National Academies Press: OpenBook

Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings (1991)

Chapter: 10. Climate Uncertainty: Implications for Operations of Water Control Systems

« Previous: 9. The Implications of Climate Change for Streamflow and Water Supply . . .
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

10
Climate Uncertainty: Implications for Operations of Water Control Systems

John A. Dracup and Donald R. Kendall

University of California, Los Angeles

Loyola Marymount University, Los Angeles

INTRODUCTION

The key question we are addressing here is how water resource systems will be operated under conditions of uncertain future climate variability and change. This question in turn raises the question of whether or not a current water resource system is adequately designed for climate variability and change. We attempt to answer these two questions using the method of climate variability analogies as suggested by Glantz (1988; 1989). Traditionally, water resources design, such as reservoir sizing, has been predicated on an assumption that the statistical parameters of historic streamflow series are stationary: that is, the parameters that characterize the historical streamflow series do not change with time. This is not to say that traditional water resource system designs assume that climate is actually stationary; traditional designs account for changing climates by considering the variance of streamflow (σ2) about its mean (Q). Stationarity assumes, however, that future variations in climate, as expressed in streamflows, will be similar to those observed in the past.

Future scenarios of climate variability of interest to water resource managers are those of long-term extreme high or low streamflows. The scenarios developed here will be based on historical analogies. We will use the high flow event that occurred on the Colorado River during the spring of 1983 as an analogy of climate variability or change causing significant increased streamflow in the Colorado River. We will use low flow events evident from tree-ring studies in the Colorado River basin as analogies of climate variability or change causing significant decreased streamflow in the Colorado River.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

The idea of forecasting by the use of historical climate analogies has been used by Glantz (1988; 1989), who poses four caveats in their use. First, the use of the climate analogy should be made clear so that it is not misleading when viewed in a different context. Second, the climate analogy should not be overextended to support unjustified conclusions. Third, one should recognize that the analogy may be inappropriate for cultural or historical reasons. Fourth, one should note that analogies can lead to the development of possible but inconsistent scenarios.

The use of historical climate analogies provides a different look into the future than that provided by large-scale computer models. Furthermore, climate analogies provide an insight into how the responsible agencies actually dealt with the climate variability as it occurred

The topic of the operation of water resource systems under climatic stress addressed here is, of course, not new. One of the best discussions of this can be found in the National Research Council's Climate, Climate Change, and Water Supply (1977) study. This study emphasizes that ''unless the exact sequence of future flows can be predicted with certainty there may be little benefit to hydrologic system design.'' The study proposes designing water resource systems with robustness (the ability to perform reasonably well under a variety of possible climates) and resilience (the ability of a system designed for one climate and set of conditions to be modified in response to persistent new climates or conditions) (Matalas and Fiering, 1977). The question to be considered here, then, is whether climate analogies can be used to determine whether or not an existing water resource system exhibits the properties of robustness and resilience.

Forecasting by analogy using the Colorado River as an example also has been studied by Brown (1988). However, she focused on the Colorado River Compact rather than on floods and droughts along the river.

THE COLORADO RIVER BASIN

If the streamflow in the Colorado River somehow could be equated with the number of words written about it, the river would constantly flow as a torrent (Dracup, 1977; Dracup et al., 1985; Hundley, 1975; Rhodes et al., 1984).

The Colorado River dominates water resource development in the seven states of the southwestern United States (see Figure 10.1).

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

It also is one of the most carefully managed river systems in the world. Its multipurpose uses of water supply, hydroelectric power, and water-based recreation compete with its management priority of flood control. With the exception of the deserts of the Great Basin, this 243,000-square-mile basin has the greatest water deficiency (average precipitation less potential evapotranspiration) of any basin in the contiguous United States. Yet, more water is exported from the Colorado River basin than from any other basin in the United States.

The basin has been divided by the Colorado River Compact into the upper Colorado basin and lower Colorado basin for purposes of interstate administration. The upper basin drainage includes the areas of Arizona, Colorado, New Mexico, Wyoming, and Utah that drain into the Colorado River above Lee's Ferry, Arizona. It is bounded on the east and north by mountains forming the Continental Divide, and on the south it opens to the lower Colorado region. The lower basin drainage includes most of Arizona, parts of southeastern Nevada, southeastern Utah, southeastern California, and western New Mexico.

WATER AVAILABILITY ESTIMATES

A wide range of climates occur in the Colorado River basin because of differences in altitude, latitude, and topographic features. In the north, summers are short and warm and winters are long and cold. In the south, the summers are longer and the winters are moderate at low altitude, but colder temperatures occur in the mountains.

About 83 percent of the water that flows in the Colorado River basin comes from the upper basin. The average annual precipitation throughout the entire upper basin is about 16 inches (40.6 cm), which amounts to 93,440,000 acre-feet per year (115 × 109 m3). Approximately 15 percent of the precipitation runs off, and most is lost to evapotranspiration.

One of the most famous and controversial hydrologic records in the United States is that of the virgin flow of the Colorado River at Lee's Ferry, Arizona. Lee's Ferry is defined as a point on the Colorado River one mile below the mouth of the Paria River. Estimates of virgin flow have been made there for the upper basin since 1896; however, runoff has been measured and recorded only since the first gaging station was established at Lee's Ferry during the summer of 1921. (The Bureau of Reclamation now uses natural

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.1 Colorado River basin upstream of the inflow to Mexico.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

flow, not virgin flow, for its operation studies.) The importance of this flow is accentuated by the Colorado River Compact, which requires that the upper basin deliver 75 million acre-feet (maf) (92.5 × 109 m3) at Lee's Ferry each 10 years. Estimates of the long-term annual average flow vary from 11.8 to 16.8 maf (14.5 to 20.7 × 109 m3) depending on the time period selected (Colorado River Board of California, 1969). Recent tree-ring analysis dating back to 1512 has indicated the long-term mean to be approximately 13.5 maf (16.6 × 109 m3) (Stockton, 1977).

The current estimates of available surface-water supply within the upper basin are less than those at the time the Colorado River Compact was negotiated. This is because of the abnormally wet period that occurred during the early part of this century. The range of annual natural flow at Lee's Ferry has varied from a low of 5.0 maf (6.2 × 109 m3) in 1977 to a high of 24.0 maf (29.6 × 109 m3) in 1917. The average natural flow from 1931 through 1989 of 14.2 maf (17.5 × 109 m3) may be closer to the long-term mean.

The laws governing the Colorado River have been presented in detail by Meyers (1966) and Hundley (1975; 1983). Only a brief summary of the major treaties, laws, and compacts will be presented here.

The allocation of Colorado River water is based on the concept of beneficial consumptive use. The allocation system operates at four levels: international, interregional, interstate, and intrastate (Weatherford and Jacoby, 1975).

