Grantee Research Project Results
2003 Progress Report: Impact of Climate on the Lower Yakima River Basin
EPA Grant Number: R827454Title: Impact of Climate on the Lower Yakima River Basin
Investigators: Vail, Lance , Stockle, Claudio , Branch, Kristi , Wigmosta, Mark , Scott, Mike , Leung, Lai-yung Ruby , Neitzel, Duane
Institution: Pacific Northwest National Laboratory , Washington State University
EPA Project Officer: Packard, Benjamin H
Project Period: April 15, 2000 through April 14, 2003 (Extended to September 30, 2003)
Project Period Covered by this Report: April 15, 2002 through April 14, 2003
Project Amount: $869,364
RFA: Integrated Assessment of the Consequences of Climate Change (1999) RFA Text | Recipients Lists
Research Category: Climate Change , Ecological Indicators/Assessment/Restoration , Water , Aquatic Ecosystems
Objective:
The objective of this research project is to develop and demonstrate a framework to assess the localized impact of climate change and climate variability on a diverse set of interdependent interests, including agriculture, water supply, water quality, fisheries, and economics. The goal is not to develop any specific new process models, but to integrate existing models to ensure that the linkages between the various models are managed appropriately. Because any such assessment is subject to considerable uncertainty, this framework explicitly considers the generation and propagation of uncertainty. This framework also communicates the tradeoffs associated with adaptation alternatives. The framework is being demonstrated on the Lower Yakima River Basin in central Washington State.
Progress Summary:
The integrated assessment framework development element of this project has three primary themes: horizontal integrated assessment, adaptation tradeoffs, and uncertainty. Progress is described relative to each of these three themes.
Horizontal Assessment
Any integrated assessment requires a computational and data management infrastructure to manage the diverse and disparate suite of data and models required to complete the analysis. A platform for integrated assessments of climate impacts requires several additional features. Based on discussions with stakeholders and decisionmakers, we have defined the critical requirements of a Climate Adaptation Management Platform (CAMP). The three critical requirements of CAMP are:
- Accountability. The primary goal of CAMP is to support the decisionmaking process. To be useful, it must ensure that the tradeoffs between multiple objectives (of interest to both decisionmakers and stakeholders) are clearly articulated. CAMP also communicates the sources and magnitude of key uncertainties.
- Accessibility. CAMP provides rapid access to models, data, and the rationale that underpins decisions. These are growing requirements of a skeptical and confused public. The ability to drill down through data and models ensures that decisionmakers are less likely to repeat past decision errors.
- Adaptability. CAMP must be able to rapidly assimilate new data and models to ensure that decisions are up to date with the most current information. Specific tools to automate the calibration process and to assimilate real-time data are requirements of the framework.
Whereas the above requirements would exist for any platform that attempts to operationalize adaptive management, there are additional unique requirements for this project. These include: (1) scaling tools to manage the considerable disparity between temporal/spatial scales of global climate models and other process models (e.g., hydrology or crops); (2) bias correction techniques to overcome the limitations of climate models to realistically reproduce historical data; and (3) data management tools to handle ensembles of climate predictions.
Status. The required models have been assembled, except for the groundwater and air quality models. The original proposed approach to represent the groundwater proved to be inadequate to properly simulate temperature and nutrients in the groundwater; therefore, we are developing a modified approach.
CAMP has been implemented as a metadata database. Integrated assessments of climate impacts involve a large (albeit often sparse) amount of diverse data. Each data set can be quite large, often involving spatial information. A credible data management system is essential for successful natural adaptation planning. CAMP does not attempt to build a single large database from the distributed databases. Instead, CAMP leaves data on their native platforms. CAMP manages the metadata and provides the pipeline linking data and models together. The CAMP metadata database will be Web accessible through the project's Web site.
Linking models to data and models to models often requires changes in required spatial and temporal scales. CAMP provides "scale filters" to transform climate and other data to the correct scale. CAMP does not attempt to provide a complete set of "scale filters." CAMP only provides a small set of such filters to manage the climate ensembles considered as examples of the protocols required to develop the filters.
CAMP also addresses the problem of structural differences between models. For instance, the salmon life cycle model employed in the integrated assessment utilizes categorical statements of the habitat, whereas the physical models used to support these categorical conclusions are deterministic process models. Fuzzy logic is used to overcome these structural differences and provide a reasonably intuitive basis for the stakeholder and decisionmaker to understand the associated processes. Fuzzy methods also are being considered in managing the climate ensembles.
Adaptation Tradeoffs
Past integrated climate assessments focused on defining the changes between historic and a future altered climate. The exact nature of the future climate is unknown, however, and unlikely to happen abruptly. The environment already is changing as a result of both climate and other forces. It is prudent, therefore, to develop methods to incrementally, but deliberately, take adaptation actions, while continuing to learn more about the efficacy of the adaptation actions and the nature of the climate change. Adaptive management encourages this deliberate experimental approach by identifying the next action in a sequence of actions whose exact sequence will be conditioned by observations occurring over time.
