Reframing Attribution: It the
previous entry I wrote about the perils and pitfalls of event attribution. In this entry I want to untangle a few issues and, then, ultimately reframe attribution. Reframe? This is in the spirit of psychology and sociology, a different way to look at something. In this case, take the word, “attribution” and think about the meaning of this word, say, from the point of view of scientists, journalists, politicians ….
To be concrete, start with this scenario.
1) There is an extreme weather event, perhaps a hurricane submerges New Orleans, or a heat wave kills 1000s in Moscow.
2) Advocates say that the event is global warming.
3) Politicians say that the event is global warming.
4) Scientists suggest that the circumstances of the event are consistent with global warming.
5) Journalists ask if the extreme event is natural or global warming.
6) Different groups of scientists hurry to investigate the event. It takes a while.
7) The scientists publish their papers and because the event was newsworthy, the journalists follow up and ask again: Was the event natural or was it global warming?
There is in this scenario entanglement. We have scientists, journalists, and politicians. I have explicitly used the plural form to suggest that there are many perspectives, many points of view, many purposes represented. Because of the presence of political interests, the question is being asked in a social environment that is more political than it is scientific.
In the previous entry, I wrote, “It is hard to see how playing the game of defining extreme events and then attributing that event to ‘climate change’ can ever be won. In fact, it seems like it is a game that necessarily leads to controversy, and controversy is the fuel of talk radio, blogs propagating around the world, and the maintenance of doubt.” The game to which I refer is described above: event, fast public attribution of the event to climate change, scientific investigation and deliberation, scientific conclusion that the event is not wholly-and-solely due to climate change. In the formal and informal media, this game devolves to:
“This event is the proof of global warming,” followed some months later by, “No it is not.”
You can read the
previous entry on why I maintain trying to attribute a single event to climate change with a yes-or-no answer or to split our weather into natural-and-changed is not scientifically sensible. That does not mean, however, that we should not study extreme events and place them into context with history, a warming climate, and how they inform our future. In fact, I have maintained that one of the most important tasks for climate scientists to take on is the quantification of variability that is “short-term” compared with the “long-term” normally associated with climate. (See
Some Jobs for Modelers, and
Ocean, Atmosphere, Ice and Land) Which brings me to “attribution.”
In the discourse described above, amongst the politicians, journalists, and scientists, “attribution” has risen to mean, “Can this event be attributed to climate change?” Sometimes it is worth going back to basics. From the
American Heritage Dictionary of the English Language attribute is “to relate to a particular cause or source.” And from the
Glossary of Terms of the
Intergovernmental Panel on Climate Change“Detection and Attribution: Climate varies continually on all time scales. Detection of climate change is the process of demonstrating that climate has changed in some defined statistical sense, without providing a reason for that change. Attribution of causes of climate change is the process of establishing the most likely causes for the detected change with some defined level of confidence.”
In fact, neither of these definitions require a yes-or-no, wholly-and-solely answer that a particular event was “caused” by the warming of the planet by increasing greenhouse gases. That requirement has risen from the quagmire of the public discourse.
In the piece
Some Jobs for Modelers I talk about “forecast busts.” These are well known to weather buffs, when weather forecasts fail. It is worst when severe weather shows up unexpectedly. In December of 1999 there was a series of Atlantic storms that hit France which were badly forecast. Detailed examination of the observations, the forecast model, and the ability of model to utilize those observations, revealed that there was adequate information to provide a better forecast. Specific failures in the forecast system were identified. (A complicated paper on those storms:
Dee et al. 2001) When I think of attribution and a single extreme event, then I think of the detailed scientific investigation of the processes that come together at the occurrence of that event.
There are many reasons to pose such a study. A basic reason is to understand the physical processes. For example, in a historic heat wave, what is the impact of regional changes in the forest, agriculture, and the urban environment? What are the specifics of the atmospheric flow that allow the development of a period of persistent heat? A perfectly legitimate question is whether or not changes in our environment related to greenhouse gases have had a discernible influence on the event.
So that becomes the question. In the complex mix of processes that are responsible for determining the temperature and winds and rain of an extreme event, is there a discernible contribution that can accounted against, attributed to, climate change? To make it more challenging, climate change is not a simple, unrelenting, uniform warming of the surface. Therefore, if there is to be a discernible signal, then it has to rise above the variability, the noise, that is implied by the complexity described in the previous paragraphs. It is not a question of whether or not an extreme event is caused by climate change, it is what influence might be attributed to the increase of greenhouse gases.
That said, there are many reasons to investigate which processes, which causes, are responsible for an extreme event. A fundamental one is to improve the ability to predict the event. Another reason is to understand the impact of the event, assess the risk associated with such events in the future, and, if warranted, develop the ways to better prepare for such events.
I want to return to the my previous blog, which was motivated by a story that originated in the
Green Blog by John Rudolf on the
New York Times website (March 9, 2011) about the Russian heat wave in the summer of 2010. The news story reported on a paper by
Randy Dole and co-authors. Within hours the Dole et al. paper was headlined on both news sites and in blogs that the paper said that the 2010 Russian heat wave had no relation to global warming. It is a source of continuing and intensifying controversy. (
from Climate Progress, recall that above, I deliberately used the plural of scientist.)
Here is the link to the abstract of
Dole et al., Was There a Basis for Anticipating the 2010 Russian Heat Wave? Dole et al. take an approach to the problem that is process-based, in the spirit of the process-based approach to a busted forecast. They search for the signal over the noise, and for the 2010 event cannot state definitively that the signal related to the increase of greenhouse gases exceeds the noise. I want to quote, however, two sentences from the “Concluding Remarks” of Dole et al.
“The results suggest that we may be on the cusp of a period in which the probability of such events increases rapidly, due primarily to the influence of projected increases in greenhouse gas concentrations.”
And looking forward.
“However, as is the case of the 2010 Russian heat waves, events will also occur that are not readily anticipated from knowledge of either prior climate trends or specific climate forcings, and for which advance warning may thus be limited.”
The Dole et al. paper does not state in any way that global warming is unreal. Quite the contrary, they work in a rigorous physics-based approach and investigate this region, at this time, for this event, and ask in the context of a forecasting problem, can a discernible contribution be attributed to human-caused greenhouse gas emissions? Their method, their analysis, their conclusions - that for some highly particular reasons - the climate change signal has not popped out of the natural variability. But as they say, it has in other places, for other phenomena.
Dole et al. provide one scientific approach to the problem of event attribution. There are other approaches. (see
Barriopedro et al. The Hot Summer of 2010: Redrawing the Temperature Map of Europe) The conclusions from these results are likely to be different, and that difference may appear inconsequential to some and enormous to others. And while these differences might appear as important to scientists, my point is that this process of event attribution is a place where the scientific investigation of the climate interfaces, strongly,with the media. Therefore, it is also a place where, by definition, scientific investigation interfaces with the political argument. Politically or in terms of informing the public, a primary result of this process is to build, amplify and maintain doubt. Here, I have tried to reframe attribution. Next, on reframing the dialogue.
r
Previous blogs on the disruptions and communications of climate science. (or how can climate scientists contribute to political discrediting of science.)
Strength in Many Peers “Have you no sense of decency, sir, at long last?”What to Do? What to Do?If Lady Chatterley’s Lover, then … Faceted Search of Blogs at climateknowledge.org