The international allocation was accomplished by the Mexican Treaty of 1944. Mexico was guaranteed an annual amount of 1.5 maf (1.8 × 109 m3) except in times of extreme shortage. However, this treaty contained no provision for water quality. Thus, joint agreements in 1965 and 1973 called for a temporary agricultural drainage water bypass and eventually a desalting plant to improve the quality of water crossing the border.

The interregional allocation was achieved when Congress approved the Colorado River Compact, which became effective in June 1929. Sectional rivalry has caused the states included in the drainage basin to agree to an equal apportionment in the use of the Colorado River system waters between the states of the upper basin and the states of the lower basin (an agreement set forth in Articles III (b) and III (d) of the Colorado River Compact).

Traditionally, the fertile lowland valleys in the lower basin states have developed economically more rapidly than have the mountain headwater "areas of origin" in the upper basin states. The upper basin states insisted that an equitable apportionment of

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

the river be made to them prior to the expenditure of large sums of federal money, which might result in a modification of equities adverse to the upper basin states. This is in essence what was achieved in the Colorado River Compact.

The intent of this landmark document was to give each basin the perpetual right to the "exclusive beneficial use of 7.5 maf (9.25 × 109 m3) of water per annum." However, the lower basin was assured that depletion in the upper basin would allow at least 75 maf (92.5 × 109 m3) of flow to the lower basin at Lee's Ferry in each successive 10-year period. Thus, the lower basin received a guaranteed 10-year, not annual, minimum flow, and the upper basin assumed the burden of any deficiency caused by a hydrologic dry cycle. It is important to note that the division of the use of water between the upper and lower basins is a fixed amount rather than a proportional amount (such as one-half of a 10-year moving average).

CURRENT COLORADO RIVER MANAGEMENT TECHNIQUES

The joint operation of Lakes Powell and Mead is subject to the following criteria, according to the Law of the River (Nathanson, 1978):

  • a Lake Powell minimum objective release of 8.2 maf per year;

  • additional releases from Lake Powell to equalize end-of-year active storages in Lakes Powell and Mead if Lake Powell would otherwise contain more water in storage; and

  • sufficient storage in the upper basin reservoirs to assure future deliveries to the lower basin without impairing annual consumptive use in the upper basin (called 602(a) storage under the Colorado River Basin Project Act of 1968).

As a result of the Law of the River, each basin currently possesses many storage facilities, including a large linchpin reservoir: Lake Powell for the upper basin and Lake Mead for the lower basin. The storage in these two reservoirs totals 51.0 maf (62.9 × 109 m3), or 85 percent of the total storage in the entire Colorado River basin. The reservoir system now stores about four times the annual flow of the river. This volume of water in storage reflects the determination of the basin states to conserve as much water as possible, providing a margin of safety in the event that a run of

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

dry years occurs. The upper basin states prefer that releases to the lower basin be the absolute minimum required by law. Furthermore, current operational regulations require that the river's manager, the U.S. Bureau of Reclamation, maintain Lakes Mead and Powell at or near equal volumes of water in storage at the end of each operating year (which occurs on September 30) (Nathanson, 1978). Each basin is thus assured "equal ownership" of the river.

Flood control protection is provided to residents, farms, and businesses below Hoover Dam. Flood control operations rest on two central elements:

  1. scheduled dedicated water storage space made available to catch the spring runoff in Lake Mead; and

  2. a forecast of how much water will enter Lake Powell from April 1 through July 31, produced by the National Weather Service (NWS) Colorado Basin River Forecast Center in Salt Lake City, Utah.

The objective of the flood control procedure, in effect, is to create enough storage space, through reservoir releases from August through January, to catch the predicted April through July runoff. The plan, in action since 1968 and slightly modified in 1982, uses a monthly streamflow forecast generated for the period January through July that predicts the spring inflow to Lake Powell. Adjustments in the forecasted storage space for the inflow can then be made to keep downstream releases below damaging levels and at the same time conserve as much water as possible.

The NWS Colorado Basin River Forecast Center uses monthly estimates to arrive at forecasts of the maximum probable and minimum probable April through July runoff. To meet these runoff estimates, the Bureau of Reclamation increases flood control space in Lake Mead starting on August 1 of each year to have 5.35 maf (6.6 × 109 m3) available by January 1. According to the flood control plan for the Colorado River, Lake Mead is the only major basin reservoir with an explicit flood control space schedule. Prior to the construction of Glen Canyon Dam, the standard flood control procedure was to have 5.8 maf (7.1 × 109 m3) of storage available on January 1, as recommended by Debler (1930). This storage requirement was increased each month until a maximum requirement of 9.5 maf (11.7 × 109 m3) was reached on April 1. These procedures were formalized by the U.S. of Army Corps of Engineers in 1955 and were continued until 1968.

The Bureau of Reclamation's scheduled outflow release rates through the dams and the storage space availability based on the

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

NWS inflow forecasts have worked well in recent decades in minimizing water lost through unneeded anticipatory releases and potential flooding below Hoover Dam. This has helped to maximize hydroelectric generation, water conservation, storage, and flood control. However, the conditions in the Colorado basin in the early 1980s were radically different from those of the 1960s and the 1970s. The changes in those conditions were significant contributors to the flooding that occurred in the spring and summer of 1983.

OPERATION DURING HIGH STREAMFLOWS: THE COLORADO RIVER SPRING FLOODS OF 1983

On January 1, 1983, there were 6.6 maf (8.1 × 109 m3) of storage space available in Lake Mead and upstream—more than the required January target of 5.35 maf (6.6 × 109 m3). Yet even with the surplus storage space available, the reservoir system was overwhelmed by the magnitude of the spring inflow to Lake Powell. Because of late precipitation and cool weather throughout the upper basin, snowpack continued to increase during April and May. Figure 10.2 shows the rapid and unusual changes in the forecasted inflow to Lake Powell from January through June 1983.

Figure 10.3 illustrates the relationship between Lake Powell inflows, Lake Powell outflows, and Hoover Dam releases from April through July 1983. Because of the massive influx of water into Lake Powell, the Bureau had to increase outflows from Glen Canyon Dam. This, in turn, obliged the Bureau to raise outflows from Hoover Dam. The releases at Hoover Dam, which historically had been held to approximately 25,000 cubic-feet per second (cfs) (708 m3 s-1), were elevated to over 40,000 cfs (1,132 m3 s-1) in July. This is a critical point, for 40,000 cfs (1,132 m3 s-1) was the targeted maximum outflow rate from Lake Mead under the 1968 revised flood control procedures. The Bureau of Reclamation operators successfully limited Hoover release rates to 40,000 cfs (1,132 m3 s-1) except for the month of July, during which Hoover releases averaged 41,854 cfs (1,184 m3 s-1). However, flooding downstream of Hoover Dam begins when the flow exceeds 19,900 cfs (538 m3 s-1).

