A Comprehensive Framework for Projecting Regional Climate Change Impacts and Analyzing Non-Stationary Statistics of Hydrologic Extremes over the Midwest and Great Lakes Region

Understanding of future climate variability and climate change impacts on regional hydrologic systems is imperative to support sustainable and resilient adaptation strategies in water resources planning and management and many other disciplines. The overarching objective of this dissertation is to develop a comprehensive framework for analyzing regional-scale climate change impacts and quantifying non-stationary risk of hydrologic extremes over the Midwest and Great Lakes region under a changing environment. Chapter 1 provides an introduction, background information, research foundation, and specific goals of this dissertation. Chapter 2 produces high-spatio-temporal resolution hydrometeorological variables over the Midwest and Great Lakes region by downscaling Global Climate Model simulations based on a statistical downscaling method called Hybrid Delta. The results show that future impacts for seasonal mean and extreme climate are substantial over this region. The study in Chapter 3 analyzes hydrologic impacts based on physically-based hydrologic simulations coupled with historical and downscaled future climate projections for the 2020s, 2050s, and 2080s, especially focusing on the changes in intensity, timing, and frequency of hydrologic extremes in 20 large watersheds throughout the region. To better address statistical problems associated with the changing behavior of hydrologic extremes over time, Chapter 4 introduces a new risk-based framework for analyzing non-stationary statistics of hydrologic extremes. A case study demonstrates that this new framework can be used to quantify risk of failure of existing infrastructure or develop robust design standards for new hydrologic infrastructure by explicitly accounting for design lifespan and gradual changes in climate. Chapter 5 focuses on an intercomparison between statistical and dynamical downscaling approaches, especially focusing on extreme high temperatures and summer extreme precipitation events associated with small-scale convective storms. The results showed that the raw simulations of 25-yr summer extremes from RCMs are fairly biased in some cases; however, they may still provide more useful information on change in small-scale dynamics than statistical downscaling because the RCMs are more physically based. For example, the RCMs can successfully account for complex decoupling effects between summer averages and summer extremes (longer droughts, but more intense P from individual storms) over the Midwest and Great Lakes region. Lastly, Chapter 6 concludes the dissertation by giving a summary of key findings and arguments as well as suggestions for future research.