During the iconic Dust Bowl, great clouds of dust swirled east out of the U.S. Northern Plains, blackening skies as far away as New York. Drought conditions set up in 1932 and didn't ease for seven years. Just a decade later, another severe drought hit, this one scorching the Southern Plains and Southwest.

ESRL scientists seeking to improve today's drought warning systems recently turned their investigative skills on better understanding the origins of those two events. In an analysis of historical records and climate model experiments, they discovered two very different setups. Their conclusions may improve scientists' ability to give accurate early warning of drought, to help communities take precautions and prepare.

"In the case of the severe 1950s drought of the Southern Plains states, it appears global sea surface temperatures were the principal cause. However, the 1930s Dust Bowl drought... was not caused by ocean conditions, but rather the evidence points to random changes in the atmosphere as the instigator," said lead author Martin Hoerling.

Hoerling, Xiao-Wei Quan, and Jon Eischeid (all of PSL) published their work, "Distinct Causes for Two Principal U.S. Droughts of the 20th Century," in Geophysical Research Letters this fall.

Drought variability over the Southern Plains was strongly forced by variability in sea surface temperatures, the researchers found, and the primary driver was the El Niño-Southern Oscillation, a pattern of ocean temperatures in the tropical Pacific Ocean that includes El Niño and La Niña. During La Niña years, characterized by unusually cold water in the equatorial Pacific Ocean, droughts in the Southern Plains were common—both in observations and in models.

But northern lands hit hardest during the Dust Bowl were farther away from weather patterns influenced by the tropical Pacific Ocean. Instead, random atmospheric variation was likely responsible.

The authors note that while an ocean observing system—a network of instruments including stationary and free-floating buoys as well as satellites—is vital to any drought early warning system, it may not adequately warn of a drought caused by other factors, and may not provide early warning for a drought over the northern Plains states, such as what occurred in the 1930s.

By Katy Human, Winter 2009