Margaret S. Hrezo
Radford University, Virginia
Carol J. Haley
Virginia Water Resources Research Center
To better understand why society is still reacting to droughts rather than planning for their eventuality, the activities of various levels of government in planning for droughts must be examined. Until very recently, the policies of most State governments for dealing with drought have been to "wait 'til it rains" and in the meantime provide some form of emergency assistance to localities and hope a catastrophe can be avoided. Many States also, as a matter of implicit policy, place primary responsibility for drought action with the Federal Government and local governments who, throughout the Nation's history, have been the primary levels of government involved when droughts occur. The policies of both the Federal Government and the local governments also have been to provide emergency relief and to try to reduce water demand to match the available supply. This governmental approach, however, has not reduced the economic losses or the level of inconvenience and suffering of the Nation's citizens. With each succeeding drought, the cycle repeats itself. As a result, the following questions arise: Do the States have a responsibility for planning for droughts? If the States do have such a responsibility, what type of planning should take place? Is there a possibility that drought-related activities of the Federal Government and the local governments are likely to change in the future, and if so, how and why?
In strictly climatic terms, a drought can be defined as an interval of time, generally months or years in duration, during which the actual moisture supply at a given place consistently is less than the climatically expected or climatically appropriate moisture supply(Palmer, 1965, p. 3). Drought generally is defined as being meteorological, hydrological, or agricultural. However, the ultimate consequences of droughts have to be placed in the context of the effects on the social and economic activities of a given region (Evan C. Vlachos, Colorado State University, written commun., January 1988). Thus, the climatic attributes of drought also need to be defined in social and economic terms because it is in these contexts that water management becomes important (Evan C. Vlachos, Colorado State University, written commun., January 1988). Yevjevich and others (1978, p. 32) have suggested that we think in terms of "sociological drought," which is defined as the meteorological and hydrological conditions under which less water is available than is anticipated and relied on for the normal level of social and economic activity of the region.
Due in part to our agrarian heritage and because water deficiencies can develop rather quickly in the root zone, most droughts are thought of as agricultural droughts. This characterization can be confusing, however, because an agricultural drought can occur in the midst of a hydrological wet period. The distribution of precipitation during a year can be such that there is a shortage of moisture (drought) during a critical growing period for a variety of crops, yet the total precipitation for the year can be greater than the historical yearly average. In this discussion of institutional and management aspects, droughts are considered as moisture shortages that seriously affect the established economy and the quality of life of a region.
The following analysis of the roles of government in planning for droughts begins with an examination of the traditional activities of government during a drought. If the traditional approach by government seems less than desirable, then the obstacles to greater involvement need to be identified and evaluated to determine if they can be removed or circumvented. The suggestions by King and others (1958) in their "Model Water Use Act" for managing water resources during droughts are examined in some detail to determine if they provide a minimum approach for States to consider in developing a plan for managing water resources for drought conditions. To better assess the status of State involvement, the drought programs and drought-related actions of the 48 conterminous States are reviewed in terms of the minimum criteria recommended by King and others (1958). Lastly, the expanded role that the Federal Government might have in planning for and mitigating the effects of droughts pursuant to existing Federal water-quality legislation is considered.
The local approach to the management of water resources during droughts is not responsive to other drought-induced issues such as minimum instream flows because these issues are not within the purview of local government. Yet water for wildlife, increased contamination due to low flows, and decreased navigation and hydroelectric-power generation are issues equally as important as the ones being addressed by local government. Water shortages cause low streamflows that have an adverse effect on fish and wildlife habitat. If there are no limits as to how much flows can decrease or for how long (the longer the period of low flow, the greater the stress), the recovery time for a habitat can be extremely long, or a habitat can be lost permanently. Low flows can increase salt-water intrusion, increase health hazards because of increased concentration of toxic substances and pathogens, decrease hydroelectric-power generation, and curtail recreational opportunities. Decreased precipitation also increases the potential for brush and forest fires and wind erosion of topsoil.
With a few exceptions, the response of government at any level to the shortages caused by decreased precipitation has been to react rather than to adopt a proactive approach to minimize the effect of droughts. Drought planning at the local level in many areas appears to be given a low priority because of the randomness of droughts, the limited resources for planning, the limited jurisdiction (local government might not be able to control streamflow levels), and the programs of the Federal Government to provide disaster relief in time of crisis. As a result, local governments are encouraged to accept an implicit policy of doing only what can be done after a crisis has occurred.
