How much water in that snowpack? Scientists seek a better gauge.

Scientists are testing a new approach for gauging the amount of water stored in mountain snows – reservoirs that supply more than 75 percent of the fresh water in the western US and that slake the thirst of some 1.5 billion people around the globe.

The aim is to measure the snow's water content more accurately and more frequently, so that water managers can mete out water stored in reservoirs more effectively. The data also are expected to improve snowmelt forecasts as a melt season progresses.

The three-year demonstration project is focusing on a watershed in California's southern Sierra Nevada Mountains that provides San Francisco with water and another watershed in Colorado that feeds the Upper Colorado River Basin.

Rapid population growth and the subsequent demand on water resources in a drought-prone region are enough to justify the effort, suggests Thomas Painter, a researcher at NASA's Jet Propulsion Laboratory in Pasadena, Calif., and the lead scientist on the project. Dr. Painter's research focuses on Earth's water and carbon cycles.

But the issue becomes more critical in the face of global warming. Taken as a whole, the West's mountains have experienced a decline in April snowpack since the 1950s, studies indicate, although some portions of the West's mountain ranges, such as the southern Sierras, have seen increases in April snowpack.

For the northern Rockies, the decline is one of the most severe seen at any time in the past 1,000 years, according to a study published in 2011. The research team attributed the trend to natural climate swings superimposed on a long-term warming trend triggered by a buildup of greenhouse gases in the atmosphere from burning fossil fuels and from land-use changes.

"In the western US, we've built ourselves around this very nice synchronicity between the mountain snowpack and human-made reservoirs," Dr. Painter says, referring to the complementary role each plays in meeting the region's water needs. At the start of a water year, the reservoirs hold what's left of last year's snowmelt. The snowpack serves as a natural reservoir that releases its supply gradually, topping off human-built reservoirs at a pace that in principle keeps water available for use during the summer and into fall.

"Once we start nudging that with climate change, we go into a very ugly scenario," he continues. With warmer temperatures, winter snows come later at lower elevations, replaced by rain early on – water that immediately flows down the slopes. This reduces the amount of water stored as snow. Rising temperatures also increase evaporation off of reservoirs and increase the rate at which snow changes into water vapor without melting into a liquid first, a process know as sublimation.

The federal government maintains a network of snow gauges throughout the US, and forecasters augment that information with measurements taken from aircraft. The data are used to build maps of snow cover and the snow's water content, as well as to feed forecast and other weather and climate-related models. For much of the country the system works well, says Andrew Rost, director of the National Weather Service's National Operational Hydrologic Remote Sensing Center in Chanhassen, Minn.

But the mountain West is a challenge. Mountain terrain makes it difficult to set up and maintain a dense network of snow-measuring stations on the ground. The sensors the center uses to measure the snow's water content from aircraft require the planes to fly no more than about 500 feet above the surface – out of the question in mountainous terrain. The sensors can't accurately measure water content with the snow when it gets especially deep, as it does in the mountains. And the aircraft fly a narrow path over set locations, largely to support forecasting needs in areas threatened by spring floods – much as federal hurricane-hunter aircraft are sent aloft only when a hurricane threatens.

All this means that "we don't know the snowpack well," Painter says. "Given how little information there actually is about it, it's remarkable" that forecasters and water managers "do as well as they do."

There is plenty of room for improvement. Citing California's American River as an example, Painter says the forecasts for total run-off expected between April and July, when the melt season typically ends, are off by at least 20 percent more than half the time. Twenty-five percent of the time, the forecasts are off by as much as 40 percent.

That's where Painter's project comes in. It uses NASA's Airborne Snow Observatory to gather key information about the snowpack over a wider area and in more detail than current approaches can deliver.

The aircraft, which fly at altitudes of between 17,000 and 19,000 feet, use a radar-like device called lidar, which sends rapid pulses of light, rather than radio waves, and receives the pulses reflected off the snow to measure snow depth. Focusing on the Tuolumne watershed in the Sierra Nevadas, which feeds the Hetch Hetchy reservoir, the team first took measurements of the land height when the ground was bare. Now they are measuring the height of the snow surface. By subtracting the two, they get the snow depth and can estimate its water content.

Another instrument measures how much light the snow's surface reflects. Past research in the Rockies has shown how dust can accelerate snowmelt by giving the dust a darker hue.

"Dust dramatically increases the amount of solar radiation absorbed by the snow," explains Jeffery Deems, a researcher with the National Snow and Ice Data Center in Boulder, Colo., and a member of Painter's team. At this stage of global warming's evolution, "the direct impact of dust on snowmelt is far greater than the direct impact of greenhouse warming on snowmelt."

For the snowpack feeding the Colorado River, the dust is coming from New Mexico, Arizona, and Utah, he notes.

Indeed, that research – which Painter, Dr. Deems, and colleagues published in 2010 – prompted the JPL scientist to wonder if snow in the Tuolumne basin might be affected by dust kicked up in the nearby San Joaquin Valley or even coming across the Pacific from Asia.

Indeed, the team's instruments have recorded "a pretty good-sized dust signal in the mountain snowpack," Painter adds. The scientists are analyzing the dust to see if they can pinpoint its origin.

The tools the team is using also cover a wider swath of the ground with each pass than do the sensors aboard the National Operational Hydrologic Forecasting Center's aircraft.

This has allowed the team to put hard numbers to a region of the Tuolumne watershed that managers had long suspected of harboring the watershed's largest volume of snow-stored water. But the region is so inaccessible that it was difficult to make the on-the-ground measurements needed to show how important that patch of the watershed is.

The use of NASA's snow observatory won't replace ground measurements, notes Frank Gehrke, a co-investigator on the project who heads the California Department of Water Resources' Cooperative Snow Surveys Program. These proved to be a reality check on the aerial measurements and can provide direct measurements of snow density, which also are key to estimating water content, he explains.

But so far, the aerial data taken during April seems to be demonstrating the observatory's worth.

The project also has federal forecasters rooting from the sidelines as they anticipate better data on a vital resource.

"We've been paying attention to this project for quite some time," says the Hydrologic Forecasting Center's Mr. Rost. "We're cheering Tom on."