Is there a Global Average Temperature?

As I write this, at least two reanalyses are showing what are still preliminary reports of record high temperatures, exceeding the warmest temperatures observed in the instrumental record. If these results hold up to scrutiny over the next couple weeks or so, this will be the hottest week on record.

I want to say more about this in a future post, perhaps after the temperatures for these days are confirmed. But events like this sometimes bring out the worst in public debate, with some of extremists on the contrarian side going to the lengths of denying that there is even such a thing as a global average temperature. Now I've become used to reading contrarians claiming that the instrumental record is in some sense unreliable, either because of claims that the data isn't sound or that it's being deliberately corrupted in support of an alarmist narrative. But the outright denial of the concept of a global average temperature, or more specifically global mean surface temperature (GMST), seems to be on the rise.

The "argument" for this, if you can call it that, appears to be that the mere concept of an "average" is a "statistical construct" that individuals to not experience. We are told that we do not experience "global average temperature;" we experience local temperatures, and local temperatures have a much larger diurnal and seasonal variability than the increase in global temperatures, so the statistical construct of global mean surface temperature is meaningless. It seems to me that there are three, overlapping components to this objection to climate science, and I'll deal with each of these in turn. 

Statistical Constructs Are Meaningful

Implicit (and sometimes explicit) in the claims of some who object to the concept of GMST is that because an "average" is a statistical construct, it's not really meaningful. To me this is the silliest component to this objection. An "average" is of course a statistical construct, but we went through the trouble of constructing it because it allows us to say meaningful things about what we average. Here's a simple illustration of why.

Imagine you're the teacher of a class of 100 students. At the beginning of the term, you give a test and grade their results. You calculate their average grade to be a 60% with a low score of 25% and a high score of 80%. This is the average grade even if no student scored precisely a 60% on the exam. Now suppose you give another test at the end of the term. Your students average 85% with a low score of 30% and a high score of 100%. This of course is the average score even if no one scored precisely an 85%. But what can we say from this? We can say that students performed better (on average) after taking the class than they did before. Their knowledge of the subject matter improves. If you analyze the data from individual students, you may find that some improved very little and a couple perhaps even performed worse on the final exam than on the preliminary exam. But on average students performed better after taking your class. That's meaningful information.

Likewise, if we calculate that GMST has warmed by 1 C in 100 years, we can say that on average the planet's surface is warming. Some places may not have warmed anywhere near 1 C; a few places may have cooled. But on average the planet warmed by 1 C. This necessarily means that there has been a long-term energy imbalance at the top of the atmosphere such that energy entering the climate system has exceeded energy escaping the atmosphere into space. The surface must warm to close that imbalance. This kind of analysis is extremely meaningful, and it depends in part on meaningful statistical constructs like "average."

There is a Global Average Temperature

The denial that the Earth's surface has an average temperature is also  truly bizarre. Temperature by definition is the average kinetic energy of the molecules that make up a substance. For a 2m surface temperature, GMST is the average kinetic energy of the molecules in the atmosphere 2m above the surface of the earth. How can that number not exist? Given that all molecules have kinetic energy, none of them are at absolute 0 K and none have an infinite temperature, the surface of the earth has a temperature. There can't not be a GMST. 

A more interesting corollary to this objection is to claim that while there is a global average temperature, we don't know what it is. There's a grain of truth in this. Scientists do not have enough weather stations to know the absolute temperature of the earth with a high degree of accuracy. NASA, for instance, estimates that the GSMT for their baseline (1951-1980) was 14 C, but with several tenths of a degree accuracy. Let's use round numbers and say NASA can be sure that GMST averaged somewhere between 13 and 15 C between 1951-1980. But generally speaking, scientists are not nearly as interested in what the absolute temperature is; they are interested in how much and how rapidly it's changing. And for reasons I'll describe in a future post, scientists can calculate this with much greater accuracy, and that's why you often see GMST published in terms of anomalies.

We Do Experience Global Temperature Changes 

The third component to this objection to the concept of a GMST is almost understandable, even if it's misguided. This portion of the objection suggests that if GMST increases by just 1 C, that means on average, your daily lows and highs shift by 1 C, and generally speaking you won't notice that. If your July 9 has an average Tmin of 75 F and a Tmax of 95 F (Tavg of 85 F), and those all warm by 2 F, you probably won't notice that much if your low changes to 77 F and your high becomes 97 F. This of course misses the point entirely; we are affected by changes in GMST, even if we can't feel it on any given days' lows or highs.

Generally speaking global temperatures during the Last Glacial Maximum (some 20,000 years ago) averaged ~6 C cooler than the Holocene Thermal Maximum. It would be trivial to characterize this by saying that July 9 at your location would cool to a low of 63 F and a high of 83 F. If you live in New York City today, GMST that is 6 C cooler would mean your house is under a thick layer of ice. If you have beachfront property today, it would mean you're now a significant drive from the ocean. It would mean there is no agriculture in Canada, and in the US, where your food is grown or raised would dramatically change. The climate system would be in an entirely different state from what it is today.

Likewise, if GMST increases by 3 C, that will have dramatic effects on sea levels, how much damage is done by storms inland, how much sea ice is near the poles, how much water mountain glaciers can provide to areas that currently depend on it, etc. It would also strongly affect where we can grow and raise food, as agricultural zones will shift poleward. What we used to call extreme heat would also happen much more frequently (it would be downright common), and the list goes on. We will experience changes in GMST in multiple ways.

Conclusion

None of these objections provide a credible argument that there is no global average temperature or that it's not meaningful. Even if you are of the opinion that AGW will not cause enough warming to warrant mitigation efforts, we still came up with the statistical construct of an "average" because this allows us to say lots of meaningful things about data. The surface of the earth is still made up of molecules with kinetic energy, so it has an average temperature, and changes in GMST still affect the climate system in meaningful ways.

This objection currently appears to be made exclusively by extremists among contrarians. Even Dr. Roy Spencer, whom I disagree with on a lot of issues related to AGW, agrees with me on this. In 2014, he published a blogpost in which he addressed contrarian arguments that "don't hold water," and this "argument" was among them. Even according to Spencer, this objection to statistics and science causes contrarians to lose credibility. This is not an objection to AGW that holds any water.