PART I-
WATER AVAILABILITY FOR AGRICULTURE IN THE SEMIARID WEST

Chapter 1-
Physical Limitations of Water Resources

by John Bredehoeft

Abstract

In considering the concept of the hydrologic cycle today one must take into account man's influence as an integral part of the functioning of the cycle. Except for the mining of groundwater, the same quantity of water is, on the average, in transit in the hydrologic cycle. Groundwater mining is extensive, especially in Arizona and the High Plains of Texas and New Mexico. Groundwater, however, is a one-time supply; to the extent that we mine it, we are faced with a shortage in the future. Both urban movement to the Southwest and energy development compete with agriculture for the available supply, especially in the areas of critical water supply, southern California and Arizona. Competition is present throughout the semiarid West; anywhere water is fully appropriated, increased urban and industrial supplies must come from agriculture. There seems little doubt that as we approach the limits of available water supply there will be increasing competition for water. In a classic economic sense increased competition implies a shortage.

The water supply of the West is nearly fully utilized. It is difficult to foresee major construction projects which will add significantly to the currently available supply. Several critical areas are now heavily dependent upon mining groundwater, a supply which will be depleted at some point in the future. Urban and energy developments, especially in the Southwest, are competing with agriculture for the available water. This competition will undoubtedly intensify, which poses two major issues for society:

1) How will society, at local, state, and regional levels, cope with the increased competition for water?

2) To what extent can the nation forego irrigated agriculture in the West without significantly decreasing its agricultural output?

---

- 18 -

It is not the intent of this chapter to address these issues; however, we will attempt to provide an overview of the current availability of water.

The Hydrologic Cycle

Traditionally, when considering the problems of water resources we hydrologists have been prone to think in terms of virgin or natural streamflow. However, it has become increasingly obvious that natural flow is a relict of the distant past. Man has impacted the water resources so dramatically, especially in the arid and semiarid West, that natural flow does not exist except perhaps in the most remote areas.

We must recognize that man's activities are today an integral and inseparable part of the hydrologic cycle. Our current understanding of the hydrologic cycle can be described in a paradigm suggested by Matalas, Landwehr, and Wolman. The three tenets of the active paradigm are:

i) human activity is inseparable from the natural system;

ii) quality is no less a concern than quantity of the water mass as it is distributed and moves through the cycle;

iii) the quantity of the water mass affects and is affected by the quality of the water.^[1]

If we accept the active paradigm as best characterizing our concept of the hydrologic cycle, then it is impossible to look at the physical and chemical limitations on water resources without looking at man's activities.

Available Water

Precipitation ultimately is the source of water resources. The average annual precipitation for the United States is depicted in Figure 1.1. That precipitation translates into runoff. West of the 100th meridian much of the land is characterized by less than one inch of runoff. The areas of abundant runoff in the West are easily identified in Figure 1.2. The relative magnitude of the average streamflow of the large rivers in the U.S. is shown in Figure 1.3. The major rivers of interest in the western states are the Columbia, the Colorado, the Sacramento, the Missouri, and

---

- 19 -

[Full Size]

Figure 1.1
Average Annual U.S. Precipitation, 1931-1960
Source: U.S. Council on Environmental Quality,  Environmental Trends,  Washington, D.C., 1981, p. 346.

- 20 -

[Full Size]

Figure 1.2
Average Annual U.S. Runoff
Source: Rickert, D.G., W.J. Ulman, and E.R. Hampton,
Synthetic Fuels Development-Earth Sciences Considerations,  U.S. Geologic Survey, 1979, p.45.

- 21 -

[Full Size]

Figure 1.3
Average U.S. Streamflow, 1941-1970
Source: U.S. Water Resources Council,  Essentials of Ground Water Hydrology
Pertinent to Water Resources Planning,  Bulletin 16, revised 1979, p.48.

their tributaries. Future large-scale surface water diversions must almost certainly come from these river systems.

Runoff comes largely from the mountains in the spring as snowmelt. The typical seasonal variation is illustrated by the long-term average monthly runoff for the Clarks Fork of the Yellowstone River near Belfrey, Montana, Figure 1.4. Storage of water, either in surface reservoirs or in aquifers, improves the timing between supply and demand, especially the seasonal demand for agriculture.

