Date of Award

12-2008

Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)

Department

Civil Engineering

Advisor

Shaleen Jain

Second Committee Member

Willem Brutsaert

Third Committee Member

Kirk A. Maasch

Abstract

Water supply reliability is vital for several competing societal objectives including hydropower production, agriculture, domestic uses, recreation, environmental health, and industry. Any changes in streamflow magnitude and timing could place extensive stress on the over-allocated water resources in western North America. The study region is comprised of the four river basins in this region: Columbia, Fraser, Sacramento-San Joaquin, and Upper Colorado. In these managed river systems, the interplay of the climatic variability and change and flow regulation imparts significant complexity to the task of managing water resources in a sustainable manner. Understanding potential streamflow alterations therefore requires an interdisciplinary framework: hydrology, climatology, and ecology - and their inter-relations - must be analyzed to achieve a comprehensive view of the system. Here, we use observational data and climate model simulations to understand the nature and causes of the 20th century hydrologic change in this region, projections for the 21st century climate, and analysis of the metrics of hydrologic variability and ecosystem services. Hydrology in western North America is significantly affected by both regulation (dam and reservoir management), and nonstationary climate. In the Columbia River Basin, daily flows for both the regulated and unregulated regime (back-calculated by models) from the 1929-2007 period are analyzed in order to: a) compare the two regimes, and thereby assess how does regulation alter the streamflow in this region and b) examine the unregulated regime over time, to study whether the background state of natural flows has changed. It is concluded that significant increases in variability of the unregulated regime has occurred, and that regulation has altered streamflow in potentially damaging ways to the environment. This change in background state may also imply that current reservoir management needs to adapt to a changing climate. Climate is studied using observations and two different types of climate models: a) Atmospheric General Circulation Models (AGCM), and b) Coupled Atmosphere- Ocean General Circulation Models. The first assessment includes a Principal Component Analysis on precipitation data in the 1952-2002 period for all basins. Spatial patterns of precipitation distribution in this region can be described by synchroneity (the tendency for the entire region to become wet and dry together) and the north-south dipole discovered in other studies (the northern basins tend to become wet, while the southern ones are dry and vice versa). Furthermore, the El Nino state influences the north-south dipole, so that during El Nino (La Nina) years the northern basins typically are dry (wet) while the southern basins become wet (dry). The nine AGCM simulations of the 20th century climate allow a careful assessment of the role sea surface temperatures (and changes therein) in engendering change in western North American hydroclimate. It is found that the changing statistics of El Nino/Southern Oscillation have resulted in an increased year-year variability in regional hydrology. The El Nino-Southern Oscillation is examined herein with regards to its teleconnections to western North America (temperature and precipitation), and possible changes over time. The nature of El Nino-Southern Oscillation teleconnections appears to vary in time, in observations, and model simulations, suggesting that careful analysis of model simulations within the context regional hydroclimate can provide some useful insights. Careful accounting of the El Nino teleconnections is used to create a new multimodel weighting scheme for assessing climate simulations from coupled Atmosphere-Ocean General Circulation Models. This nonparametric method allows a robust characterization of the uncertainty in projected climate change on regional scales, where the fidelity with which individual models reproduces the joint densities of two climatic variables (such as an El Nino index and temperature). Overall this research finds that a careful characterization of the hydroclimatic variability and change in the western North American region is a critical first step toward: (a) understanding the nature and causes of the hydroclimatic change in the second half of the 20th century, (b) developing projections specific to each of the four river basins for the 21st century climate, and (c) providing useful climatic information to support adaptation within the water resources management sector.

Download

Files over 10MB may be slow to open. For best results, right-click and select "save as..."

Share