CAMP evaluates each adaptation alternative as a sequence of actions. Each action has an associated timescale (e.g., building additional reservoir capacity takes many years, whereas changing reservoir operating rules can be implemented immediately). This helps the decisionmaker clearly understand the priorities for allocation and scheduling resources for adaptation.
Status. A Web-based (Java and XML) application has been developed to allow a clear visualization of multiobjective tradeoffs resulting from the integrated assessment. The application runs through a Web browser for more maximum accessibility and provides ability to thoroughly drill down through the supporting data and models. Assessment endpoint metrics have been discussed and modified based on stakeholders' comments. Once stakeholders see the tradeoffs, it is expected that the endpoint metrics will change. Optimization functions within CAMP are provided with a Pareto genetic algorithm. The Pareto genetic algorithm was selected because of its robustness and its ability to work well in distributed computing environments. Pattern detection functions to assist in detecting robust alternatives are provided within CAMP using artificial neural networks and simple heuristics.
Uncertainty
Uncertainty is inevitable. Climate processes are notoriously complex and chaotic. Uncertainty starts with the climate and will continue to propagate through the integrated assessment. CAMP focuses the user on the uncertainty in selecting the correct "next action" and not on the process level uncertainties. Often the "next action" is appropriate for a large number of feasible future climate scenarios. The set of climate scenarios is captured as an ensemble of individual climate realizations.
Status. Where appropriate, random (probabilistic) considerations have been embedded in the assessment endpoints (e.g., mean time between junior users having less than 75 percent proratable water allocation). Entities for narrative descriptions of uncertainties (process, model, and data) have been included within the metadata database. The role of artificial neural networks in identifying robust alternatives has been defined. The metadata database includes entities for the explicit management of climate ensembles and ensembles downstream in the integrated assessment.
Framework Demonstration
The framework was demonstrated by evaluating the ability of reservoir managers and farmers to adjust to an altered climate. Reservoir managers can adjust to an altered environment by altering their reservoir release policies. Farmers can alter the crops they plant to adjust to the altered climate. Reservoir managers must consider the tradeoffs between instream flows to preserve aquatic habitat for fisheries and meeting demands for irrigation water diversions. Farmers must consider the tradeoffs between crop yields and the risks of water prorationing.
Surface water storage reservoirs are designed and operated to reshape the incoming hydrograph to meet multiple needs, such as irrigation, recreation, hydropower generation, and flood control. Reservoir operating policies are designed to reflect changes in the inflow hydrograph associated with historically observed climate variability. The inflow variability that a reservoir can successfully reshape is a measure of its adaptive capacity. Climate change may result in changes in both the general pattern of the inflow hydrograph and the associated needs to be served by the regulated flows. The change in the adaptive capacity of the reservoir system in the Yakima River Basin as a result of a systematic change in climate was evaluated. The adaptive capacity is expressed as the change in the ability of the reservoir system to serve multiple objectives.
Status. A spatially distributed hydrology model (DHSVM) was driven with 19 years of downscaled daily estimates of precipitation and air temperature for each of four scenarios. One scenario represented historical conditions. The three other scenarios were members of an ensemble of meteorological conditions predicted for the period of 2040 to 2059 under a "business-as-usual" projection of atmospheric CO2 change. The four meteorological scenarios were provided by a regional climate model. The runoff predicted by DHSVM was routed into regulated and unregulated streamflow portions. The regulated portion of the streamflow passes through a reservoir, whereas the unregulated portion is unaffected by the reservoir operation. The storage of the hypothetical reservoir was based on the combined storage of the five reservoirs in the Yakima Basin.
Operational simulations of a range of reservoir properties and water management policies were performed for each of the scenarios. To ensure the simulations were consistent with operational constraints, reservoir release decisions were based on imperfect forecast knowledge. An ensemble of regulated and unregulated streamflow forecasts was developed every 5 days based on the current conditions (notably snowpack) and meteorological time series for each of the 19 years from the current day to the end of the same water year. Prorationing required to ensure that nonproratable water rights were met was estimated for each member of the forecast ensemble. The rank of the magnitude of the proration level selected reflects the degree of risk aversion.
The Crop Systems (CropSys) model was used to evaluate the change in yields for the baseline and ensemble climates. Alfalfa, apples, sweetcorn, irrigated wheat, dryland wheat, and wine grapes crop yields were evaluated for each of the climate scenarios and over a range of water prorationing levels.
