UC Davis San Francisco Estuary and Watershed Science Title Delta Subsidence Reversal, Levee Failure, and Aquatic Habitat—A Cautionary Tale Permalink https://escholarship.org/uc/item/9pp3n639 Journal San Francisco Estuary and Watershed Science, 11(1) ISSN 1546-2366 Authors Bates, Matthew E. Lund, Jay R. Publication Date 2013 DOI https://doi.org/10.15447/sfews.2013v11iss1art1 License https://creativecommons.org/licenses/by/4.0/ 4.0 Peer reviewed eScholarship.org Powered by the California Digital Library University of California MARCH 2013 Delta Subsidence Reversal, Levee Failure, and Aquatic Habitat—A Cautionary Tale Matthew E. Bates1, 2,† and Jay R. Lund1 ABSTRACT we prioritize islands for investment based on trade- offs between anticipated outcome and lost agricul- Various schemes are often suggested to reverse the tural revenues. This approach might help integrate subsidence of lands below sea level in California’s subsidence-reversal activities into long-term Delta Sacramento—San Joaquin Delta, an area protected planning under a range of flooding, land use, and by levees (dikes) that have significant probabilities habitat management scenarios. of failure. Elementary modeling is used to estimate the probability distribution of land elevations at time of failure for 36 of these subsided islands, KEYWORDS assuming a reasonable potential subsidence rever- Subsidence reversal, levee failure, Sacramento–San sal rate. Given estimated annual probabilities of Joaquin Delta, flooded islands, aquatic habitat, levee failure, elevation gains at this rate are not agricultural revenue. expected to exceed 1 to 2 m before flooding, which would be insufficient to restore most subsided islands to mean sea level (msl). However, under INTRODUCTION some circumstances 1- to 2-m gains are significant. Like many coastal and inland lowlands, California’s A framework is introduced for evaluating islands Sacramento–San Joaquin Delta (Delta) is an as promising candidates for subsidence reversal often unstable landscape whose fate is commonly based on elevation goals other than msl, as dem- debated. Far from the dynamic tidal estuary of onstrated though a hypothetical aquatic habitat pre-European times, today’s Delta is a fixed sys- example. Here, we recommend relevant subsidence tem of islands and levees (dikes) built by various reversal strategies by comparing an elevation goal groups and individuals, adhering to no uniform with each island’s anticipated flooded depth, and standard (Thompson 1957; Jackson and Paterson 1977; Hundley 2001; Lund and others 2010). † Corresponding author: [email protected] Though levee dependability is crucial for local 1 Center for Watershed Sciences, University of California, Davis, Davis, CA 2 Environmental Laboratory, Engineer Research and Development Center, agriculture, local flood protection, and statewide U.S. Army Corps of Engineers, 696 Virginia Rd., Concord, MA 01742 water supply, the possibility of levee failure is ever-present (Matthew 1931; Houston and Duncan 1978; Duncan and Houston 1983; Finch 1985; SAN FRANCISCO ESTUARY & WATERSHED SCIENCE Prokopovich 1985; CDWR 1995; Kelly 1998; mean sea level (msl) (Mount and Twiss 2005; Torres and others 2000; USGS 2003; URS and J.R. Deverel and Rojstaczer 1996; Drexler and others Benjamin & Assoc. 2009a). 2009b). Changes in land use or shallow flooding can halt subsidence from soil oxidation and com- Unplanned levee breaches are costly, with dewater- paction (Tate 1979; Ingebritsen and others 1999). ing and repair costs ranging from $43 million to Island subsidence can be reversed through land- $240 million per island (URS and J.R. Benjamin & management practices that foster the accumulation Assoc. 2009a), yet frequently occur (Florsheim and of sediments and organic material (e.g., Miller and Dettinger 2007)—approximately 160 Delta levees others 2008). breached and flooded during the 1900s (URS and J.R. Benjamin & Assoc. 2009c). If breached islands Natural subsidence reversal (accretion) often occurs are sufficiently subsided, repair is often uneco- in marshes and wetlands as mild flow rates and nomical (Weir 1950; Logan 1989, 1990a, 1990b; vegetation encourage sediments and plant litter Suddeth and others 2008, 2010). Flooded islands to collect. Thick layers of organic soils, like those such as Clifton Court Tract (1907), Lower Sherman found throughout the Delta, form when these Island (1925), Big Break (1927), Donlon Island deposits accumulate faster than they decompose (1937), Franks Tract (1938), Mildred Island (1983), (Gorham 1957; Boelter and Verry 1977). Many Little Franks Tract (1983), Little Mandeville Island studies have analyzed natural accretion under (1994), and Liberty Island (1998), were abandoned various settings, finding elevation-gain rates that completely when their owners faced repair costs range from a few millimeters to a few centimeters that exceeded land values (CALFED 1998; URS and per year, depending on location and management J.R. Benjamin & Assoc. 