A notable exception to inaction at the local level to plan for drought conditions occurred in the Washing-ton, D.C., area. The leadership role in this case was taken by the Interstate Commission on the Potomac River Basin (ICPRB), but the implementation of the water resources plan was by local government. The plan finally adopted saved between $200 million and $1 billion compared to longer scale structural solutions previously proposed (Sheer, 1986, p. 106). Implementation of the plan was through eight separate but interlocking contracts executed in 1982 (Sheer, 1986, p. 106).
With the exception of eight States-Arkansas, California, Connecticut, Delaware, Florida, Minnesota, New Jersey, and South Carolina-the activities of the State governments in managing water resources during droughts have been minimal. Most State governments have not passed legislation providing for additional drought planning beyond slight modifications in their water laws. Governors, on occasion, will declare counties or designated areas as disaster areas in order to make individuals eligible for Federal relief.
Although governments can plan effectively for droughts, fundamental problems that deter action need to be examined and understood before drought planning can become a reality. Five obstacles to planning for droughts-specificity, randomness, drought phenomenon, cost of droughts, and political considerations-are discussed here.
Planning for and management of hazardous events presuppose that those events are well defined and discernible to all. The planning necessary to reduce the effects of most natural hazards is difficult because the intensity and frequency of the events are unknown, although there is never any question as to their eventual occurrence. Although no technical expertise is required to determine when floods, volcanic eruptions, or earthquakes have been experienced, considerable uncertainty exists as to when droughts start and end. A drought is almost a "non-event"(Evan C. Vlachos, Colorado State University, written commun., January 1988). Any discussion about planning for drought conditions and management of water resources, therefore, requires some definition as to what constitutes a drought. This lack of specificity can be a major contributing factor, although unstated, as to why the planning for droughts and the managing of water resources during droughts have received less attention than they deserve.
Droughts produce the same type of situation. For individuals experiencing a drought, options to deal effectively with water shortages are limited. Collective action by all the individuals provides the best solution. The use of ground water during times of drought in many Eastern States provides a comparable situation to the one described by Hardin in the case of common pastures. The common law in many Eastern States deal with ground water as a common resource that is appropriated under the Rule of Capture (Cox, 1982, p. 115). What you capture is what you get, and those having the deepest wells and the largest pumps get the most water. Given these circumstances, the solution for an individual during drought conditions might be to drill a well. If all individuals act in the same manner, a variety of consequences can occur to the detriment of each. A shallow aquifer eventually can be depleted, the individuals can be competitors for water and cause larger and larger cones of depression as deeper wells and larger pumps are utilized, or the increased pumping can cause saltwater intrusion, which will destroy the quality of the water in the aquifer for all. The best solution for all parties might be an agreement between the individual well owners, or restriction by the local government, to curtail the time and rate of pumping. Under these circumstances, an individual who pumps water from a common aquifer cannot plan effectively for droughts. If an individual well owner seeks to con-serve ground water or plan for a water shortage, he or she needs to be aware that the water he or she does not pump will probably be pumped by others.
Because droughts affect larger geographic areas than those occupied by single communities, the ability of an individual community to respond effectively is affected by the actions of similar communities in the drought area. The position of each community in this larger arena can be analogous to that of the individual in the community. Individual actions by each community can be counterproductive to the policy best for the region as a whole. For one community, the solution may be the building of a reservoir on a stream that is the water source for other communities downstream. As each community opts to resolve its water needs without regard to its neighbors, the stream can become an inadequate water source for all. The development of a regional water supply for all communities might be the best solution, but this will require the consensus of all of the communities.
The length of a drought also has a significant effect on the total costs; long droughts are more costly than shorter droughts. A sustained drought, such as the one in the 1930's, can have economic and social costs that are never quantified. During the 1930's drought, for example, agriculture was abandoned in some sections of the Nation; this abandonment, in turn, caused dislocation of people and severe impairment of the economic substructure that supported agriculture. The 1930's drought had an effect on a whole generation of Americans, wherever they lived and however they made their living (Harrison, 1977, p. 34). Even if the value of human life is ignored, the total economic losses from droughts can be staggering.
