The Pitfalls of General Reporting: A Case Study

Cross posted from ClimateSight

Today’s edition of Nature included an alarming paper, indicating record ozone loss in the Arctic due to an unusually long period of cold temperatures in the lower stratosphere.

On the same day, coverage of the story by the Canadian Press included a fundamental error that is already contributing to public confusion about the reality of climate change.

Counter-intuitively, while global warming causes temperatures in the troposphere (the lowest layer of the atmosphere) to rise, it causes temperatures in the stratosphere (the next layer up), as well as every layer above that, to fall. This pattern of a warming surface and a cooling upper atmosphere is due to an enhanced greenhouse effect: since more heat is trapped near the surface, less is radiated back up through the stratosphere.

This pattern was observed in the Arctic this year. As the Nature paper mentions, the stratosphere was unusually cold in early 2011. The surface temperatures, however, were unusually warm, as data from NASA shows:

Mar-May 2011

Dec-Feb 2011

While we can’t know for sure whether or not the unusual stratospheric conditions were caused by climate change, this chain of cause and effect is entirely consistent with what we can expect in a warming world.

However, if all you read was an article by the Canadian Press, you could be forgiven for thinking differently.

The article states that the ozone loss was “caused by an unusually prolonged period of extremely low temperatures.” I’m going to assume that means surface temperatures, because nothing else is specified – and virtually every member of the public would assume that too. As we saw from the NASA maps, though, cold surface temperatures couldn’t be further from the truth.

The headline, which was probably written by the Winnipeg Free Press, rather than the Canadian Press, tops off the glaring misconception nicely:

Record Ozone loss over the Arctic caused by extremely cold weather: scientists

No, no, no. Weather happens in the troposphere, not the stratosphere. While the stratosphere was extremely cold, the troposphere certainly was not. It appears that the reporters assumed the word “stratosphere” in the paper’s abstract was completely unimportant. In fact, it changes the meaning of the story entirely.

The reaction to this article, as seen in the comments section, is predictable:

So with global warming our winters are colder?

First it’s global warming that is destroying Earth, now it’s being too cold?! I’m starting to think these guys know as much about this as weather guys know about forecasting the weather!

Al gore the biggest con man since the beginning of mankind!! This guys holdings leave a bigger carbon footprint than most small countries!!

I’m confused. I thought the north was getting warmer and that’s why the polar bears are roaming around Churchill looking for food. There isn’t ice for them to go fishing.

People are already confused, and deniers are already using this journalistic error as evidence that global warming is fake. All because a major science story was written by a general reporter who didn’t understand the study they were covering.

In Manitoba, high school students learn about the different layers of the atmosphere in the mandatory grade 10 science course. Now, reporters who can’t recall this information are writing science stories for the Canadian Press.

Copyright retained by author.


  1. "...since more heat is trapped near the surface, less is radiated back up through the stratosphere."

    I think that explanation is incorrect. The heat radiated up through the stratosphere is approximately the same, at equilibrium. The reason for the difference is that the lapse rate is negative in the stratosphere.

    Adding non-condensing GHGs moves the entire temperature profile upwards (approximately), so each level experiences temperatures that were previously at a slightly lower altitude. A positive lapse rate causes warming, a negative lapse rate causes cooling. Zero lapse rate would leave the temperature unchanged.

    [ This is the disputed posting. I still think it is simply wrong. -mt ]

  2. "Nullius in Verba" writes in to contest the article's claim that "This pattern of a warming surface and a cooling upper atmosphere is due to an enhanced greenhouse effect: since more heat is trapped near the surface, less is radiated back up through the stratosphere."

    I agree that this explanation doesn't hold water.

    However, Nullius provides an alternative explanation wherein lapse rates are fundamentally tied to altitude or pressure (unclear which), which is also wrong.

    For the purposes of the present article, it is something of a distraction - what matters here is that a stratospheric cooling trend is in fact predicted along with a tropospheric warming, and that this prediction is strongly confirmed by observations to date.

    I once (long ago in my usenet days) had a simple explanation of my own for this which, has also been shot down, so I won't venture an alternative. But it's probably a good idea to look for something definitive and accessible on the question.

    Any suggestions?

  3. Here's my explanation: there's little heat flow from convection across the tropopause, so stratospheric temperature is governed by its radiative balance. Ozone absorbs incoming solar ultraviolet, leading to heating, which is balanced by thermal radiation from the greenhouse gases in the stratosphere. Ozone has decreased and other greenhouse gases have increased, both of which shift the balance to lower temperatures.

    Kate's phrasing might be better put as that the effect of greenhouse gases means "more heat is re-radiated within the lower atmosphere making it warmer, and more is radiated out of the stratosphere making it cooler" or something like that.

    It's a clear consequence of the physics whenever you apply the math of the problem, but the conceptual explanations as always can be tricky.

  4. ...

    I suggest Ramanathan 1998 "Trace-Gas Greenhouse Effect and Global Warming" p192.

    "As we mentioned earlier, in our explanation of the greenhouse effect, OLR reduces (with an increase in C02) *because of the decrease in temperature with altitude*. In the stratosphere, however, temperature *increases with altitude* and as a result the cooling to space is larger than the absorption from layers below. This is the fundamental reason for the CO2 induced cooling."

    That is to say, the lapse rate is negative in the stratosphere.


