Chapter 1 - DWR Drought Report, July, 2000

This chapter briefly summarizes California hydrology and water supplies and describes hydrologic conditions associated with past droughts. It is important to remember that California hydrologic data cover a limited period of historical recordonly a few stream gages have a period of record in excess of 100 years, and likewise only a few precipitation records extend as much as 150 years. Efforts to go beyond the historical period of record to evaluate the occurrence of earlier droughts, or to forecast future droughts, are described at the end of this chapter.

The water supplies used by Californians come from several sourcessurface water released from reservoirs, surface water directly diverted from unstored streamflows, and groundwater. Supplies derived from desalting and water recycling are also important to individual agencies relying on these sources, but they collectively represent less than one percent on California's water supply.

Roughly three-quarters of California's runoff occurs north of Sacramento, while about the same proportion of water needs occurs south of Sacramento. Figure 1 shows the extensive system of conveyance infrastructure constructed in response to the imbalance in the locations of supplies and demands. Access to this conveyance capacity has important implications for water transfers, as discussed in Chapter 3.

Much of California enjoys a Mediterranean-like climate with cool, wet winters and warm, dry summers. An atmospheric high pressure belt results in fair weather for much of the year, with little precipitation during the summer. The high pressure belt shifts southward during the winter, placing the State under the influence of Pacific storms bringing rain and snow. Most of California's moisture originates in the Pacific Ocean. As moisture-laden air moves over mountain barriers such as the Sierra Nevada, the air is lifted and cooled, dropping rain or snow on the western slopes. This orographic precipitation is important for the State's water supply.

Average annual statewide precipitation is about 23 inches, corresponding to a volume of nearly 200 million acre-feet over California's land surface. About 65 percent of this precipitation is consumed through evaporation and transpiration by plants. The remaining 35 percent comprises the State's average annual runoff of about 71 maf. Less than half this runoff is depleted by urban or agricultural use. Most of it maintains ecosystems in California's rivers, estuaries, and wetlands. Available surface water supply totals 78 maf when interstate supplies from the Colorado and Klamath Rivers are added. Figure 2 shows the distribution of California's average annual precipitation and runoff.

On average, 75 percent of the State's average annual precipitation of 23 inches falls between November and March, with half of it occurring between December and February. A shortfall of a few major storms during the winter usually results in a dry year; conversely, a few extra storms or an extended stormy period usually produces a wet year. An unusually persistent Pacific high pressure zone over California during December through February predisposes the year toward a dry year. Figure 3 compares average monthly precipitation in the Sacramento River region with precipitation during extremely wet (1982-83) and dry (1923-24) years.

Water agencies such as the Department or the U.S. Geological Survey report hydrologic data on a water year basis. The water year extends from October 1st through September 30th. This report, for example, was published in water year 2000 (October 1, 1999September 30, 2000). Hydrologic data presented throughout this report are presented in terms of water years. The (water year) 1987-92 drought corresponds to the calendar period of fall 1986 through summer 1992. Water project delivery data (e.g., State Water Project deliveries) are presented on a calendar year basis.

The influence of climatic variability on California's water supplies is much less predictable than are the influences of geographic and seasonal variability, as evidenced by the recent historical record of precipitation and runoff. For example, the State's average annual runoff of 71 maf includes the all-time low of 15 maf in 1977 and the all-time high (exceeding 135 maf) in 1983. Floods and droughts occur often, sometimes in the same year. The January 1997 flood was followed by a record-setting dry period from February through June; the flooding of 1986 was followed by six years of drought (1987-92).

Figures 4 and 5 show estimated annual unimpaired runoff from the Sacramento and San Joaquin River Basins to illustrate climatic variability. Because these basins provide much of the State's water supply, their hydrologies are often used as indices for water year classification systems.

