fter the levees broke in New Orleans and flooded most of the Big Easy, it was not difficult for the media and flood control experts to identify California as the most likely site for another levee-related disaster. More than 2,400 mi (3,800 km) of aging, deteriorating levees protect two key regions in the Golden State—regions that might become the next New Orleans. The two regions of concern are the Central Valley, which comprises the basins of the Sacramento and San Joaquin rivers, and the delta shared by those two rivers, which is about 50 mi (80 km) upstream of San Francisco Bay.

As was the case in New Orleans, the possibility of catastrophic floods in either the Central Valley or the delta is not a new concern; major floods occur every decade or so. In the Central Valley, three people died during the last major flood, in January 1997, when a levee failed on the left bank of the Feather River, a floodwater bypass called the Sutter Bypass failed without warning, and more than 30 levees failed on the San Joaquin River. More than 120,000 people were forced from their homes, and 55,000 of them were sent to temporary shelters.

This January, California’s Department of Water Resources (DWR) warned in a report entitled Flood Warnings: Responding to California’s Flood Crisis that levee failures during the Central Valley’s annual winter and spring flood season posed a “ticking time bomb for flood management in California.”

Meanwhile, in the delta—where levees protect some 60 islands below sea level—there have been approximately 160 breaches within roughly the past century, notes Les Harder, Ph.D., P.E., M.ASCE, the acting deputy director for public safety for the DWR. Harder describes the delta levees as “the hardest working” flood control structures in California because they must constantly hold back water, whereas the Central Valley levees generally protect only against seasonal flooding. The cause of breaches in the delta is sometimes unclear, he adds. In June 2004, for instance, a levee broke in a part of the delta known as Upper Jones Tract, an island roughly 10 mi (16 km) west of Stockton, causing nearly $100 million in emergency response and water pumping costs, damage to private property, crop loss, and levee repair. But no reason for the break has been discovered. “All the evidence got washed away,” Harder explains.

The delta levees especially worry Raymond B. Seed, Ph.D., M.ASCE, a civil engineering professor at the University of California at Berkeley, who envisions an “Armageddon-like scenario” if an earthquake strikes the delta levees. That’s because roughly two-thirds of Californians—more than 20 million people—obtain at least some of their drinking water from the delta. Much of California’s agriculture industry also relies on delta water for irrigation. But that water could be rendered nonpotable if an earthquake breached enough levees simultaneously to draw in vast quantities of salt water from San Francisco Bay. Parts of California “would cease to be a modern society,” warns Seed, who helped write the report Seismic Vulnerability of the Sacramento–San Joaquin Delta Levees. Released in April 2000, the report was sponsored by the CALFED Bay-Delta Authority, a landmark partnership between state and federal agencies involved in protecting the ecosystem encompassing San Francisco Bay and the delta. Compared with the pre-Katrina situation in New Orleans, where 129 mi (206 km) of earthen levees protected fewer than half a million residents in a single metropolitan area, the area in jeopardy in California encompasses more than 2,400 mi (3,800 km) of levees in two distinct geographic regions, with a nearly New Orleans–sized population endangered in Sacramento alone. Another 10 million or more Californians might be forced to abandon their homes in the worst-case scenario—that is, if a delta catastrophe cut off their access to freshwater from that region. Moreover, while the levees in New Orleans faced their greatest threat from a large hurricane—which is often tracked for days before making landfall—the Armageddon scenario for California’s delta levees involves that most unpredictable of natural disasters, an earthquake.

Seed and another leading researcher on the California levees put the potential for a West Coast catastrophe into starkly post-Katrina terms. According to Seed, an earthquake that is strong enough to damage dozens of the delta’s levees—thus rendering the region’s freshwater undrinkable for a year or more—can be expected roughly every 200 years. “That’s an interesting number,” he notes, “because once every two hundred years was the expected frequency of how often New Orleans would have its levees breached by a category-four or bigger hurricane—stunningly the same level of risk in terms of likelihood.”