- 22 -

[Full Size]

Figure 1.4
Average Monthly Runoff, Clarks Fork of Yellowstone River
Source: Rickert et al.

Groundwater forms an additional resource. The important aquifers of the western United States are shown in Figure 1.5.

Depletion of Water

Given our picture of surface and groundwater, how much is utilized? Relative water depletion is depicted in Figure 1.6. Depletion is defined as "the total consumptive use plus any water exported from each basin, divided by the total supply". Groundwater mining has been excluded from the long-term supply. This is perhaps the most important single illustration in this paper. Several critical areas show up on the map of depletion:

1) Most of the lower Colorado River basin, southern California, and most of Nevada, where the depletion

- 23 -

[Full Size]

Figure 1.5
Extensive Aquifers of the U.S.
Source: U.S. Water Resources Council, 1979.

- 24 -

[Full Size]

Figure 1.6
Relative Water Depletion in the U.S.
Source: Rickert et al.

- 25 -

exceeds 100 percent. The differences are made up from mining groundwater.

2) South-central California, including the San Joaquin and Owens Valleys, where the depletion exceeds 75 percent.

3) The High Plains of Colorado and west Texas, where the depletion exceeds 75 percent.

4) Much of New Mexico, where the depletion exceeds 75 percent.

The depletion map is somewhat misleading, since instream flow requirements are not accounted for, and they are important constraints on water availability.

Groundwater constitutes an important additional source of water. Groundwater withdrawals are shown in Figure 1.7. California and Texas are the two largest users of groundwater, accounting for 37 percent of the total withdrawn nationwide, closely followed by Nebraska, Idaho, Kansas, and Arizona, which together account for an additional 26 percent of the total. These six states account for almost two-thirds of the groundwater withdrawn in the United States.

The relative importance of groundwater as a source of water in the semiarid West is depicted in Figure 1.8. Groundwater constitutes the major source of water, exceeding approximately 50 percent in much of the High Plains, a large portion of Arizona, and parts of California.

Much of the groundwater withdrawn is being mined. The Second National Water Assessment of the U.S. Water Resources Council^[2] identified areas of groundwater overdraft-"mining" in my terminology-as shown in Figure 1.9. The principal areas of overdraft identified west of the 100th meridian are (1) the high plains of Texas, New Mexico, Colorado, Oklahoma, and Kansas, and (2) large areas of Arizona. Moderate overdrafts occur over much of the area west of the 100th meridian.

Water Use

How is the water used? Figure 1.10 is a graph of water withdrawals for the period 1950 through 1975 for the entire U.S. The

- 26 -

[Full Size]

Figure 1.7
U.S. Groundwater Withdrawals, 1975 (million gallons per day)
Source: U.S. Water Resources Council, 1979.

- 27 -

[Full Size]

Figure 1.8
U.S. Groundwater Withdrawals, 1975
(percent of fresh water used from groundwater sources)
Source: CEQ, 1981.

- 28 -

[Full Size]

Figure 1.9
U.S. Groundwater Overdraft
Source: CEQ, 1981.

- 29 -

[Full Size]

Figure 1.10
U.S. Water Withdrawals, by Use, 1950-1975
Source: CEQ, 1981.

largest withdrawals are for power plant cooling and irrigation. Consumptive use, on the other hand, presents a very different picture. Figure 1.11 shows nationwide water consumption. Irrigation accounts for by far the largest fraction of consumption. In the western states irrigation accounts for more than 90 percent of the consumptive use.

Groundwater use is also interesting; the growth in groundwater withdrawal over the last 25 years has been almost exclusively for irrigation, as is shown in Figure 1.12. In 1977 42 million acres were irrigated, for which the consumption was approximately 82 billion gallons a day (92 million acre-feet per year). Something approaching one third to one half of that water came from groundwater, much of which was mined, as Figure 1.9 indicates.

- 30 -

[Full Size]

Figure 1.11
U.S. Water Consumption by Use, 1950-1975
Source: CEQ, 1981.

[Full Size]

Figure 1.12
U.S. Groundwater Use, 1950-1975

- 31 -

Eighty-four percent of the fresh water consumed in the coterminous United States is consumed in the 17 western states; most is utilized for agriculture. The acreages irrigated in the 17 western states are given in Table 1.1. California accounts for 23 percent of the total acreage; together, Texas and California account for 42 percent of the total.