2009c). Furthermore, rates (e.g., Patrick and DeLaune 1990; Callaway and of levee failure persist (Florsheim and Dettinger others 1996; Goman and Wells 2000; Lane and 2007), and continued island subsidence and sea others 2006; Drexler and others 2009a; Deverel level rise, increasing seismic tension, and the and Leighton 2010). In engineered subsidence effects of climate change are expected to increase reversal, the natural accretion of organic matter the frequency and consequences of levee failure and sediments is expedited in protected and man- (URS and J.R. Benjamin & Assoc. 2009a; Lund and aged marshes (Ingebritsen and others 1999; Miller others 2010). Annual risks of levee failure have and others 2008). Earth-moving, controlled levee recently been calculated for most islands in the breaches, deposition of dredged sediments, and Delta (URS and J.R. Benjamin & Assoc. 2009a). similar techniques may also be plausible for raising elevations or augmenting natural accretion (Ford Land subsidence worsens the consequences of and others 1999; Ingebritsen and others 1999; Ray island failure, and is primarily caused by the aero- 2007). bic microbial oxidization of soil carbon, and the shrinking and compaction of peat soils drained of Subsidence reversal has implications for aquatic their natural water content. Other factors—includ- habitat. Much of the Sacramento—San Joaquin ing the historical burning of peat to remove weeds Delta is now inhabited primarily by non-native and pests and to prepare the soil for planting, aquatic plants and animals, which complicates wind erosion, compaction from heavy farm equip- Delta management. Limiting and mitigating the ment, and the general geologic subsidence of the effects of non-native plant species has become a area—also are relevant (Weir 1950; Broadbent major concern for Delta ecologists (e.g., Moyle and 1960; Atwater and others 1977; Prokopovich 1985; others 2010) and the California Department of Rojstaczer and others 1991; Rojstaczer and Deverel Boating and Waterways (2001), which has a leg- 1995; Ingebritsen and others 1999; Drexler and islative directive to manage specific aquatic weeds others 2009b). Farming and other soil disturbances in the Delta. Even if subsidence reversal cannot continue to transfer Delta topsoil into the atmo- restore islands to mean sea level, it may be suit- sphere, leaving island elevations farther below able for tailoring flooded-island habitats that favor 2 MARCH 2013 or discourage particular native or invasive species. Probabilities of Failure Other purposes (e.g., recreation, water quality) also may benefit from depth-dependent, subsidence- Current annual probabilities of levee failure have reversal activities. been estimated for most Delta islands as part of the Delta Risk Management Strategy (DRMS) study This study examines subsidence reversal’s poten- commissioned by the California Department of tial role in the Delta by (1) modeling each island’s Water Resources (CDWR). These failure probabilities expected elevation over time with a reasonable include the effects of earthquakes, floods, and other subsidence-reversal rate and probability of levee causes (URS and J.R. Benjamin & Assoc. 2009a), failure, (2) estimating the likely extent that subsid- with the methods, models, assumptions, and back- ence reversal can restore Delta islands to mean sea ground data for these calculations described by level before flooding, and (3) introducing a frame- URS and J.R. Benjamin & Assoc. (2008). For the work for evaluating expected outcomes in terms of 36 subsided islands considered here, the annual elevations other than mean sea level. This approach probabilities of failure range from 1% to 10% is demonstrated through a hypothetical applica- annually, with a mean of 5% (Figure 1; URS and tion to avoid depths dominated by a submerged J.R. Benjamin & Assoc. 2009a). The failure prob- invasive waterweed. Islands are then ranked based abilities reflect a longstanding literature on the on cost and probability of achieving this outcome. weakness of island levees in the Sacramento–San The results are analyzed for sensitivity to alterna- Joaquin Delta (Houston and Duncan 1978; Duncan tive subsidence reversal rates. In modeling subsid- and Houston 1983; Finch 1985; Kelly 1998; Torres ence reversal, we examine ambitious engineering and others 2000; URS and J.R. Benjamin & Assoc. projects with rates of elevation gain that surpass 2009a). natural accretion and sedimentation. The most promising methods will be those that produce the