Jeffrey Mount, Ph.D., a professor of geology at the University of California at Davis’s Center for Integrated Watershed Science and Management, delivered a lecture at the university last month entitled “The Sacramento/San Joaquin Delta: The Next New Orleans?” In an interview for this article, Mount noted that New Orleans had two to two and a half times the level of flood protection currently guarding Sacramento and other major cities in the Central Valley.

In a sign of how complicated the situation in California can be, even the experts don’t agree on what constitutes the greatest threat to the Golden State. Seed clearly sees a seismic event in the delta region as the worst-case scenario. Mount believes that the greater problem lies in the fact that the Central Valley levees—originally built to protect agricultural land—are now forced to protect urban development in one of the fastest growing regions in the country. “The [Central Valley] levee system was not designed to support widespread urbanization,” Mount says. “Yet that is exactly what’s happening, further taxing a decaying infrastructure.”

What the Central Valley and the delta share, it seems, is the possibility that each is a major disaster waiting to
happen.

he State of California and the U.S. Army Corps of Engineers are currently most focused on the Central Valley, which drains an area of more than 43,000 sq mi (111,300 km¾). In September, alarmed by what happened to New Orleans, California’s governor, Arnold Schwarzenegger, its two U.S. senators, Diane Feinstein (D) and Barbara Boxer (D), and a bipartisan group of U.S. House members from northern California joined forces to seek $92.6 million in federal funding for flood protection efforts in the Golden State. According to a September 14 letter Schwarzenegger sent to Feinstein and Representative Richard Pombo (R), the “Sacramento metropolitan area has the lowest level of flood protection for any large urban area in the nation. Other river cities such as Tacoma, Dallas, St. Louis, and Kansas City have 500-year flood protection. Even New Orleans had a 250-year level of flood protection.” But Sacramento’s protection is only for a 100-year event, which Schwarzenegger called “woefully inadequate.”

Most of the money sought would be applied to Corps projects in or near the Central Valley, primarily along the Sacramento and American rivers and at Folsom Dam (located downstream of the confluence of the American River’s north and south forks); more than $4 million would reimburse the Sacramento Area Flood Control Agency for work already completed. Only $3.5 million would be appropriated for the delta, and those funds would only be for studying the problem.

Delta and Bay Region

State of California, DWR

This focus on the Central Valley is not surprising, given that a “major flood in the Sacramento area alone would put at risk over 400,000 people, 170,000 structures, 117 schools, and a potential for $7 [billion] to $15 billion in damages,” explains a 2004 paper, “A Review of Corps of Engineers Levee Seepage Practices in the Central California Flood Control System,” presented at the 24th annual conference of the United States Society on Dams. The paper was written by John R. Hess, P.E., M.ASCE, the chief of geotechnical and environmental engineering for the Corps’s Sacramento district, and George L. Sills, P.E., A.M.ASCE, a geotechnical engineer with the Corps’s Engineering Research and Development Center, in Vicksburg, Mississippi.

The Sacramento River area is most at risk every decade or so from major floods that are caused by heavy winter rains and warm winter weather that causes rapid snowmelt; the San Joaquin River area faces major flooding on a similar schedule from both winter rains and snowmelt.

Unfortunately for the Central Valley, its flood control system of levees, channels, and weirs is large, old, and “significantly deteriorated, partly due to deficiencies in the original design and partly due to deferred maintenance,” concedes the DWR’s January 2005 report, Flood Warnings: Responding to California’s Flood Crisis. The system comprises roughly 1,600 mi (2,570 km) of earthen levees built primarily by local farmers and large agricultural interests over a period ranging from California’s Gold Rush, in the mid-1800s, to the early 1960s. Most of these levees were built “using hydraulic fill, using no real design practices or construction standards, which resulted in highly pervious, low-density embankments subject to stability, seepage, and erosion problems,” explain Hess and Sills in their paper. Moreover, “minimal data concerning construction or the nature of the foundation and levee materials” are available for these structures.