- 32 -

Looking at statistics for the nation as a whole may appear to be somewhat misleading. However, since the 17 western states dominate the consumptive use, consuming 84 percent, the statistics for the nation are strongly influenced by the West, where agriculture is the primary consumer of water.

The Lower Colorado Basin

In any overview of the water resources of the semiarid West, the lower Colorado River basin and southern California stand out as the most critical areas for water. Another look at the depletion map, Figure 1.6, indicates that the water supply is more than 100 percent depleted in these areas. This is substantiated by the overdraft of groundwater shown in Figure 1.9.

The Colorado River is the principal long-term source of water for much of this area. Stockton and Jacoby,^[3] utilizing tree-ring data, reconstructed Colorado River streamflow back to 1512. Using this record they estimated the mean annual flow at 13.5 million acre-feet. This is approximately 2 million acre-feet less than anticipated when the water rights were divided in the 1922 Colorado River Compact. Unfortunately, the 1922 Compact was based on records of flow during a series of unusually wet years from 1906 to 1920. The availability of water from the Colorado is further complicated by a number of Indian claims upon the river which are as yet unresolved.

A synthesized record of the flow of the Colorado River below all major diversions, in Figure 1.13, portrays the outflow of the river into the Gulf of California. The downward trend of the residual flow, which is caused by an increasing use of water from the Colorado River, is evident. Usage by Mexico as well as by the United States is reflected in the residuals. (Under the terms of a treaty between the United States and Mexico in 1944, supplemented by various "minutes" and negotiations, Mexico is allotted an annual quantity of 1.5 million acre-feet.)

Diversions from the Colorado began considerably before 1900. However, prior to that year, annual net diversions generally were less than 1.0 million acre-feet. The residual flows during 1935-39 were unusually low, largely because of the initial filling of Lake Mead. Low flows from 1960 to 1978 reflect nearly complete use of the river. In 1979 and 1980, major floods in the Lower Colorado River basin downstream from the principal reservoirs resulted in larger outflows.

- 33 -

[Full Size]

Figure 1.13
Annual Flow of Colorado River
Above All Major Diversions, 1910-1980

Clearly all the water in the Colorado is currently utilized. The consumptive use within the basin is compared with entitlements from the river in Figure 1.14. The large consumptive use in Arizona is made up in part by groundwater mining.

The water in the Colorado is also plagued by an increasing load of dissolved salts. This load comes from a number of natural sources and from sources which are the result of man's actions. Approximately one third of the total salt load is the result of irrigation. Another 10 percent or so comes from Flaming Gorge Reservoir and from Lake Mead, where salts are being leached from geologic deposits inundated by the reservoirs. Figure 1.15 attempts to summarize both the concentration of dissolved solids as well as the total salt load.

- 34 -

[Full Size]

Figure 1.14
Consumptive Uses and Losses of Water in the
Colorado River System, 1971-1975 Averages

Water is in short supply in the Lower Colorado River basin. Population statistics indicate a growth in urbanization both in Arizona and southern California. If urban growth is to continue, there will undoubtedly be pressure to shift water away from agricultural use.

- 35 -

[Full Size]

Figure 1.15
Salt Load in the Colorado River, 1941-1978 Averages

Alternatives for Additional Water Supplies

A number of alternatives have been discussed for increasing the water supply. These are categorized for the purpose of discussion into: (1) increased surface storage; (2) increased groundwater development; (3) more efficiency of water utilization; and (4) large-scale interbasin transfers of water.

Increased Surface Storage

Surface storage is the traditional method of providing additional available water. Additional reservoir sites exist in some parts of the western states. Langbein^[4] has reviewed historic trends in reservoir development in the U.S. Table 1.2, taken

- 36 -

from Langbein, shows the reservoir capacity currently available in a number of the major river basins of the country. Langbein has suggested that a unit capacity of approximately 400 acre-feet of storage per square mile of drainage area represents a potential limit for reservoir development; the Colorado has a potential unit capacity of 400 acre-feet per square mile.

Langbein also plotted the historic trend of reservoir capacity; this plot is shown in Figure 1.16. The growth in capacity for all purposes and for withdrawal has flattened out since 1960. The question is whether this reduction in reservoir construction will continue, or if it is simply an aberration in long-term growth curve.