Water year classification systems provide a means to assess the amount of water originating in a basin. The Sacramento Valley 40-30-30 Index and the San Joaquin Valley 60-20-20 Index were developed by the State Water Resources Control Board for the Sacramento and San Joaquin River Basins as part of SWRCB's Bay-Delta regulatory activities. Both systems define one "wet" classification, two "normal" classifications (above and below normal), and two "dry" classifications (dry and critical), for a total of five water year types.

The Sacramento Valley 40-30-30 Index is computed as a weighted average of the current water year's April-July unimpaired runoff forecast (40 percent), the current water year's October-March unimpaired runoff (30 percent), and the previous water year's index (30 percent). A cap of 10 maf is put on the previous year's index to account for required flood control reservoir releases during wet years. Unimpaired runoff (calculated in the 40-30-30 Index as the sum of Sacramento River unimpaired flow above Bend Bridge, Feather River unimpaired inflow to Oroville Reservoir, Yuba River unimpaired flow at Smartville, and American River unimpaired inflow to Folsom Reservoir) is river production unaltered by water diversions, storage, exports, or imports. A water year with a 40-30-30 index equal to or greater than 9.2 maf is classified as "wet." A water year with an index equal to or less than 5.4 maf is classified as "critical." Unimpaired runoff from the Sacramento Valley, often referred to as the Sacramento River Index or the Four River Index, was the dominant water supply index used in SWRCB's Decision 1485. The SRI, while still used in SWRCB's Order WR 95-6 as a water supply index, is no longer employed to classify water years. By considering water availability from storage as well as from seasonal runoff, the 40-30-30 Index provides a more representative characterization of water year types than does the SRI. However, no indexing scheme can be a perfect representation of water year type. For example, the inability to store large volumes of wet year runoff (due to reservoir flood control requirements and the relatively low ratio of storage capacity to wet year runoff volumes for most California rivers) distorts the 40-30-30 Index value for the year following a very wet year.

The San Joaquin Valley 60-20-20 Index is computed as a weighted average of the current water year's April-July unimpaired runoff forecast (60 percent), the current water year's October-March unimpaired runoff (20 percent), and the previous water year's index (20 percent). A cap of 4.5 maf is placed on the previous year's index to account for required flood control reservoir releases during wet years. San Joaquin Valley unimpaired runoff is defined as the sum of unimpaired inflow to New Melones Reservoir (from the Stanislaus River), Don Pedro Reservoir (from the Tuolumne River), New Exchequer Reservoir (fromthe Merced River), and Millerton Lake (from the San Joaquin River). A water year with a 60-20-20 index equal to or greater than 3.8 maf is classified as "wet." A water year with an index equal to or less than 2.1 maf is classified as "critical."

Although not used to classify water years, the Eight River Index is another water supply index employed in Order WR 95-6. The Eight River Index, defined as the sum of the unimpaired runoff from the four Sacramento Valley Index rivers and the four San Joaquin Valley Index rivers, is used to define Delta outflow requirements and export restrictions. Key index months for triggering Delta requirements are December, January, and February.

Under average hydrologic conditions, about 30 percent of California's urban and agricultural water needs are supplied by groundwater. This percentage increases in dry years when water users whose surface supplies are reduced turn to groundwater, if available. Figure 6 shows the total number of well construction/modification reports received annually by the Department, illustrating the relationship between groundwater use and hydrologic conditions. Well drilling activity increased during the 1987-92 drought and was at a minimum in wet years such as 1982 or 1983.

The amount of water stored in California's groundwater basins is far greater than that stored in the State's surface water reservoirs, although only a fraction of these groundwater resources can be economically and practically extracted for use. Figure 7 shows major areas of current and potential groundwater development in California. The greatest amounts of groundwater extraction occur in the Central and Salinas Valleys and in the Southern California coastal plain. At a 1995 level of development, California's estimated developed groundwater supplies were about 12.5 maf under average hydrologic conditions. This amount is exclusive of groundwater overdraft, estimated at about 1.5 maf annually. More than 1 maf of this estimated annual overdraft occurs in the San Joaquin Valley.

