I recently had a friend tell me that the number of hot days in the United States has declined over the last eighty years, thus proving that if the increased carbon dioxide in the atmosphere is having an effect on climate, it must be cooling the United States. To prove this claim, he showed me the graphic I include below.
A quick Google image search revealed the graphic was taken from climate science denier Tony Heller, aka “Steve Goddard,” via his conspiracy website ironically titled “RealClimateScience.” (I won’t link to it because I don’t want to increase his income from views to the ads on his site.)
The graphic above isn’t the one my friend showed me. I’ll present that one below, if you want to read more.
Note this graphic purportedly compares the number of weather stations reporting temperatures over 105 degrees F so far this year to the number reporting temperatures over 105 degrees F in 1936. I haven’t checked whether his data is accurate. Let’s take it as given.
There’s quite a lot wrong with this graphic, and with the claim made by my friend. In the interest of showing how dishonest climate science denial memes work, let’s list a few of the problems.
1) Incompleteness. The summer of 2016 isn’t over yet, so even on the surface, it’s not a good comparison.This is only a minor problem however, compared to the other ways in which this graphic is misleading and dishonest.
2) Oddity. The year 1936 was in the midst of the Dust Bowl era, a time when much of the United States was experiencing an historic drought which was caused (or at least, substantially worsened) by incredibly bad land management in the American southwest. Overfarming, overgrazing and deforestation turned much arable land into near-desert, leading to dry and overheated conditions in much of the continent. Comparing the U.S. today to any time in the 1930s is like comparing a vibrant city to a place that has been firebombed. It’s not a valid comparison. It’s a comparison that is so obviously faulty it had to be made on purpose.
Incidentally, notice also that this rather defeats any argument that human influence can’t affect climate. The extreme temperatures in the continental U.S. in 1936 were caused, in some measure, by human activity, and were not entirely a function of natural variation. But let’s move on.
3) Cherry picking. The comparison here is of one (1) year (1936) to one (1) other year (2016). No sensible statement can be made about climate trends by comparing the current year to some cherry-picked previous year.
Imagine a worker (call him Tony) whose wage income increases smoothly by $1000 every year over a period of forty working years. But other things affect his income as well. Maybe in 2005 Tony made an extra $50,000 by selling his dad’s old vintage Studebaker. Comparing his 2005 income to his 2016 income will not give you a good feel for how his wages have changed over the last ten years, and certainly won’t tell you where his income is likely to go in the future. Nor will it tell you what affect his work performance is having on his income. 2005 was an unusual year, so it makes a poor base for comparisons.
In the same way, 1936 was unusually hot in America, by random natural fluctuation as well as because of human activity. Comparing that year to this one won’t say anything about underlying climate trends. As implied by Item 2) above, this unusual year was specifically chosen to give the result Heller wanted. It was a dishonest and intentionally misleading choice.
4) Inappropriate yardsticks. The point that is implicit in Tony Heller’s graphic, and made explicit by my denier friend’s interpretation of that graphic, is that “the number of hot days” is a good proxy for “what is happening to the climate.” That’s just wrong. It’s an inappropriate measure.
Climate is driven by (among other factors) the total heat energy in the system. This affects rainfall, wind speed and direction, severity of storms, and a host of other factors that impact plant and animal life, growing seasons, the need for heating and air conditioning, and just about every other aspect of our lives. Total heat content can be approximated by looking at average temperatures, but not by looking at the number of days above an arbitrary limit.
As a simple example, say that one year the temperatures over a 5-day summer period in Sample City are 65, 70, 103, 105, and 72. The average for these five days is 83. Say in another year, the temperatures over the same 5-day period are 95, 96, 98, 99, and 97. The average temperature of the second series is 97. The heat energy available in the second year is much greater than in the first year, even though the first year had higher maximum temperatures.
Not only has the average substantially increased (by 14 degrees), but so has the minimum temperature in the series (from 65 to 95, a jump of 30 degrees). Again, this is despite that fact that the maximum temperatures is lower in the second series than in the first (99 as compared to 105). A region that habitually exhibits temperatures in the first range will have wildly different characteristics (vegetation, animals, rainfall, severity of storms and so on) from what you’d see in a region with the second range.
Heller’s tactic of counting the number of weather stations that measure temperatures above some arbitrary limit says very little about how the overall climate characteristics of the region have changed. Looking at minimum temperatures (or, even better, the average over reasonable time spans) is far more useful and informative.