- 37 -

[Full Size]

Figure 1.16
Usable Major Reservoir Capacity in the U.S. since 1920
Source: Langbein, 1982.

Our assessment is that surface reservoirs will continue to be increasingly difficult to develop. Recent legislation such as the National Environmental Protection Act (NEPA) makes it easier for environmental groups to voice their interests. Every major new reservoir project seems likely to receive some resistance from opposing groups. Major conflicts will, in many instances, be settled politically. In arid regions such as the lower Colorado River basin, where water is particularly critical, additional reservoirs may evaporate as much or more water as is made available, thereby further concentrating the dissolved salts. Increasing surface storage in the lower Colorado is a losing proposition.

Increased Groundwater Development

Groundwater is already heavily utilized, as has been pointed out, much of its development resulting in mining of water. The increased costs of pumping imposed by increased energy costs have reduced groundwater pumping, especially in areas such as Arizona.

The one area with apparent potential for a major increase in groundwater development is Nebraska. Table 1.3 is a compilation of the water in storage in the Ogallala Aquifer, the result of an ongoing U.S. Geological Survey study of the system. Approximately two thirds of the water in storage is in Nebraska, an enormous reserve of groundwater. Only in Texas and New Mexico has more than 10 percent of the water initially in storage been depleted. The depletion statistics may be somewhat misleading, since it is economically impractical to remove all the water initially in storage; perhaps 50 to 70 percent is a reasonable estimate of what might be removed under favorable economic conditions.

These data indicate that only a small percentage of the water in the Ogallala has been removed. Obviously an enormous quantity of groundwater is still present for development in Nebraska.

More Effective Water Utilization

A number of measures have been suggested to effect better utilization of water available. Among these, increased irrigation efficiency, weather modification, reuse of wastewater, conjunctive use of groundwater, desalination, and increased use of saline water have been considered.

Increased efficiency of irrigation has obvious advantages. But a major nagging question is: what happens to the salts in the system when one increases the efficiency? A study of a reach of the Arkansas^[5] suggested that following an initial two-to-three-year period after increasing irrigation efficiency, groundwater in the shallow aquifer along the Arkansas River would become more saline. This increase in salinity of the groundwater would increase the salinity of the flow in the river.

Pillsbury,^[6] in an article in Scientific American entitled "The Salinity of Rivers", argues that salt buildup is a major problem for all irrigation projects. His thesis is that sufficient water must be applied to continually remove salt from the soils. Salt buildup seems to pose some limit on possibilities for increasing irrigation efficiency.

- 39 -

In such systems as the South Platte, or the Arkansas in Colorado, or the lower Colorado, most of the water goes to support beneficial transpiration. It seems questionable that increased efficiency can materially add to the useful supply.

Weather modification has received considerable attention. The data, although not totally conclusive, suggest that cloud seeding could increase precipitation locally, with a 10 percent increase in supply possible. The question remains as to what happens downwind-does cloud seeding reduce rainfall? This issue remains to be settled. However, it appears that some local increase in available supply is possible from weather modification.

- 40 -

Reuse of wastewater is another possible source of water. Reuse is already practiced in a number of places. In irrigation, reuse occurs through return flow, which replenishes the streamflow. Municipal wastes have been purchased in such areas as Phoenix, Arizona, for utilization in irrigation. The city of Irvine, California, reuses all of its wastewater, mostly for municipal irrigation.

Major metropolitan areas along the coast continue to discharge some wastes to the sea. Some of this water could be reused beneficially. However, the costs of cleaning it up may be such as to preclude it for use in agricultural irrigation.

The shallow aquifers in the earth provide an enormous fresh water reservoir. Many of these are already utilized extensively as active storage reservoirs. The conjunctive water use developments along the Platte, the Arkansas, the Rio Grande, and the Snake rivers are classic examples of utilization of the groundwater system as a storage reservoir.

In certain areas such as the southern San Joaquin Valley in California, groundwater reservoirs can be utilized to store water in periods of abundance. Already a number of such developments are well established elsewhere in California, particularly in Orange County and the Santa Clara Valley.

