A Multi-Scale Study of U.S. Drought Risk and Predictability

Abstract

This thesis proposes a multi-scale study of U.S. drought risk and predictability in order to better understand 1) how land, atmosphere, and oceans interact during the onset of drought, 2) how the processes involved in generating mechanisms of drought can be better identified by a multi-scale study, rather than a single dimensional study, and 3) how reliable known drought mechanisms are in explaining and predicting recent and future drought events.

In this thesis, there are seven chapters. Chapter 1 summarizes the objectives and goals of the thesis research. Chapter 2 investigates changes in the low flow regime over the eastern U.S. region due to climate change and variability. It also assesses the association between low flows and large-scale atmospheric circulations. In Chapter 3, the multi-scale driving mechanisms of droughts and floods over the southeastern U.S. are studied using a recently published regional reanalysis dataset. It establishes favorable conditions of the southeastern U.S. droughts and floods on land, and in the atmosphere and oceans. In Chapter 4, the influence of tropical cyclones (TCs) on the eastern U.S. droughts is assessed as a protective factor by comparing two land surface model simulations that are forced with or without TC-related precipitation. Chapter 5 investigates recent changes in drought risk influenced by the Pacific and Atlantic Oceans. Chapter 6 examines the current predictability of U.S. drought using a multi-model ensemble forecasting system. This study tests whether current climate forecasting models can reproduce the observed associations between global sea surface temperature and U.S. precipitation. Chapter 7 provides a brief summary of this study and discusses the next steps and future work of this study.

By crossing the spatial-temporal scales in studying drought, this thesis finds that there are previously hidden processes from a fixed scale (one dimensional) approach. These hidden processes are found via a multi-scale study by investigating interactions of different scale processes on land, and in the atmosphere and oceans. This thesis also finds that some of the recent drought events are inconsistent with the known drought mechanisms found from previous studies (e.g. strong associations of global oceans with U.S. drought occurrences). It is the gap between what we know and what we have that causes the limited predictability of recent drought events. To fill the gap, this thesis suggests that the design of a muli-scale study for drought-generating mechanisms is necessary. An update of our understanding of the mechanisms taking account of their historical changes will lead to a more robust improvement in our predictability and provides a more actionable information for water resources management and policy.