SOURCE: Columbia University

DATE: June 12, 2019

SNIP: Over the last few years, scientists have started to recognize the potential for climate change to cause multiple societal impacts close together in space or time. In light of this emerging threat, Columbia University’s Initiative on Extreme Weather and Climate hosted a three-day workshop at the end of May to focus on planning for when climate extremes get complicated.

The Correlated Extremes workshop discussed the odds of extreme events happening together, how those odds are changing as the planet gets hotter, and the potential consequences of these linked events. Radley Horton, conference organizer and climate scientist at Columbia’s Lamont-Doherty Earth Observatory, said the gathering drew academics as well as policy experts and representatives from government and business.

The workshop examined three types of “correlated extremes” scenarios. The first is when multiple variables interact during a single event — for example, when humidity makes a heatwave more deadly, or when hurricane flooding sparks electrical fires. “If you put all these factors together, the stats look different, and the societal impacts look different,” Horton explained.

The second type of correlated extreme is when events occur in sequence in a given place — like when a hurricane strikes and then knocks out power, or blocks roads and thus hampers rescue and recovery efforts in advance of a heat wave.

Finally, the third type of scenario is when multiple places experience extreme events at the same time. For example, if several of the world’s breadbaskets are struck with drought or destructive rains at the same time, it could wreak havoc on the global food system.

“Just pushing up greenhouse gas concentrations a little more could profoundly increase the probability of some types of correlated extremes,” Horton warned. To him, correlated extremes show that “the climate system and our human and ecological systems may be more vulnerable than we realize, and the societal protection against correlated extremes may not be as large as we realize either.”