January 25, 2015, by Brigitte Nerlich
Hottest year on record
Media reporting on climate change has heated up a little bit over the last ten days or so, after an announcement by NASA and NOAA which read like this: “NASA, NOAA Find 2014 Warmest Year in Modern Record”.
This was taken up by the media and, between 15 January and 25 January 2015, the phrases ‘warmest year on record’ and ‘hottest year on record’ were used 450 and 373 times respectively in All English Language News (data base: Lexis Nexis; search terms were these phrases and climate change or global warming). (I have not looked at these news items in detail and at how they used these phrases)
This information was welcomed by some and contested by others. What’s interesting is that the ‘warmest/hottest year’ meme collided with the ‘pause’ meme. Again, I have not yet had the time to disentangle how the two memes have fared, that is to say, how they were used argumentatively in the climate change debate and how, for example, an associated idea, namely statistical significance, was deployed in this context and by whom.
As one can imagine, there have been quite a few blog posts on the hottest year issue. This morning I saw one by Victor Venema, where he reflected on what the word ‘record’ may mean in the context of this debate (he also lists posts by Andrew Revkin, Gavin Schmidt and others).
This made me think a bit more about the semantics of it all. I first went to the Oxford English Dictionary to look up the word ‘record’, but as the OED is not accessible to people outside universities, I’ll stick to the really nice Google entry on the word. Interestingly, ‘record’ seems to have emerged from an amalgamation of the Greek word for ‘heart’ and the Latin word for ‘remember’ (it’s a bit more complicated, but still). More interestingly still, when you look at the various meanings of the noun ‘record’, you’ll see almost immediately that in debates about ‘the hottest/warmest year on record’ people sometimes seem to conflate the two senses, namely sense 1 and sense 3:
Sense 1: “a thing constituting a piece of evidence about the past, especially an account kept in writing or some other permanent form.”
Sense 3: “the best performance or most remarkable event of its kind”
I have the impression, but I might of course be wrong, that the debate triggered by the NOAA/NASA announcement flickered between the two meanings which caused a bit of confusion. And, of course, depending on the purpose of your article, blog post, tweet and so on and on the audience you want to reach, you can focus more on one than the other, and also use either ‘warmest’ or ‘hottest’ and so on.
Be this as it may, it is important to ‘record’, ‘remember’ and even ‘take to heart’, it seems to me, that we are dealing with a long-term upward trend or pattern with random (short-term) noise added onto it. It’s probable that 2014 is the warmest year on record, but it’s really this long-term trend that matters and 2014 is entirely consistent with this trend. It’s therefore important as a piece of evidence in a longer and larger story, and one should probably accept it as such.
Image: Old thermometers, wikimedia commons
It is quite common in the climate data community to refer to the datasets with the observations and measurements as the “climate record” or the “temperature record”.
This might be because these are the people that have access to the paper copies, where the past observers have recorded their observations by hand. It emphasizes how much work went into creating these holy datasets. Maybe it also points to the regrettable fact that much of these paper records are still not digitized and can thus still not be used to study the changes in our climate system.
Thus we had a temperature record record. I thought this would be a bit confusing. Thus in my post on the 2014 record, I have tried to avoid using the term “temperature record” for the data.
That’s looks like a sensible linguistic strategy to me! We’ll see whether it catches on or not.
Yes, your last paragraph is right on the button in terms of assessing the significance of this record surface temperature.
Since you’re interested in semantics it’s worth thinking about this:
1. there is an element of “random noise” in the temperature record which arises from the imprecision in measuring the surface temperature anomaly. In fact this may be the source of some of the angst in reporting of the record temperature – some naysayers use this uncertainty in attempting to dispute that 2014 is a record temperature – as you very nicely indicate this doesn’t matter terribly since it is the trend that is important and we wouldn’t have “record” temperatures if there wasn’t an upward temperature trend.
2. However much of the short term “noise” in the record isn’t “random” noise. It would be better to describe this as “stochastic” variation since it is unpredictable but deterministic – i.e. the short term variability results mostly from El Nino/La Nina and volcanic eruptions which are not predictable (as yet, anyhow) but are real influences on short term temperature that can be understood in hindsight….
i.e. an interesting semantic distinction between “random” and “stochastic”!
OK well having written that I see that Wikipedia defines a stochastic process as one that is non-deterministic (random)! However we do need a word for a contribution that is unpredictable (as far as we are able to say) but a result of a causal influence that we can understand (volcanic eruption/El Nino etc.). I use “stochastic” for that but obviously not everyone does!
Perhaps a better term for those causal but unpredictable events is “contingent”. 🙂
Ah yes, I was agonising over ‘random’ (as over so many words). But if I had used stochastic, people would have known immediately that I was deeply out of my depth and just pretending! I like contingent though… Next time…
I would say that in principle both words mean the same.
In case of noise, it is more likely to be not correlated. For example, random numbers. But you also have correlated noise, red noise, etc.
In case of stochastic, it is more likely to be correlated. For example, a stochastic process in general describes a process that has correlation in space and or time. But it would naturally also include the special case of no correlation.
The term noise also has a connotation of being unwanted. Thus I personally like the term Variability most. In case of describing a climate signal, the advantage furthermore is that variability can also include the annual and daily cycle, which are not stochastic.
[…] Hottest year on record […]
Just wanted to make two comments as a “consumer” of climate change news.
1)It winds me up no end when news reports say “hottest year since 19xx” and then don’t say whether 19xx was hotter or whether that was the year records began.
2) Just as climate change deniers read a lot into a single data point and ignore the trends and context, so I think it is a bit dangerous for climate scientists to make a fuss over “hottest year” etc – the trend is the key thing, presumably.
I agree on both points, especially the first, as this has happened to me as a consumer again and again, also about floods etc. But good newspapers do sometimes say, hottest/wettest year since records began to be taken in 1910 or something like that. I saw something like this recently and thought, ah, for once! But I can’t remember now where!
Yes, the trend is the thing! So shouting about the hottest year without mentioning the trend might just be digging your own ‘warmist’ rhetorical trap!