The critical water shortage from which Californians got a brief respite thanks to a major El Niño last winter is the result of a relatively rare multi-year severe drought. But even the unlikely possibility that such a drought will not happen again soon won’t come close to solving the state’s future problems, for the population of California is expected to reach 60 million by mid-century, about double its 1990 population. Desperate for the water that the state water project was unable to provide the past few years, farmers were drilling deep wells and pumping ever deeper, draining aquifers around the state even more dramatically than they have in the past. Even prior to the severe recent conditions, pumping has caused the ground to subside over twenty-five feet in some areas of the state. As that ground sinks, the subsurface areas that were once saturated with water become compressed, effectively assuring that the water to recharge such aquifers will be far less able to do so, even in good years with plenty of water.


Like it or not, the future of California will entail an ever-greater reliance on surface water sources, portending even greater economic and social disruption when future droughts inevitably occur. The state water project that has made California the garden of the USA is a marvel of engineering, a testament to the foresight of those policymakers who laid the groundwork for what is arguably the most impressive such system in the world. But for all its virtues, the system is insufficient to meet serious droughts, and woefully incapable of supplying the water needs of California’s increasing population and agricultural productivity in the decades to come.

California has already tapped virtually every surface water source that might possibly be channeled into the state’s system. Over 700 miles of pipelines and aqueducts channel water from far-flung sources into the thirsty Central Valley and the cities of the Bay Area and the southern part of the state. Even before the most recent drought, the demands of agriculture and cities pushed the system to such an extent that salmon, smelt, and other aquatic species saw serious impacts due to water levels sometimes too low to provide the environment that they require to maintain their life cycles. The plan currently being considered to dig two water diversion tunnels under the Sacramento/San Joaquin Delta with the intention of lessening that environmental impact has been estimated to cost as much as $67 billion[1], yet it would not add a single drop to the state’s water supply. In a candid admission of that fact, the director of the California Department of Water Resources commented when discussing the water tunnel project, that “We’re not going to drought-proof California.”

But we can, with a different plan.

The last few years saw an unusually high number of acres lying fallow for want of water, and some fruit and nut groves, which take years to develop, died for lack of water. The decline in productive farms as a result of insufficient water will result in the loss of many thousands of jobs, with just the direct costs of this agricultural disruption estimated to exceed $5 billion, but the knock-on effects just within the farming communities will also be severe. And since California normally produces about half of the fruits, vegetables, and nuts for America’s tables, the inevitable rise in prices for these products have an economic impact that ripples across the nation.

As bad as this sounds, it is but a microcosm of a global situation that is far worse in many countries. Billions of people already live in water-stressed circumstances today, and aquifers from the Great Plains to the Middle East are being pumped dry to provide for an ever-increasing population. With over 7.2 billion people on the planet today, demographers predict a mid-century global tally of at least nine billion and, very probably, ten billion or more. Providing water for not only the personal needs but the agricultural needs of a further three billion people is a challenge that can only be met by desalination on a scale hitherto undreamed of.

If one were to ask an agriculturalist if it’s possible to provide food for ten billion people, the answer will almost certainly be, “Yes, if you can provide the water.” Similarly, if one asks a hydrologist how to provide the water required by ten billion people, the answer is likely to be, “Yes, if you can provide the energy.” In the end, it all comes down to energy. With sufficient affordable energy, desalination on a massive scale is certainly possible. But with the threat of climate change being one of the probable causes of severe drought in many parts of the world, it’s critical for that energy to be provided by carbon-free sources.

In just the past decade or so, desalination technology has improved dramatically. Middle Eastern nations such as Saudi Arabia, the UAE and Israel have pioneered ever-larger and more efficient desalination plants. Experts from those countries are assisting in the construction of a desalination plant north of San Diego, as that city hopes to make itself less dependent on the state’s water system. Saudi Arabia, foreseeing an ever-increasing demand for water, is building the world’s largest desalination plant that is expected to produce as much as 1.5 million cubic meters of water per day once it’s complete.[1] There, as in so many other areas of the world, increased pumping from aquifers has resulted in estimates that their deep groundwater reserves will be gone in 25 years.

This is the shape of things to come. Barring some unforeseen cataclysm, we’ll have ten billion people on the planet within a few decades, and providing for their necessities—much less a comfortable standard of living—will be an unprecedented challenge of epic proportions. California’s current and future dilemma is but a foretaste of the sort of challenges that have to be met on a global scale.

