February 3, 2017 – You need to understand atmospheric rivers – airborne water streams that deliver as much rain as hurricanes or tornadoes – if you live in the West. Fortunately, scientists are developing a wealth of new tools to predict and explain these storms.
Atmospheric rivers are California’s drought-busters. As we saw in the recent series of storms between January 7 and 10, a single wet weekend can dramatically reverse the state’s water accounts.
The state still hasn’t completely pulled out of the drought. But just a few days after those storms, the California Department of Water Resources increased its water delivery forecast for 2017 from 45 percent to 60 percent for water agencies that rely on the State Water Project.
Atmospheric rivers are just what they sound like: a column of water carried aloft by a narrow band of wind. The moisture usually originates in the tropics of the eastern Pacific Ocean and travels all the way across the sea in a narrow band before striking the U.S. coast – usually somewhere in California.
These events can deliver as much as 50 percent of California’s water supply in as few as eight storms every year. But until recently, predicting them has been difficult. Meteorologists had no way to tell where an atmospheric river would strike, how wet it would be, or for how long. Now they have a variety of tools that help provide those answers. Here’s a look at some of the progress:
1. Rock you like a hurricane? Yes, atmospheric rivers really do.
Experts on atmospheric rivers sometimes use the phrase “horizontal hurricane” to explain how these storms work. And it’s apt, because atmospheric rivers usually manifest as a long, narrow band of high-intensity rainfall that reaches across the Pacific Ocean before targeting California.
Now we know these storms are also just as wet as a hurricane or tornado.
The big storms that hammered California from January 7-10 amounted to at least three and possibly four separate atmospheric rivers. The precipitation they delivered approached 20in (51cm) at some locations along the coast and in the Sierra Nevada, said Michael Dettinger, a research hydrologist at the U.S. Geological Survey who studies atmospheric rivers. That equals the rainfall often delivered by major hurricanes in the southeast or tornadoes in the Midwest.
“Very often in California, we have a big storm and it’s sort of like, ‘Well, I’m sure somebody in the Midwest wouldn’t think much about this storm. We’re just these Californians who are used to sunny weather,’ and all that,” Dettinger said. “But factually, our biggest storms are hurricane-scale storms. And there’s no place other than the hurricane belt that you get storms this big. So really, when we have a big, bad storm here, there’s no reason to apologize. Our big, bad storms are as bad as anyone else’s.”
2. How big? Scientists now have a way to rate them.
Weather experts use the Saffir-Simpson scale to rate hurricane intensity and the Fujita scale to rate tornado strength.
Now there’s a way to rate extreme precipitation events like atmospheric rivers. It’s called the R-Cat scale, short for “rainfall category.” Dettinger helped develop it as a way to objectively rank rainfall events, no matter where they occur.
If more than 8in (200mm) of rain falls at any measuring station over a three-day stretch, that earns an R-Cat 1 rating. The scale steps up from there with every additional 4in (100mm) of rain. An R-Cat 3 event, for instance, means a weather station got 16-20in of rain over three days. R-Cat 4 (the biggest, at least so far) means more than 20in over three days.
During the January 7-10 storms, many areas on the coast and in the Sierra Nevada saw R-Cat 2 rainfall levels, or more than 12in of rain. A handful saw R-Cat 3 levels, or more than 16in, including the town of Venado, near the coast in Sonoma County; and Downieville in Sierra County.
One location, Strawberry Valley in Yuba County, near Sly Creek Reservoir, saw 20.51in of rain, or 521mm, in three days. That puts it in R-Cat 4 territory, one of the largest rainfall events ever recorded in California.
“The extremity and rarity of the largest events is quite comparable to hurricanes and tornadoes,” Dettinger says. “So when we say it’s an R-Cat 4, that’s a big thing. It could be as far back as 2006 that we last had one of these show up.”
3. Extreme rainfall is strongly linked to atmospheric rivers.
The number of R-Cat 3 or 4-rated rainfall events every year roughly matches the number of major hurricanes that occur annually in the Atlantic or the number of extreme tornadoes in the Midwest.
Dettinger used the new rating scale to look back at previous storms over the past 60 years. He found there have only been about 48 events that ranked as big as R-Cat 3 or 4. Of these, Dettinger said, 92 percent were associated with atmospheric rivers, and 90 percent were in California.
“They almost always happen in California, and they’re almost always associated with landfalling atmospheric rivers,” Dettinger said.
4. Those ratings, and more, may soon show up in forecasts.
The rainfall ratings aren’t just for scientists. They can also help the public gauge the intensity of an approaching storm. And they could show up in your nightly weather forecast soon.
Ten years ago, the National Weather Service was reluctant to forecast rainfall amounts more than three days out. There was too much uncertainty to ensure useful information.
Now, that window has grown quite a bit. Forecasters can now estimate storm intensity and rainfall as much as 10 days out. And within five days, they can start to tell us something about where the atmospheric river will strike the coast. That’s a big stride from just a few years ago, when forecasters would often liken an atmospheric river to a “loose fire hose” flailing around and gushing water unpredictably.
“We’re beyond that now,” said Marty Ralph, director of the Center for Western Weather and Water Extremes, a branch of the Scripps Institution of Oceanography at U.C. San Diego and the leader in atmospheric river research. “We have a pretty good sense of when they’re coming, and the hose isn’t flapping as much.”
Ralph said researchers can now predict within 500-600 miles (805-965km) where an atmospheric river will make landfall, and they’re working to shrink that range. For instance, their computer models can now plot the likelihood that an atmospheric river will strike each degree of latitude on the Pacific Coast, from Baja California to Alaska.
The next step is to include R-Cat ratings in regular forecasts. We might hear, for example, that a storm bound for California has “R-Cat 3 rainfall potential.” This could help vulnerable areas prepare for flooding, mudslides and road closures.
“The pieces are all there to express forecasts in these terms,” said Dettinger, who is also a Scripps research associate and a principal investigator on Ralph’s team.
5. The next frontier is duration and elevation.
Two other factors determine how wet and wild an atmospheric river will be: The snow level, and how long it lasts.
A warm storm with high snow levels means more of the watershed is contributing to streamflow, which can increase flood risk. It also melts a portion of any existing snowpack, boosting streamflow beyond what the clouds deliver. A 2,000ft (610m) rise in snow level, Ralph said, can triple the amount of runoff.
“What happened as the big atmospheric river came in on Saturday [January 7] was the snow level jumped from 5,000ft (1524m) above sea level all the way up to 10,000ft (3048m) in about four hours,” Ralph said. “None of the models predicted that, but we saw that in real time.”
That’s because Ralph’s team has 10 vertically pointing weather radar units installed at important reservoirs around the state. These record precipitation as it changes from rain to snow, and the elevation at which that occurs. They’ve been in place for several years. But because of the long drought, they are only now getting regular use in atmospheric river research.
“This winter is allowing us to see how they perform much better, and to start to see how they might be useful for people who are affected by the snow level,” Ralph said. “Nowhere else in the world has this kind of data.”
Duration is a similar concern. How long an atmospheric river stays parked over a particular region determines how wet things get. And it’s not a linear relationship: An atmospheric river that lasts for 40 hours, Ralph said, can deliver seven times more rainfall than a 20-hour event.
The Center for Western Weather and Water Extremes has an enormous amount of data on its website to help predict, monitor and measure atmospheric rivers. A lot of it is not comprehensible to the layman. But they are working to bridge that gap with improved graphics, and new tools like duration and snow-level predictions.
“I suspect by next winter, if not sometime this winter, they’ll become part of our bag of tricks we put online for everyone to see,” Dettinger said.