Absolute Certainty ... Elusive

That Storm, That Cold Day, That Drought … How Scientists Try to Evaluate Links to Warming

These days it seems as if a single hot or cold day is all it takes to inspire a reporter or politician to blame the mercury’s position on global warming, or, alternatively, claim it as proof that global warming doesn’t exist. More extreme events such as hurricanes or floods inspire even more headlines and comments … and political punditry.


View larger image
NCAR’s Kevin Trenberth

“The question is usually: ‘Is this caused by global warming,’” says Kevin Trenberth, head of the Climate Analysis Section at the National Center for Atmospheric Research (NCAR), and a lead author for the Intergovernmental Panel on Climate Change.

“But it’s just not a helpful question. It doesn’t have a yes or no answer, but everyone wants a yes or no answer.”

So, how much, if anything, can accurately be said about correlating long-term climate change and the various weather patterns we routinely experience, especially the extreme ones? Quite a lot, as it turns out, but in most cases, what can be said accurately is very different from what is most often said, at least in the popular media.

That’s a problem, because both underestimating and overestimating the influence of climate change clearly complicates public debate on the topic (which for some may be a goal in the first place).

Could it Have Happened Without Human Influence?

It’s key to recognize that absolutely nothing can be said with absolute certainty. That uncomfortable reality complicates life for those wanting to send a clear message. But weather has so much natural fluctuation that virtually any weather event we see could have happened without any human influence on the climate.

“It’s very difficult to establish a causal link between a weather event and climate change,” says Myles Allen, head of the Climate Dynamics groups at Oxford in the UK. “That doesn’t mean there are no causal links, but it is undoubtedly hard to establish what they are and to quantify them.”

To overcome that difficulty, climate scientists are focusing increasing effort and multiple approaches on studying those potential causal links. By understanding those efforts, journalists and educators can be better equipped to help the public discern what we do and do not know, without sensationalism in either direction.

With Us or Without Us

One method for teasing out the influence of global warming on a particular event, such as the brutal European heat wave of 2003, is to take a virtual look at the climate over the past century or so with and without human influences incorporated. While dependant on complex models and mind boggling computing power, the basic principle is simple. It’s something like the climatological equivalent of “It’s a Wonderful Life,” but instead of looking at what the world would be like with only George Bailey’s influence removed, scientists aim for a view of what it would be like if none of us had been here.

Though simpler methods were used in the past, researchers can now run advanced climate models hundreds of times with human influences such as carbon dioxide emissions included, and then run the models separately, with those influences removed. Then they look at the virtual climate that evolves during each model run and count the number of times conditions similar to an event of interest occurred. This approach can give a relative measure of the degree to which human influence increased the likelihood or risk that the extreme event would happen.

If, for instance, it showed up in 80 percent of the runs with human influences and 20 percent without, scientists could conclude that human influence made the event four times more likely. The same basic strategy can be used to look at events such as floods and droughts, and also to look at how global warming might decrease the likelihood of certain patterns such as abnormally cold temperatures across a season.

The more complicated the event, the more difficult to identify climate change connections. A heat wave is, of course, a straightforward predicted outcome of global warming. With hurricanes, however, there are aspects of global warming that could lead to increased intensity, and others that might reduce intensity. Ergo the subject of global warming’s impact on hurricanes remains a topic of great debate.

“There is an answer to the question of how much – if at all – humans increased or decreased the risk of Katrina,” says Allen, “We just don’t know it.”

Loading the Dice

Climate models are not sufficiently sophisticated to allow scientists to answer questions about the increased or decreased likelihood of a few hot or cold days, says Allen. Even if they were, any reliable measure of change in risk would almost certainly be vanishingly small because of weather’s perpetual variability. That reality is not likely to prevent TV newscasters, editorialists, politicians, political pundits and others from drawing connections anyway, at times to support their preconceived biases.

Allen likens anthropogenic influences on the climate to loading the dice. The loaded dice won’t always come up sixes, but they will much more often than unloaded dice. Similarly, the argument from modeling is that human influences increase the likelihood of extreme events that cause everything from discomfort to death. In the case of the European heat wave, Allen’s group showed that the likelihood it would occur in 2003 was at least doubled as a result of human factors. His group now is completing a study of how human influences may have increased the likelihood of recent European flooding.

While dealing with risk percentages, rather than absolutes, can be confusing at times, most people are quite used to this type of thinking. In fact, we routinely deal with relative risk decisionmaking as a part of normal life.

Many people, for instance, may choose to eat less to reduce their chances of heart attack, even if they recognize that a heart attack can still hit the healthiest of eaters and the most committed exercisers. The same goes for smoking and the risk of lung cancer, even though it’s quite possible to get lung cancer having never smoked a cigarette in your life.

According to Allen, suggesting that demonstrably increased risk from human factors had nothing to do with an extreme event because it could have happened anyway is akin to suggesting a lung cancer victim’s heavy smoking is just pure coincidence.

Considering Underlying Factors

Trenberth, at NCAR, says that because the public at times has difficulty digesting statistical discussions, he believes it important to highlight some of the well-understood, underlying factors related to global warming that are pushing weather toward extremes.

Chief among these is that ocean temperatures have already warmed up about 1 degree Fahrenheit over the past century or so, a problem strongly tied to global warming. “That’s a pervasive influence on every single weather system we have,” he says, because warm ocean water can affect everything from hurricane formation to the amount of rain that falls in Mississippi.

Trenberth and other researchers have a different strategy for addressing questions of climate change impacts. “In NOAA and in my group, we may put less emphasis on sorting out the human component and be more concerned about understanding what happened and why.”

Rather than look at risk percentages based on analyses of total human impacts, they instead take a closer look at the factors controlling extreme events such as major droughts. To examine potential connections between rising ocean temperatures and weather events, researchers are running increasingly complex climate models with and without recent ocean warming trends included.

Both Trenberth and Allen are part of a new organization called the International Group on Attribution of Climate-related Events that aims to increase overall understanding of connections between climate change and weather patterns. The group advocates establishment of an effective attribution service that will accelerate the process of analyzing human-climate connections, which can currently take years for a given event. The hope would be to make information available quickly to better inform the public and improve decision makers’ abilities to manage the risks of extreme weather.

“My view personally is that it is no longer enough for a weather service to just predict weather,” says Allen, “We have to explain it as well.”

Mark Schrope is a freelance science writer living in Melbourne, FL.

Bookmark the permalink.

Comments are closed.