The rapid sequence of meteorological events occurring late in the spring, coupled with the problem of attempting to move massive amounts of water through Lakes Powell and Mead in a short period of time, resulted in streamflows greater than those

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.2 Forecasted inflow to Lake Powell from January through June 1983.

Source: Dozier and Brown, 1983.

FIGURE 10.3 Relationship among Lake Powell inflows, Lake Powell outflows,  and Hoover Dam releases from April through July 1983.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

experienced during the previous two decades by lower basin residents and businesses. The unusual and unexpected flooding along the Colorado River during 1983 was the result of three converging factors: the sudden required operation of a full river system, an encroachment into the downstream flood plain, and climate variability in the basin. Each of these is discussed below.

The Full River System

The filling of Lake Powell behind Glen Canyon Dam began in 1963 and was completed in 1980. During this 17-year period, there were virtually no required flood control operations on the Colorado River. Runoff in excess of downstream water supply and hydropower generation was easily stored.

However, in 1980 Lake Powell became full, which required that the river now be operated in a careful, prudent manner; there became little room for forecast error. The forecasted inflow to Lake Powell had to be not only accurate, but also carefully monitored on a real-time basis. Monitoring inflows allows corrective management responses if conditions permit. However, in 1983, Lake Powell inflows rose so rapidly that there was no time for mitigating responses. For example, on May 24, 1983, the unregulated inflow to Lake Powell was approximately 37,000 cfs (1,047 m3 s-1). Eight days later, on June 1, the unregulated inflow was 102,000 cfs (2,887 m3 s-1) (U.S. Bureau of Reclamation, 1983).

Even the availability of real-time data may not have been sufficient to manage a wet year such as 1983, since it takes substantial time to move water through dams with structurally limited release rates. The 1983 April through July inflow into Lake Powell was more than 14 maf (17.3 × 109 m3). Approximately 140 days would be required to discharge that quantity at a rate of 50,000 cfs (1,415 m3 s-1).

Physical Encroachment into the Flood Plain

Physical encroachment into the lower basin flood plain is a function of the defined flood plain boundaries, the relative stability in the annual streamflows, and societal decisions. Thus, encroachment into the lower basin flood plain, which would not have been possible in the absence of the upstream storages, was encouraged by a combination of technological fixes and lax zoning

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

practices in counties bordering on the Colorado River (Arizona Republic, June 22, 1983; July 1, 1983).

The two major dams on the Colorado River have performed as planned in controlling the variability of streamflow rates (Figure 10.4). Glen Canyon and Hoover dams have consequently provided substantial protection to the lower Colorado flood plain in terms of their ability to reduce the river's meanderings and the flooding associated with high spring streamflows. Even in 1983, releases at Hoover Dam did not significantly exceed 50,000 cfs (1,415 m3 s-1).

The history of development in the flood plain roughly began with the construction of earthen levees in the area around Yuma, Arizona. The levee system was constructed to protect agricultural land (fertile flood plain soil) from the annual rush of spring snowmelt (U.S. Army Corps of Engineers, 1982). With the completion of Hoover Dam and the subsequent decrease in spring streamflow variability, more flood plain acreage became available for development. This resulted in the construction of residential and commercial structures in these areas. As the 1982 review of flood control operating procedures notes (U.S. Army Corps of Engineers, 1982):

Few, if any, structures were located in the 40,000 cfs (1,132 m3 s-1) flood plain in the lower Colorado River at the time of the closure of Hoover Dam (1935) and for some years thereafter. For many years the flood control operation plan for Hoover Dam has incorporated a "target maximum" flood control release of 40,000 cfs (1,132 m3 s-1). With the completion of Glen Canyon Dam in 1962, streamflow variability was sharply narrowed. This coincided with the period of the greatest physical encroachment into the flood plain, including construction and development within the streamflow profile of less than 28,000 cfs (792 m3 s-1).

The period when the Colorado reservoir system was filling with water constituted a time during which true exposure to climatic impacts, such as precipitation variability, did not exist. It was not representative of a new climatic regime in the basin, but only of anthropogenic interference with the flow of the river. The encroachment into the flood plain was possible because water was in storage upstream and also because the filling of Lake Powell was drawn out for almost two decades. Two decades are more than sufficient to affect societal perceptions of climate stability.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.4 Mean monthly flows downstream of Hoover Dam in the Colorado River.

Climate Variability

Perceptions of climate stability in the Colorado River basin are not borne out by historical data. In fact, the third factor that contributed to the 1983 lower basin flooding was the variability of the climate in the arid American Southwest. As Figure 10.5 demonstrates, the variability of the river's streamflow is substantial. As recently as 1977, the western United States was hit with a severe drought that resulted in a significant drop—to 5.0 maf (6.2 x 109 m3)—in the estimated virgin flow of the Colorado (Upper Colorado River Commission, 1982).

It is interesting to note that the 1977 drought also played a key role in the decision among the Colorado River basin states to defer any action regarding revision of the river management scheme (Broadbent, 1983). The concern in the Colorado River basin was again reinforced in terms of the major adverse climatic impact anticipated: i.e., drought is of concern, while a wet year is not. Because of the effect of a full system and a variable climate, a

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.5 Estimated virgin flow in the Colorado River.

Bureau of Reclamation official had noted in 1979 that ''the present operation strategy . . . involves an 85 percent risk that damaging floodflows will occur between 1980 and 1984'' (Freeny, 1981). Furthermore, because of climate variability, it has been estimated in a preliminary study by the U.S. General Accounting Office that if the river continues to be managed by keeping reservoirs full or nearly full, controlled flooding similar to that which occurred in 1983 can be expected to be repeated once every 10 to 15 years (Arizona Republic, Sept. 3, 1983).

The flooding on the Colorado River during the spring of 1983 also was exacerbated by a late spring snow, rapid warming, and rain on snow. Furthermore, the antecedent moisture conditions were higher than predicted (J. Lease, Bureau of Reclamation, personal communication, November 1990).

Thus, one can argue that it is not the streamflow variability in itself that caused the 1983 spring flooding and the associated damages. Flood plain encroachment and a full reservoir system, in conjunction with streamflow variability, converged to create appropriate conditions for the events of 1983.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

Conclusions for River Basin Management Under High Flow Conditions

The issue of dedicated flood control storage space is a critical point in the management of the Colorado River. Accepting the Bureau of Reclamation's assumption that the Central Arizona Project will help to alleviate the flooding problem by providing an additional water diversion point, a temporary change in the dedicated flood control space could help to protect property in the lower basin (Broadbent, 1983; Freeny, 1981). However, since the precipitation and runoff of 1983 were so abnormal compared with recent years, proposals for increasing storage space will not necessarily be received warmly (Broadbent, 1983).

