(11-12) 04:00 PST Sequoia National Park -- Coping with climate change takes sophisticated analytical tools. In the mountain environments of the West, it also takes mules, shovels and plenty of sweat.

A pack train loaded with tools and equipment was due in an hour. But daylight and decent weather are precious at these elevations. Molotch, part of a weekend work crew of five scientists and student helpers, had come up with Kim ahead of the mules for an early start building a new kind of mountain observatory -- what may well be the world's most elaborate snow gauge.

They are installing a unique network of ground sensors, weather gear and other equipment to measure how much snow and ice build up each winter in the 400-mile Sierra range -- and then see where the snowmelt goes.

Warming during this century is expected to shrink the snowpack by pushing snow lines to higher elevations. Even if the same amount of precipitation falls, it's expected to come more as rain, less as snow, while spring snowmelts are likely to arrive sooner. That would mean less effective water storage during the long Sierra winters and, possibly, an increased risk of floods from fast mountain runoff.

So tracking Sierra snowfall is hardly an academic exercise.

First step: Dig some holes.

Kim took one shovel, Molotch another. The student struggled against roots and chunks of random granite buried inches beneath the duff. The holes had to be 2 feet deep. Molotch, 33, shared insights gained from his lengthy training for a doctorate in mountain hydrology and climate science.

"Throw a little more arms into it," he suggested.

Groundwork for observatory

Molotch was laying the foundations for an observatory designed not to peer outward at stars or distant planets, but rather at the very ground he and Kim were digging.

It's called the Sierra Nevada Hydrologic Observatory, the brainchild of about 20 scientists brought together by a UC Merced engineering professor, Roger Bales, one of the country's most respected mountain-snow experts.

"We knew we wanted to do something big," Bales said in an interview. "It was really a consensus that the priority for hydrological research right now is the mountains."

The scientists figure it will take about $2 million a year for 10 years to install the necessary equipment, including devices in trees to monitor sap flow and towers to detect changes in moisture and carbon dioxide in the air.

The National Science Foundation kicked in a $600,000 grant to help start the project, which also has drawn support from UC and state water agencies. Some of the first instruments are going in here, an out-of-the-way drainage rarely glimpsed by tourists drawn to the huge trees for which Sequoia National Park is named.

Molotch and Kim were digging holes to contain some of the first soil-moisture and temperature sensors in the network.

"We're certainly entering into some uncharted territory with this," Molotch said. "We want to understand how water moves through the system -- Mother Nature's plumbing -- rather than just treat it as a statistical problem. And to do that, we need data."

If the plans are realized, instrument clusters would extend from the foothills, at 1,200 feet, where most of California's major reservoirs are located, up to the Sierra crest, 11,000 feet or more above sea level.

Instruments would be strung out along transects -- more-or-less straight lines drawn up the western Sierra flank -- as far north as the American River basin outside Sacramento, or even farther up to the Feather River watershed. Instruments are also planned along Tioga Road and possibly elsewhere in Yosemite National Park, bolstering many years of research on the impact of climate change in the park.

Shrinking snowpack

In ambition, at least, the project seems a worthy match for the grandeur of the Sierra and its critical role as the font of California's economic lifeblood.

"We want to do some representative sampling all across the landscape," Bales said. "And it's rough country."

The scientists hope their electronic networks will be rugged, too, at least enough to produce year-round data in the most vicious High Sierra winters. The goal is to track how water builds up in snow and ice, then gets taken up into trees and plants, soaks into the ground or runs off into streams and reservoirs.

Or, as Bales likes to put it, to "capture the pulse of the Sierra Nevada."

The pulse may be fading. The outlook suggests the state may lose the equivalent of 4 million to 5 million acre-feet of water in a shrinking snowpack this century, about 10 percent of California's total water storage capacity. (An acre-foot is the amount of water it takes to cover 1 acre with water 1 foot deep.)

Given the magnitude of the potential loss, water experts want to keep much closer watch on Sierra snowfall and winter water storage.

"We need better information," said Frank Gehrke, chief of snow surveys for the California Department of Water Resources, adding that as global warming continues, the need becomes "critically more important."

Water officials have been tracking Sierra snowfall for decades so that reservoir managers can be told ahead of time how much runoff to expect after storms and during the spring snowmelt. The information guides last-minute management decisions involving the storage and release of millions of gallons of water.

The traditional system, though much improved in recent years, yields estimates that tend to be off by as much as 20 to 30 percent -- a margin of error likely to get wider with each year of increased variability in the weather.

