The Question
You cannot see it from your window, but on a stormy winter day a river may be flowing directly over your head. Not a river of water you could touch — a river of vapor, a ribbon of moist air hundreds of miles wide and thousands of miles long, streaming out of the warm tropics toward colder land. Scientists call these atmospheric rivers. When one slams into a coastline and rises over mountains, all that vapor squeezes out as rain and snow, sometimes for days.
Here is the number that makes people blink. A strong atmospheric river can transport up to 15 times the amount of water that flows out of the mouth of the Mississippi River — through the sky, in a single band of air. The Pacific version that soaks California has a nickname: the "Pineapple Express," because it draws moisture from near Hawaii. And here is the twist that makes forecasting them so urgent: the very same storms that drown towns also fill the reservoirs that keep those towns alive. The question is whether we can learn to read these rivers precisely enough to catch the water without drowning in it.
What the Evidence Shows
Atmospheric rivers are both the West Coast's greatest water gift and its greatest flood threat, and that dual nature is now measured, not guessed. Researchers at Scripps Institution of Oceanography estimate these systems deliver a large share — often cited as up to half — of California's total water supply in a typical year. Yet they are also the West's main flood threat. A 2019 study in Science Advances, led by Tom Corringham at Scripps, used 40 years of National Flood Insurance Program data and found that more than 84% of the nearly $51 billion in flood damage in Western states was caused by atmospheric rivers — about $1 billion a year — with each step up the intensity scale bringing roughly ten times more damage. When several arrive back to back, the ground saturates, rivers overtop and hillsides slide.
To make the danger legible, scientists built a scale. In 2019, a team led by Marty Ralph at Scripps introduced a five-step rating — AR1 to AR5 — that works much like the categories used for hurricanes. A low-end AR1 is "mostly beneficial," a welcome drink of rain. A high-end AR5 is "mostly hazardous," a days-long deluge. The winter of 2023 showed the top of the scale in action: a relentless parade of atmospheric rivers struck California, and by one widely reported estimate dumped more than 32 trillion gallons of water on the state in about three weeks from late December 2022, killing at least 20 people while refilling reservoirs and ending the state's worst drought categories.
"A small number of extreme ARs cause most of the flood damages in the West, and even modest increases in intensity could significantly increase their impacts."
— Tom Corringham, Scripps Institution of Oceanography, lead author of a 2019 study on atmospheric river flood damageThe forecasting is improving fast because scientists now fly straight into these storms. Under a program called AR Recon, aircraft cross the Pacific during atmospheric river events and release dropsondes — small sensor packages that parachute down through the storm, radioing back readings of temperature, moisture, and wind on the way. Those measurements, fed into weather models, have measurably sharpened predictions of where and how hard an atmospheric river will hit. That new precision is already changing how a single dam is run at Lake Mendocino in California, under an approach called Forecast-Informed Reservoir Operations, where managers release or hold back water based on what the incoming storms are forecast to do.
The same storm that saves a city and the one that drowns it are separated by a single day of forecast error.
Why This Is Happening
The stakes are simply enormous, on both sides of the ledger. Water scarcity and flood disaster are two of the costliest problems the American West faces, and atmospheric rivers sit at the heart of both. Any tool that shifts the odds — capturing more supply, dodging more damage — pays for itself many times over. That economic gravity pulls research money, aircraft, and political attention toward the problem year after year.
The science matured just as the observing tools arrived. The AR category scale gave forecasters and the public a shared language, the same way hurricane categories did decades ago. At the same time, reconnaissance flights and better satellites started filling the ocean's data gaps. When a clear framework meets fresh measurements, forecast skill tends to climb quickly — that is exactly the path hurricane tracking followed.
It is not only a Californian story, which spreads the effort. Atmospheric rivers strike the United Kingdom and Western Europe too, driving major floods there, and they matter for the Pacific Northwest, Chile, and beyond. Because so many regions share the threat, research collaborates across borders, and improvements in one place quickly benefit others. A global problem attracts a global scientific push.
What Could Happen
Category ratings appear in ordinary weather reports, reconnaissance flights become standard, and models predict landfall and intensity days ahead with hurricane-grade skill. Reservoirs across the West are run on forecasts, capturing more water and pre-releasing before floods. The public learns to read AR1-to-AR5 the way it now reads storm categories.
Forecasters nail the extreme AR4 and AR5 storms that threaten cities, yet the finer details — exactly which valley floods, how much a given town gets — stay fuzzy. Warnings improve lives and reservoirs, but neighborhood-level precision remains just out of reach, arriving later than 2035.
Reconnaissance programs lose budget, or the atmosphere's chaos caps how far ahead these storms can be predicted. Forecasting improves modestly but never reaches hurricane-level routine. This is the disappointing path — possible, but against the momentum already built into the science.
What Can We Do
You cannot steer a river in the sky, but you can learn to live with one — and the coming forecasts only help people who know to use them.
Learn the AR scale before it appears in your forecast. Just as you know a Category 4 hurricane is serious, get familiar with what an AR4 or AR5 means: a prolonged, dangerous soaking. When these ratings enter mainstream weather reports, you will already understand the warning, and you will act sooner than a neighbor who is hearing the term for the first time.
Know your local flood and landslide risk now. Atmospheric rivers do their worst damage on saturated ground and steep slopes. Find out whether your home sits in a floodplain or below an unstable hillside, and make a plan for evacuation and documents. The forecast buys you time only if you already know what to do with it.
Support forecast-informed water management. The Lake Mendocino model — using storm forecasts to decide when to store and when to release — turns a flood risk into a water supply. Back its expansion to other reservoirs in your region. It is one of the rare policies that fights drought and flooding at the same time, using nothing but better information.
Value the flights and satellites that make it possible. Reconnaissance missions and Earth-observing satellites are the unglamorous backbone of this progress. When science budgets are debated, remember that a single well-forecast atmospheric river can save more than the program costs in a year. Public support keeps the planes flying into the storm.
- NOAA — What are atmospheric rivers?
- ScienceDaily (UC San Diego) — Atmospheric river storms create $1 billion a year flood risk (Corringham et al., Science Advances, 2019)
- CW3E, Scripps — Water Year 2019 atmospheric rivers rated on the new AR Scale
- CW3E, Scripps — Atmospheric River Reconnaissance (AR Recon) overview
- CW3E, Scripps — Forecast-Informed Reservoir Operations (FIRO)
- FOX Weather — California atmospheric rivers drop 32 trillion gallons of water over 3 weeks, January 2023