The majority of California's groundwater production occurs from alluvial materials in the large basins indicated in Figure 7. Groundwater levels in such basins typically decline during droughts due to increased extractions. For example, groundwater extractions were estimated to exceed recharge by 11 maf in the San Joaquin Valley during the first five years of the 1987-92 drought. Drawing down groundwater reserves in drought years is analogous to surface reservoir carryover storage operations. The extent to which groundwater levels recover depends on the amount of subsequent extractions and recharge. Figure 8 shows hyrographs for two wellsone located in a basin experiencing long-term overdraft and the other in a basin not experiencing long-term overdraft. Both hydrographs show the effects of increased extractions during the 1976-77 and 1987-92 droughts, followed by post-drought rebound.

Droughts exceeding three years are relatively rare in Northern California, the source of much of the State's water supply. Historical multi-year droughts include: 1912-13, 1918-20, 1923-24, 1929-34, 1947-50, 1959-61, 1976-77, and 1987-92. The 1929-34 drought established the criteria commonly used in designing storage capacity and yield of large Northern California reservoirs. Table 1 compares the 1976-77 and 1987-92 droughts to the 1929-34 drought in the Sacramento and San Joaquin Valleys.

One approach to supplementing California's limited period of measured data is to statistically reconstruct data through the study of tree rings. Information on the thickness of annual growth rings can be used to infer the wetness of the season. A 420-year reconstruction of Sacramento River runoff from tree ring data was made for the Department in 1986 by the Laboratory for Tree Ring Research at the University of Arizona. The tree ring data suggested that the 1929-34 drought was the most severe in the 420-year reconstructed record from 1560 to 1980. The data also suggested that a few droughts prior to 1900 exceeded three years, and none lasted over six years, except for one period of less than average runoff from 1839-46. John Bidwell, an early pioneer who arrived in California in 1841, confirmed that 1841, 1843, and 1844 were extremely dry years in the Sacramento area. The Department is currently funding the University to expand tree ring data for the Sacramento River watershed to cover approximately the past 1,000 years. Similar tree ring studies covering the period between 1550 and 1977 were conducted for the Colorado and Santa Ynez Rivers. According to these studies, the most severe drought on the Colorado River occurred during 1580-1600, and the most severe drought on the Santa Ynez River occurred during 1621-37.

A 1994 study of relict tree stumps rooted in present-day lakes, rivers, and marshes suggested that California sustained two epic drought periods, extending over more than three centuries. The first epic drought lasted more than two centuries before the year 1112; the second drought lasted more than 140 years before 1350. In this study, the researcher used drowned tree stumps rooted in Mono Lake, Tenaya Lake, West Walker River, and Osgood Swamp in the central Sierra. A conclusion that can be drawn from these investigations is that California is subject to droughts more severe and more prolonged than anything witnessed in the historical record.

The historical record of California hydrology is brief in comparison to the time period of geologically modern climatic conditions. The following sampling of changes in climatic and hydrologic conditions help put California's twentieth century droughts into perspective, by illustrating the variability of possible conditions. Most of the dates shown below are necessarily approximations, since the dates must be inferred from indirect sources.

11,000 years before present Beginning of Holocene EpochRecent time, the time since the end of the last major glacial epoch

6,000 years before present Approximate time when trees were growing in areas now submerged by Lake Tahoe. Lake levels were lower then, suggesting a drier climate.

900-1300 A.D. (approximate) The Medieval Warm Period, a time of warmer global average temperatures. The Arctic ice pack receded, allowing Norse settlement of Greenland and Iceland. The Anasazi civilization in the Southwest flourished, its irrigation systems supported by monsoonal rains.

1300-1800 A.D. (approximate) The Little Ice Age, a time of colder average temperatures. Norse colonies in Greenland failed near the start of the time period, as conditions became too cold to support agriculture and livestock grazing. The Anasazi culture began to decline about 1300 and had vanished by 1600, attributed in part to drought conditions that made agriculture infeasible.