Throughout the West, climate experts say snow-anchored water systems like California's will need to be much more precisely managed to effectively balance such competing interests as urban growth, agriculture, flood control and the needs of wildlife.

Average Sierra snowmelt runoff accounts for about 35 percent of the state's total usable annual surface water supply -- about 15 million acre-feet of runoff each year. No one can really say how the numbers may change, but a decline seems virtually certain.

"There's no chance the average winter snowpack in the West is going to be in the next 50 years what it has been the last 50 years," said Philip Mote, a prominent Western climate researcher at the University of Washington.

Temperatures this century are projected to increase by about 2.5 to 9 degrees Celsius. Rising temperatures push snow lines to higher elevations, where the steep mountain peaks present less ground for snow to cover. A 3 degree Celsius rise could cause the snow line to rise about 1,500 feet.

Besides more rain and less snow, climate change promises earlier spring snowmelts, which reduce the effectiveness of Sierra water storage.

To cope with this, Bales and his colleagues seek a much deeper understanding of how the water gets into the mountains, and where it winds up.

"It's essentially an accounting problem," said Robert Rice, a UC Merced research scientist working on the project. "All the water that comes in has to go somewhere. So if we can close the water balance, or account for where it all goes, then we can begin to get better at predicting things."

Taking its pulse

Some climate-change scenarios suggest storms could become more violent, raising the risk of catastrophic floods even while reservoir managers struggle to maintain enough storage for dry seasons and extended droughts.

"Your goal is to have a full reservoir when the snowmelt runoff stops, so if that snowmelt volume is decreasing, or the timing changes, you have to be more cautious," said Bruce McGurk, operations manager for Hetch Hetchy Water and Power, a unit of the San Francisco Public Utilities Commission that provides water for 2.4 million people.

"The global warming issue raises the tension level for a water manager," he said. "You're always worried about coming up dry, but you can't keep the reservoirs full all the time, because you have to accommodate storms. So we will need some improved tools as the risk factors and my headache levels rise."

Molotch and Kim worked to install the prototype instrument cluster at a site known as Wolverton, after an old ski lodge complex where the local roads end, at the Marble Fork of the Kaweah River. By taking the pulse of this watershed, scientists figure they can get enough data to make estimates for many other similar watersheds in the steep southern Sierra.

Two others from UC Merced soon showed up: Peter Kirchner, a doctoral student in Bales' engineering department, and another undergraduate, Stephanie Tanverakul. Rice also was on his way, but first he had to make a trip to the nearest hardware store, two hours away, because a supply of bolts turned out to have the wrong type of threads.

The mule train arrived bearing coils of wire, plastic housing for electronic gear, and sections of metal scaffold and posts.

Kirchner and Tanverakul went off to set up a weather tower bolted onto a garage-size granite boulder. The three mules, led by hired cowboys Rowdy Fitzgerald and Junior Ocampo on horseback, departed for another load.

Preserving soil makeup

Molotch and Kim kept digging.

They were making slow progress, partly because each spadeful had to be set aside in order, spread onto gray tarps stretched out beside the pits, so that the dirt could be put back the same way it came out.

The scientists didn't want to risk skewing the moisture and temperature readings by altering the composition of the soil too much. They also had to take care not to walk on the ground uphill of the holes, to avoid tamping down the dirt, because that, too, might affect how water ran down or soaked into the ground.

A dozen of these 60-centimeter-deep holes had to be dug, not counting many false starts where the roots and rocks proved impassable, including six here and six at a site higher up, closer to the crest near a place called Panther Meadow.

The locations were chosen to span the so-called rain-snow transition zone over several thousand feet in elevation, to capture the expected up-slope movement in the snow line as average temperatures rise.

An area surrounding two trees was to be sampled at each site. Molotch and Kim, eventually relieved by Kirchner, had to dig out three pits at each tree, one at the base of the trunk, one in the "drip zone" beneath the outer branches, and one out in the open. Each pit would contain four rod-shaped sensors, about 18 inches long, evenly spaced in parallel to measure at increasing depths down to about a yard below the surface.

"It's very much a cutting-edge, state-of-the-art thing they are doing," Mote said from his vantage point in Washington state. "I wish we had something like it here in the Northwest, or even half as good."

California's traditional snow surveys -- which include field treks every winter by crews on cross-country skis and snowshoes -- won't be replaced anytime soon.

But classic methods are best at measuring the amount of snow only at a particular location. Statistical models used to assemble broader estimates show signs of breaking down already, because of warming that has already occurred.

The rationale driving the new system is to produce the first ground-tested, complete accounting of the Sierra water system -- right down to the roots of the trees.