The groundwater aquifer has obvious advantages for storage as only small surface areas are affected, evapotranspiration is greatly reduced, and in many places the aquifer serves as an excellent filter for the water. On the other hand, aquifer storage has the disadvantage that it is sometimes expensive to recharge groundwater, especially if one has to utilize wells. How much impact conjunctive use will have in the overall water management in the West is difficult to forecast at this time.

The cost of desalinating water makes it too expensive, in most instances, for agriculture. However, the use of desalination for municipal and industrial use may reduce the competition for water currently utilized in agriculture. Saline water can also be utilized for industrial purposes such as cooling, and for special purposes such as slurrying coal. There is abundant saline groundwater over much of the West, and use of these resources could reduce the competition for water.

How effective more efficient water utilization measures will be in making water available is anyone's guess. If collectively they could make available 10 percent of the water currently used in agriculture, this would approximately equal all of the other consumptive uses. Ten percent may be an achievable goal.

Large-Scale Interbasin Transfers of Water

Large-scale interbasin transfers, particularly to the lower Colorado River basin, have been proposed as a source of water for some time. The major interbasin transfers are shown in Figure 1.17. The two really significant transfers occur in the Colorado basin and in California. By far the largest of these transfers occurs in California.

Traditionally, the states have primacy with respect to utilization of water. Large-scale interbasin transfers cannot take place without a change in state primacy. As water is perceived to be a critical commodity, state primacy will be harder and harder to change. We are pessimistic that this policy can be changed significantly to allow further large interbasin transfers between states. In fact the magnitude of the transfers in California has only been possible, in our judgment, because they occurred within a single state. Interbasin transfer continues to be a sensitive issue even in California, as witnessed by the 1982 referendum over the Peripheral Canal.

It seems problematical that major quantities of water are available for interbasin transfer. For example, Whittlesey and Gibbs,^[7] who reviewed the utilization of water in the Columbia for hydropower, concluded that water for irrigation in central Washington costs the general public $150 per acre per year in increased energy costs. This cost comes from lost hydropower downstream and from large quantities of energy to supply supplemental irrigation water which is provided irrigators at very low rates. Under such circumstances it seems highly unlikely that Washington would allow additional water to be diverted for irrigation within the state, and certainly it would fight a major interbasin transfer to another state. Similar situations exist in other western states which, at first glance, appear to have "surplus" fresh water.

Conclusions

It is increasingly difficult to effect major structural changes which would provide large quantities of water to those areas where water is in critical supply-southern California, Arizona, and the High Plains of Texas and New Mexico. Outside California, large interbasin transfers must face the issue of state primacy, a particularly difficult issue to overcome.

- 42 -

[Full Size]

Figure 1.17
Major Interbasin Water Transfers in the Western U.S.
Source: Modified from Geraughty, J.J., D.W. Miller,
F. Van Der Leeden, and F.L. Troise,  Water Atlas of the
United States,  Water Information Center, Port Washington,
N.Y., 1973.

- 43 -

One must turn to other measures to utilize more effectively the water that is currently available. Increased efficiency, weather modification, reuse, and conjunctive use, while perhaps not dramatic, have the potential to make better utilization of the available water supply. If collectively these measures could make available 10 percent of the water currently consumed by agriculture, that quantity would approximately equal the total of all other consumption in the West.

On the average, the quantity of water in transport in the hydrologic cycle remains unchanged. Except for the fact that we are mining groundwater, no less water is available than heretofore. The fact that we are approaching the limit of the water which can be developed means that there is, and will continue to be, ever-increasing competition for that water. Increased competition implies a higher value for the commodity. While as a society we rarely make large-scale water decisions purely on economic grounds, higher value also implies a higher price. Thus, in the context of increased competition, we have a shortage, at least of inexpensive water.

A number of areas in the West depend heavily upon groundwater for their supply. The areas of largest overdraft of groundwater are Arizona and the High Plains of Texas and New Mexico. Much of this water is a one-time supply, obtained by a "mining" operation. Although that is not necessarily bad, the supply is finite, and at some point, perhaps in the distant future, will be gone. Arizona has recently moved to strengthen its groundwater law to protect the resource.

The drought of the mid-70s in California motivated farmers to drill many new wells to tide themselves through a period of shortage. Now that the wells are drilled, they continue to be pumped, demonstrating that additional supplies of surface water do not always ease the overdraft of groundwater. In many instances, new supplies bring more land into production. To the extent that we are mining groundwater, we are running out of water.

