The Climate CIRCulator is a monthly newsletter covering climate science and the Northwest written by scientists and communicators.

Linking Extreme Weather to Climate Change

  • Jun 30, 2016

One of the most frequent questions climate scientists are asked after an extreme weather event such as a drought or hurricane is: “Was it caused by climate change?” For a long time, scientists would usually say that no single event could be attributed to climate change with any degree of confidence. That is no longer the case.

In climate science, linking a given climate phenomenon or weather event to a variety of external factors (greenhouse gases, solar variability, El Niño) is called attribution. A still up-and-coming field, attribution of extreme weather events has nonetheless developed by leaps and bounds in recent years. For proof one need look no further than a recent report focusing on attribution of extreme weather events from the National Academies of Sciences.

tropical storm isaac

A GOES-13 infrared satellite image of Tropical Storm Isaac provided by the U.S. Naval Research Laboratory in Monterey, Calif.

The report is the result of an analysis of the field by an expert panel of 10 members, including OCCRI Director Dr. Philip Mote. In their report, the panel gives readers a kind of guided tour of the history and development of attributing extreme events, and what can and cannot be said when attempting to attribute an extreme event to anthropogenic climate change.

The history goes something like this: The first attempt to attribute an extreme event to climate change came in 2004. The event under scrutiny was the 2003 European heat wave, a devastating episode that’s estimated to have killed tens of thousands of people, mostly elderly individuals caught by surprise in housing without air-conditioning. Since the publication of that initial analysis, the field of attribution has taken off, with arguably the largest launch occurring between 2012 and 2015 when the number of submitted studies concerning attribution grew by a factor of five, according to the report’s authors.

Not surprisingly, as the number of studies tackling the problem of attribution has grown, so too has the field’s sophistication. Which brings us to the report’s next big highlight: Why can some events be attributed to climate change with more confidence than others? To picture this, let’s use an oft-used analogy: baseball.

In recent decades, professional baseball, that great American pastime, has been plagued by a number of doping scandals involving anabolic steroids and other performance-enhancing drugs. In tandem with the drug use has been an extreme increase in the number of home runs, a rise far surpassing what could be expected from past performance. Understandably, many suspect a connection between the performance-enhancing drugs and the athletes’ enhanced performance. The problem is that while the overall number of home runs can be correlated to the rise of drug use, attributing individual home runs to drug use is impossible. In fact, you can’t say for certain if a particular home run was due to steroids. Instead what you get is a number representing the probability that a given home run occurred in the presence versus the absence of steroids. This, in a nutshell, is a problem of attribution.

Now take the same idea and consider climate change and extreme event attribution. Here the steroids are carbon dioxide (CO2) and other greenhouse gases, while home runs represent extreme events. To paraphrase Jerry Meehl of the National Center for Atmospheric Research, we now have extreme events occurring in a climate that is on steroids.

By trapping heat from the sun, CO2 adds energy to the climate system, energy that — given what we now know about the mechanisms of climate, which is quite a lot — should increase the number of extreme events, such as heat waves and droughts. As with home runs, attributing individual extreme events to a single cause is difficult, even if the overall trend is clear.

As with steroids and home runs, what you get when you try to ascertain anthropogenic climate change’s fingerprints on a given extreme event is a number representing the probability that a given event is more or less likely to have occurred given the presence or absence of anthropogenic climate change. What’s more, note the report’s authors, these probabilities occur on a spectrum that ranges from events that can be attributed to climate change with a high degree of confidence to events whose connection to climate change can be less confidently attributed.

How extreme events fall on the spectrum is pretty straightforward: events with a direct link to temperature increases can be attributed with a higher degree of confidence because there exists a pretty straightforward and well-understood link between increased CO2 levels and rising temperatures. So it’s not surprising that extreme heat events, such as heat waves, can be linked to climate change with a high degree of confidence. But as that initial connection to rising temperatures weakens as other compounding factors are added, so too does the confidence level.

What results is a spectrum that stretches from extreme heat and cold events (whose frequency is directly influenced by shifts in the temperature distribution) on one end of the spectrum, followed by droughts (which can result from warm winters, a lack precipitation, or both) and extreme rainfalls (warm air holds more moisture, but factors other than just a direct temperature link are also key), and on down the line to the low confidence end of the spectrum which includes wildfires, cyclones, and severe convective storms, which can produce hail storms and tornadoes.

Regardless, this spectrum, however intuitive it might seem, represents the extent to which the science of climate attribution has grown in its sophistication. This means that when the question “Was that event caused by climate change?” comes up, climate researchers no longer have to give the old noncommittal answer. Instead, as the report’s authors put it, they can say:

“In many cases, it is now often possible to make and defend quantitative statements about the extent to which human-induced climate change (or another causal factor, such as a specific mode of natural variability) has influenced either the magnitude or the probability of occurrence of specific types of events or event classes.”

In other words, we are that much closer — if not exactly, then at least probabilistically — to saying what’s been hitting so many balls out of the park.

This originally appeared on the Climate CIRCulator.