Looks Like Rain Again. And Again

http://www.nytimes.com/2014/05/13/science/looks-like-rain-again-and-again.html?_r=0

May 12, 2014
Justin Gillis

The acid test of a scientific theory is whether it makes predictions that eventually come true. So consider this old prediction, from a pair of researchers in Australia and New Zealand. They were summarizing the results of then-primitive computerized forecasts about global warming:

“The available evidence suggests that a warmer world is likely to experience an increase in the frequency of heavy precipitation events, associated with a more intense hydrological cycle and the increased water-holding capacity of a warmer atmosphere.”

That was published in 1995, and it was based on research going back to the 1980s. Fast forward to 2014.

In the National Climate Assessment, published last week, researchers in the United States reported that “large increases in heavy precipitation have occurred in the Northeast, Midwest and Great Plains, where heavy downpours have frequently led to runoff that exceeded the capacity of storm drains and levees, and caused flooding events and accelerated erosion.”

-----

But scientists can justly claim to have demonstrated some predictive skill about many of the potential implications of the human release of greenhouse gases.

Their track record actually goes back to 1896, when a Swede named Svante Arrhenius first predicted that emissions of carbon dioxide would cause the planet to warm. It took more than 80 years to be sure he was right. At roughly the same time that realization was taking hold, climate scientists running computer models of the atmosphere began to focus on the likelihood of heavier rains in a future climate.

-----

In the National Climate Assessment, the experts reported huge increases since the mid-20th century in the amount of precipitation falling in very heavy rainstorms: up 71 percent in the Northeast, 37 percent in the Midwest and 27 percent in the Southeast. The effect was seen on a smaller scale west of the Mississippi River, too, even in parts of the country where the climate is drying out over all.

What led the researchers to expect this long before it actually happened?

Initially, the forecast was based on simple physics from the 19th century. As we pour carbon dioxide into the air, the lower atmosphere has to warm. As it does, it is able to hold more moisture, and as the surface of the ocean also warms, more moisture tends to evaporate from it.

In the United States, the increase in water vapor has been on the order of 3 percent or 4 percent since the 1970s (most of the human-caused global warming has occurred since then). That may not sound like a big jump, but the effect is enormous.

Two leading scientists, Kevin E. Trenberth at the National Center for Atmospheric Research and David R. Easterling at the National Oceanic and Atmospheric Administration, ran some calculations and agreed that the warming has, on average, put more than a trillion gallons of extra water into the air over the contiguous 48 states, probably closer to two trillion.

That extra water has to fall as rain or snow. But from the elementary physics, it was long unclear whether this would mean more rainy days over all, or more intense rains, or both.

It was the computer models of the climate that suggested, starting in the late 1980s, that the answer would be the latter, and so it has turned out. One way to think of it is that even with a lot of moisture in the air, conditions are not always right for rain, but when they are right, the skies have a lot more water to dump.

-----

Researchers sponsored by the Australian government were the first to really drill into the implications of the finding. In their 1995 overview paper, A. M. Fowler of the University of Auckland in New Zealand and K. J. Hennessy of Australia’s national research program warned that society needed to start thinking about the risks. They suggested toughening standards for the designs of levees and dams, and hardening roads and culverts against the possibility of more flash floods.

-----