The one bright spot in the water picture in the West is Nebraska, where a huge supply of groundwater is present in the aquifer. The figures on the Ogallala Aquifer in Nebraska suggest that this is probably the largest virtually untapped supply of water present in the 17 western states.

There can be little doubt that we are entering an era of continually increasing competition for water. In the Southwest, where water shortage threatens most critically, increasing urbanization and increasing energy development both compete

- 44 -

with agriculture, now the largest water consumer. Steve Reynolds, State Engineer of New Mexico, aptly states the current water situation when he says, "Water flows uphill toward money." To what extent agriculture in the West can accommodate the competition is the issue.

Discussion:
Harold C. Fritts

I see no significant weakness in Bredehoeft's lucid and concise discussion except that his projections are based upon relatively short hydrologic records. More specifically, paleoclimatic data indicate that worldwide climate changes occurred around the turn of the century-measurements such as Bredehoeft has used, which are confined to the 20th century, are likely to be biased by these changes.

I have used tree-ring widths as proxy climate records (substitutes for instrumented data) to estimate the magnitude of this bias. The ring widths of approximately 1000 trees from sites throughout the West were calibrated with the 20th-century instrumented climatic record throughout the United States. The

- 45 -

calibration equation was then applied to past ring-width growth to estimate past variations in climate.^[1] The estimates of climate were then verified with independent instrumented data^{[1], [2] available prior to the time period used for calibration. Finally, optimal reconstructions were selected based upon the best calibration and verification statistics.}

When these procedures are applied to California precipitation^[3] (Figure 1.18), pre-20th century precipitation is reconstructed to be below the 20th century mean; when a line is drawn through the plot, long periods of extended drought are evident.

Figure 1.19a shows another analysis^[4] in which the means for 1901-1970 temperature and annual precipitation in 11 North American regions were compared to the reconstructed means for 1602-1900. The 20th century was slightly cooler than the 17th-19th centuries for five regions in the West, and warmer for the remaining regions. It was 19 percent wetter in California (Region 2), above average in four additional southwestern regions, and dryer elsewhere.

Thus one can see that when expectations for precipitation are based solely on this century they would overestimate the long-term expectations for moisture because of recent anomalous trends in precipitation, particularly in California. Similarly, temperature projections would underestimate conditions west of the Rockies and overestimate them east of the Rockies.

Figure 1.19b shows the standard deviations of the reconstructions in the West for the 20th century, compared to the standard deviations for three prior centuries. They indicate a lower variability in 20th-century climate, especially in the amount of precipitation.

In addition, reconstructions of surface pressure^[4] suggest that coastal storms became more southerly displaced around the turn of the century, bringing higher moisture into California and the Southwest during winter. These storms appear to have traveled on the average in a northeast direction through the Great Lakes. The resulting southerly air flow brought less moisture and warmer temperatures to the eastern portions of the country. Prior to the 20th century storms apparently entered the country more often over the Pacific Northwest, passed over the Rockies, and traveled eastward or southeastward, bringing colder temperatures and more moisture to the East. However, this pattern was more variable, more severe storms were reported in the East,^[5] and plains droughts occurred that were as severe, if not more severe, than those in the 1930s.^[6]

- 46 -

[Full Size]

Figure 1.18
Average Annual Precipitation for 18 California Stations Reconstructed from 52 Western
Tree-ring Chronologies Dots represent eight-year weighted averages used to smooth
out the annual values. The horizontal line corresponds to the 1901-1961 mean value.

- 47 -

[Full Size]

Figure 1.19
Differences in Climate between the 20th Century and Three Prior Centuries Averaged within 11 Different Regions in
North America Figure 1.19a shows the change in means for 1901-1970 compared to 1602-1900.
Figure 1.19b shows the percent change in standard deviation for 1901-1961 compared to 1602-1900. The upper
value in each case is for the reconstructed annual temperature in degrees Centigrade; the lower value is for the
reconstructed annual precipitation, in percent.

- 48 -

Stockton and Boggess^[7] point out that the consequences of a dry and warm climatic change would be greatest in many areas of the arid Southwest, especially in the Lower Colorado, Missouri Arkansas-White-Red, and Texas Gulf, where groundwater is already extensively used.