The response to the 1977 to 1978 drought indicated that there is substantial support for maintaining full reservoirs upstream and that the flood control for lower basin residents and other interests is not given high priority by the other beneficiaries of the river. Water resources in the American Southwest are managed for dry years, not for extremely wet years such as 1983. This is precisely the justification for the reservoir system to be maintained at a nearly full level.

Actions have been taken by the United States government since 1983 to reduce the potential for and the impact of flooding. A Colorado River flood warning system was developed through congressional appropriations in 1984 and 1985. A broad band of uncertainty was incorporated into the runoff forecasts in determining reservoir flood control space. The Federal-State Colorado River Management Work Group developed annual operating plans that provide for more effective river management under full and near full reservoir conditions.

Public Law 99-450, the Colorado River Floodway Protection Act, was enacted by Congress and approved by the President in 1986. The act established a floodway from Davis Dam to the Mexican border. The floodway was designed to accommodate a once-in-a-hundred-year river flow consisting of controlled releases and tributary inflows, or a flow of 40,000 cfs (1,132 m3 s-1). Except for limited purposes, no new expenditures or new federal financial assistance can be made available for construction within the floodway. No new flood insurance may be provided for new construction within the floodway. No new leases of lands within the floodway can be granted that are inconsistent with the operation and maintenance of the floodway. A floodway task force was formed to advise the Secretary of the Interior and the Congress on the restoration and maintenance of the floodway.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

During the past eight years the natural flow of the Colorado River continued to be highly variable. The 1983 to 1986 recorded streamflows were greater than 20 maf (24.7 × 109 m3) per year. The 1988 to 1990 streamflows were the lowest three consecutive streamflows on record (Colorado River Board of California, 1990).

SEVERE AND SUSTAINED DROUGHT IN THE COLORADO RIVER BASIN

The second climate analogy we can use as a climate change scenario is the supposition of a severe and sustained drought in the Colorado River basin. Since the Colorado River has a storage capacity of more than four times the average annual natural runoff and an elaborate system of laws governing water allocation and system operation, some entities believe that the basin has been drought-proofed. However, a close examination of the hydrologic reconstructions of unimpaired flows at Lee's Ferry based on tree-ring analysis indicates that droughts may be much more severe than analysis of conventional streamflow data indicates.

A study of severe, sustained drought affecting the Southwest involves establishing three factors: (1) streamflow characteristics and probabilities of drought events, (2) the effects of storage systems on water availability at key points in the system over time given streamflow characteristics during drought, and (3) relationships between available supplies and projected demands at key points during the period of analysis.

The Analytical Model

Severe and sustained drought along the Colorado River will be buffered by the massive storage provided in the Colorado River basin. For this reason, this drought analysis focuses on annual operation of the two major reservoirs in the upper and lower basins—Lakes Powell and Mead—and related effects on water availability relative to aggregate demand. Unimpaired measured streamflow at Lee's Ferry was taken to be the "base case" hydrology and compared against reconstructed sequences of equal length from tree-ring data dating back to 1520 (D. Meko, University of Arizona, personal communication, 1989).

The Colorado River Annual System Regulation Model (CRASR) (Metropolitan Water District of Southern California, 1980) was used

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

to simulate the basin operation. It was substantially modified and rewritten by the authors to reflect current operating regimes. The model simulates annual regulation of Lake Powell and Lake Mead. Seasonal operation is not critical in this study because of the large amount of available storage capacity compared to annual runoff.

Nature of the Deficit

Streamflow is a useful hydrologic variable for drought studies at the basin scale, because it integrates a variety of processes over a watershed, including runoff, soil moisture, and evapotranspiration. Changes in climate are thus reflected as changes in streamflow volumes over a specified time period. Typically in the western United States, the record consists of a period of streamflow measurements dating back 60 to 80 years. At Lee's Ferry, the Colorado River streamflow records date back to 1906. In contrast, streamflow records at Lee's Ferry based on tree-ring studies have been reconstructed back to 1520, offering a record of greater variability than the measured record.

Under conditions of climate change, the magnitude of the change in the statistical parameters is unknown. However, we do know that climate will cause a change not only in the mean (Q), but also in the variance (σ2) and the skew (γ). Rather than performing a sensitivity analysis with these parameters and creating what-if scenarios, streamflow reconstructions provided by tree-ring analyses from the basin's recent past are used as a basis for this climate analogy study of droughts.

A comparison of some reconstructed sequences with hydrologic streamflow records of equal length is shown in Figure 10.6. What is most striking is the smaller mean annual flow for the reconstructed sequences—which ranges from 12.97 to 13.78 maf (16.0 to 17.0 × 109 m3)—compared to the 1906 to 1985 record of 15.06 maf (18.6 × 109 m3). Stochastic flows generated from an autoregressive model of order 1 (AR(1)) were also analyzed in this drought study (Kendall and Dracup, 1991). It has been shown that hydrologic sequences provided by tree-ring reconstructions were the most appropriate choice for this type of analysis (Kendall and Dracup, 1990).

Time Interval and Threshold Level of the Data

Drought analyses are typically considered on a monthly or annual time scale. For this analysis, a time scale of one-year

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.6 Eighty-year mean flows for the Colorado River at Lee's Ferry are nonstationary.

Source: Adapted from data obtained from D. Meko, personal communication, October 1990.

intervals was chosen, as this is consistent with the time resolution period of tree-ring reconstructions.

Threshold levels in drought analyses are typically taken to be a streamflow's mean or median. Other appropriate thresholds may include average annual demand. For this analysis, the threshold level was taken to be the mean annual measured unimpaired flow at Lee's Ferry, important because comparisons using this threshold level will be made with annual unimpaired flows at Lee's Ferry reconstructed from tree-ring data.

Drought Parameters for the Colorado River

A study of the Colorado River using streamflow indicates that there is a tendency for low flows to follow low flows and for high flows to follow high flows. One possible reason for this behavior

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

is that the process that governs streamflow has a nonstationary mean. Incorporation of tree-rings into this type of analysis gives a measure of support to this idea. An anthropogenically-induced climate change would exacerbate this process. Evidence of nonstationarity in the mean is shown in Figure 10.6.

Previous analyses performed by a panel on water and climate for the National Research Council (1977) argued against a stationarity-independent or short-memory process for the Colorado River streamflows. Furthermore, the panel estimated that the long-term mean for flows at Lee's Ferry was 13.5 maf (16.7 × 109 m3) (Dracup, 1977).

