Our work focuses on decadal-multidecadal drought variability – a scale neither well constrained by observations nor well represented in models. We focus on key vulnerable regions, where natural variability and anthropogenic change combine to amplify the risk of prolonged, severe drought with large consequences: southwestern North America (US and Mexico), Australia, Amazonia, South Asia, the Caribbean, and West Africa/Sahel. Our strategy takes advantage of several unique observational, model and stakeholder resources: (1) an unprecedented ensemble of simulations of the past millennium from a state-of-the-art Earth System Model (NCAR CESM1.0; in addition to the CMIP5 archive); (2) an expanding set of published and emerging paleoclimate datasets from multiple proxies that reveal long observational histories of decadal-multidecadal hydroclimate variability; (3) a longstanding network of stakeholders and collaborators in the southwestern US and Mexico with whom we can develop best practices in applying drought risk estimates to real-world problems across a broad social context; and (4) a long history of working on drought variability and stakeholder-driven science in other drought-prone regions worldwide. Our research strategy addresses the following questions: What is the observed nature of decadal-multidecadal drought in vulnerable regions of the globe? How well do the NCAR ESM and other CMIP5 models simulate observed decadal-multidecadal drought? How do we combine observational and model-based perspectives to quantify estimates of prolonged drought risk in key regions over the next few decades? How can knowledge about drought risk be optimized for use by diverse stakeholder communities in developed and developing countries?
Drought is among the most ruinous of natural disasters and is expected to become increasingly prevalent in a warming world. Superimposed on anthropogenically-forced changes will be natural decadal to multidecadal hydroclimatic variability that is poorly quantified for most of the globe, larger than the forced changes for the near term, and incompletely represented in coupled models. Among the greatest challenges of decadal prediction are the quantification of prolonged drought risk in vulnerable regions and the integration of knowledge about this risk into the decision-making processes of the many resource managers and other stakeholders who deal with drought. We are overcoming these two challenges through interdisciplinary collaboration between leading climate institutions (NCAR and the University of Arizona) with exceptional strengths in the integration of climate modeling, knowledge about climate variability, and stakeholder partnerships.