Two primary future climatic projections have been made by climatologists today.^[8] The most popular is that the climate is likely to warm, due to the burning of fossil fuel and an increase of atmospheric CO/d/s-22/s+2/u. The second is that the climate was anomalous for the first half of the 20th century and that it is now likely to revert to the state of prior centuries. In either projection, climate in the semiarid West is likely to be drier, perhaps warmer, and more variable. This would indicate that the existing projections of water resources for the West based on the 20th-century hydrologic record are in all likelihood overestimates of what the water resources may be in the future.

Acknowledgement

The research reported here was supported in part by NSF Grant ATM75-22378 Climate Variability, Climate Dynamics Program and by the California Department of Water Resources, Agreement No. B53367.

Discussion:
Parry D. Harrison

Mr. Bredehoeft presents a rather gloomy picture of the water supplies in the West. Although much of what he says is correct, some of it tends to be a little misleading.

I do not fully agree that the water supply of the West is nearly fully utilized. Some river basins like the Colorado could be said to be fully utilized. However, an example of underdeveloped water supply is the Columbia River at The Dalles, with an average flow of over 140 million acre-feet per year; and the Willamette River at Portland averages over 23 million acre-feet per year. I could name at least ten other rivers that discharge between 3 and 15 million acre-feet per year.

While it is true that not all of these vast water supplies can be utilized and storage projects are very difficult to construct, many worthwhile storage projects have yet to be constructed. The problem with most of these rivers is that they are far from the heavy demand areas of California and the Southwest.

Runoff Predictions. A most difficult problem, and yet a paramount need, is accurate prediction of streamflows for the next six months, year, two years, and five years. Much has been written about the hydrologic cycle, the correlation of precipitation and runoff with sunspots, wind patterns, volcanic activity, effect of air pollution on weather, and effect on weather of atomic explosions. Nevertheless, the ability to predict precipitation and hence runoff with any degree of accuracy has not been demonstrated. The theory has been that the key lies in history; hence, studies of tree ring data, runoff records, stochastic analysis with the aid of computers-and we are still a long way from an acceptable solution.

Irrigation. Somewhere between 30 and 75 percent of water diverted for irrigation is a direct depletion and is consumed by evapotranspiration. The remainder either percolates into the ground and becomes part of the groundwater resource or returns to the stream and becomes available for reuse. Return flow usually is of poorer quality than the source. Many significant groundwater resources have been the result of or enhanced by irrigation. (Examples: Columbia basin in central Washington, Snake Plain aquifer in south-central Idaho, and the Sacramento and San Joaquin valleys in California).

Groundwater. Groundwater pumping from an aquifer that is being mined is a depletion of that resource, whatever its use. There may be some reuse or secondary use of the water pumped;

- 50 -

but unless it is reinjected, groundwater that is being mined is not a renewable resource. When it is used up, it is gone. In contrast, streams provide a renewable supply which comes every year, with some fluctuations depending on the weather.

Some things can be done to enhance or prolong the life of groundwater resources. These may include (1) artificial recharge: this may be feasible if there is an available resource; (2) limitation or restriction of groundwater pumping; and (3) efficient use of available supplies.

Competition for Water. Severe competition for limited water supplies in some areas may make it necessary to choose between irrigation, streamflows for fish, or domestic needs. Abundant and cheap water supplies enhance the quality of life in the West, but in the future some locations may not be able to enjoy them. Most countries do not have the luxury of abundant, high quality water supplies to the extent that we do in the United States and Canada. In the past cities have had to restrict the watering of lawns or filling of swimming pools to ensure adequate supplies for drinking, washing, and fire protection.

Many of those vying for control of water supplies have a single-track approach. Some typical comments have been:

"My need is paramount."

"Irrigation provides food; do you want to watch the fish swim upstream or would you rather eat?"

"Fish have been nearly eliminated by diversions and pollution for nearly 100 years; this has to be rectified now!"

"Recreation needs are increasing by leaps and bounds; water-based recreation must be given a high priority."

"Water is needed for power production. Power is the basis for our high standard of living. It means jobs!"

Narrow, unyielding approaches make it all the more difficult to find solutions to water supply problems facing the West. The competition is becoming keener every year. Cool heads and clear vision are needed to make good decisions that will influence the quality of the western lifestyle for years to come.