 

Typical Channel Cross Section on the Sacramento River Corridor

Jones & Stokes

The system developed as a result of “dueling levee” incidents: A farmer on the east side of the Sacramento River would raise up a shovel-built levee 5 ft (1.5 m) high to protect his land, but this levee would flood the farm on the west bank. So the west bank farmer would build an even taller levee and that would flood the east bank farm or other farms downstream. To make matters worse, during the Central Valley flood season—which begins each November—one farmer might deliberately destroy another farmer’s levees in an effort to save his own land from flooding, notes Peter D. Rabbon, P.E., F.ASCE, the general manager of California’s Reclamation Board. The Reclamation Board was established in 1911 precisely to end such chaos and develop and oversee a single flood control plan for the Central Valley, says Rabbon. Though the board is administratively part of California’s DWR, it maintains separate and independent decision-making powers, Rabbon says.

The Corps became involved in 1917, when the Central Valley was one of the first two areas designated by Congress for federal flood control projects (the other being along the Mississippi River). But while the Corps has built reservoirs and dams in the Central Valley over the years, it has primarily just enhanced or rehabilitated preexisting levees after floods or other damage; it has not built many new levees, notes Jim Sandner, the chief of the operations and readiness branch of the Corps’s Sacramento district.

Additionally, the Water Resources Development Act of 1986 requires the Corps to share the costs of most flood control projects with a nonfederal or local sponsor. (The Reclamation Board is the nonfederal sponsor for most Central Valley levees.) Under this arrangement, the Corps generally pays 65 percent of the costs, the Reclamation Board 25 percent, and the local levee district 10 percent. Operation and maintenance costs are turned over to local entities entirely, says Sandner. The Corps generally designs such projects but contracts out the actual construction, he adds.

Most levees in the Central Valley system were once privately owned, but gradually they have been grandfathered into the federal flood protection program. Thus while the federal government exercises legal control over these “project levees,” the structures themselves are owned, operated, and maintained by local levee or reclamation districts or by the State of California itself; the Reclamation Board, however, believes that actual ownership is an unresolved issue. The Reclamation Board works with roughly 130 such local districts throughout the state, roughly half in the Central Valley and half in the delta, says Rabbon.

 

The Sacramento River and the San Joaquin River Flood Control System Project Levees and Channels

State of California, DWR

The Corps still inspects project levees to determine whether they meet the minimum standards for assistance under Public Law 84-99—which prescribes whether the Corps will help repair a levee in the event of damage—and it also inspects a small number of privately owned levees in the Central Valley whose owners meet the qualifications for inclusion in the PL 84-99 rehabilitation program. The Corps also has the authority to fight floods to save lives or protect property that has been developed for human habitation or commercial use whenever the Sacramento district commander issues a declaration of emergency. Following such emergencies, damaged levees eligible under PL 84-99 can be repaired to their preflood condition without cost to the local sponsor of a federal flood control project, says Sandner; a nonfederal project sponsor must pay 20 percent of the cost. After major flooding in 1997, for instance, the Corps helped repair some 600 levee sites in the Central Valley. But the local sponsor must pay the costs of any improvements to the levees.

For a levee to be eligible for PL 84-99 assistance, the minimum standards include a 3:1 slope on the side facing the water and a 2:1 slope on the land side, Sandner explains. The top of the levee would need a crown at least 20 ft (6.1 m) wide that would support an all-weather road so that the structure could be more easily inspected during a flood, and there would have to be at least a 10 ft (3 m) easement on the land side that was kept clear for inspections and flood-fighting activities, he adds.

A well-designed levee eligible for PL 84-99 assistance will also include a minimum 3 ft (0.9 m) freeboard (the difference between the height of the levee and the project design water elevation during a 100-year flood), Sandner says.

The height of levees in the Central Valley varies considerably, mostly in a range of 12 to 25 ft (3.7 to 7.6 m), although some areas with shallow flooding can get by with levees of just 3 to 5 ft (0.9 to 1.5 m), says Rabbon. Crowns can be wide enough to accommodate a state highway and in some cases even buildings. But many Central Valley levees are actually relatively small and narrow compared with levees along the Mississippi River, which can have bases of several hundred feet. California levees in some cases may measure 100 ft (30.5 m) at the base, notes Sandner. And since many of these levees were constructed with material dredged right out of the adjoining river, they often suffer from “a sandy foundation that is fairly permeable and will allow seepage,” explained Hess in an interview for this article. “We certainly prefer levees that are built out of clay, but we’re stuck with the foundations that Mother Nature gives us.”