The good news is that California can again take the lead in demonstrating a solution to the problem of water and agricultural plenty. We already have the technologies to drought-proof California. Such an accomplishment would stabilize not only the economy of the nation’s garden state, but the entire nation’s food supply. And that accomplishment can demonstrate a path forward that will enable every nation’s standard of living to rise even as we finally reach what is projected to be the planet’s peak population. The fact that this can be accomplished in an environmentally benign manner is all the more encouraging.

Making Use of What We Have

The California Water Project (hereinafter CWP) is an invaluable resource that would need only modest modification to manage a great deal more water. Seemingly out of desperation, there are calls for more dams and water storage in order to corral more water in wet years for use in the dry ones. But it doesn’t take a prophet to imagine the hue and cry that would arise from even the suggestion of building more dams. And even with more reservoirs, when the snowpack in the mountains doesn’t materialize, having more reservoirs just means having more empty reservoirs.

The anchor of the entire CWP lies in the north. Three reservoirs there—Oroville, Shasta, and Trinity Lakes—comprise half the water storage capacity of the entire CWP (and some of the smaller ones are simply holding water that’s been channeled from those three). This makes the entire system overly

dependent on the snow and rain that fall in the northern part of the state, an area that in 2014 and 2015 was at a level less than 20% of normal as of mid-March, near the end of winter. Even if we had twice the storage in those reservoirs, it would have done no good then, or in most other years. The winter of 2015/2016 fortunately has finally brought some of the reservoirs (including the two biggest) back to normal, though the snowpack in the mountains that feed those reservoirs through the summer still is less than normal.


The CWP has long-term delivery contracts for about 4 million acre-feet of water per year, though in most years it is incapable of providing that much. Shasta Lake, the largest of the CWP reservoirs by far, holds about 4.5 million acre-feet when it’s completely full, which isn’t often. We can see from the illustration, though, that this year the reservoirs are being maintained at historically high levels thanks to the El Niño bailout, though even with that Trinity Lake hasn’t been able to recover. Trinity Lake and Lake Oroville together hold another six million acre-feet. Of course much of the water in those reservoirs necessarily runs down the Sacramento River to the sea. In drought years especially, the allotment of what water is there creates an almost inevitable conflict between agriculturalists and thirsty cities on one side and environmental necessity on the other. Unless we can provide a lot more water in the future—no matter the precipitation—something’s got to give.

The logical approach to the stabilization of the system would be to “create” a substantial amount of fresh water and use our existing reservoirs to impound it for the use of both farmers and cities. It so happens that this can be done without building a single new dam.

When Shasta Dam was partially built, World War II intervened. Suddenly both steel and manpower were being diverted to the war effort. So even though Shasta Dam was designed to be 800 feet high, it was decided to build it 600 feet high. There has been much discussion about raising the dam either 100 or the full 200 feet to its designed 800-foot height. Raising it 100 feet would approximately double its capacity. 200 feet would triple it.

Since the CWP contracts to deliver (when possible) four million acre-feet of water per year, raising Shasta Dam just 100 feet would add more than that much extra capacity. If Shasta Lake could be reliably at or near full by the end of every winter, it would guarantee that the entire agricultural sector of California would be supplied while assuring plenty of flow in the Sacramento River for environmental stability. The dependability of water supply would not only provide economic stabilization for agricultural enterprises and their many employees and owners, but it would smooth out the roller coaster of price swings due to produce shortages, which our whole country experiences when the agricultural system is disrupted.

When the CWP started delivering water to the farms in the state, groundwater pumping diminished considerably and the subsidence of the land consequently slowed. Land subsidence has increased in drought years and slowed in good water delivery years. But because the aquifers get compacted when the land subsides, years of heavy pumping such as we’ve seen recently and can expect to see again in future low-water years will only exacerbate the problem that can’t really be reversed. Already the subsidence is so serious that it’s affecting the flow of the CWP aqueducts in some areas. They’ve been designed with a sufficient gradient to move a substantial amount of water at a certain speed. If the system begins to sink in places, it’s not going to be easy (or cheap) to re-route the aqueduct. In the short term, we may well be facing groundwater pumping restrictions in order to minimize the subsidence problem, even though such regulation would almost certainly result in the loss of orchards and vineyards, a loss of decades of effort and investment in each case. In the long term, we have to plan to provide the needed water from the surface to minimize pumping to an extent that subsidence can be stopped and aquifers can recharge where possible, though many of them have been irredeemably shrunken.