A drought may be characterized as any year or series of consecutive years during which average annual streamflow is continuously below some specified threshold level, which is typically taken to be the long-term mean (Yevjevich, 1967; Dracup et al., 1980a). A drought event is considered to be composed of three defining attributes: duration (D); severity (S), which is the cumulative deficit; and magnitude (M), which is the average water deficit, such that S = M × D. The parameters are interrelated; two are necessary and sufficient to completely define a single drought event (Dracup et al., 1980b). Duration and severity are the most correlated parameters and may be considered the two primary parameters dependent on streamflow values. Magnitude is a secondary parameter; duration and magnitude are weakly correlated.

Twenty hydrologic droughts occurred between 1906 and 1985 (see Table 10.1). Of these, four had a severity of over 15 maf (18.5 × 109 m3), with a duration ranging from 3 to 5 years. It is apparent that a severe and sustained drought in the Colorado River basin is not necessarily produced by a single continuous drought event but may arise from a series of events separated by 1 or 2 years. For the period of record, a series of droughts, which account for the critical period, began in 1954 and continued through 1971.

Two sets of 80-year sequences from tree-ring reconstructed streamflows at Lee's Ferry that exhibited different types of severe droughts were chosen for the purpose of comparison. The periods selected are 1520 to 1599 (Figure 10.7) and 1600 to 1679 (Figure 10.8). These periods display different types of droughts with regard to their number, duration, and severity. The 1520 to 1599 sequence displays a series of droughts that contribute collectively toward severe and sustained drought conditions, while the 1600 to 1679 sequence displays a single severe drought.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

TABLE 10.1 Historic hydrologic droughts on the Colorado River at Lee's Ferry, 1906 to 1985.

Drought No.

Duration (Years)

Starting Year

Severity (maf)

Magnitude (maf/Year)

1

1

1909

2.954

2.954

2

2

1911

0.665

0.333

3

1

1914

0.769

0.769

4

1

1916

1.103

1.103

5

1

1920

2.290

2.290

6

2

1925

2.432

1.216

7

1

1932

6.609

6.609

8

5

1934

16.520

3.304

9

2

1940

9.384

4.692

10

1

1944

1.617

1.617

11

2

1946

5.506

2.753

12

2

1951

4.678

2.339

13

4

1954

20.092

5.023

14

3

1960

15.267

5.089

15

2

1964

10.422

5.211

16

4

1967

9.176

2.294

17

2

1972

2.971

1.486

18

1

1974

1.945

1.945

19

3

1977

15.351

5.117

20

1

1982

7.410

7.410

FIGURE 10.7 Streamflow sequence at Lee's Ferry, 1520 to 1599.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.8 Streamflow sequence at Lee's Ferry, 1600 to 1679.

Maintaining the same threshold level of 15 maf (18.5 × 109 m3) (the long-term mean from 1906 through 1985) for all sequences, 15 drought events were identified for both tree-ring reconstructed streamflow records. The drought characteristics differ significantly. The period from 1520 through 1599 exhibits four droughts, beginning in 1573 and extending through 1597. These are shown in Table 10.2. Conversely, the period from 1600 through 1679 displays a single severe drought with a 15-year duration, as shown in Table 10.3.

Index-Sequential Hydrologic Sequences

In order to develop probability estimates of water supply, many U.S. agencies use the index-sequential method to create a set of realistically probable streamflows (Kendall and Dracup, 1991; U.S. Bureau of Reclamation, 1985; Metropolitan Water District of Southern District, 1989). Each sequence differs from the next only by the fact that the first streamflow year is incremented by one. For example, in hydrologic sequence one, the first historic streamflow

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

TABLE 10.2 Hydrologic droughts on the Colorado River at Lee's Ferry, Arizona, from 1520 through 1599, as reconstructed from tree-ring analysis.

Drought No.

Duration (Years)

Starting Year

Severity (maf)

Magnitude (maf/Year)

1

1

1523

6.000

6.000

2

2

1528

2.300

2.300

3

5

1532

27.20

5.440

4

2

1538

8.000

4.000

5

7

1542

33.10

4.729

6

1

1552

2.100

2.100

7

1

1556

0.600

0.600

8

2

1559

6.500

3.250

9

3

1562

7.400

2.467

10

5

1567

9.400

1.880

11

5

1573

16.30

3.260

12

3

1580

13.90

4.633

13

5

1584

29.90

5.980

14

7

1590

34.60

4.943

15

3

1598

6.600

2.200

TABLE 10.3 Hydrologic droughts on the Colorado River at Lee's Ferry, Arizona, from 1600 through 1679, as reconstructed from tree-ring analysis.

Drought No.

Duration (Years)

Starting Year

Severity (maf)

Magnitude (maf/Year)

1

2

1601

5.800

2.900

2

2

1608

4.100

2.050

3

3

1612

5.100

1.700

4

1

1617

0.500

0.500

5

1

1620

1.900

1.900

6

7

1623

18.80

2.686

7

3

1631

20.80

6.933

8

6

1635

15.90

2.650

9

2

1642

1.100

0.550

10

6

1645

18.80

3.133

11

2

1654

14.80

7.400

12

1

1657

1.200

1.200

13

15

1659

52.10

3.773

14

4

1675

8.100

2.025

15

1

1680

1.100

1.100

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

year is assumed to occur in the first demand year of a study. In sequence two, the second historic streamflow occurs in the first demand year, with the first historic streamflow being wrapped to the bottom of the deck, so to speak. Consequently, there are as many sequences as there are years of hydrologic data. In order to develop probability distribution functions, this approach was used to generate pseudo-likely hydrologic traces of both the measured and tree-ring reconstructed streamflows.

Matalas and Fiering (1977) noted that a bias exists in geochronological records such as tree rings when used to estimate streamflows. Long-term mean estimates based on streamflows constructed from tree rings are reasonable, but tree-ring indices are more normally distributed and more highly correlated than recorded historical streamflows. However, it has never been determined whether these differences in statistical properties are significant enough to offset the utility of the reconstructed flows in water resource systems analyses.

Demand Assumptions

System regulation studies to determine future water supplies are performed over a series of years to accommodate growing demands. For this study, a demand period from 1989 to 2020, or 32 years, was simulated to accommodate the large drought periods observed in the tree-ring reconstructed hydrologic sequences. Demand data are from data used by the Bureau of Reclamation as the base case for its Colorado River System Simulation model (U.S. Bureau of Reclamation, 1985). Upper and lower basin demand schedules used in the analysis are shown in Tables 10.4 and 10.5.

The model output is directly related to the hydrologic sequence input. For water system design purposes, streamflow is recognized as a random process, of which the historic hydrologic record is one expression.

