Extreme weather forensics

Not every climate sleuth uses the same methods

30 January 2012  •  Because the connection between greenhouse gases and local weather is so complex, scientists are using a number of techniques to determine whether a connection can be made or not. Many of these techniques can be grouped into three general categories, outlined below. For more on how the research is carried out, see the introduction to this in-depth report, Doping the atmosphere?, and the article Is this climate change?.

Leading practitioners of these approaches, including...

Randall Dole (NOAA) Kevin Trenberth (NCAR) Peter Stott (UK Met Office Hadley Centre)

...have framed the problem like this

Increased greenhouse gas concentrations contributed X% to the magnitude of this particular event.
 
The influence of human-produced greenhouse gases is now part of our climate’s background state. It plays some role in every weather or climate event.
 
Increased greenhouse gas concentrations changed by X% the odds of reaching an event of this magnitude.
 
Research examples July temperatures across western Russia over the last century show large natural variability, but we cannot find evidence of climate trends during that period that would cause a heat wave on par with the disaster that struck Moscow in 2010. [research summary]
 
On average, the most intense rain events—like 2011’s Hurricane Irene across the mid-Atlantic—can be expected to produce 5–10% more rain than before, because there is up to 7% more water vapor in the air for every 1°C rise in temperature (4% for every 1°F). [research summary]
 
It’s likely that the accumulation of human-produced greenhouse gases to date has at least doubled the odds of a heat wave in Europe as intense as the one in 2003, when tens of thousands of people perished. [research summary]

Behind the research  This approach comes at the problem from a forecaster’s point of view: what can the past tell us about today’s weather? In this view, if natural variability can largely explain a recent extreme event, and if recent climate patterns are not trending toward such an event, then climate change may not have played a significant role in creating that extreme (though in some cases, the chance of an influence will grow in the coming decades).
 
Rather than analyzing a particular event with a global climate model, this approach combines long-term global and local observations with long-established rules of physics that explain how temperature and water vapor respond to greenhouse gases.
 
Global climate models are employed to test the presence of a human influence on a particular process and then to extrapolate that result to extreme events, thus estimating their frequency. Although the results can be complex to interpret and communicate, this technique is becoming the most widely used way to quantify the odds that climate change is altering high-profile disasters.
 
Some teams using this approach  NOAA’s Interpreting Climate Conditions research team has analyzed several high-profile events using this approach, including the 2010 Atlantic hurricane season and 2011’s rash of U.S. tornadoes.
 
Kevin Trenberth (NCAR) and colleagues, who analyze the global budgets of moisture and heat, use this approach to put heavy rains and snows into a long-term global perspective. Trenberth also proposes that, rather than proving that climate change has influenced an event, scientists should now attempt to prove that it hasn’t. [research summary]
 
Peter Stott (UK Met Office) and colleagues continue to refine this approach, which they pioneered with their study of the 2003 European heat wave. Other scientists in Britain and elsewhere have followed suit.
 
What they’re saying “If we’re going to provide science-based input to help improve early warning about weather disasters, and provide information useful for adapting to them, then quantitative assessments of both natural and human contributions to extreme events are vital.”
Randall Dole, NOAA
 
“It’s not the right question to ask if this storm or that storm is due to global warming or [due to] natural variability. Nowadays, there’s always an element of both.”

Kevin Trenberth, NCAR
 
“It can still be problematic to blame a specific individual extreme weather event on climate change, because there have always been extremes of weather around the world. However, if the likelihood of a particular extreme weather event has changed, it is possible to say something.”

Peter Stott, UK Met Office Hadley Centre