Hess says that a well-designed and well-engineered levee should be tightly compacted in 6 in. (152.4 mm) layers to produce a very high density. But many of the Central Valley’s levees were originally built as quickly as possible by just dumping material in place with no compaction. “So the density is extremely low,” he says.

To make matters worse, the original flood control system in the Central Valley was intended as much for scouring out mining debris—sediment that clogged rivers as a result of hydraulic gold mining—as for flood protection, explains Harder. “Now those sediments [from mining] are all scoured out, but the river still wants to scour, wants to erode. So it’s difficult to maintain the levees when you have a highly erosive scouring action.”

Indeed, a 2003 field reconnaissance report for the Corps documented 184 bank erosion sites along the Sacramento River and tributaries, up from 183 sites the year before. But many “of the sites showed an increased amount of erosion and the number of critical and potentially critical sites has increased from 24 to 37,” noted the Corps’s contractor, Ayres Associates, of Sacramento, which conducted the survey using a portable Global Positioning System receiver.

The floods of 1986 and 1997 revealed that “underseepage” poses a major problem for Central Valley levees, notes Hess. Underseepage occurs in the levee’s foundation—rather than through the levee itself—and was blamed for failures in 1997 in one levee that had been strengthened to withstand seepage and in another that had not previously exhibited problems.

Deferred maintenance also is a problem for the Central Valley levees, mainly because of the expense involved. It costs California an average of $17,700 per mile ($11,000/km) to maintain roughly 152 mi (245 km) of levees in the Central Valley, notes Rabbon. The most expensive area is a roughly 20 mi (32.2 km) stretch in the city of Sacramento, which costs $62,900 per mile ($39,093/km).

Though many local levee districts have lower maintenance costs, the overall expense of such work is rising, Rabbon says. Thus, some districts, especially in rural areas, put off the necessary work, he adds.

Tough federal and state environmental laws add to maintenance costs. Many endangered species of plants and animals inhabit the Central Valley river basins, driving up operation and maintenance bills, notes a 2002 report, Sacramento and San Joaquin River Basins Comprehensive Study, prepared jointly by the Reclamation Board and the Corps.

And remember the problems associated with scouring from the days of hydraulic mining? To fight the resulting erosion, “we used to dump rock on the scour sites, but we can’t do that [now] because of its impact on habitats and fisheries,” explains Harder. “So today we have to do a lot of environmental mitigation, and if we are going to repair erosion sites we have to do it carefully and put in some kind of restoration in terms of plantings and so on, which makes things more expensive, more difficult to do, more time consuming.” Such efforts, Harder adds, consume resources “that were never envisioned a hundred years ago.”

Each spring and fall, the DWR inspects maintenance conditions throughout the Central Valley’s levees and issues a report. The 2003 report, published in November 2004, found more than 14 mi (22.5 km) of levees along the Sacramento River and its tributaries to be “non-compliant,” which meant that maintenance work departed significantly from federal and state guidelines, for example, by failing to keep levee crown roadways passable during the flood season or failing to remove undesirable growth from levees, says Rabbon.

In addition, approximately 102 mi (164 km) of Sacramento River levees and 19 mi (30.6 km) in the basin of the San Joaquin were found to be in need of improvement. In other streams throughout the Central Valley system, roughly 14 mi (22.5 km) were deemed noncompliant, and roughly 123 mi (198 km) needed improvement.

