The spatiotemporal evolution of severe droughts in coupled global climate models: drivers and impacts in the past and future

Abstract

Persistent and severe droughts have occurred regularly over the late Holocene, however, few of these events have been directly observed and thus their impacts and drivers are not well understood. To overcome the challenges associated with a short observational record, coupled global climate models have been used to establish that internal modes of atmosphere-ocean variability, intrinsic atmospheric variability, and land-atmosphere feedbacks all play a role in the life cycle of droughts in North America, in addition to external forcing. We build off this prior work, quantifying the full range of large-scale drivers and local-scale impacts of meteorological droughts in climate models, establishing the relative importance of different types of uncertainty for these ranges, and then determining how and why the drivers and impacts of meteorological drought will change in the future under anthropogenic climate change. This dissertation develops a framework through which the spatiotemporal characteristics of simulated meteorological droughts (i.e. precipitation deficits), as well as their drivers and impacts, can be compared to an observed event. The drought in central and western North America between 2011 and 2014 is used for comparison throughout due to its location, severity, and a high availability of observational data. In Chapter 2, we establish the framework, identifying meteorological droughts based on their precipitation deficits and spatiotemporal characteristics. Chapter 3 utilizes this framework to determine the consistency of the local-scale impacts and large-scale drivers of meteorological droughts in 65 coupled global climate model simulations. We found that a range of atmosphere-ocean conditions can lead to droughts with similar spatiotemporal characteristics, and that the precipitation deficits during these droughts can have a range of hydrological and ecological impacts. In Chapter 4, we utilized initial condition large ensembles to quantify the relative roles of intrinsic climate variability and structural uncertainty in the range of atmosphere-ocean conditions from Chapter 3. We find that while both intrinsic variability and structural uncertainty play a role, structural uncertainty is substantially larger. In Chapter 5, we compare the spatiotemporal characteristics, local-scale impacts, and large-scale drivers of historical and future meteorological droughts. We found that increased precipitation will lead to fewer meteorological droughts in the future, however their impacts will become more severe for certain aspects of hydrology and ecology. Additionally, the large-scale atmosphere ocean drivers of meteorological droughts are becoming more consistent in the future, with implications for their predictability.