It seems impossible in this land of rivers and rain to imagine a time when the Cascades' mighty snowpack could be stunted by global warming.

But that's the warning of climatologists, who forecast that within 20 years even a slight warming could dramatically — and with surprising speed — shrink the snows that blanket Northwest mountains.

And since that snowpack plays a crucial role in dispensing precious water in dry summer months, salmon, farms and people could compete even more for something the region often takes for granted.

Edward Miles, who oversees the University of Washington's Climate Impacts Group — one of the nation's premier climate-change research bodies — is leading a campaign to urge those who oversee the region's signature resource to plan now for a future of restricted river flows.

Yet while resource managers say his forecasts might be right, they struggle with how — and if — to respond.

"I can't think of a way we're incorporating climate changes into salmon-recovery planning," said Brian Brown, an assistant regional administrator at National Marine Fisheries Service. "I don't get a sense that people are dismissing it, but we have a hard enough time trying to understand what (fish) came back last year."

Welcome to the politics of global warming, where the issues are so complex, the science so imprecise and controversial, and the outcomes — from flooding to droughts to no change at all — so variable, few are eager to do anything but wait.

"There are a lot of other burning issues — rates, financial situations," said Nancy Stephan, a weather forecaster with the Bonneville Power Administration (BPA). "Climate change is a lot harder to get your arms around than something that's more immediate."

This much is known: The Earth has warmed by 1 degree Fahrenheit in the past century with accelerated warming during the past two decades, according to the National Academy of Sciences, which advises the federal government on scientific matters.

But scientists still argue over how much we understand natural climatic changes. Experts like Miles rely on intricate computer modeling to make their conclusions, and their work relies on assumptions, such as how much man-made greenhouse-gas emissions may increase.

To Miles and other climate experts, in the region, the implications are clear. Their projections aren't proven, but if change happens it could come soon, and the best insurance is planning. Doing nothing could come with a cost all its own.

"It's hard to change without a severe crisis, where everybody sees the wolf at the door," Miles said. "But when you get to that point, it's too late."

A harrowing prediction

Miles' group is not the only one whose forecasts call for big changes. Bill Pennell, director of the global environmental change division at the Pacific Northwest National Laboratory in Richland, said his lab has reached the same conclusions for at least the past five years using independent forecasts.

"We were seeing these kinds of changes even when we were first looking at cruder models five and six years ago," Pennell said. "As we've refined the models, our predictions have become pretty robust."

They paint a harrowing picture.

Snow gives life to Washington rivers, powers its cities, supplies drinking water for millions and irrigates thousands of acres of farmland. But those heavy Cascade snows, according to Miles, are so vulnerable to warmth, even a slight rise in average temperatures can mean the snowpack will form at higher elevations — nearly 500 feet higher — and there will be less of it.

His group forecasts that by 2020, 17 to 29 percent of snow will instead fall as rain, which will increase winter stream flows. But that means there will be less water available later during dry months. Late-summer flows in the Columbia Basin streams could drop 12 percent.

By 2040, summer stream flows could be reduced by 25 percent, according to Miles.

More frequent and pronounced late-summer droughts would spell trouble for farmers, particularly in the Yakima and Columbia river basins. In low-water years, some crop producers already are taking buyouts because they can't get the needed water.

"In Idaho, marginal farms are going out of existence and reverting to grasslands," said Philip Mote, a research scientist with Miles' group. "There simply isn't enough water to preserve agriculture over the number of irrigated acres we have now."

It's still too early for anyone to understand how changes in stream flow would affect water temperature and siltation — both of which are important to salmon.

"Salmon, over thousands of years, have learned to adapt to certain types of flow conditions, when peak runoff occurs in May and June, but that runoff peak could be shifted a couple of months earlier," said Pennell of the Pacific Northwest National Laboratory.

For the Northwest's electrical grid, more winter rains would help meet demand during the peak heating months. But for the rest of the West, peak loads arrive during the summer, when air conditioning reigns. And even a 10 percent decrease in Northwest stream flows during the summer could result in the loss of generating capacity equivalent to what is needed to keep the lights on in Seattle.

In Western Washington, Miles contends, the most conservative estimates show population growth will increase demand and competition for drinking water, and "it takes 20 to 25 years to change a water system."

The two men have cornered water officials and given presentations to Gov. Gary Locke. They've buttonholed staff with the Army Corps of Engineers and BPA and Eastern Washington farm groups. Miles testified before Congress. Pennell briefed committees of the state Legislature.

But it's been a struggle to get the attention of decision-makers.

"Nobody has really started to think about this seriously," Pennell said. "Most people are driven by shorter-term goals and this is kind of out there on the horizon. The reaction has been, 'This is interesting, but it's not happening on my watch.' But it's happening sooner than people might think."

Truth or fiction?

Critics of global-warming theory, such as Oregon state climatologist George Taylor, argue that not enough is understood about existing cycles of warming and cooling, and that the modeling is not accurate enough to forecast weather patterns with precision.

"By their very nature, the models can't be trusted," he said. "My general feeling is that human impacts do occur, certainly locally — perhaps globally — but I happen to think that the natural variations are much bigger than the human impacts."

Still, Taylor said, whether global warming is real or not, population growth alone will increase conflicts over water, and "hoping for the best and planning for the worst is probably a good way to handle that."

Some government agencies only now are beginning to consider a slightly warmer future. In the next year, for the first time, the Northwest Power Planning Council will try to gauge how stream flows would change with warmer conditions.

Meanwhile Seattle Public Utilities is trying to identify how climate change could affect the city's snowmelt-dependent drinking-water supply. The agency is working with Miles' group to study in more detail how lower snowpack could affect water supplies in the Cedar and Tolt watersheds.

"We're trying to take the next step ... and downscale the research to see what it means for us," said George Schneider, water resources manager for the utility.

The Benton County Commissioners are firmly behind Miles' and Pennell's work — though they admit that's in part because it may give them political leverage to push for a long-sought $1 billion-plus reservoir called Black Rock proposed east of Yakima.

"People tend to read into this data whatever they want," said Mote, the UW scientist. "But at least they're talking."

Other agencies aren't sure what to make of the research.

On a visit earlier this year to the Pacific Northwest National Lab, Locke listened to a presentation by Pennell and his staff. But while Locke is trying to enforce limits on carbon-dioxide emissions on new power plants to stem the growth of greenhouse gases, the state is only now beginning to talk about what happens if river flows drop further.

"Some of these situations, if there's simply less snowpack, there's going to be less water, so there may be only so much we can do," said Ron Shultz, a policy analyst for Locke. "There's so much that's unknown."

And BPA's Stephan said she's aware of nothing at BPA that incorporates climate change, but recognizes the inherent complications.

"I back what they're doing completely," she said of Miles and Pennell. "They've come a long way. But the complexity of these models is enormous. I'm not sure they've got it yet. I have confidence that in the next 20 years they'll have better estimates. But, unfortunately, you'd have let 20 years go by."

That, Miles said, is precisely the point. Even people who question their projections need to weigh the impact of doing nothing.

"What's the probability that we're correct?" he asked. "At what point do we need to buy some insurance? For us, that insurance is planning."

Craig Welch can be reached at 206-464-2093 or cwelch@seattletimes.com.