U.S. mega-droughts and associated sea surface temperatures in coupled ocean-atmosphere runs

Consensus climate projections of some 20 Coupled Model Intercomparison Project (CMIP3) global simulations indicate that substantial drying of Southwestern U.S. climate is likely if greenhouse gas (GHG) emissions continue to increase in the 21st century. Whether more severe droughts also are to be expected in a greenhouse-warmed world is not obvious, however, since effects of persistent “mega-droughts” are found in proxy records of preindustrial-era climate. An assessment of the future hydroclimatic risk thus should begin with a study of Southwestern U.S. mega-droughts in CMIP3 “control runs” that are radiatively forced with preindustrial GHG concentrations. Because land-ocean interactions play an important role in Southwestern U.S. hydroclimate, identification of coherent inter-model patterns of sea surface temperatures (SSTs) associated with mega-drought occurrences is also a goal. In this study, several precipitation-based metrics are used to identify occurrences of megadroughts in each CMIP3 control run. In addition, composites of the corresponding annual SST anomalies are constructed, after removing artifacts of coupled-climate drift from some simulations. A qualitative agreement in tropical SST patterns is found among the models, with anomalously cold East Pacific temperatures being especially conducive to triggering mega-droughts in the Southwest U.S. Moreover, both mega-drought occurrences and associated SST anomaly patterns are relatively insensitive to the choice of drought metric. The drought/SST relationships are now analyzed in simulations of the 20th and 21st century climate change.