The hydrologic sequences comprise what can be taken to be year-to-year water supply. Coupled with this are year-to-year water demands. Lower basin demands include all uses from Lake Mead and downstream. The demands include those for California, Arizona, Nevada, and Mexico. Water losses due to evaporation or seepage from the conveyance system can be regarded as an additional demand.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

TABLE 10.4 Upper Colorado River Basin Water Demands (maf).

Water Year

Moderate Demand

1988-89

3.540 acre feet

1989-90

3.550

1990-92

4.040

1991-92

4.050

1992-93

4.060

1993-94

4.070

1994-95

4.080

1995-96

4.090

1996-97

4.200

1997-98

4.310

1998-99

4.410

1999-2000

4.520

2000-01

4.630

2001-02

4.640

2002-03

4.660

2003-04

4.680

2004-05

4.700

2005-06

4.720

2006-07

4.740

2007-08

4.760

2008-09

4.780

2009-10

4.800

2010-11

4.830

2011-12

4.840

2012-13

4.860

2013-14

4.860

2014-15

4.890

2015-16

4.900

2016-17

4.910

2017-18

4.930

2018-19

4.940

2019-20

4.960

 

SOURCE: U.S. Bureau of Reclamation, 1985.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

TABLE 10.5 Normal Lower Colorado River Basin Water Demands (maf).

Water Year

Normal Demands (inc. Mexico and losses)

1988-89

9.370

1989-90

9.460

1990-92

9.480

1991-92

9.490

1992-93

9.490

1993-94

9.500

1994-95

9.500

1995-96

9.510

1996-97

9.510

1997-98

9.520

1998-99

9.530

1999-2000

9.540

2000-01

9.550

2001-02

9.550

2002-03

9.550

2003-04

9.550

2004-05

9.550

2005-06

9.550

2006-07

9.550

2007-08

9.550

2008-09

9.550

2009-10

9.550

2010-11

9.550

2011-12

9.550

2012-13

9.560

2013-14

9.560

2014-15

9.560

2015-16

9.570

2016-17

9.570

2017-18

9.570

2018-19

9.580

2019-20

9.580

 

SOURCE: U.S. Bureau of Reclamation, 1985.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

Comparison of Severe, Sustained Drought Simulations

Simulations were made for the hydrologic periods 1906 to 1985 (the base case), 1520 to 1599, and 1600 to 1679. Each simulation produced results from 80 index-sequential hydrologic sequences and included reservoir storages, releases, and imposed shortages in the upper and lower basins. These results were ranked using a Weibull plotting position in order to form empirical cumulative distribution functions. Average storages, releases, and shortages were computed. The simulations were based on one interpretation of the Law of the River and the 602(a) storage requirement.

Lake Powell

Results were derived from simulating 80 index-sequential sequences for each of the three 80-year hydrologic periods studied. Each simulation included streamflow at Lee's Ferry and reservoir storages and releases given water demands as specified in Tables 10.4 and 10.5 for periods beginning in the water year 1988-1989.

Exceedance probabilities developed for Lake Powell indicate that a repeat of a hydrologic period like that of 1520 to 1599 or 1600 to 1679 would be significant. Figure 10.9 indicates that at about the 90 percent exceedance level (i.e., the level at which there is a 90 percent probability that the reservoir storage will be equal to or greater than the indicated amount), the difference between the base-case period 1906 to 1985 and the period 1520 to 1599 is about 11 maf (13.6 × 109 m3) for the year 2000. Conversely, there is a 10 percent probability that the reservoir storage will be equal to or less than the indicated amount. Furthermore, Lake Powell is at minimum power pool level at about the 90 percent exceedance level, as shown in Figure 10.9.

For the hydrologic period 1520 to 1599, the minimum power pool probability exceedance decreases to about 80 percent by the year 2010, as shown in Figure 10.10, and to about 75 percent by the year 2020, as shown in Figure 10.11.

A comparison of the two drought scenarios shows that they have similar probability curves with maximum differences at the higher exceedance levels. While both sequences exhibit quite different types of droughts, preceding or subsequent high flow periods tended to produce an equalizing effect between them, when considering the entire 80-year sequence.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.9 The amount of storage Lake Powell would contain in the year 2000 under conditions like those that have occurred in three historic hydrologic periods: 1520 to 1599, 1600 to 1679, and 1906 to 1985. The curves show the chance that storage would exceed the indicated level.

FIGURE 10.10 The amount of storage Lake Powell would contain in the year 2010 under conditions like those that occurred during the periods 1520 to 1599, 1600 to 1679, and 1906 to 1985.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

Figure 10.11 The amount of storage Lake Powell would contain in the year 2020  under the three historic hydrologic scenarios.

Lake Powell average storage levels were calculated by taking the average from 80 wrapped sequences for each simulation year. Results are shown in Figure 10.12. Note that the average storage level difference under severe sustained drought conditions ranges from approximately 5 to 6 maf (6.2 to 7.4 × 109 m3) from 1998 to 2020.

Lake Mead

Lake Mead simulation results are similar, with the difference between the base case and both drought sequences on the order of 6 maf (7.4 × 109 m3) for the year 2000 at about 90 percent exceedance. Minimum power pool exceedance probabilities are about 80 percent and are shown in Figure 10.13. For the most severe drought scenario (1520 to 1599), minimum power pool exceedance levels decrease to about 50 percent by the year 2010, while those for the sequence 1600 to 1679 decrease to about 70 percent (see

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.12 Lake Powell average storage comparison: the average storage levels the lake would contain if history repeats itself.

FIGURE 10.13 The amount of storage Lake Mead would contain in the year 2000 under conditions like those from the periods 1520 to 1599, 1600 to 1679, and 1906 to 1985. The curves show the chance that storage would exceed the indicated level.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

Figure 10.14). By simulation year 2020, these exceedance levels decrease to about 40 percent and 55 percent respectively (see Figure 10.15). The minimum power pool exceedance probability predicted by the base case is greater than 95 percent.

Average storage levels for Lake Mead are shown in Figure 10.16; the figure indicates a maximum between the base case and severe and sustained drought scenarios.

Upper Basin Shortages

Water shortages in the upper basin were allocated according to the interpretation of the Law of the River and storage provision requirements. Following the same procedure used for the development of exceedance probabilities for reservoir storages, similar cumulative distribution functions were developed for upper basin

FIGURE 10.14 The amount of storage Lake Mead would contain in 2010 under hydrologic conditions like those that have occurred in the past.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.15 The amount of storage Lake Mead would contain in 2020 under  hydrologic conditions like those from three past periods.

