Rivers in the sky

Scientific conferences usually don’t physically experience their subjects. But during a session on “atmospheric rivers” last December at a geophysics meeting in San Francisco, one of those very rivers was barreling down on meeting attendees.

Like freight trains loaded with water vapor, atmospheric rivers are long, narrow bands whose winds funnel huge amounts of moisture through the sky. When they hit coasts, these rivers can drop their moisture as rain and cause destructive flooding, as in January 2005 when more than 20 inches of rain soaked southern California, killing 14 people and causing hundreds of millions of dollars in damage.

Scientists (and San Francisco) managed to escape December’s atmospheric river without such harm, but the storm dumped more than 10 feet of snow in parts of the Sierra Nevada, putting the mountains on track for their wettest recorded season. That sort of impact underscores why researchers have recently become fascinated with atmospheric rivers. Completely unknown just over a decade ago, these rivers turn out to be not only a key factor in Western flooding and water supply, but also a major player in the planet’s water cycle.

“Water is life, and atmospheric rivers provide water,” says Paul Neiman, a meteorologist at the National Oceanic and Atmospheric Administration’s Earth System Research Laboratory in Boulder, Colo. New research is revealing how these rivers work, as well as helping forecasters better predict their consequences.

At any given time, somewhere between three and five atmospheric rivers are typically ferrying water in each hemisphere. More than 1,000 kilometers long, they are often no wider than 400 kilometers and carry the equivalent, in water vapor, of the flow at the Mississippi River’s mouth. “That has really captured the imagination of scientists,” says Marty Ralph, also a meteorologist at the Boulder lab. “There are only a handful of these events, and yet they do the work of transporting 90-plus percent of water vapor on the planet.”

Ordinary clouds don’t carry lots of water vapor long distances; they rain out as soon as water droplets coalesce and get heavy enough to fall as precipitation. In the 1990s, MIT researchers calculated from wind and moisture data that jets in the atmosphere, which the scientists termed atmospheric rivers, must exist to help ferry water around the planet.

Since then researchers have gotten a better look at the rivers, using microwave-sensing instruments carried on polar-orbiting satellites. Solar radiation bouncing off Earth’s surface in microwave wavelengths is affected by the amount of water vapor between the ground and the satellite, but microwaves aren’t affected by clouds the way visible and infrared radiation are. So microwave instruments are able to photograph ribbons of water vapor coursing through the atmosphere.

In the early days of atmospheric river research, scientists weren’t sure that the bright bands of water vapor in microwave satellite images really translated to super-soggy conditions. So teams flew research airplanes into storm systems, some of which spawned atmospheric rivers, to measure how wet things got. “You could really sense the juiciness,” Ralph says. “You could smell it in the cockpit.”

Atmospheric rivers are born because of temperature differences between Earth’s tropics and its poles. During winter, a pole cools compared with the equator, creating a strong temperature gradient across the hemisphere, a difference that causes low-pressure storms to spin off in the midlatitudes. Winds within the storm can funnel moisture into a narrow band at its leading edge — the atmospheric river. At the San Francisco meeting, George Kiladis of the Boulder lab described a March 2005 river that apparently sucked moisture into the Pacific Northwest all the way from the tropics, in the “intertropical convergence zone” where winds from both hemispheres meet. Kiladis and colleagues describe the river in a paper to appear in Monthly Weather Review.

People living on the West Coast are familiar with atmospheric rivers such as the famous “Pineapple Express,” which occasionally ferries moisture directly from Hawaii. But the rivers can also come up through the Gulf of Mexico or along the eastern seaboard.

Scientists are now moving from spotting atmospheric rivers to understanding them and trying to predict their impact. Leading the way is California, which is setting up four atmospheric river observatories along its coast to track the rivers as they arrive. Each river can have strikingly different effects depending on the angle and speed at which it approaches mountain ranges and watersheds.

“To be able to nail down specific water basins that are most prone to flooding, you really need to know precisely where that atmospheric river will make landfall,” Neiman says. “That’s the tough part.”

During a November 1994 storm, now known to be an atmospheric river, for instance, forecasters predicted that less than one-tenth of an inch of rain would hit some parts of the San Francisco Bay area. In places, more than 11 inches fell, David Reynolds, a meteorologist with the National Weather Service’s office in Monterey, Calif., said at the geophysics meeting. How that moisture is distributed within the river and how long it sits in one location determine what areas will see the most flooding.

Not all atmospheric rivers are devastating — in fact, most of them are weak — but they cause many of the most extreme West Coast floods. In one study, Ralph and colleagues looked at all seven floods that occurred on California’s Russian River between 1997 and 2006. All were due to atmospheric rivers, the researchers found. The amount of intense rainfall the West Coast gets from atmospheric rivers over time, says Ralph, is comparable to the soakings the Gulf Coast and southeastern United States receive from major landfalling hurricanes.

In January, California emergency planners met in Sacramento to run through a doomsday scenario dubbed ARkStorm. Officials tested how they would respond if a series of atmospheric rivers hit the coast one after the other. That scenario was modeled on the rivers that hit in the winter of 1861–62 and flooded the state’s central valley. The capital had to be temporarily moved from Sacramento to San Francisco, and the governor took a rowboat to his inauguration.

To better predict such disasters, researchers at NOAA and the Scripps Institution of Oceanography in La Jolla, Calif., are working with state officials to pinpoint the most vulnerable areas. Many times, the river of humidity runs into a mountain, is forced upward, and rains out its water. Other times the river hits the base of a range and shifts to flow around it. Figuring out which process dominates in which locations will help officials better prepare, Ralph says.

For instance, in 2009 planners in Washington faced a crisis when the Howard Hanson Dam, on the Green River above Seattle’s southern suburbs, began leaking just as an atmospheric river was aiming right at it. The scientists analyzed how the river would run into the watershed and dump its water, and predicted it would not cause lots of rainfall above the dam. The U.S. Army Corps of Engineers decided not to assume emergency control, and the rain did taper off quickly.

Atmospheric rivers may become even more relevant as global temperatures rise. Researchers aren’t sure exactly how climate change will affect the rivers, but warmer air generally means that the atmosphere can hold more water vapor, says Neiman. On the other hand, winds may weaken in a globally warmed world, meaning the rivers might carry more water but be less effective at delivering it.

More answers may come within the next few months, as NOAA scientists plan to fly unmanned aircraft into several storms to try to learn even more about atmospheric rivers.