The 2004 flood on the Upper Jones Tract damaged private property as well as farm crops. No precise cause was ever determined for the initial break in the levee. State of California, DWR

Rabbon notes that it is possible to provide good maintenance to a deteriorating levee and poor maintenance to a healthy structure. Still, when looking at the overall Central Valley, Sandner sees it this way: “I would say there probably is not really an ideal stretch of levee” anywhere in that 1,600 mi (2,570 km) system. Some of the levees are in such poor condition that they might even be declared “inactive” by the Corps under PL 84-99, he adds.
In addition to the threats posed by poor construction, seepage, erosion, and lack of maintenance, levees in the Central Valley must contend with increasingly threatening weather. The American River’s Folsom Dam—which provides flood control space and the capability of releasing excessive flows downstream—was constructed by the Corps in the 1950s on the basis of hydrology records from the first half of the 20th century, notes Hess. But since then “we’ve had five storm events, each of which was greater than anything in the previous fifty years,” he notes. “So the hydrology is changing—we’re getting more frequent and more severe storms, so that means the level of protection [needed] is changing in California.”

As a result, meeting the Federal Emergency Management Agency (FEMA) requirements for 100-year flood protection under the National Flood Insurance Program “is a significant challenge in the Central Valley,” note Hess and Sills in their seepage paper. The Central Valley levees were never intended to provide any designated level of flood protection, notes Rabbon. Instead, their original purpose was to pass along a certain flow of water and, in some cases, intentionally flood agricultural land. The Sacramento River flood control system, for example, has a design flow of about 600,000 cfs (17,000 m„/s), Rabbon says. But during periods of high water about 80 percent of that flow—roughly 500,000 cfs (14,160 m„/s)—is taken out of the Sacramento River and sent to the Yolo Bypass, a system of levees the Corps built on low-lying agricultural land in the 1950s, Rabbon explains.

“If you were to drive over [the Yolo Bypass] in the summertime, you would just see flat fields, lots of row crops, lots of agriculture,” Rabbon says. “But if you go over it in the winter and the bypasses are flowing, you would see water up to, say, eight feet deep,” stretching for as much as a mile from one levee to the other.

Unfortunately for the levee system, urban development is on the rise in the Central Valley, turning land that was once agricultural into residential and business areas. This means that the levees are now shouldering much more responsibility than their builders intended. Sacramento County, for instance, is among the 10 fastest-growing counties in the United States, according to 2000 census data. The county’s population grew by nearly 15 percent between 1990 and 2000, reaching more than 1.2 million residents, the Census Bureau reports. The city of Elk Grove, California—a Central Valley community heavily protected by levees—grew at the second-fastest rate in the nation last year, increasing by 10.6 percent from 2003, the Census Bureau says.

Such rapid development does not itself damage the levees, but it can hinder efforts to fight floods. “Along the American River there are houses that are built right up to that ten-foot line” required for easements on the land side of a levee, notes Sandner. “So it’s very, very hard to actually go in and do a flood fight in some of those areas.”

Although the Central Valley does not sit below sea level—in contrast to the case in New Orleans and the Mississippi’s delta—the levee system itself creates conditions that are comparable to being that low, notes Rabbon. “When you put in a levee system, you’ve taken water that used to spread out—and in the Central Valley it used to spread out for miles—and you have now confined that water in between levees, which means that same water all of a sudden has to sit much higher,” Rabbon says. His own home, in south Sacramento about half a mile (0.8 km) from the river, is not below sea level, he says. But during the winter the water behind the levees rises higher than the land. “If the levee breaks I can be flooded up to about ten feet deep,” he predicts.

California also recently took on a greater legal liability for its levees, according to the DWR’s January 2005 report, Flood Warnings: Responding to California’s Flood Crisis. As that report explains, a California appellate court decided in 2003, in Paterno v. State of California, that “when a public entity operates a flood control system built by someone else, it accepts liability as if it had planned and built the system.” The decision makes it possible for the state to “ultimately be held responsible for the structural integrity of much of the Central Valley flood control system,” which protects more than half a million people, 2 million acres (809,000 ha) of cultivated land, and approximately 200,000 structures with a total estimated value of $47 billion, the report says.

There is, however, one bit of good news for the Central Valley: it is not as threatened by flooding from earthquakes as is the delta of the Sacramento and San Joaquin rivers, notes Rabbon. That’s because the seismic event would have to strike exactly during a period of seasonal high water—a combination of catastrophes with a “very small probability,” he says.

he situation in the delta is quite different regarding earthquakes and other factors. For one thing, most of the 1,100 mi (1,770 km) of levees there are still privately owned and maintained. Moreover, these structures protect mostly agricultural land, much of it owned by corporate interests. Though some 400,000 people live on the outskirts of the region, few actually live on the delta’s islands, which number around 60.

