Mountains are the foundation of water in the western United States, natural infrastructure that captures snowfall during the winter and releases snowmelt over the spring and summer. In California, the snowpack holds nearly as much water on average as all the reservoirs put together, effectively doubling the state’s surface storage. But, as the world warms, we may not be able to rely on this ecosystem service much longer. A new study projects that snowpack shrinkage will likely disrupt the West’s water system well before the end of the century. This finding adds urgency to water managers’ efforts to adapt to climate change.
“This is not a problem for later,” says Michael Anderson, state climatologist at the California Department of Water Resources. “This is a problem for now.”
Other climate change-driven impacts to water in the West include earlier snowmelt, which exacerbates the impact of snowpack loss by exhausting the water supply before the dry season ends, as well as stronger atmospheric rivers, the winter storms that deliver much of the West’s rain and snow. Snowpack loss will also limit the hydropower that California depends on to meet its zero-emission energy goals.
Snowpack declines are driven mostly by rising temperatures, which push the snowline higher up mountains. “Snowpack loss is like the canary in the coal mine for climate change — temperature has direct impacts on snow,” says Erica Siirila-Woodburn, a Lawrence Berkeley National Laboratory (Berkeley Lab) hydrologist who co-led the 11-member interdisciplinary team that contributed to the new study, which was reported in October in Nature Reviews Earth & Environment. Study co-leader Alan Rhoades, a hydroclimate scientist who is also at Berkeley Lab, puts it this way: “The freezing point of water is non-negotiable — it just won’t snow if the temperature gets too high.”
To get a more definitive handle on Western snowpack losses, Woodburn, Rhoades and colleagues combed through decades worth of relevant published research. The team considered hundreds of studies, ultimately selecting 18 with snowpack projection data that they could extract and combine for further analysis. Snowpacks in the West have already declined by 20% since 1950, but projections of the timing of future losses varied widely.
“Surprisingly, there was no consensus on the timeframe,” Woodburn says. “We wanted to pull that together.” The researchers found that at current greenhouse gas emission rates, water released from snowpacks in the West will likely diminish another 20% by 2050. This is the equivalent of the storage capacity of Lake Mead on the Colorado River, which distributes Rocky Mountain snowmelt to seven states including California and is the largest reservoir in the country. Looking further into the future, the researchers project that by 2100 the water released from snowpacks across the West could decline a jaw-dropping 50%.
California will be hit particularly hard. Breaking the analysis down by mountain range showed that the Cascades and the Sierra Nevada will likely suffer snow losses of about 45% by 2050. In contrast, snow losses in interior mountains like the Rockies are projected to be less severe, at 20% to 30% by mid-century.
Drilling deeper into the data revealed that the West could see a sharp increase in low-to-no-snow winters within decades, with “low snow” defined as one-third or less of the historical high. California could undergo low-to-no-snow stretches persisting five years in a row as soon as the 2040s, and 10 years in a row as soon as the 2060s.
Knowing what to expect and when will help managers prepare for this impending water disaster. “Water managers are all very worried about snowpack loss,” Rhoades says. “We want to help them be more proactive rather than reactive to snow loss.”
Rethinking Return Rates
California water managers are also worried about atmospheric rivers. These storms are relatively warm to begin with because they transport water vapor from the tropics, and are becoming even warmer with climate change. “Up to 40% of the storms are at or near freezing already,” Rhoades says. Warmer storms deliver more of the water they carry as rain than snow. Atmospheric rivers provide up to half of the West’s precipitation, including about half of the Sierra Nevada snowpack, and cause more than 90% of insured flood losses in California with average annual damages of $300 million.
As the world warms, these storms will likely become even more potent. “Atmospheric rivers are near saturated and a warmer climate increases their ability to hold moisture — they become wetter,” says Alexander Gershunov, a climate scientist at Scripps Institution of Oceanography in La Jolla. Planners need to know how often to expect these wetter storms, which drop heavier precipitation and heighten the risk of floods.
In 2019 Gershunov published a study in Nature Scientific Reports on how climate change will affect atmospheric rivers in West, including projected precipitation extremes in California. Soon after, he estimated return periods for major storms in the Bay Area at the request of David Behar, who directs the climate program at the San Francisco Public Utilities Commission. “It was a surprise to me,” Gershunov says. “I thought someone had already done it.” Behar needed the results fast so Gershunov and Tamara Shulgina, also at Scripps, crunched the numbers over a weekend.
First Gershunov and Shulgina evaluated daily precipitation in the Bay Area between 1950 and 1999 to get a historical baseline of major storms. Then they used climate models to project how often these major storms will occur in the future. Their analysis, which Gershunov stresses was preliminary yet sound, showed that major storms will return at shorter intervals in the second half of this century than would have been expected historically. The coming changes will be most pronounced for storms of the greatest magnitude. By 2070, for example, historical 20-year storms will occur approximately every 10 years, while historical 50-year storms will occur about every 15 to 20 years. These estimates assume that current greenhouse gas emission rates will continue.
Now Gershunov is working with colleagues at the University of Nevada to sharpen major storm predictions for California and elsewhere in the West, and to extend the timeframe of storm-return projections further into the future. This information is needed for planning water infrastructure from reservoirs to neighborhood drainage systems. “We can’t just use the old data to plan for structures that will last 100 years or more,” he points out.
Adapting the West’s water system to climate change will entail a combination of old and new techniques. High on the list is managed aquifer recharge, which uses flood water to infiltrate and replenish groundwater basins. Municipalities in dry parts of California have long used this strategy to augment their water supply. Notably, the Orange County Water District refills its groundwater basin with both stormwater and recycled water. Over the past 50 years, this program has boosted the basin’s sustainable yield three-fold.