FIGURE 10.16 Lake Mead average storage comparison: the average storage levels  the lake would contain if history repeats itself.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

shortages. Year 2000 shortage probabilities are shown in Figure 10.17. The base case hydrologic sequence (1906 to 1985) indicates no shortages. The sequence 1520 to 1599 indicates shortages of about 300,000 acre-feet (0.37 × 109 m3) at the 75 percent exceedance level. The 1600 to 1679 sequence indicates upper basin shortages of about 300,000 acre-feet (0.37 × 109 m3) at about the 35 percent exceedance level.

Shortages of about 300,000 acre-feet (0.37 × 109 m3) at the 95 percent and 85 percent exceedance levels are indicated by year 2010 for the two drought sequences (see Figure 10.18). The base case sequence indicates no shortages. Similar results are shown for simulation year 2020, except that exceedance probabilities have increased in all cases, as shown in Figure 10.19. That is, shortages are present at the 98 percent and 88 percent exceedance probability levels for the two drought scenarios, while base case shortages are indicated at 50 percent exceedance probability.

Average shortages for the 32-year simulation period are shown in Figure 10.20. Upper basin shortages are not shown as exceeding 300,000 acre-feet (0.37 × 109 m3), which is consistent with the prevailing Bureau of Reclamation operation policies.

FIGURE 10.17 Possible upper basin water shortages in the year 2000 under hydrologic conditions like those from 1520 to 1599 and 1600 to 1679.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.18 Possible upper basin water shortages in the year 2010  under two historic hydrologic scenarios.

FIGURE 10.19 Possible upper basin water shortages in the year 2020  under three historic hydrologic scenarios.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.20 The average water shortage that would occur in the upper basin if history repeats itself.

Lower Basin Shortages

Calculated lower basin shortages are much more severe than those for the upper basin. For simulation year 2000, no shortages are predicted by the base case hydrology, although 1.45 maf (1.8 × 109 m3) shortages are calculated at the 70 and 30 percent exceedance levels for the drought sequences 1520 to 1599 and 1600 to 1679, respectively, (see Figure 10.21). For the year 2010, the base case hydrology yielded 1.45 maf (1.8 × 109 m3) shortages at about the 50 percent exceedance level, while the 1520 to 1599 drought sequence shows the same shortages at the 95 percent exceedance level. The drought sequence 1600 to 1679 indicates 1.45 maf (1.8 × 109 m3) shortages at 85 percent exceedance, as shown in Figure 10.22. For simulation year 2020, the base case exceedance levels increase to about 60 percent, as shown in Figure 10.23. Note that for the hydrologic periods 1520 to 1599 and 1600 to 1679, the exceedance probabilities stay approximately the same. Lower basin average shortages are shown in Figure 10.24.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.21 Lower basin water shortages in the year 2000 under hydrologic conditions like those during the periods 1520 to 1599 and 1600 to 1679.

FIGURE 10.22 Lower basin water shortages in the year 2010 under hydrologic conditions like those during the periods 1520 to 1599, 1600 to 1679, and 1906 to 1985.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

FIGURE 10.23 Lower basin water shortages for the year 2020 under hydrologic  conditions like those from three historic periods.

FIGURE 10.24 Lower basin average water shortages that would occur  if history repeats itself.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

Conclusions of Drought Analysis

A repeat of the hydrologic periods 1520 to 1599 or 1600 to 1679 on the Colorado River would have significant impacts on water availability and shortage allocations in the upper and lower basins. It should be noted that results presented here are not speculative what-if climate scenarios. They represent outcomes of hydrologic sequences that have been reconstructed from the historical past. Observation of Figure 10.6 indicates that the river has had above-average flow of 15.06 maf (18.6 × 109 m3) over the past eighty years as opposed to its previous average of about 13.5 maf (16.7 × 109 m3) per year. The potential for anthropogenic climatic change effects could exacerbate a situation in which the equilibrium level of the Colorado River over the long term (its mean flow) is lower than that measured over the past century. Shortages on the order of 300,000 acre-feet (0.37 × 109 m3) (under current system operation regimes) could occur in the upper basin as a result of hydrologic conditions similar to those revealed in the study of tree-ring data. Lower basin shortages on the order of 1.5 maf (1.9 × 109 m3) have probabilities of occurrence of over 90 percent and about 85 percent, respectively, for the 1520 to 1599 and 1600 to 1679 droughts.

Conclusions and Implications for Climate Change

Flooding of the Colorado River in the lower Colorado River basin caused substantial damage to homes and businesses in the spring and summer of 1983. Abnormal meteorological events—the greater than normal precipitation—contributed to the flooding but were not solely responsible for it. Two other factors that contributed to the flooding were: (1) the practice of maintaining a system of full reservoirs to satisfy demands from consumptive users, hydroelectric generators, and recreational interests; and (2) physical encroachment into the flood plain, made possible by the dams along the river. Although these factors could be managed physically, thereby averting the flood risks seen in 1983, there are many contrary interests that may interfere with steps to mitigate or prevent flood damage in the future. Although priorities for Colorado River management are mandated by U.S. law, such management has historically been the product of political and economic constraints created by the river's many beneficiaries.

If the business-as-usual approach has been an optimistic one, then areas of future study need to focus on activities that can help

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

mitigate negative impacts in the upper and lower Colorado River basins. These include conservation practices, the conjunctive management of surface and ground water, and systems for orderly and mutually agreeable movement of water to higher-valued uses, including temporary water transfers under drought conditions.

The merits of the drought study lie in the use of data drawn from actual climatic events. While there is a degree of bias in the tree-ring reconstructed sequences, estimates of the streamflow mean are considered reliable.

It is doubtful that Lake Powell and Lake Mead would be allowed to drop to minimum power pool levels before other types of shortage procedures were invoked, since in a real-time operation there is no way to forecast whether a drought has ended. This analysis distributes shortages in the upper and lower basins in a manner that would be consistent with existing laws and operating criteria. This analysis is not intended to be a statement of how shortages would actually be allocated. An area for future research is a critical assessment of potential shortage allocations and strategies that might be invoked by the Secretary of the Interior in the event of a severe, sustained drought. It is in the political arena that decisions will be made about ways of responding to, mitigating, and avoiding drought effects.

The extraordinary amount of storage along the Colorado River gives the region robustness and resilience. The lesson of importance to policymakers is to develop flexibility in the operating procedures that matches the variability of the resource being controlled: namely, stremflow. For example, rigid operating procedures may need to be set aside and new procedures developed that allow for changing the current power pool and base flow requirements.

REFERENCES

Arizona Republic. June 22, 1983. The flood facts. P. A2.

Arizona Republic. July 1, 1983. Residents who build near river should have expected trouble, watt says. P. A2.