The thin population also means that, except for approximately 275 mi (443 km) of federal project levees along the navigable channels, the Corps of Engineers has not inspected or rehabilitated the private delta levees because they are not part of the federal levee system in California. Additionally, the owners of private levees are not able to share the costs of levee improvements with the Corps unless they form a levee district or persuade the Reclamation Board to act as their sponsor, Sandner says. The Reclamation Board does, however, exercise some review over private levees in the delta through a program that pays roughly $5 million a year toward maintenance, as well as in situations in which a private levee owner wants to make changes that would affect a federal project levee, says Rabbon.

Once a sea-level marsh, the delta now encompasses an area of roughly 1,150 sq mi (3,000 km¾) reclaimed for agricultural use. Levee building began in the late 1800s, the work done mostly by the Chinese laborers who had helped build the western part of the Transcontinental Railroad. Pumps then lowered the water table to produce islands composed mostly of compressible peat, which produced exceptional farmland, notes Seed. Unfortunately, peat naturally decomposes and consolidates, which makes the islands subside. As a result, delta farmers had to keep raising their levees, eventually using barge-mounted dredges. But the extra weight of these taller levees compressed the soil, causing the levees to sink and forcing the farmers to build even higher ones.

Roughly two-thirds of California—more than 20 million people—obtain at least some of their drinking water from the delta protected by such levees as the one damaged in 2004 on the Upper Jones Tract. State of California, DWR

As pointed out in Seismic Vulnerability of the Sacramento–San Joaquin Delta Levees, the CALFED report released in 2000 that Seed and Harder helped to write, these levees “were generally built of nonselect, uncompacted materials without engineering design and without good construction methods.”

The western and central delta islands have been sinking at a rate of roughly 3 to 4 in. (76 to 102 mm) a year for 140 years, and much of that land is now 15 to 25 ft (4.6 to 7.6 m) below sea level. The original levees, in turn, were less than 5 ft (1.5 m) high at first but now rise to heights of 35 to 40 ft (11 to 12 m), notes Seed. “You can stand at high tide and see the water if you’re on the crest [of the levee] and then turn and look behind you and there, twenty-five feet below you, is this essentially flat piece of land, and at the very far edge of the horizon . . . is the other end of the island—it appears like the river has been lifted up above the land,” he says. “It’s very daunting.”

Historically, California earthquakes have not severely damaged the delta levees, which are located more than
16 mi (26 km) from the major strike-slip faults in the San Francisco Bay Area, namely, the San Andreas, the Hayward, and the Calaveras. There are, however, faults that are less active within 9 mi (15 km) of the delta, as well as “small but significant local faults in the Delta region,” notes the CALFED report. Thus, the “lack of historic damage” does not necessarily mean the levee system can withstand moderate earthquake shaking, the report warns. Instead, the “current levee system simply has never been significantly tested.”

The report also examined various seismic scenarios, including a magnitude 7.1 event along the Hayward Fault and a 6.0 earthquake immediately northwest of Sherman Island, which has some of the most seismically vulnerable levees in the delta, according to Seed. In the simulation described in the report, the Hayward earthquake produced 0 to 4 levee failures, a relatively low number, the report says. But the seismic event near Sherman Island produced a “fairly high” number of failures—as many as 32.

The report describes the annual likelihood of the Sherman Island scenario as lower than that of other scenarios in the study. Still, the Sherman Island levees are especially susceptible to seismic events because their foundations are “unusually sandy” and thus most likely to liquefy during an earthquake, Seed says. When a bowl-like delta island is breached, “it fills up and becomes a big pond, miles and miles across,” Seed explains. The wind blowing across such a newly made pond generates significant waves that would strike the inboard face of the levees on the other side of the former island. In contrast to the outboard faces, the inboard faces do not enjoy riprap protection against erosion because they are supposed to be on the dry side. Thus, they would erode very quickly. That is exactly what happened to a section called Frank’s Tract, which was an island “but became a huge lake in the middle of the delta” after its levees failed in 1938, Seed notes.

