Freshwater is a crucial resource in the Northwest, and the effects of climate on water resources strongly influence and link with many other potential impacts. Despite the region's reputation as a wet place, this actually only applies west of the crest of the Cascade Mountains, and even there the summers are dry. Most of the region receives less than 20 inches of precipitation a year, and dry summers make freshwater a limiting resource for many ecosystems and human activities. Water supply, availability, and quality are already stressed by multiple growing demands. The stresses are likely to increase especially when you consider that the maximum precipitation occurs in January while the maximum water use usually occurs in July or August.

The Cascades largely divide rivers partly or entirely controlled by rainfall, whose flow peaks in winter, from those controlled by snowmelt, whose flow peaks naturally (because the operation of reservoirs changes this peak) as late as June. The Columbia, a snowmelt-dominated river, is one of the nation's largest, draining roughly three-quarters of the region and carrying 55-65% of its total runoff. The Columbia is the region's primary source of energy and irrigation water. It is managed by multiple agencies for multiple, often conflicting uses, including electricity production, flood control, fish migration, habitat protection, water supply, irrigation, navigation, and recreation.

Agriculture takes the largest amount of Columbia River water. Other demands, though, are growing, for example, the amount of water needed for in-stream flows necessary to protect salmon. With more than 250 reservoirs and 100 hydroelectric projects, the Columbia River system is among the most developed in the world and has little unclaimed water remaining for additional demand. At the same time, regional population growth and changing priorities are intensifying competition for water. Because its watershed is so large, the amount of water in the Columbia reflects an average of weather conditions over large areas and whole seasons. Because of this, the effects of climate on its flow can be detected and projected with more confidence than for smaller river systems, which respond most strongly to shorter-term and more local events (e.g., a single storm in a small location).

The water cycle in the Columbia watershed shows a strong signal from both ENSO and PDO. Because the warm phases of both of these oscillations tend to make winters both warm and dry, they have strong effects on snowpack, streamflow, and regional water supply. Warm-phase years accumulate less snowpack, and shift from snow accumulation to melting earlier in the season. The warm phases of ENSO and PDO each reduce average annual flow in the Columbia by roughly 10%, with an even larger reduction of peak spring flow. The effects of the two oscillations are nearly additive, so years with both in their warm phase have brought the lowest snowpack and streamflow, and the highest incidence of droughts.

The future probably holds increases in winter flooding and increases in summer drought. The following discusses potential impacts on the region's freshwater resources and strategies that might assist in addressing the potential impacts from climate changes conclude the discussion.

While all the climate models studied agree that Columbia River streamflow is likely to increase in winter and decrease in summer, the models disagree on the size of the winter and summer changes. As a result of that disagreement they also disagree on whether the total annual flow increases or decreases. Projections for annual flow in the 2020s range from a 22% increase to a 6% decrease and for the 2050s range from a 10% increase to a 19% decrease.

Basic characteristics of river basins help determine their likely response to these climate changes. In basins with cold winters like the Columbia, whose flow is snowmelt dominated, the very likely effect of these changes of temperature and precipitation will be to increase winter flow (mainly because more of the precipitation falls as rain) and decrease summer flow (mainly because less water is stored as snow). Both precipitation and temperature matter when looking at future water availability in the Columbia. When changes in temperature and precipitation are considered separately, temperature -- which climate models project with greater confidence than precipitation -- has the larger effect on summer streamflows. In warmer basins like most of the basins west of the Cascades, higher winter run-off increases the likelihood of wintertime flooding.

Even if future streamflow were not expected to change as the climate changes, the region would still experience severe difficulty in satisfying water demands during the next century. Much of this problem would be due to rapidly growing population and other increasing demands such as for expansion of irrigated farmlands. This projected growth in demand, even without any problems or reductions in supply, implies bitter conflicts among the various water users -- irrigated agriculture, fish protection, municipal and industrial supply, and hydropower. Now, add to these stresses, reductions in summer water supply because of warming climate.

Though projected changes in total annual flows are relatively small, the projected seasonal shifts are likely to bring large effects. Projected warmer, wetter winters are likely to bring increases in winter flooding risk in low elevation river basins, while continuing growth in population will very likely increase the property vulnerable to such flooding. Impacts on human health are also possible, particularly where urban storm-sewer systems are inadequate to handle the increased runoff during high-flow events.

In contrast to the smaller river basins, the Columbia River system has an extensive network of dams that store water for summer use and guard against flooding. There is little increased risk of flooding because spring peak flows are not expected to increase much and because the Army Corps of Engineers manages reservoirs for flood control and has clear authority to prevent flooding. The system is much less able to function with low flows than high because there is no single clear authority to mitigate droughts.

When droughts occur, hundreds of entities including states, irrigation districts, fisheries managers, and tribes all assert their rights to scarce water. Reduced summer flows are likely to bring substantial reductions in both hydropower and freshwater availability by mid-century. Any reductions in available water would aggravate already-sharp water allocation conflicts due to population growth, expansion of irrigated farmland, and increasing priority for maintaining salmon habitat. Droughts, like the 1994 drought in the Yakima Valley and the regional 2001 drought, cause bitter disputes and huge economic losses for those who depend on summer water and do not receive it. In a future with drier summers, such conflicts and losses will become more frequent.

Strategies for providing adequate water resources under changing climate conditions vary in focus and in cost. Some options, listed below, relate generally to water resource issues that are felt today as well as are likely in the future. This list is not exhaustive and should be used as a starting point.

Managing under projected future scarcity is highly likely to require some combination of reducing demand, increasing supply, and reforming institutions to increase flexibility and regional problem-solving capacity.

Demand for water can be reduced through various means:

• technical methods e.g., more efficient irrigation methods, changes in agricultural land management, or high-efficiency plumbing fixtures in new construction; and
• policy approaches e.g., tax incentives for conservation investments; revision of rate structures; and development of institutions to allow reallocation of water to higher-value uses, particularly in times of shortage.

Supply can be increased through various methods:

• technical options such as: developing groundwater sources or using groundwater recharge for water storage; improving system management by using seasonal climate forecasts; and
• policy considerations, including: promoting optimal use of existing lower-quality supplies, e.g., by delivering reclaimed non-potable water for some uses; or developing non-hydro electric generating capacity; or if permitted by both governments, negotiating water purchases from Canada.

Increasing institutional flexibility is an essential component of response, but is difficult because of the fragmentation of the current system, however, developing a shared information resource such as a regional water management database could improve cooperation among federal, state, and academic communities.

Planning decisions for investment in infrastructure with expected lifetimes of many decades should incorporate long-term climate change projections.