CALVIN: Clarifying California’s Old and Murky Water Problems

By Gordy Slack  

A lot of important opinions and decisions are based on “safe” assumptions that turn out to be dead wrong. For example, in California’s perennial state of crisis over one of its most precious resources, water, who would have guessed that O’Shaughnessy Dam and the 360,000-acre-foot Hetch Hetchy Reservoir that stands behind it, might be superfluous? Debate has long focused on whether the dam was more important as a great engineering accomplishment (which it surely is) or as an abomination in a National Park (many think so), but never on whether the water it holds is adding value to the California’s chronically short supply. It took a computer model to do that.

Professor Jay Lund and colleagues at the UC Davis Center for Watershed Sciences have developed software to model California's water storage and distribution system.

The CALifornia Value Integrated Network (CALVIN) is an economic-engineering water model that employs seven decades worth of hydrological data, integrated from many disparate sources, to simulate the state’s baroquely-complex hundred-reservoir water storage and distribution system.

Created by UC Davis professors Jay Lund, Richard E. Howitt, and Marion W. Jenkins, it models the engineering structures of California's water system as well as the economic demands for water, allowing users to evaluate the consequences of changing either economic or engineering parameters.

The model analyzes all of the state’s water supply and delivery systems, and projects the impacts of changes in the systems. What, for example, would be the consequences for farmers in the Central Valley in the case of a prolonged drought?  Or what if a levee were to break in the Sacramento-San Joaquin Delta? Or, most famously, what if O’Shaughnessy Dam were removed and Hetch Hetchy Valley restored to its original Yosemite-Valley-like grandeur?

One of Lund’s graduate students, Sarah Null, put the latter question to CALVIN for her PhD thesis and revealed a whole new way of envisioning Hetch Hetchy’s future.

“Null’s model results showed that water supply would not be much of a problem if you removed the dam,” he says. “But it would still be an economic problem.”

O’Shaughnessy adds a little storage capacity to the system, which has many other downstream reservoirs, on the 5 –10 percent of the years when there is not enough water in the system. In those years, CALVIN revealed, it should be possible to buy water from farmers, mostly right in the watersheds where it was needed, says Lund.

The dam is not coming down any time soon, though, predicts Lund. Not because of the water itself, though; “It is the six-to-twelve million dollars worth of hydropower provided by turbines along the system and the avoidance of a need to construct a billion-dollar filtration plant that give Hetch Hetchy its real value today,” says Lund.

“At some point—especially if we need to start filtering that water—we might decide that a place like Hetch Hetchy Valley is such a scarce recreational resource, we would rather have it be like Yosemite.”
Fuelling CALVIN, or any model, with good data is key. That is where innovations in accurate sensor technology enter the CALVIN picture.

CALifornia Value Integrated Network (CALVIN) provides valuable information to policy makers for the state's water storage and distribution system .

Lund and his colleagues at UC Davis are applying CALVIN’s data-crunching power to helping the state figure out how to solve the tangled riddle of what to do about the San Francisco Bay Delta. Currently, freshwater is drawn from the Delta and pumped south, over the Tehachapi Mountains, to feed southern California’s thirst. But the system is not sustainable. “The transition for the Delta is going to be forced upon us by sea level rise,” says Lund.

As sea level rises, and as the levees that keep salt water from getting to the pumps age and become more vulnerable, millions of southern Californians who rely on the water are exposed to unacceptable risk. If the levees were to break—quite possible in a major earthquake—the pumps would start drawing saltwater and the system would shut down. Furthermore, the pumps change the natural tidal flow in the Delta and are bringing at least five species of fish dangerously close to extinction. The Delta smelt is the best know of these.

We have increasingly good data on the fish, says Lund, partly because of tiny identification radio chips that have been inserted just under the skin of some of them so that the fishes’ whereabouts and survivability can be quantified. Knowing how many fish there are, where they are, and how they respond to different levels of uptake and runoff will be key to fashioning a recovery plan for these species.
Other sensors measure and report the depth and density of snowpack in the Sierra Nevada Mountain Range, California’s biggest water storage unit, the rates of flow on rivers and streams, evaporation rates, and other factors.

But all of this data poses problems, too. “This revolution in censors has buried us in numbers,” Lund says. “And the number are meaningless, unless you organize them so they bear insights into what’s really going on and how it’s likely to play out in the future. That’s what is really exciting and important about the work. We’re trying to develop insights, not numbers.”

“We often get excited at the low level of the widget or the software algorithm and then again at the high level of the problem,” Lund says. “But the stuff in between—the stuff that connects the fundamentals and the problems—often gets neglected.”

Lund has had to wrestle a lot with the “stuff in between.” Because CALVIN’s data resources span decades the challenge of coordinating all the information was particularly acute. Lund and his team had to “assemble lots of data never meant to be in the same room together.” That data was collected from different places, by different people, for different purposes, at different times, often to accommodate ancient punch-card computers.

The group took software developed at the Army Corps of Engineers Hydrologic Engineering Center and the University of Texas at Austin that was built for a six-reservoir system and constructed around it a shell that would allow them to model California’s huge and intricate system of more than a hundred reserves.  And around that, they employ modeling that will allow analysis of the economic and water-quality ramifications of different scenarios.

“We are always kluging something together carefully,” says Lund.

Relying on CALVIN and other inputs, Lund and his colleagues from Davis and the Public Policy Institute of California last year published a report (now published as a book with UC Press) on the Delta, suggesting that a peripheral canal that would move water around the Delta, instead of through the Delta as it does today, would be best option for the Delta’s water-supply, economic, and environmental challenges.