Designing the Future

Shasta Lake is the key to the future of California agriculture. Trinity Lake—third largest reservoir in the state at almost 2.5 million acre-feet, would likewise be involved, albeit without any modification. Lake Oroville would not be affected, since its location at the east side of the Central Valley restricts its catchment to the Sierra Nevada Mountains.

The west side of Trinity Lake is approximately 120 miles from the Pacific. Its surface water level is over a thousand feet higher than Shasta Lake. A desalination system could produce fresh water at desalination plants along the coast in the vicinity of Eureka, then pump the water up over the Coast Range and into Trinity Lake. With Trinity and Shasta Lakes only about twelve miles apart at their closest points and Shasta Lake so much lower, a simple overflow canal without any pumps would be able to fill Shasta Lake as water is pumped into a full Trinity Lake.

Whenever a major infrastructure project is built, it’s incumbent upon its designers to assure that it will be used to its maximum potential. It should be quite possible (albeit expensive) to produce and transport up to 4.5 million acre-feet per year. That’s the same amount as Shasta Lake’s volume will increase if the dam is raised one hundred feet, as described earlier. It would guarantee that in most years the Eureka Project (an apt name both for the location of the water source as well as the allusion to Archimedes’ bathtub) could run full-out and maintain both Trinity and Shasta Lakes at or very near their maximum volume.

It’s important to look at how the CWP moves water today. The major pumping stations are all modular. In other words, they have multiple pumps and multiple pipelines that transport the water. So the Eureka Project would not have to be built all at once, so long as the pumping stations are designed with expansion in mind. The only portions of the system that would be built at maximum capacity would be the gravity-flow aqueduct sections (including any water tunnels that might be cut through ridges in the Coast Range).

The pumping stations and pipelines that serve the CWP were mostly built 30-50 years ago. Several pumps have been replaced since, undoubtedly with greater efficiency. No new technology would be required for such a system, but the logistics—both physical and political—would be daunting. The production of 4.5 million acre-feet of water per year would require the construction and continuous operation equivalent to ten of the largest desalination plants ever built. The sheer amount of power to operate the desal plants and to pump the water over the Coast Range would be considerable, as would the other costs of operation. And of course the political battle that would be waged over the desal plants can easily be imagined.

A Better Idea

At this point one is tempted to look at a map to imagine the geographical aspect of such a plan, and if that map extends a bit northward from California it’s nearly impossible to ignore that just about 400 miles to the north there is a vast quantity of perfectly good fresh water constantly flowing to the sea via the Columbia River that delineates the border between Oregon and Washington. It’s easy to visualize that very water flowing down the coast and then being desalinated at great expense to pump into California’s reservoirs.

Neither the Columbia’s distance from Lake Shasta nor the technology needed to divert water directly from the Columbia to Shasta are deal-breakers. The length of the system would be a bit over half the 700-mile length of the existing California water system aqueducts, and the pumping stations to move that water the approximately 400 miles south would likewise be substantial. But once in place, the operational costs of a Columbia diversion project would be trivial compared to that of a massive desalination system as described above. And let’s remember that the desal proposal would require at least 135 miles of aqueducts and a lot of pumping power in addition to the power that would be needed for the desalination itself.

It’s no secret that Oregonians and Washingtonians often have a low opinion of Californians, and politically that nonsense alone would be a factor in such a plan. But let’s just look at the actual environmental and other challenges and leave state chauvinism out of the equation for the moment. Would a Columbia diversion project make sense?

Any time a water diversion project is contemplated there’s always the legitimate concern about how it affects whatever and whoever is downstream. Since the Columbia River has many dams in place, diverting water is going to affect the reservoirs and dams downstream, at least to some degree. Even if that effect is small, the very perception of someone downstream getting shortchanged will make upstream diversion a difficult political prospect, so the logical solution to that potential problem would be to divert the water from below the last dam on the Columbia, after which the river flows unhindered to the sea. The last dam on the Columbia is the Bonneville Dam, located just west of the Cascade Range and about 40 miles east of Portland. It is an unusual dam, as can be readily seen in this aerial view:

Contrary to the usual type of dam that holds back an entire river, Bonneville is what’s called a run-of-the-river dam. You can see the turbine spillways releasing their white water in the center. Besides this, though, locks and fish ladders allow most of the flow of the Columbia to flow on either side. The 20 turbines of the Bonneville power station each carry about 1.25% of the flow of the Columbia River, leaving 75% left unaffected by the turbines.