In contrast to a lack of public support, special-interest groups at the State level might oppose activities that are essential for an effective water-management plan applicable to droughts. For example, farm groups in Virginia strongly oppose any Federal, State, or regional water-management plan because they believe there should be no regulation of water apart from the Riparian Doctrine (Mark Tubbs, Virginia Farm Bureau, commun., 1981). Water management. of necessitv, must be at the core of any program to mediate the effects of water shortages that occur during droughts, but political factors can substantially dampen the interest in managing water resources even during droughts.
The responsibility for managing water resources during droughts, once assumed by the State, needs to be vested in such a manner as to require timely action and not be vulnerable to legal challenges by groups who do not favor an approach taken by the State. Although expanding the Governor's powers to deal with disasters by including droughts might be expeditious, the action taken by most States generally is to group management activities with the authority primarily designed to respond to disasters after they have occurred rather than to undertake planning activities to reduce the cost of droughts in advance of their occurrence. Colorado, Delaware, New Jersey, and North Carolina are examples of States that have used executive power to develop statutory guidelines that define droughts and delineate interaction among State agencies responsible for water resources (Hrezo and others, 1986a, p. 162-163).
The authors of the "Model Water Use Act"(King and others, 1958) developed five planning tools-identification of drought indicators, designation of government authority, notification of the public, curtailment of water use and maintenance of revenues, and monitoring of water-user compliance-to cope with planning for water shortages associated with droughts. These tools addressed the following fundamental questions:
Because of the difficulty of deciding when droughts start and end, specific drought indicators must be used to decide when to implement a water-management plan. When such indicators have been identified, water users can formulate contingency plans and make decisions on future economic investments(Hrezo and others, 1986b, p. 47). The drought indicators must be precise and susceptible to little, if any, subjective decision making. The latter makes the indicators vulnerable to court action by those who oppose advanced planning.
A variety of drought indicators can be used, including the Palmer Index (a drought-severity index),in stream flows, historical data on the present and anticipated needs for water, the degree of subsidence or saltwater intrusion, the potential for irreversible adverse effects on fish and wildlife, and reservoir or ground-water conditions relative to the number of days of water supply remaining (Hrezo and others, 1986b,p. 47). Usually it is desirable to select a number of drought indicators to reflect the seasonal relation of sup-ply versus demand. The Delaware River Basin Com-mission (Hrezo and others, 1986b, p. 51) relies on five drought indicators-precipitation, ground-water levels, reservoir storage, streamflow, and the Palmer Index. Ranges of values for each of these indicators are assigned to one of four drought stages-normal, drought watch, drought warning, and drought emergency (Hrezo and others, 1986b, p. 51). To activate any one of the drought stages, three of the five drought indicators must indicate a given drought stage (California Department of Water Resources, 1988, p. 34). The drought indicators should not be so complex as to cause uncertainty about whether some stage of the water-management plan should be activated; for example, if precipitation and reservoir storage are two drought indicators, the decision to activate the plan will be unclear if the precipitation is less than normal while reservoir storage is normal.
When there is only one source of water supply, one drought indicator may be sufficient; for example, when a city's only source of water is a reservoir, the water-management plan can be activated if reservoir storage, expressed as a percentage of normal seasonal capacity, decreases below a specified percentage. The phasing criteria used by Manchester, Conn., are an example (table 1)-a drought watch goes into effect when reservoir storage is at 70 percent of normal seasonal capacity, and stage I of the water-management plan becomes operational when reservoir storage is 57 percent of normal seasonal capacity.