Mid-1500s A.D. Severe, sustained drought throughout much of the continental U.S., according to dendrochronolgy. Drought suggested as a contributing factor in the failure of European colonies at Parris Island, South Carolina and Roanoke Island, North Carolina.

1850s A.D. Sporadic measurements of California precipitation began.

1890s A.D. Long-term streamflow measurements began at a few California locations.

Accurate long-term weather forecasting would be extremely valuable for water project operations. Currently, predictions sufficiently detailed to be useful for project operations are limited to about two weeks at best, and these predictions have perhaps a 50 percent accuracy rate. Had water project operators known in advance that 1987-92 would be dry, project operations could have been modified to increase carry-over storage and to equalize deliveries over the six years of drought.

Long-term forecasting remains in its scientific infancy. The National Weather Service issues 30 and 90-day forecasts. Academic institutions, such as the Scripps Institution of Oceanography in San Diego, have attempted experimental seasonal forecasts. The accuracy and level of detail of these efforts remains insufficient for water project operations. It is only recently, for example, that researchers have had sufficient understanding of global weather patterns and atmospheric/oceanic interactions to be able to identify conditions associated with the El Niño Southern Oscillation in the Pacific Ocean. That understanding has yet to be translated to forecasts of runoff, partly because ENSO events affect different parts of California differently.

Using global weather models to predict future climatological conditions requires collection of massive amounts of data and access to substantial computational power (i.e., supercomputers). Although electronic data processing capabilities have increased exponentially since the early days of mainframe computers, data collection will remain a limiting factor into the foreseeable future, due to the sheer volume of information needed to represent global atmospheric/oceanic conditions. Atmospheric conditions themselves may furthermore be inherently too variable to support long-range forecasts of sufficient reliability for short-term water project operations. A more realistic expectation might be the ability to forecast shifts in global conditions, such as potential global warming or decadal oscillations in ocean temperatures in the equatorial Pacific. It can be safely said that the ability to accurately predict dry conditions will remain elusive within this report's short planning horizon.

One dry year does not constitute a drought in California, but does serve as a reminder of the need to plan for droughts. California's extensive system of water supply infrastructureits reservoirs, groundwater basins, and inter-regional conveyance facilitiesmitigates the effect of short-term dry periods. Defining when a drought begins is a function of drought impacts to water users. Hydrologic conditions constituting a drought for water users in one location may not constitute a drought for water users in a different part of the state or with a different water supply. Individual water suppliers may use criteria such as rainfall/runoff, amount of water in storage, or expected supply from a water wholesaler to define their water supply conditions

The percent of average values are determined for measurement sites and reservoirs in each of the State's ten major hydrologic regions. Snowpack is an important indicator of runoff from Sierra Nevada watersheds, the source of much of California's developed water supply.

The Department used two primary criteria to evaluate statewide drought conditions during the 1987-92 droughtrunoff and reservoir storage, either actual or predicted. A drought threshold was considered to be runoff for a single year or multiple years in the lowest ten percent of the historical range, and reservoir storage during the same time period at less than 70 percent of average. These were not hard and fast values, but guidelines for identifying drought conditions.

Drought is a gradual phenomenon. Although droughts are sometimes characterized as emergencies, they differ from typical emergency events. Most natural disasters, such as floods or forest fires, occur relatively rapidly and afford little time for preparing for disaster response. Droughts occur slowly, over a multiyear period. There is no universal definition of when a drought begins or ends. Impacts of drought are typically felt first by those most reliant on annual rainfallranchers engaged in dryland grazing, rural residents relying on wells in low-yield rock formations, or small water systems lacking a reliable water source. Criteria used to identify statewide drought conditions do not address these localized impacts. Drought impacts increase with the length of a drought, as carry-over supplies in reservoirs
are depleted and water levels in groundwater basins decline.

CHAPTER 1 Hydrology And Water Supply

CHAPTER 1

Hydrology And Water Supply