It was not a good year for people, weather and climate. The winter was strangely warm in many places and the summer ridiculously hot. As a large fraction of the country suffered through extreme or even extraordinary drought many folks naturally wondered, "Is this climate change?" Then along came a presidential election in which the words "climate change" disappeared from the dialogue. Now, just a week or so before voting day, the convergence of westbound Hurricane Sandy with a eastbound cold front is creating a massive storm, a Frankenstorm even, that is threatening millions of Americans. Weird weather is making yet another appearance in our lives and once again we ask, "Is this climate change?"
The hyper-charged political landscape we are crossing now creates its own sparks when trying to answer that question. In a world looking for "wake-up calls" and "smoking guns," how do scientists address the thorny issue of attribution? Did anthropogenic climate change cause the storm that rained out your picnic yesterday? Is it causing the terrifying storm crawling up the East Coast now? There are deep, powerful and potent issues here that touch on both science and the relationship between science and politics.
Let's start with the science.
For years, most climate scientists would say it's impossible to link an individual weather event with climate change. That, in fact, is the difference between weather and climate. Climate is all about long-term trends — not the 5-day forecast.
Recently, however, some researchers have taken the issue of attribution seriously. Using a variety of techniques, they are attempting to quantify the role human-driven climate change plays in particular events. This is science at the bleeding edge, where framing their questions correctly so that they might lead to meaningful answers is still a hot issue.
Researchers like Randall Dole of NOAA, for example, might ask what percentage of an extreme event's magnitude came from a changing climate. Peter Stott of the UK Met Office frames the question differently. He looks at the odds for a given extreme weather event to occur given human-driven climate change. Kevin Trenberth of NCAR takes a third view, asking: Given a changed background climate, how should we expect weather to change?
All of these different perspectives (sometimes framed as "Weather on Steroids") have led to new quantitative explorations of climate change's role in what is happening now, not 30 years in the future. In an early example of attribution science Peter Stott and colleagues took on the extraordinary heat waves that struck Europe in 2003 (killing thousands). Their conclusion?
"...we estimate it is very likely (confidence level >90%) that human influence has at least doubled the risk of a heat wave exceeding this threshold magnitude."
This kind of science has allowed researchers to get a much better handle on attributing climate change as a game changer for events like this summer's killer heat and drought.
So how about the Frankenstorm?
Here the waters get muddied. There is a hierarchy of weather events which scientists feel they understand well enough for establishing climate change links. Global temperature rises and extreme heat rank high on that list, but Hurricanes rank low. As the IPCC special report on extreme events put it "There is low confidence in any observed long-term (i.e., 40 years or more) increases in tropical cyclone activity (i.e., intensity, frequency, duration), after accounting for past changes in observing capabilities."
The reasons for "low confidence" are manifold. Some part of the caution comes from the complexity of the problem, and some part comes from the lack of good data before the satellite era (about 1970). Thus, many climate scientists will not want to go out on a limb for hurricanes. They just don't have the tools to make strong inferences.
This is not to say progress isn't being made. One thing that does seem clear is that warmer oceans (a la global warming) mean more evaporation, and that likely leads to storms with more and more dangerous rainfall of the kind we saw with Hurricane Irene last year. In addition, a paper published just last month, used records of storm surges going back to 1923 as a measure of hurricane activity. A strong correlation between warm years and strong hurricanes was seen. Thus if you warm the planet, you can expect more dangerous storms.
Which brings us to our bottom line. The science of climate attribution is very exciting and full of cool, new ideas. It has already provided us with first steps towards more precision in understanding how climate change is changing climate now, already. For hurricanes, however, sticking to the science means it is still hard to point to an individual storm and say, yes! Climate change! A more reasoned approach is to take the full weight of our understanding about the Earth and its systems and go beyond asking if any particular event is due to global warming or natural variability. As Kevin Ternbeth of NCAR says "Nowadays, there's always an element of both."
Finally there is the issue of science and politics. Tania Lombrozo wrote a beautiful piece here on Thursday asking if "Scientists Should Promote Results Over Process." Speaking directly to climate change, she concludes
"Overstating confidence in scientific claims may ... miss a long-term benefit for a short-term advantage: rhetorical oomph comes at the cost of an opportunity to educate people about how science works and why the products of science are our most reliable guides to the natural world."
I could not have said it better myself.