|Drought stage||Drought stage initiating conditions||Demand-reduction objective|
|Drought watch||Reservoir water levels at 70 percent of normal seasonal capacity.||Informational only, raise public awareness|
|Stage 1||Reservoir water levels at 57 percent of normal seasonal capacity.||Cut back withdrawals from reservoirs by 5 percent or reduce total system use by 3.8 percent.|
|Stage 2||Reservoir water levels at 40 percent of normal seasonal capacity.||Cut back withdrawals from reservoirs by 30 percent or reduce total system use by 20 percent.|
|Stage 3||Reservoir water levels at 0 percent of normal seasonal capacity||Eliminate withdrawals from reservoirs and reduce total system use by 70 percent.|
|Demand-reduction amount, in million gallons per day|
|Stage 1 Minor shortage potential||Total system storage is not filled to capacity as of June 1. Streamflow and snowmelt forecasts indicate that inflows will be inadequate to fill storage facilities before the beginning of the peak-use season.|| Water-system management |
|Stage 2 Moderate shortage potential||Total system storage is predicted to fall below the level required to meet expected demands during a 1-in-50-year drought. System inflows continue to be low. Weather forecasts predict a continuing trend of warmer, drier than normal conditions.|| Water-system management|
|Stage 3 Serious shortage||Total system storage drops below the level required to meet expected demands during a 1-in-50-year drought. System inflows continue to be low. Weather forecasts predict a continuing trend of warmer, drier than normal conditions.|| Water-system management|
|Stage 4 Severe shortage||Conditions described for stage 3 occur near the end of the peak-use season.|| Water-system management |
|Stage 5 Critical emergency||Customer demands and system pressure requirements cannot be met.|
|Demand-reduction amount, in million gallons per day|
|Stage 1 Minor shortage potential||Total system-storage levels are dropping due to the increased use associated with a warm, dry summer. Weather forecasts predict a continuing trend of warmer, drier than normal conditions.|| Water-system management|
|Stage 2 Moderate shortage potential||Total system storage is expected to fall below the level required to meet expected demands during a 1-in-50-year drought.|| Water-system management|
|Stage 3 Serious shortage||System inflows continue to be low.|| Water-system management |
|Stage 4 Severe shortage||Weather forecasts predict a continuing trend of warmer, drier than normal conditions.|| Water-system management|
|Stage 5 Critical emergency||Customer demands and system pressure requirements cannot be met.|| Water-system management |
A sliding scale for drought indicators also can be shown graphically in terms of the storage in a reservoir, as shown by the operation curves for three reservoirs in New York that also are part of the Delaware River basin (fig. 1). When the actual reservoir level drops below the drought-warning zone or drought zone, schedule of reduced diversions from the basin to the various localities takes effect. The Pennsylvania Drought Contingency Plan for the Delaware River Basin is based on these criteria. Where ground water is one of the main sources of water supply, drought indicators based on ground-water levels can be used; the Alameda County Water District in California has such a plan (California Department of Water Resources, 1988, p.38).
Figure 1. Operation curves for three New York City reservoirs (Cannonville, Pepacton, and Neversink) in the Delaware River Basin. (Source: U.S. Geological Survey, 1986, p. 31.)
Most water-management plans have correlated successive stages of a drought strategy to certain deficit-reduction goals (California Department of Water Resources, 1988, p. 39). Fewer than three stages in a plan can result in marked differences in the actions to be implemented between the first and the second stages. More than five stages in a plan, however, can cause frequent transitions between stages, which can decrease the effectiveness of the plan. An example of a workable plan is the Seattle Water Department five-stage plan for reducing water use (table 1). Agencies having water-management plans have determined that fall droughts have a lesser probability of occurrence, but; if they do occur, they are likely to develop more quickly and be more severe.
A reduction in water use will cause a reduction in the revenues of the water suppliers. The reduced revenues come at a time when costs are greater because of expenditures made to deal with the drought. In the absence of a water-revenue reserve or a drought-emergency account, water suppliers need to either increase water rates or impose a drought surcharge(California Department of Water Resources, 1988,p. 49). The use of a drought surcharge has several advantages compared to a simple increase in water rates. For example, the drought surcharge is easier to ad-minister, and the amount of revenue to be generated is more predictable. This surcharge probably is more acceptable to the customer because it is a one-time charge that is understandable and allays the fear that a water-rate increase to make up revenues lost during droughts will continue when the drought is over.
Figure 2. State-level methods for managing water resources during drought conditions. (Source: Information from Hrezo and others, 1986a, p.146-157).
The eight States that have comprehensive water-management plans incorporate all of the basic concepts suggested in the "Model Water Use Act." However, the approach by each State has been different and reflects individual State needs and existing water-allocation systems. The drought indicators that are used to determine when the plans are to begin and to end probably were the least precise component in the plan of each of the eight States. All the water-management plans are activated at the State level except in Florida, where activation is the responsibility of existing water-management districts, and in California, where activation is in the charge of the intrastate districts mandated by statute.
