While we’re on the subject of droughts, Ed Cook and his colleagues and Lamont-Doherty have a new review paper (preprint, 10 meg pdf) of North American drought that makes some relevant points about the differences between what happens here and the current African drought/famine I mentioned below.

Point one is the distinction between the effects of the Dust Bowl of the ’30s and the Southwest drought of the 1950s:

The 11-year 1950’s Southwest drought was likewise extreme, with it being centered primarily over Texas and New Mexico. The environmental impact of the Southwest drought was severe, but it had less socioeconomic impact than the Dust Bowl drought because of irrigation, improved agricultural practices, better governmental support, and a much stronger underlying economy.

Point two relates to my comment about a society’s resilience on all relevant time scales. In this case, “relevant” could mean on much longer time scales than we’re currently prepared for:

For example, the 16th century megadrought identified by Stahle et al. (2000) is now seen to have affected large areas of North America, especially in the West and northern Mexico, and was much more prolonged than any of the 20th century droughts. The drought reconstructions also provide clear evidence for a much drier climate across the West and Great Plains during Medieval times, a drought that lasted with few interruptions for a few hundred years and which greatly taxed both hunter-gatherer and agriculturalist populations (Jones et al., 1999). Such “no analog” megadroughts are scary because the modern-day agricultural and hydrologic systems that depend upon adequate water supplies to produce and function may not have the resilience to survive much beyond the observed “worst case scenario” droughts of the past 100-150 years, e.g., the Dust Bowl drought.