5) Sample size. My denier friend tried to use Heller’s graphic to say something about world climate. (This is, of course, exactly what Heller wants people to do.) World climate is a complex thing, affected by geography, wind patterns, ocean currents, and a host of other factors. Not all places on Earth are affected the same way by this complex system. That’s why (for example) the Sahara Desert tends to be hotter and drier than Seattle, Washington. The United States accounts for less than 5% of the total world’s surface area. One can’t expect this tiny cherry-picked sample to say anything sensible about climate trends that affect the whole world.
Suppose I note that my favorite baseball team won one of its last twenty games. One out of twenty is 5%, just as the continental U.S. is 5% of the globe. This fact by itself doesn’t mean my team won all of those twenty games, and it certainly doesn’t mean they’ll win the World Series. The sample size of 5% is not large enough to tell us anything meaningful about the other games, and says nothing about the challenges my team faces in playing teams other than the one it played against in that one game.
To claim that “what happened to the continental United States” is a good proxy for “what is happening to the climate of the whole world” is senseless. Tony Heller wants people to draw unsupportable conclusions from this senseless sample.
6) Global. So let’s look at what happened to the whole world’s temperatures in the years before and after 1936, up to the present day. The graph below is the NASA GISS land-ocean average annual temperature index for the period 1880 to 2015 (in black), with a 5-year running mean (in red).
As you can see, there was a peak in world temperatures in the early 1940s, so Tony Heller’s graphic which includes particularly hot days in the continental U.S. might be expected to show some unusual warmth in 1936. You can also see, global temperatures have risen substantially since then, particularly in a rapid rise since about 1970. Since around 1980 the running mean has exceeded what it was c. 1936, and has never dropped back to that level. It is now far beyond where it was back then.
This reinforces what we said above about sample sizes and cherry-picking specific years and specific regions. If we want to draw conclusions about the world’s climate (as my friend tried to do), we have to look at the world’s climate. Tony Heller is encouraging his readers to avoid most of the data we have at our fingertips. He is engaging in dishonest misdirection.
7) Natural variation. Although it isn’t mentioned in Heller’s graphic, my denier friend often draws another conclusion from memes such as Heller’s. Note that in the NASA GISS graph, the world’s temperature increased markedly from c. 1910 to 1940. Yet human production of greenhouse gasses (other than coal burning for household heating and industrial processes) didn’t really take off until after that time. My friend argues (and with some justification) that the temperature variations prior to 1940 were mostly due to natural processes, and didn’t require human activity. He then claims we therefore can’t really know that the immense increase in global temperature after that time is caused by humans. It could well have been more “natural variation.”
This is equivalent to saying that we can’t calculate the effects of an airplane hitting a skyscraper until we know the limits of the effects of earthquakes and hurricanes on skyscrapers. Natural variation sometimes causes buildings to collapse. How do we know that the World Trade Center was brought down by being stuck by airliners?
We know because A) there were no earthquakes or hurricanes in New York on September 11, 2001, and B) we have film of the airliners hitting the World Trade Center’s Twin Towers on that date. Also C) we can calculate the amount of kinetic energy those airliners imparted to the buildings when they struck, along with the effects of exploding fuel. We can also calculate the effects upon the buildings of the natural variations in weather and seismic activity on that day. It’s pretty easy to tell which set of causes had more impact.
In other words, we can look at the actual causes that affected the events. We have data and we have physics. We know the concentration of CO2 and other greenhouse gasses has increased markedly in the last century. We know the effect CO2 has on the ability of the atmosphere to retain heat. We know the various natural influences that were in effect in the 1930s, and the influences that are impacting the climate now. This isn’t hard.
For this “natural variation” argument to be sensible, we have to examine the actual factors that are in force. To compare the heat waves of the 1930s to the increase in global temperatures today, we need to look at the conditions that occurred in those two time periods. We know what they are. We can calculate the results.
It’s not enough to say that since something happened before, it can happen again. We have to look at what caused it to happen before, and see whether the same thing is causing events today. It’s not honest to say that we need to know the limits of natural variation before we can calculate the impact of new influences. We already know quite a lot more about both natural variation and human intervention than deniers pretend we know.
The sorts of errors I describe above are common when climate science deniers try to dispute the impact that human activity has on planetary climate. (These are, in fact, only a few of the many errors in this one graphic.) Science disinformers like Tony Heller intentionally mislead unwary victims who don’t skeptically examine the dishonest claims and false memes.
It is the responsibility of all of us to help inform the public about the errors and disinformation being spread by science deniers. Rational decisions cannot result from dishonest propaganda.