitation averaged 0.18 meters, or 18 centimeters, per year and reached 44 centimeters in some locations. Those are large additions, but they remain outputs from CESM2 at roughly one-degree atmospheric and ocean resolution. The authors explicitly call for similar experiments in higher-resolution models.
How Atlantic cooling reaches the Pacific
The AMOC carries warm surface water north through the Atlantic and returns colder, denser water south at depth. As greenhouse warming and added freshwater weaken that overturning, less heat is transported northward and a relatively cool region develops in the North Atlantic.
The model links that cooling to changes in temperature gradients, storm tracks and westerly winds. It also produces a stationary Rossby wave train, a sequence of large atmospheric pressure disturbances that carries the influence eastward across Eurasia and into the North Pacific.
Over the Pacific, the resulting westerly jet becomes stronger. Atmospheric rivers are long, narrow corridors of concentrated water vapor, and their movement depends heavily on winds. Stronger prevailing winds make more of them reach the North American coast, especially in winter.
What I find most useful here is that the paper does not suggest Atlantic water somehow travels to California and feeds the storms. The connection is atmospheric. A circulation change in one ocean alters the planet-scale pattern of winds that governs weather over another.
California depends on the storms it fears
Atmospheric rivers are not simply disasters. According to the U.S. Geological Survey, they can supply up to half of California’s annual precipitation. A study of West Coast drought endings found that landfalling atmospheric rivers ended roughly 33 to 40 percent of persistent California droughts examined between 1950 and 2010.
The same systems cause much of the state’s worst flooding. A small number of very wet hours can determine whether a year ends with adequate reservoirs or overflowing rivers. When heavy rain lands on saturated soil, burn scars or a warm snowpack, the consequences can include flash floods, landslides and debris flows.
California already pays for both ends of that volatility. NOAA’s historical billion-dollar disaster database lists 14 drought events and six flooding events affecting the state from 1980 through 2024, each causing at least $1 billion in inflation-adjusted losses.
More atmospheric-river rain is therefore not equivalent to a simple gain in water supply. Reservoir capacity, storm sequence, freezing level, soil conditions and where the rain falls determine whether the water can be stored or becomes a hazard.
The UC Riverside release about the study points to forecasting and expanded storage as possible ways to handle additional water. The model itself did not test reservoirs, flood-control systems or groundwater recharge, so those are adaptation implications rather than findings from the simulation.
The paper models a slowdown, not a collapse
The AMOC’s future strength remains uncertain. Climate models broadly project weakening under continued warming, but they disagree about its size, and the question of whether a full collapse is possible this century remains contested.
The California result does not require a collapse. It comes from a gradual weakening in one model under SSP5-8.5. A 2025 study in Nature found that the AMOC weakened but did not collapse under extreme forcing across 34 climate models, partly because Southern Ocean upwelling kept some overturning going. That paper did not eliminate uncertainty about the amount of weakening.
The title’s strongest claim is best read as shorthand for an extra source of intensification, not as a measured property of every future storm. Emissions, internal climate variability, the pace of AMOC weakening and the way a model represents narrow weather systems all affect the size of the result.
The next useful test is clear: repeat the free-versus-fixed AMOC experiment across several climate models and at resolutions fine enough to represent the most extreme atmospheric rivers and California’s coastal mountains more realistically. That would show how much of the projected increase belongs to AMOC weakening, and how much belongs to the particular structure of CESM2.