Water laws of 27 States have been modified to establish systems of water-use categories (fig. 2) .Although having these systems in place eases the management of water supplies once a drought occurs, most of the changes are designed to meet emergency water shortages rather than to provide a strategy to mitigate the effects of droughts before an emergency develops (Hrezo and others, 1986a, p. 149). These 27States can be subdivided into those that have modified the riparian doctrine with some form of a permit system; that use a modified appropriation system; or that rely on civil-defense, disaster, or emergency legislation that incorporates droughts within its definitions (Hrezo and others, 1986a, p. 148). A description of these sub-divisions follows.
Six States (fig. 2) in the East have modified then common-law riparian doctrine by requiring water-use permits for some uses. Many of the permits are subject to water-use restrictions during times of severe water shortages. Georgia has water-use restrictions that are implemented during water emergencies. Iowa has made all use of surface and ground water subject to a state-wide permit system; permit users are denied water whenever the minimum instream-flow standard is reached. In Kentucky, permits give no guarantee of a water right during droughts. Conservation programs are required under all permits in Maryland when the safe is threatened by existing or projected water demand. North Carolina's permit system is limited to designated capacity-use areas; other laws in North Carolina grant water-emergency powers that affect the capacity-use areas and the rest of the State. Pennsylvania requires permits only of public water suppliers using surface water; permits are conditioned on there being an emergency water plan in place. All these States have taken some initial steps to manage their water resources during droughts, but they utilize few of the planning tools suggested by King and others(1958) in their "Model Water Use Act" for water management (Hrezo and others, 1986a, p. 155).
Twelve States (fig. 2) in the West have attempted to deal with severe water shortages by modifying the appropriation doctrine for allocating water. Under this doctrine only the most senior appropriation in time will receive water during water shortages. All these States have modified the appropriation doctrine to some degree to accommodate water shortages caused by droughts. New Mexico, for example, provides for changing the place of diversion, storage, or use of water if an emergency exists (Hrezo and others,1986a, p. 152). In Oregon, the Director of Water Resources can order State agencies and political subdivisions to develop water-conservation and water-use-curtailment plans that encourage conservation, reduce nonessential water use, prevent waste, provide for reuse of water, and allocate or rotate the supply to domestic, municipal, and industrial uses(Hrezo and others, 1986a, p. 152). Utah allows its State Engineer to use regulatory authority to prevent waste in order to mandate rotation of irrigation water where no use will benefit from a diversion of the water supply (Hrezo and others, 1986a, p. 152). In these 12 States, legislation provides some management tools to assist in decreasing the effects of a water-shortage crisis. Little of the modifying legislation includes the planning tools suggested by King and others (1958)in their "Model Water Use Act" to manage water shortages due to droughts in a comprehensive way.
Nine States (fig. 2) in the Midwest and Mid-Atlantic region have left the management of droughts, from a State perspective, almost entirely to the Governor. These States perceive water shortages as primarily the responsibility of local government. The State has a role when the shortage becomes extensive in terms of time and scope and affects such a large area of the State that the water shortage is perceived as a severe crisis approaching a disaster. These States choose to deal with water shortages due to droughts on an ad hoc basis. New York passed a law that will take effect in 1990 and will require a conservation plan for surface withdrawals.
When droughts affect more than one State but are not national in scope, interstate commissions and compacts can provide the management function to mitigate drought effects. Their effectiveness is predicated on having well-publicized plans and specific rules for planning purposes so that all users know how they will fare when the river flows cannot accommodate all the withdrawal demands (Hrezo and others,1986a, p. 164). The Delaware River Basin Compact(Public Law 87-328, 75 Stat. 688, 1961) and the 1978Potomac River Low Flow Agreement are examples of interstate compacts that have been used to plan for the problems associated with droughts. Such commissions or compacts require, however, the approval of all of the involved State legislatures, which could be a difficult task politically. The fact that the number of such entities is small is ample evidence of the difficulty and time required to establish them.
Any plan prepared under such process shall include, but not be limited to, (1) a process to(i) identify, if appropriate, salt water intrusion into rivers, lakes, and estuaries resulting from reduction of fresh water flow from any cause, including irrigation, obstruction, ground water extraction, and diversion, and (ii) set forth procedures and methods to control such intrusion to the extent feasible where such procedures and methods are otherwise a part of the waste treatment management plan***.
