By Robin Meadows
In the late 1990s, hardly anyone had heard of the storms called atmospheric rivers. That includes Marty Ralph, founding director of Center for Western Weather and Water Extremes (CW3E) and a leading expert on these relatively recently recognized streams of water vapor in the sky.
Atmospheric rivers are long, narrow plumes of exceptionally wet air that shoot across the ocean and drop rain or snow when they hit land. These storms deliver much of the precipitation and cause most of the flooding in the Western U.S., with economic damages averaging $1.5 billion a year.
Ralph was already investigating the storms now known as atmospheric rivers back in the ‘90s. He just didn’t realize it until he read the paper that coined their name.
“The light bulb literally went off: that’s what I’d been studying,” he recalls. “It was published in 1998 and I came across it in 2003, and it changed the course of my work and helped create what we’re doing today at CW3E.”
CW3E, based at the Scripps Institution of Oceanography at the University of California, San Diego, aims to boost resilience to extreme weather events that cause flooding and affect the water supply. Partners include the California Department of Water Resources, the National Weather Service, and the U.S. Army Corps of Engineers.
Today, atmospheric rivers are widely known. And, sparked in large part by Ralph’s initial curiosity about them, these storms can now be forecast far enough in advance to help protect people and property from floods as well. Atmospheric river forecasts also help preserve the water supply via a new approach to managing reservoirs that is fittingly called Forecast Informed Reservoir Operation (FIRO).
Advances in atmospheric river forecasting are the result of the federally-required aerial Atmospheric River Reconnaissance (AR Recon) program, which is led by Ralph and Vijay Tallapragada of the National Oceanic and Atmospheric Administration (NOAA). AR Recon includes flights that release weather instruments into storms to predict where, when, and how much rain will fall.
Ralph recently traced the arc of atmospheric river science and solutions in the keynote address at the 2024 California Extreme Precipitation Symposium, where the theme was Anticipating and Planning for California Floods.
To learn more, Robin Meadows spoke with Ralph about the satisfaction of doing basic research that makes a difference in people’s lives, the importance of observational weather studies, and future research needs. This conversation has been edited for conciseness and clarity.
What drew you to meteorology?
Marty Ralph: As a young teen, I moved from Tucson to Nogales, Arizona, which is just over the border from Mexico. You might expect Nogales to be dry in the summer but the monsoon makes it much wetter than Tucson. You can see from space how lush Nogales is during the summer.
We lived on a hill so I could see the tops of thunderclouds 50 miles away. Right behind our house was the Santa Cruz River, which was normally dry, and I watched thunderstorms form and get closer, and eventually the river was full. It flooded up to six feet deep during the summer. The land went from brown to green in a couple of weeks.
The smells, sights and sounds of these storms were an earthy experience that inspired my interest in precipitation and flooding. I’ve been hooked on this stuff ever since! I still make summer visits to Nogales to see the monsoons.
Why did you start studying atmospheric rivers before you really knew what they were?
I knew from personal experience that the storms we now know are atmospheric rivers were super productive for rain. As a graduate student at UCLA, I had a rain gauge in my yard and in 1986 we got seven inches in two days―that was half the rain for the entire year.
I was studying clouds and extratropical cyclones, and put two and two together. I knew the storms we now know are atmospheric rivers were those that produced most of the flooding.
Atmospheric rivers are literally rivers in the sky. They’re rivers of water vapor pushed by the wind instead of liquid moved by gravity. In the mid-latitudes and polar regions, atmospheric rivers move the greatest amount of water vapor by far. The average atmospheric river transports 2.6 times the average discharge of the Amazon River―the biggest terrestrial river ―and 25 times the Mississippi River discharge.
How did Atmospheric River Reconnaissance (AR Recon) get started?
Marty Ralph: 1998 was a super big year for us in getting to where we are today. First of all, there was a big El Niño that year. It was predicted there would be big impacts from it and that turned out to be pretty true.
And―by absolute serendipity―as a young scientist at NOAA in 1996, I had proposed with Dave Reynolds of the National Weather Service a crazy aircraft-borne experiment off the West Coast to study low level jets ahead of a cold front and extratropical cyclones. You know, you can’t even get those words out very easily. We now know that’s an atmospheric river but I didn’t have that realization until later.
In 1998, we took NOAA’s Hurricane Hunter aircraft out of Monterey to do a research study in a strong storm. We didn’t have real-time communication from the plane but after we landed, we were able to convey the details of what we saw. That turned out to be helpful for the Monterey forecast office.