Environmental Impact

Any time water is pulled out of a river in substantial quantities, there’s always a concern about the environmental impact, the fish and other marine creatures that will be sucked in and killed in the process. For the Columbia River this issue has added salience, since the river is home to a substantial commercial salmon fishery. This fact led the builders of the Bonneville Dam to ameliorate the impact by putting an impressive fish ladder system in place for the salmon, and to this day the fish ladders are a tourist attraction during the summer salmon migration.

The Bonneville Dam has been operational for many decades and clearly the Columbia River salmon fishery is none the worse for it. Despite the fact that a certain amount of salmon fry meet their end in the turbines on their way to the sea, the efforts to minimize that problem is greatly aided by the fact that the turbines only have a maximum of 25% of the river flow running through them. Those turbines are the key to eliminating the environmental impact of diverting water to an aqueduct system to California.

Since the output of the turbines doesn’t contain live fish, the trick is to divert water directly from a few of the turbines’ outflow into a pumping station that would send the water into the aqueduct system. That way there would be no question of such a diversion affecting the salmon run or any of the other creatures that might otherwise be impacted by pumping water out of the river in the way it’s done from the Sacramento River as some of its water is diverted into the California Water Project. This environmental impact has been a constant source of grief, and by pulling water directly from some of the Bonneville turbines that problem would be eliminated. Unless such an environmentally-benign method is employed, the impact on salmon and other river life would certainly be one of the biggest objections to this project and would likely make it a non-starter.

Number Crunching

As we saw earlier, the current capacity of Shasta Lake is about 4.55 million acre-feet. If Shasta Dam would be raised another 100 feet, that capacity would exceed 9 million acre-feet. Building the dam up to its original 800’ intended level would increase that capacity to a whopping 13.9 million acre-feet. Even if Shasta Lake could be guaranteed to be full every spring at just its present capacity, it would go a long way toward stabilizing agricultural and other water demands in California. At present, water contracts with the State Water Project (SWP) entitle their holders to 4.2 million acre-feet, pretty close to a full Shasta Lake. But they generally receive only 50-90% of their contracted amount. On average, the SWP delivers between 2.5 and 3.5 million acre-feet, though in the drought of 1991 only a little over half a million were delivered, and during the most recent drought the state announced that it wouldn’t deliver on any of its contracts.

These statistics clearly indicate that if California were to raise Shasta Dam just 100’ and manage to keep it full in the spring, California’s water troubles would be history. We learned from the past few years that a diversion of 4.5 million acre-feet wouldn’t be enough in some years to fill the expanded reservoir, though combined with the other reservoirs in the system it would have to be a very serious drought indeed to substantially stress the system.

Let’s assume that we raise Shasta Dam by just 100 feet, increasing its holding capacity to about 9 million acre-feet. Could we realistically divert 6 million acre-feet per year from the Columbia River without a substantial environmental impact? If so, we could pretty much guarantee water stability to the entire California system, both agriculture and cities, even in multi-year drought conditions. And given the impact of climate change that we’re already experiencing, we would be foolish not to plan for such droughts. Look at this past winter, 2015/16. This was a year of an exceptionally strong El Niño, yet when the snowpack was measured on the usual April 1 measurement day, it was slightly below normal for that date. One would have assumed that the El Niño would have dumped substantially more snow than normal that winter. So if less than normal snowpack is the result of a powerful precipitation driver like we saw this year, California will be in for some more long-term droughts in the not too distant future. But six million acre-feet, drought or no drought, would do the trick.

So let’s look at the numbers. As mentioned earlier, each of the 20 turbines on the Bonneville Dam carry 1.25% of the Columbia’s flow. That amounts to a flow rate of a little over 100 cubic meters per second (cm/s). For the project to move 6 million acre-feet per year to California, it would require a flow rate of about 235 cm/sec. That means that the output of just slighty more than two of the Bonneville Dam’s 20 turbines (just a shade more than 2.5% of the Columbia’s flow at this last dam before it flows to the sea) would be all that California would need.

How does this flow rate compare to elements of the California system already in use these past decades? The Banks Pumping Station, the first one in the CWP, pumps at a rate of 302 cm/s, enough to handle the output from three turbines at the Bonneville Dam. The California aqueduct has a capacity of 370 cm/s, close to the output from four Bonneville turbines. The pumping stations and the aqueduct system they serve in California were built decades ago. There is certainly no technical or engineering issue that would make this project impossible if the politicians decided to do this.