Now water managers seek to use fallowed farmland and floodplains for managed aquifer recharge, too. “People are excited about connecting headwater flows with Central Valley agriculture,” Siirila-Woodburn says. The benefits of redirecting the rush of early snowmelt and the floodwaters from atmospheric rivers to fields could be huge.
Aquifers are natural reservoirs, and California’s vast groundwater basins far outstrip the capacity of its surface reservoirs. Statewide, aquifers offer roughly one billion acre-feet of storage, while surface reservoir storage averages 23 million acre-feet of water per year and tops out at 50 million acre-feet.
Above: A 1997 downpour including an early atmospheric river event led to unprecedented releases from the Oroville Dam, damage to the spillway, and flooding downstream. Video: KOVR 1997.
Newer approaches to climate change adaptation include forecast-informed reservoir operations (FIRO), which optimizes surface water storage. Many California reservoirs are operated for both water storage and flood control, and balancing these two functions can be tricky. Currently, water is released to make room for stormwater according to a schedule that is based on historical weather patterns, and is dictated by a U.S. Army Corps of Engineers Water Control Manual specific to a given reservoir.
The schedule in the Corps’ manual doesn’t always fit with reality, however. In anticipation of storms that may never arrive, reservoir operators can make releases unnecessarily, losing water that could have been used during droughts. In contrast, FIRO combines the latest in atmospheric river forecasting with real-time runoff data, letting operators know exactly when to release or retain water.
In a test of FIRO in Lake Mendocino during the winter of 2019–2020, storage increased safely by 19%, a welcome outcome during the subsequent severe drought. And a 2020 modelling study in Water Resources Research suggests that FIRO could up Lake Mendocino’s storage even more, by as much as 33%.
Now FIRO’s potential to lessen flood risk is being tested in the Yuba and Feather river watersheds. This system includes two reservoirs: Lake Oroville, which is operated by the Department of Water Resources (DWR), and New Bullards Bar Reservoir, which is operated by the Yuba Water Agency. The region has suffered five major floods since 1950, resulting in 41 deaths and devastating economic impacts.
Atmospheric rivers are the primary drivers of flooding in the Yuba-Feather watershed. In 2017, an atmospheric river dropped so much rain there that it damaged Lake Oroville’s spillway, raising fears that a torrent of water would gush from the rapidly filling reservoir and inundate downstream communities. While these fears were ultimately unrealized, more than 188,000 people were placed under evacuation orders and had to leave their homes.
Yuba-Feather flooding also depends on how soggy the ground is when a storm hits; the elevation where rain turns to snow; and whether rain falls on top of snow. “If there’s a snowpack and a really warm atmospheric river comes, it can potentially melt the snow,” explains John James, who manages water operations at the Yuba Water Agency and is a co-chair of the Yuba-Feather FIRO Steering Committee. Other partners on this project include DWR and the Center for Western Weather and Water Extremes (CW3E), a pioneer in atmospheric river research.
The Yuba-Feather FIRO instrumentation includes multiple atmospheric river monitoring stations, which collect meteorological data including air temperature, precipitation, and wind speed and direction. Some of the stations also measure soil moisture to depths of three feet, which helps determine the potential for surface water runoff. Some also include radar that points upward, which helps determine the altitude in the atmosphere where rain turns to snow and ice.
The results so far are encouraging. Compellingly, FIRO could have mitigated the impacts of a catastrophic flood in the Yuba-Feather system on New Year’s Day 1997. After weeks of steady rain and snowfall, a powerful atmospheric river brought heavy rain and rapid snowmelt, breaking levees on the Feather River.
But if FIRO had been in place, reservoir operators could have released water much earlier than they did. Releases 48 to 72 hours before the storm struck could have lowered the water level on levees near the Yuba-Feather confluence by several feet. This finding assumes coordinated operations of Lake Oroville and New Bullards Bar Reservoir, as well as a second spillway on the latter.
“FIRO lets reservoir operators make decisions ahead of time instead of waiting for the water to show up,” James says. “Otherwise, if they wait and then start to release, the water downstream is already filling up those channels.”
In a happy coincidence, the Yuba-Feather FIRO assessment is concurrent with — and so can inform — updates of the Corps’ manuals for the two reservoirs in the system. These updates are on track for completion by the end of 2024.
While Siirila-Woodburn and Rhoades expect managed aquifer recharge and forecast-informed reservoir operations to be heavy lifters in adapting to the future of water in the West, these strategies probably won’t be enough on their own. “I think it’s going to be a buckshot approach rather than a silver bullet to creating a resilient water portfolio,” Rhoades says. Other likely components of this portfolio include decreasing water demand via conservation as well as increasing the supply via, for example, recycled water, potable reuse, and desalination.
Thinning forests, which are overgrown after decades of fire suppression, could also extend the water supply. Plants transfer water from the ground to the atmosphere in a process called evapotranspiration. “Each tree or plant is like a straw into the ground — over half of the water budget is lost to evapotranspiration in arid regions such as the western United States,” Siirila-Woodburn says. “Forest management like prescribed burns and thinning are ways to retain some of that water in the system.”
Despite the scale and immediacy of the coming climate change-driven disruptions to the West’s water system, Siirila-Woodburn and Rhoades are optimistic. “Little or no snow can be gloom and doom but there are paths forward,” Rhoades stresses. “We have a lot of smart folks working on this. Hopefully in 50 years we can say we overcame low-to-no-snow — or that it wasn’t as bad as expected due to proactive climate mitigation and adaptation measures.”