Sandner puts it this way: “If you have just one delta location that’s damaged it can cause a domino effect throughout the islands. If one levee fails then you have that whole huge island flooded, which puts more pressure on the other levees adjacent to that area, and that could cause those levees to fail and just keep working through the entire delta area.”

If the delta levees fail en masse, say, several dozen at a time in a seismic event, the result could be Seed’s Armageddon scenario: the breaches could not be repaired before enough salt water was sucked in from San Francisco Bay to contaminate the drinking water for 20 million people around the bay and in Southern California. If that happened, it might take several years to repair the damage and for spring runoffs to flush the salt out of the watershed. Other water sources could probably sustain only about half the population that now uses delta water, Seed adds. Thus, an estimated 10 million people might be forced out of the affected areas for well over a year, causing catastrophic economic and social disruption.

he problems with levees in the Central Valley and the delta are numerous. The potential for catastrophe is enormous and well known. The solutions, however, are more elusive and, in some cases, quite long term or prohibitively expensive. The 2002 report prepared jointly by the Reclamation Board and the Corps of Engineers, Sacramento and San Joaquin River Basins Comprehensive Study, tried to examine the flood control system “more holistically” instead of undertaking any number of “patchwork fixes,” notes Sandner. But that report’s broad recommendations for systemwide as well as regional measures were “probably overly ambitious” in that they required decades to implement at a cost of billions of dollars, he adds.

That report was also very general, discussing the construction of new levees or raising and realigning levees without actually naming the sites where such work needed to be done or describing the details of such projects. Indeed, it noted that while it was “widely anticipated that the Comprehensive Study would ultimately recommend a ‘Master Plan’ for the development of integrated flood damage reduction and ecosystem restoration projects throughout the Central Valley . . . there is no agreement at this time where the various measures should take place.”

In the delta, a CALFED “levee system integrity” report published in July 2000 estimated that 520 mi (837 km) of levees would need “major rehabilitation or reconstruction work,” the costs ranging from $600 million to $1.3 billion. This work, however, would not include upgrades for seismic stability.

By comparison, then, the roughly $90 million in federal flood control assistance being sought by Schwarzenegger and California’s U.S. Senate and House members will begin to address only the problems with the state’s levees—and will do so only if the money is actually spent; a roughly comparable amount was authorized by Congress in 2004 but was never appropriated. What’s more, the money will probably be designated for the most statically unstable levees—those with foundations suffering from cracks and seepage—but will do nothing to address the seismic issues in the delta, says Seed.

Various studies are under way to determine more precisely which levees throughout the state need to be improved, including a flood insurance mapping project being conducted by FEMA and a database containing information about the health of the entire levee system that is being developed by the DWR. There is also a two-year study on managing risk in the delta in the works to determine the threats that the delta levees face from “earthquake, flood, or sudden unanticipated failures,” explains Harder, who is leading the study.

Meanwhile, the Corps continues its efforts to repair and improve levees when floods occur or when local sponsors are willing to share costs. Following floods in the Central Valley in 1986 and 1997, for instance, the Corps improved some 90 mi (145 km) of structures at a cost of roughly $140 million, notes Hess. Other projects also are under way or are planned.

Along the Sacramento and American rivers, for instance, this work involves three main approaches designed to strengthen the levees there and prevent seepage, especially underseepage. One approach is to construct a slurry trench cutoff wall through the center of a levee. These underground walls are filled with a mix of soil, cement, and bentonite (a type of clay) to block seepage. According to Recommendations for Seepage Design Criteria, Evaluation, and Design Practices, a report released by the Corps of Engineers in 2003, they extend down through the levee into the subsurface soils to reach a less pervious layer and can reach depths of 80 ft (24.4 m).

Building seepage berms on the land side of a levee is a second key approach. These earthen berms are designed to control seepage and prevent erosion and to provide additional weight to balance upward water pressure. They are typically designed as semipervious structures incorporating a drainage layer of clean sand or a drainage rock layer with adjacent filter layers. Seepage berms should have a minimum width of four times the maximum levee height in a reach, with a maximum width of roughly 300 ft (91 m). The height of the berm typically varies from 5 ft (1.5 m) to 3 ft (0.9 m).