Cui Bono?

A glance at the precipitation map illustrates in no uncertain terms that the Cascade Range forms the western border of what may not technically be a desert that makes up about 60% of the state of Oregon, but which is certainly semi-arid. But the Central Valley of California is also semi-arid, yet it is the garden of the USA. It’s all about the water.

Under normal conditions, California’s agriculture industry is worth about $40 billion or more. Is there any reason why Oregon’s semi-arid land east of the Cascades couldn’t bloom like California’s Central Valley? Imagine the countless jobs and tens of billions of dollars in annual revenues that could utterly transform Oregon into a garden state, turning previously barren land fruitful just as we’ve seen the CWP transform California.

If the output from just over two turbines at the Bonneville Dam can keep California’s bountiful agriculture industry thriving, why not build the Columbia Diversion Project to divert the output from four of the Bonneville’s 20 turbines? The dam is located just slightly west of the Cascade range. Yes, it would take a lot of electricity to pump the water back east a ways so that it could flow through an aqueduct system along the eastern Cascade foothills. But from that aqueduct system, Oregon farmers (or those who would soon be farmers) could run spurs off to the east to create Oregon’s new booming agriculture industry. If California and the federal government offered to pay for the main aqueduct system while permitting Oregon landowners to build (at Oregon’s expense) spurs off of it capable of using nearly half the water, it’s hard to imagine that the project couldn’t win the support of the majority of Oregonians. After all, the environmental impact would be minimal, and atop all the jobs that would materialize in the new agricultural industry, there would be thousands of good jobs created just building the water system in the coming years. As Oregon’s climate warms in the coming years, a lush and bountiful farmland would replace a potential desert.

Of course it’s impossible to get too specific about the costs of all this, since the route would have to be determined and surveyed. It’s possible that the dam at The Dalles—the next dam upriver from Bonneville—would be used as the source instead of the Bonneville Dam. That would undoubtedly save a lot of money since it’s already on the eastern side of the Cascade Range, and at a higher elevation. Like the Bonneville Dam, it is a run-of-the-river dam, with 22 turbines of higher capacity than the Bonneville turbines, averaging 169 cm/sec. If these turbines were to be used, the output from just three of them would be about equal to that of five at the Bonneville Dam. From a logistical and economic standpoint, The Dalles Dam would be a better choice. The possible political perceptual downside, however, is that you can’t claim that “it’s all just flowing unused to the sea” from there, since the Bonneville Dam is downstream. But the arguments for using The Dalles would be pretty easy to make.

As for the electricity to run the pumping stations, that could originally be generated by the dams themselves. In the future, however, if economical and ultra-safe modular nuclear reactors (such as the molten salt systems currently under development) would be built to power the system and surrounding communities, they could be built right along the aqueduct. The water used to cool the reactors—often a point of complaint from environmentalists—could be the same water, which as we know from its source would have no fish in it. Here again, the environmental impact would be nil.

During the recent drought, farmers in California’s Central Valley were contracting for scarce water at prices as high as $1,000 per acre-foot. Even at 1/10th that price, six million acre-feet per year being sluiced down from the Columbia to California would bring in $600 million per year, not to mention whatever money Oregon might extract from its farmers who would be using their share. Sure, the project would be expensive to build, but not nearly as expensive to operate as a desalination project like the one described in the beginning of this document, and the long-term benefits to both California and Oregon would be staggering.

What Other Options Are There?

The USA can hardly afford to let its garden state dry up. Using this plan, it would instead add another garden state to fill its larders. There has been a lot of talk about building more reservoirs and dams in California, channeling water this way or that, but none of those plans actually create any new water. When the snowpack doesn’t materialize, the state will be as desperate as ever. The fact that over-pumping is making California more dependent than ever on surface water is all the more reason to recognize and embrace a plan that will bring in dependable water in great volume even during droughts, ensuring the state’s verdant future.

The fraction of water from the Columbia that this would divert is very likely within the range of normal fluctuations. Even if the Columbia is lower than normal due to drought, the argument is the same: from here it just flows to the sea. It is highly unlikely that even up to 6% of its flow being diverted would make a meaningful difference to the salmon runs or other aquatic life, though of course the environmental impact report accompanying such a plan would have to clarify and confirm that.

The best time to plant a tree was 25 years ago. The second-best time is today.

Tom Blees

President, The Science Council for Global Initiatives
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.
Tel: 530-848-6546