The legislative history amplifies on this point (1972U.S. Code Congressional and Administrative News, p. 3706) (emphasis added):
Salt water intrusion no less than point sources of discharge, alters significantly the character of the water and the life system it supports. Salt water intrusion often devastates the commercial shellfish industry. It must be accounted for and controlled in any pollution control program. It makes no sense to control salts associated with industrial or municipal waste point sources and allow, at the same time, similar effects to enter the fresh water as a result of intrusion of salt water. Fresh water flows can be reduced from any number of causes. The bill requires identification of those causes and establishment of methods to control them so as to minimize the impact of salt water intrusion.
Droughts can be one of the major causes for reduced freshwater flows, and the law requires that methods be established to control or to minimize the causes of reduced freshwater flows that allow saltwater intrusion.
The Siting Requirement under the Safe Drinking Water Act (Public Law 93-523) contains language(emphasizes added) that can be construed to include drought conditions and, thus, impose on States the requirement to adapt siting criteria that include drought conditions:
Before a person may enter into a financial commitment for or initiate construction of a new public water system or increase the capacity of an existing public water system, he shall notify the State and, to the extent practicable, avoid locating part or all of the new or expanded facility at a site which: (a) Is subject to a significant risk from earth-quakes, floods, fires or other disasters which could cause a breakdown of the public water system or a portion thereof; or ***.
Droughts could conceivably be "other disasters" provided for in the Safe Drinking Water Act. A combination of these water-quality considerations might, in the future, cause the Federal Government to give greater consideration to managing the effects of droughts, at least with respect to those effects that impinge on water-quality issues.
Lastly, the quality of surface-water bodies is affected markedly by the runoff that occurs when precipitation increases after a drought. During the drought, pollutants accumulate on the land surface and on other surfaces, such as pavement and structures. It is not uncommon for droughts to be followed by a period of abnormally high precipitation that tends to aggravate the already existing water-quality problems by rapidly flushing large loads of pollutants into surface-water bodies. After the drought in England in the 1970's, the nitrate concentration in the Thames River increased to the point where the public-supply intakes had to be closed (Blackburn, 1978, p. 54). This kind of post drought problem may be reflected in what the Federal Government requires States to do to meet water-quality standards. Some drought planning may occur at the State level as a byproduct of the action taken to address this water-quality problem.
Blackburn, A.M., 1978, Management strategies-Dealing with drought: American Water Works Association Journal, v. 1978, p. 51-59.
California Department of Water Resources, 1988, Urban drought guidebook: California Department of Natural Resources, Department of Water Resources, Office of Water Conservation, Water Conservation Guidebook 7,144 p.
Comptroller General of the United States, 1979, Federal response to the 1976-77 drought-What should be done next: Washington, D.C., Government Printing Office,25 p.
Cox, W.E., 1982, Water law primer: American Society of Civil Engineers, Water Resources Planning and Management Division, Proceedings, v. 18 (WRI),p. 107-122.
Harrison, Robert, 1977, Response to droughts: Water Spectrum, v. 9, no. 3, p. 34-41.
Heathcode, R.L., 1986, Drought mitigation in Australia: Great Plains Quarterly, v. 6, p. 225-237.
Hrezo, M.S., Bridgeman, P.G., and Walker, W.R., 1986a,Integrating drought planning into water resources management: Natural Resources Journal, v. 26p. 141-167.
-1986b, Managing droughts through triggering mechanisms: American Water Works Association Journal, v. 1986, p. 46-51.
King, D.B., Lauer, T.E., and Zieglar, W.L., 1958, Model Water Use Act with comments, in Water resources and the law: Ann Arbor, University of Michigan, p. 533-614.
Palmer, W.C., 1965, Meteorological drought: U.S. Weather Bureau Research Paper 45, 64 p.
Sheer, D.P., 1986, Managing water supplies to increase water availability, in National water summary 1985-Hydrologic events and surface-water resources: U.S. Geological Survey Water-Supply Paper 2300,p. 101-112.
U.S. Geological Survey, 1986, National water Summary 1985-Hydrologic events and surface-water resources: U.S. Geological Survey Water-Supply Paper 2300,506 p.
Yevjevich, Vujica, Hall, W.A., and Salas, J.D., eds., 1978,Drought research needs-Conference on drought research needs, Colorado State University, Fort Collins, Colo., December 12-15, 1977, Proceedings: Fort Collins, Colo., Water Resources Publications, 288 p.