A third approach frequently used is to install relief wells along the toe of the levee on the land side to collect seepage and channel it away so that it does not erode the levee itself. Typically, these wells are made of stainless steel and have diameters of roughly 8 to 12 in. ( 203 to 304 mm) and depths of 50 to 100 ft (15 to 30.5 m), notes Hess. Generally they are spaced at about 50 to 200 ft (15 to 61 m) on center for miles along the levee.

Some levee heights are increased for better flood protection by adding concrete floodwalls. These walls—which generally range in height from 1 to 5 ft (0.3 to 1.5 m)—make the levee higher but do not require a larger footprint. Thus they do not encroach upon private property, explains Mohsen Tavana, a civil engineer in the operations technical section of the Corps’s Sacramento district.

Possible engineering solutions in the delta range from strengthening key levees to deliberately breaching other levees so that the islands they protect can “gradually return to their natural condition of being a sea-level marsh,” notes Seed. This would prevent the levees from failing and drawing in salt water from San Francisco Bay, he explains.

California officials must also prepare for a major breach of delta levees by stockpiling large rocks, empty sandbags, and miles of filter fabric so that emergency response teams could quickly close levee breaches and armor the inboard sides of breached levees, Seed says. And the state must make certain that heavy barges are ready to help flood crews gain access to breaches around islands that are accessible only by boat or where the road across the levee becomes too damaged to use.

All of this must be arranged in advance to ensure that time is not lost when a major disaster strikes, Seed stresses. Such preparations will cost “peanuts compared to the kinds of risk we’re talking about” if the delta’s water is heavily contaminated, he says.

Another proposal, this one aimed at preserving freshwater supplies for the 20 million Californians who depend on the delta, involves bypassing the delta entirely. The idea is to build a 40 to 50 mi (64 to 80 km) canal—a peripheral canal—that would carry water from the Central Valley around the delta and down to consumers in the central and southern parts of the state. The canal concept dates to the early 1960s but was overwhelmingly defeated by California voters in 1982; people in northern California worried that the canal would siphon off too much water from their region, and environmentalists feared that Southern California would no longer support measures to protect the delta if their supplies of water were guaranteed through other means.

Since those objections remain viable, Seed suggests building something smaller and less expensive—a peripheral aqueduct, he calls it—that would involve a series of pipes to safely carry around the delta roughly 40 percent of Southern California’s current water usage. The aqueduct could give the state several more years to fix a damaged delta without forcing millions of people to relocate in search of water, he explains. But the pipes would never take away as much water from northern California as the canal, and people living in Southern California would still need to support repairs along the delta. Seed estimates that the peripheral aqueduct and the emergency preparation measures he advocates could cost $4 billion to $5 billion, versus as much as $3 billion for just the peripheral canal.

Some do not see the answer in engineering solutions at all. Mount, for instance, stresses the need for regional planning mechanisms with enough clout to discourage urban growth in floodplains, growth that he believes is pushing into the delta. Others wonder if the political will exists to make the tough and expensive decisions necessary—even in a post-Katrina world. California’s flood control system underwent a major change in late September when Schwarzenegger replaced all seven members of the Reclamation Board. Such critics as the editors of the Sacramento Bee raised the possibility that the board was ousted because it “recently had taken a more aggressive stance to challenge urban development behind levees,” which was “unpopular with developers and local governments.”

But Mount, one of the ousted board members and a vocal opponent of such urbanization, downplays that accusation. Still, other observers feel that important time will be lost as the new board gets accustomed to its responsibilities.

A key question, of course, is, how imminent is the threat? In other words, is the clock really ticking on a flood disaster “time bomb”? Earthquakes strong enough to devastate the delta cannot be predicted with any accuracy. But major floods hit the Central Valley every 7 to 10 years, notes Sandner. Since the last two major floods were in 1986 and 1997, he predicts that “we’re coming up on the next big flood here pretty quick.”