There had been a similarly strong storm 10-15 years before when, sadly, many lives were lost. This time, our measurements ahead of the 1998 storm led to flood warnings and pre-positioning of hundreds of rescuers. There was very little loss of life.
That really inspired me. CW3E launched in 2014 and the first genuine AR Recon was in 2016.
Atmospheric rivers used to be hard to predict. How did you figure out what was going on in these storms?
Marty Ralph: Our paper on our 1998 aircraft-borne experiment led to the first observational definition of atmospheric rivers―a way to define them that could be applied to observations.
Another reason 1998 was big for getting where we are today is there were enough microwave satellites to stitch images together.
You can see in this infrared satellite image that there’s not much going on in the central Pacific, there’s really nothing remarkable to a meteorologist:
But look what happens when we look at that new sensor, the microwave―or water vapor― channel. We know now that’s an atmospheric river and it jumps out at you clear as day, even though there weren’t a lot of clouds with it at that point. Here it is in the red outline, shooting to the West Coast:
So what does this mean, and what did that light bulb moment of realizing I was studying atmospheric rivers mean for me?
Here’s an extratropical cyclone, a big one. We now know what’s going on down where the atmospheric river is. But there’s a low center near the top and, to a large degree, the field of meteorology spent the last century literally studying the low center and the structures around it. And the sense was, if you get the low right, everything else is going to be OK. The forecast will be OK.
But what we now know is when we look at the winds, we see these streamlines coming in from the northwest, wrapping around this beautiful bullseye. It’s a bit like a hurricane structure, you’d think, but it’s different: the purple near the bottom is water vapor and you can see the winds coming from the southwest. And that is what we now know of as an atmospheric river.
And it turns out that the atmospheric river is typically 1,000 kilometers from the low center―so if you get the low center right, it doesn’t mean you get the atmospheric river right. So I’m like we’ve got to study this, that’s where the action is.
How did atmospheric river observations help launch Forecast Informed Reservoir Operation (FIRO)?
Marty Ralph: Guerneville is the most flood prone community in the West Coast, and Lake Mendocino is upstream of it on the Russian River. The rule for this reservoir was to make sure about half of it was available for flood control in the winter and then it could refill in the spring, depending on, of course, the storms. A lot of years, there weren’t enough storms to refill it.
In water year 2013, there was a big atmospheric river followed by a reservoir release to be ready for another storm. But no more storms came for the rest of the winter, and that unnecessary water release left the whole watershed in a terrible situation that summer.
Here’s what we hypothesized in our first FIRO meeting: what if we had kept, say, 10,000 acre-feet extra, and each day the reservoir operator could then look at the forecast and say, “Is there an atmospheric river coming or not?”
We had already proved in a 2006 paper that atmospheric rivers are the storm type that matters for flooding on the Russian River. We showed that all seven Russian River floods over eight years corresponded to landfalling atmospheric rivers. That made the connection between them: there’s an atmospheric river offshore and when it hits the shore, it makes heavy rain and can also be the cause of all the floods on the Russian River.
Our analysis showed that FIRO would have left us with about 11,000 acre-feet more, and that is exactly what we managed to pull off a few years later in reality. And this year Lake Mendocino had an extra 35,000 acre-feet of water going into the summer compared to what it would have had without FIRO.
What’s next for AR Recon and FIRO?
Marty Ralph: The Atmospheric River Reconnaissance, Observations and Warnings Act (ARROW) was signed by Biden into law in December 2023. The ARROW Act includes authorizing annual AR Recon in the northeast, central and northwest Pacific, with flights originating from the West Coast, Hawaii and Guam. The storms come west to east, so expanding out to Guam is going to give us more lead time as we head into a stormy winter.
The Act also authorizes AR Recon in the Gulf of Mexico and East Coast. These are places where atmospheric rivers have not been widely recognized but in 2022, an intense storm over the Gulf of Mexico crossed the southeastern U.S. and fueled a dangerous nor’easter across the Northeast. Nor’easters, the big ones, have atmospheric rivers in them so they’re actually an atmospheric river phenomenon.
AR Recon is also getting new aircraft, the G550, which can fly higher, faster and longer.
We’re also looking to expand FIRO nationally, and Congress is supporting that. But FIRO needs at least three days of lead time, and storms are harder to predict in the rest of the country. This is especially true for clusters of thunderstorms in the Great Plains, and for tropical storms and hurricanes in the Southeast. There’s some hope in New England, though, and this is a really important thing for our FIRO community.
We invented a playbook for better storm prediction―better observations, better understanding of storm behavior, better weather models―with our atmospheric river work. Now we need to focus on better prediction for other kinds of storms.