The lifecycles of fall Chinook salmon, spring Chinook salmon, and bull trout were compared to the predicted changes in the water temperatures and timing of hydrograph events to assess the impact of climate change on their survival and the ability of water management policies to adapt to these changes.
Preliminary Conclusions
A computational infrastructure providing accessibility, adaptability, and accountability is essential. Uncertainty in climate forecasts is inevitable. To overcome the inertia because of uncertainty against adopting adaptation programs, it is necessary to focus decisionmakers and stakeholders on decision uncertainty and away from process model and climate forecast uncertainty. Decisionmakers and stakeholders have an appreciation of the value of robust and reversible actions supported by a clear articulation of tradeoffs. Scale issues are difficult to communicate.
Climate assessments do not stand alone and should be integrated into existing planning processes to have any real impact in the near term. Climate effects must be shown relative to other processes; for instance, in some settings the impact of land use changes may exceed the impacts of climate change.
While little change (~10 percent) in precipitation is predicted between the historic and the future climate, a significant increase in temperature (1.9°C) is projected. This increase in temperature results in considerably less snowpack being maintained through the winter months. The change of the winter-dominated precipitation from snow to rain results in a significant shift in the timing of the runoff. The loss of the snowpack also reduces the ability to predict future water supplies. The total number of days requiring some prorationing increases significantly from the historical climate to the future climate. There also is an increase in the number of years requiring significant prorationing.
The Yakima River Basin's economy and cultural resources are dependent on an adequate water supply. Reservoirs provide a limited capacity to adapt to the inevitable variability in water supply because of climate variability. A systematic change in climate can further tax the existing water supply infrastructures’ adaptive capacity. The Yakima River Basin is extremely sensitive to climate change, particularly changes in winter temperatures. The loss of snowpack reduces the effective storage of the existing reservoir system while simultaneously making water supply forecasts increasingly dependent on the precipitation forecasts.
The fertilization provided by increased atmospheric CO2 levels increases the yield of most crops. The warmer temperatures associated with this increase in atmospheric CO2 levels, however, increases the demand for irrigation water. The impact of increased atmospheric CO2 levels is more than offset by the decreased yield predicted by the altered climate for dryland wheat.
Fall Chinook salmon likely will be impacted by a shorter period in their natal waters (typically in mainstem reaches). Fall Chinook enter their natal waters to spawn once the water temperatures have dropped in the fall and the flows have increased. The salmon leave their natal waters on the way to the ocean on the spring freshet. Warm stream temperatures are predicted later into the fall, delaying the salmon's upstream migration. The unregulated spring freshet is predicted to occur significantly earlier. These two changes leave the salmon far less time as eggs fry and smolt in their natal waters. Spring Chinook salmon overwinter in tributaries of the mainstem. Lower flows and warmer temperatures will reduce their survival.
Adfluvial bull trout stocks spawn and, in the early stage, rear in streams with most of the growth and maturation occurring in lakes or reservoirs. Adults enter mainstream rivers early in the summer, often holding near their natal tributaries for months before migrating upstream. Bull trout are strongly influenced by temperature. The temperature of these high-elevation, spring fed tributaries is not expected to change significantly, as snow is still predicted to form and melt to recharge the groundwater feeding these tributaries. Dramatic changes, however, in pool elevations of the reservoirs in which brook trout reside are predicted adversely to impact the brook trout. These changes may be required to meet other downstream water needs, such as irrigation targets and instream flow needs for other fish species.
Future Activities:
As this project comes to a conclusion, the framework and the results of the demonstration will be thoroughly documented and presented to the regional and local stakeholders and decisionmakers.
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other project views: | All 25 publications | 1 publications in selected types | All 1 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Scott MJ, Vail LW, Jaksch J, Stockle CO, Kemanian A. Water exchanges: Tools to beat El Nino climate variability in irrigated agriculture. Journal of American Water Resources Association 2004;40(1):15-31. |
R827454 (2001) R827454 (2003) |
not available |
Supplemental Keywords:
water, watersheds, groundwater, land, soil, sediments, global climate, precipitation, ecological effects, ecosystem, scaling, aquatic, habitat, integrated assessment, sustainable development, decisionmaking, community-based, cost benefit, nonmarket valuation, socioeconomic, environmental assets, engineering, social science, ecology, hydrology, modeling, general circulation models, climate models, Pacific Northwest, Washington, WA, EPA Region 10, agriculture., RFA, Scientific Discipline, Air, Geographic Area, Hydrology, Environmental Chemistry, climate change, State, Ecological Risk Assessment, EPA Region, integrated assessments, environmental monitoring, fish habitat, watershed, Yakima River Basin, economic models, socioeconomic indicators, Washington (WA), climate models, agriculture, environmental stressors, water quality, Region 10, aquatic ecology, climate variability, groundwater, air qualityRelevant Websites:
http://yakimaclimate.labworks.org/ Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.