Andrew Revkin has this commentary at the New York Times: How ‘Warmest Ever’ Headlines and Debates Can Obscure What Matters About Climate Change.
I will argue below that Revkin has, inadvertently or not, linked a science denialist trope to the important scientific finding that 2014 is the warmest year on record, as part of his presumably well intentioned effort to focus on trends rather than individual points. (See his comment on this blog below.) Yes, the trend is more important than a given data point, but the headline does not really obscure, but rather, underscores.
I’m afraid the devaluing of 2014, or any year, as a new data point in measuring global surface temperature change will become yet another climate science denialist claim. The strength of this claim would lie entirely in absurd idea that one data point in set a of data demonstrating a trend, with N=130, is not important if it is not “statistically significant” in difference from nearby points. That is actually what we expect when adding new data to the end of a trend, when the trend itself is real and there are already sufficient data points to lend a high degree of confidence. Indeed, the temperature anomaly estimate for 2014 is exactly what we would expect under conditions of continuous global warming.
In his State of the Union Address, President Obama said, “2014 was the planet’s warmest year on record. Now, one year doesn’t make a trend, but this does?—?14 of the 15 warmest years on record have all fallen in the first 15 years of this century.” The President refers here to last Friday’s announcement by the key US government agencies that study climate and weather, NOAA and NASA, that 2014 was the warmest year in a record of temperatures that goes back to 1880. President Obama correctly notes the milestone of 2014 being the warmest year, and that the larger, more important, fact, is that all of the warmest years we’ve experienced since industrialization have been very recent.
Revkin notes the NOAA/NASA report and links this to criticisms from the Daily Mail (one of the most notorious rags at the fringes of journalism), the right wing “The Federalist,” and an extreme climate science denialist site. This has the effect of creating a balance between established government scientists and agencies who say one thing, and critics providing an alternative view. What this really is, of course, is a false balance between mainstream science on one hand and rather extreme science denialism on the other. By uncritically providing their point of view along side the findings of NOAA and NASA he effectively elevates cranks to the status of expert. This slight of hand seems to have had the function of allowing Revkin to note that “it’s a distraction to focus on records … given how year-to-year differences in global temperature are measured in a few hundredths of a degree Fahrenheit, and given the implicit uncertainty in such measurement.”
So. 2014. Shiny! Since we are capable of thinking about only one thing, lets’ avoid thinking about 2014? No. That would be wrong.
Revkin previously claimed that when writing about science he tries to only refer to scientists, in a 2006 interview conducted by Paul Thacker:
Thacker: Scientists consistently complain that the journalistic practice of “balance” allows skeptics to gain an unfair toehold in media coverage, which ignores consensus in favor of controversy. Do you agree, and do journalists need to rethink their approach to covering complex scientific issues?
Revkin: Balance is a necessary evil, a crutch, particularly in daily journalism, but only works with coverage of the science –policy interface if the journalist works hard to label the voices in a story to reflect what they represent (a consensus or knowledgeable minority) and certainly to reflect their motivation or potential conflict (paid by industry? on staff at an environmental group?). When I’m writing strictly on a scientific finding, I avoid voices from anyone other than scientists. When I’m writing on policy, I’ll quote those with an agenda, but only if I label their agenda.
He did not do this in his latest New York Times commentary.
Much of the rest of Revkin’s post is a foray into probability and statistics that ends up not being very helpful, mainly because the meaning of 2014’s surface temperature estimate was not appreciated.
Here is what is going on. The value of a given year when it is one of 130+ data points is limited. That was true before 2014 turned out to be the warmest in that data set, it would have been true if 2014 had turned out to have been a bit cooler or warmer. It applies to each and every one of the years for which surface temperatures have been estimated by a half dozen or so agencies or research groups. What we are really measuring here is change over time, over decades of time. We do not decide if there is a trend of change over time in such a data set by looking at a given year. We look at the trend itself, incorporating all of the data.
At the same time, every single point counts. The trend is an accumulation of annual averages of individual years arranged across time. We are ultimately asking two questions of these data. First, we are asking about patterns in the past, since humans started releasing copious amounts of the greenhouse gas CO2 into the atmosphere. We are seeking other relationships over that time frame as well, looking at the strength of the sun over time, the effects of atmospheric dust spewed by volcanoes or human activities, and complex interaction between “surface temperature” (the thing these data sets measure) and the actual global temperature which includes the ocean (containing well over 90% of this heat). In order to examine this record in light of all of those variables, we have to pay attention to each and every year. And, the most important year is always the one we just got, because it serves as a tiny but helpful test of our running hypothesis. And, it is more data! It is like finding a new friend.
The second question we are asking is what will happen in the future. Right up until the start of January, the year 2014 was in the future. The temperature value estimated for 2014, or more exactly, the degree (literally) and direction of this year as an anomaly reflected against a long term average, is the very point of this research, and of course, this applies to 2015, 2016, and so on. The relative position of 2014 in relation to the long term data is of great interest. Had that value been substantially less than the trend predicts, there would be some ‘splainin to do. Had the value been way higher than the trend predicts, there would be some ‘splainin to do. As it turns out, 2014 couldn’t have been much closer to the exact value the long term trend line predicted. It was dead on under the assumption of continued warming. That, ladies and gentleman, is a data point of worthy note.
2014’s temperature value is significant for the very fact that it is not statistically different from expectations under the widely accepted model of anthropogenic global warming. That is the meaning of 2014’s surface temperature estimate.
Please remember that the “surface temperature” is only the measurement of the air near the ground and the surface of the sea, combined. This is less than 10% of the total heat containing portion of the planet affected by global warming. The surface temperature is an important indicator of change over time, and for historical reasons this is a measurement we use. But it is like the tail of the dog, where the dog is the ocean, where the other 90% or so of heat resides. John Abraham just posted a commentary on the most recent data, just updated, on ocean heat content, and it has been rising apace. (See also this paper on that topic, and this post by Joe Romm.)
2014’s surface temperature estimate obscures nothing, it reveals. It is not a distraction. It is the point.
I think it is important to note that the reality of global warming, and tracking it, is complex. This was, in fact, the warmest year on record. The fact that a new warmest year will almost always be only a little warmer than a previous recent warmest year does not diminish the importance of the new top rank year, but rather, underscores it, since all the “warmest years” are recent. If this year beat out the next warmest year by only a small amount, and that next warmest year was from 1881, we would not be impressed. But since this new warmest year’s friends, the other previous winners of this particular numerical beauty contest, are all very recent, we are impressed. And we should be worried.
It would be odd to not acknowledge a new warmest year when the data come out, and it would be odd to not recognize its significance.