Arizona Republic. Sept. 3, 1983. $80 million flood was ''unavoidable.'' P. B1.


Benjamin, J. R., and C. A. Cornell. 1970. Probability, Statistics and Decision for Civil Engineers. New York: McGraw Hill.

Broadbent, R. (Commissioner, U.S. Bureau of Reclamation). 1983. Prepared statement for the U.S. House of Representatives, Com-

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

mittee of Interior and Insular Affairs, September 7, Yuma, Arizona.

Brown, B. G. 1988. Climate variability and the Colorado River compact: implications for responding to climate change. Chapter 12 in Glantz, M. H., ed., Societal Responses to Regional Climate Change: Forecasting by Analogy. Boulder, Colo.: Westview Press.

Colorado River Board of California. 1969. California's Stake in the Colorado River. Los Angeles: The Board.

Colorado River Board of California. 1990. Executive Directors Report, October. Los Angeles: The Board.


Debler,? 1930. Hydrology of the Boulder Canyon Reservoir with Reference Especially to the Height of Dam to Be Adopted. Denver, Colo.: U.S. Bureau of Reclamation.

Dozier, L., and J. Brown. 1983. Management of Colorado River Dams. Presented at the Eighth Annual Workshop of the Rural Communities Institute, Gunnison, Colo., August 1-3.

Dracup, J. A. 1977. Impact of the Colorado River basin and Southwest water supply. In J. Wallis, ed., Climate, Climatic Change, and Water Supply. Washington, D.C.: National Academy Press.

Dracup, J. A., K. S. Lee, and E. G. Paulson. 1980a. On the definition of droughts. Water Resources Research 16(2):297-302.

Dracup, J. A., K. S. Lee, and E. G. Paulson. 1980b. On the statistical characteristics of drought events. Water Resources Research 16(2):289-296.

Dracup, J. A., S. L. Rhodes, and D. Ely. 1985. Conflict between flood and drought preparedness in the Colorado River basin. Pp. 229-244 in J. Lundquist, U. Lohm, and M. Falkenmark, eds., Strategies for River Basin Management. Dordrecht, Netherlands: D. Reidel Publishing, Co.


Freeny, G. B. 1981. Managing conflicts on the lower Colorado River system. In Proceedings of the National Workshop on Reservoir Systems Operations. New York: American Society of Civil Engineers.


Glantz, M. H., ed. 1989. Forecasting by Analogy: Societal Responses to Regional Climate Change. Boulder, Colo.: National Center for Atmospheric Research.

Glantz, M. H., ed. 1988. Societal Responses to Regional Climatic Change: Forecasting by Analogy. Boulder, Colo.: Westview Press.


Hundley, N., Jr. 1986. The west against itself: the Colorado River--an institutional history. Pp. 9-49 in G. D. Weatherford

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

and, F. L. Brown, eds., New Courses for the Colorado River: Major Issues for the Next Century. Albuquerque, N.M.: University of New Mexico Press.

Hundley, N., Jr. 1975. Water and the West. Berkeley and Los Angeles, Calif.: University of California Press.

Jacoby, G. C., Jr. 1975. Lake Powell Effect on the Colorado River Basin Water Supply and Environment. Lake Powell Research Project Interim Report. University of California at Los Angeles: Institute of Geophysics and Planetary Physics.


Kendall, D. R., and J. A. Dracup. 1991. A comparison of index-sequential and AR(1) generated hydrologic sequences. Journal of Hydrology 122:335-352.


Matalas, N. C., M. B. Fiering. 1977. Studies in geophysics. In Climate, Climatic Change, and Water Supply. Washington, D.C.: National Academy Press.

Metropolitan Water District of Southern California, Resources Division. 1989. State Water Project Simulation Model.

Metropolitan Water District of Southern California. 1980. Colorado River Annual System Regulation Model, Report No. 941.

Meyers, C. J. 1966. The Colorado River. Stanford Law Review: pp. 1-75.


Nathanson, M. N. 1978. Operating Criteria. Chapter 7 in Updating the Hoover Dam Documents. Washington, D.C.: U.S. Bureau of Reclamation.

National Research Council (NRC). 1977. Climate, Climatic Change, and Water Supply. Washington, D.C.: National Academy Press.


Rhodes, S. L., D. Ely, and J. A. Dracup. 1984. Climate and the Colorado River: the limits of management. Bulletin of the American Meteorological Society 65:682-691.


Stockton, C. W. 1977. Interpretations of past climatic variability from paleoenvironmental indicators. In J. Wallis, ed., Climate, Climate Change, and Water Supply. Washington, D.C.: National Academy Press.


Upper Colorado River Commission. 1982. Thirty-Fourth Annual Report. Salt Lake City, Utah: The Commission.

U.S. Army Corps of Engineers. 1982. Colorado River Basin-Hoover Dam: Review of Flood Control Regulations, Final Report. Los Angeles, Calif.: The Corps.

U.S. Bureau of Reclamation. 1983. Status of Reservoirs: Colorado River Storage Project. Washington, D.C.: U.S. Department of the Interior.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×

U.S. Bureau of Reclamation. 1985. Colorado River Simulation System. Washington, D.C.: U.S. Department of the Interior.

Weatherford, G. D., and G. C. Jacoby. 1975. Impact of energy development on the Law of the Colorado River. University of New Mexico School of Law National Resources Journal 15:171-213.


Yevjevich, V. M. 1967. Objective Approach to Definitions and Investigations of Continental Droughts. Hydrology Paper 23. Ft. Collins, Colo.: Colorado State University.

Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 177
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 178
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 179
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 180
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 181
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 182
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 183
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 184
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 185
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 186
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 187
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 188
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 189
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 190
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 191
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 192
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 193
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 194
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 195
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 196
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 197
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 198
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 199
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 200
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 201
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 202
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 203
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 204
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 205
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 206
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 207
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 208
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 209
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 210
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 211
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 212
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 213
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 214
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 215
Suggested Citation:"10. Climate Uncertainty: Implications for Operations of Water Control Systems." National Research Council. 1991. Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings. Washington, DC: The National Academies Press. doi: 10.17226/1911.
×
Page 216
Next: 11. Economic Consequences of Climate Variability on Water in the West »
Managing Water Resources in the West Under Conditions of Climate Uncertainty: A Proceedings Get This Book
×
Buy Paperback | $50.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The question of whether the earth's climate is changing in some significant human-induced way remains a matter of much debate. But the fact that climate is variable over time is well known. These two elements of climatic uncertainty affect water resources planning and management in the American West.

Managing Water Resources in the West Under Conditions of Climate Uncertainty examines the scientific basis for predictions of climate change, the implications of climate uncertainty for water resources management, and the management options available for responding to climate variability and potential climate change.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!