    (Much better. Nevertheless, edited to remove snark)

  5. Jeff Masters explains well in this article from a few days back:

    Greenhouse gases cause stratospheric cooling

    When ozone absorbs UV light, it heats the surrounding air. Thus, the loss of ozone in recent decades has helped cool the stratosphere, resulting in a feedback loop where colder temperatures create more PSCs, resulting in even more ozone destruction. However, in 1987, CFCs and other ozone-depleting substances were banned. As a result, CFC levels in the stratosphere peaked in 2000, and had fallen by 3.8% as of 2008, according to NASA. Unfortunately, despite the fact that CFCs are falling in concentration, the stratosphere is not warming up. The recovery of the ozone layer is being delayed by human emissions of greenhouse gases like carbon dioxide and methane. These gases trap heat near the surface, but cause cooling of the stratosphere and increased formation of the PSCs that help destroy ozone. We need only look as far as our sister planet, Venus, to see an example of how the greenhouse effect warms the surface but cools the upper atmosphere. Venus's atmosphere is 96.5% carbon dioxide, which has triggered a hellish run-away greenhouse effect. The average surface temperature on Venus is a sizzling 894 °F, hot enough to melt lead. Venus's upper atmosphere, though, is a startling 4 - 5 times colder than Earth's upper atmosphere. The explanation of this greenhouse gas-caused surface heating and upper air cooling is not simple, but good discussions can be found at Max Planck Institute for Chemistry and, for those unafraid of radiative transfer theory. One way to think about the problem is that the amount of infrared heat energy radiated out to space by a planet is roughly equal to the amount of solar energy it receives from the sun. If the surface atmosphere warms, there must be compensating cooling elsewhere in the atmosphere in order to keep the amount of heat given off by the planet the same and balanced. As emissions of greenhouse gases continue to rise, their cooling effect on the stratosphere will increase. This will make recovery of the stratospheric ozone layer much slower.

    Greenhouse gases cause cooling higher up, too

    Greenhouse gases have also led to the cooling of the atmosphere at levels higher than the stratosphere. Over the past 30 years, the Earth's surface temperature has increased 0.2 - 0.4 °C, while the temperature in the mesosphere, about 50 - 80 km above ground, has cooled 5 - 10 °C (Beig et al., 2006). There is no appreciable cooling due to ozone destruction at these altitudes, so nearly all of this dramatic cooling is due to the addition of greenhouse gases to the atmosphere. Even greater cooling of 17 °C per decade has been observed high in the ionosphere, at 350 km altitude. This has affected the orbits of orbiting satellites, due to decreased drag, since the upper atmosphere has shrunk and moved closer to the surface (Lastovicka et al., 2006). The density of the air has declined 2 - 3% per decade the past 30 years at 350 km altitude. So, in a sense, the sky IS falling due to the greenhouse effect!

    Since any increase in solar energy would heat both the lower and upper atmosphere, the observed drop in upper atmospheric temperatures in the past 30 years argues against an increase in energy coming from the sun being responsible for global warming. The observed cooling of the upper atmosphere is strong evidence that the warming at Earth's surface is due to human-emitted greenhouse gases that trap heat near the surface and cause compensating cooling aloft. It should also give us additional confidence in the climate models, since they predicted that this upper atmospheric cooling would occur. Keep in mind, also, that 2010 was tied for Earth's hottest year on record, and the amount of energy coming from the sun during 2009 - 2010 was the lowest since satellite measurements began in the late 1970s. There has been no long-term increase in energy coming from the sun in recent decades, and the notion that global warming is due to an increase in energy coming from the sun simply doesn't add up.

  6. Oddly, Masters' explanation is the same one I came up with.

    One way to think about the problem is that the amount of infrared heat energy radiated out to space by a planet is roughly equal to the amount of solar energy it receives from the sun. If the surface atmosphere warms, there must be compensating cooling elsewhere in the atmosphere in order to keep the amount of heat given off by the planet the same and balanced.

    I am reasonably convinced that it is wrong.

    Someone (William Connolley?) questioned me on this and if I remember right I shamefully slunk off without coming to any firm conclusion. Someone else said it had "something to do with the ozone layer", which is consistent with NiV's Ramanathan reference. I'll try to dig it up on the usenet archives.

    Masters' discussion of the temperature at the top of Venus' atmosphere is a bit baffling, but it doesn't really make much sense. We are discussing stratospheric cooling as an effect of increased near-surface concentrations of greenhouse gases. The base state is not direct evidence either way, I think.

    The reason no part of the atmosphere needs to get cooler to compensate for the warmer lower atmosphere is exactly the same as the reason the lower atmosphere got warmer - the decreased transmissivity of the atmosphere in the infrared. All else equal, the surface that is in radiative balance moves up as GHGs increase. If the lapse rate is positive, i.e.

    dT/dz < 0 then that surface is cooler in the base state, and hence may not need to warm up from the base state. At which point, the devil in the details shows up.

  7. The answer to why increasing CO2 causes both stratospheric cooling and tropospheric warming is laid out in Chapter 4 of my book, Principles of Planetary Climate. The simple remark is that increasing concentration of a greenhouse gas does two things: it increases absorption of infrared coming from warm layers below (a warming effect) but it also increases emission of infrared from the layer of atmosphere in question. The issue is which effect wins? The triumph of CO2 cooling in the stratosphere comes about because CO2 emits strongly in the 15 micron band, but the upwelling infrared is depleted in this band because it has all been absorbed by lower layers. A substance (like sulfate aerosol) which has broadband absorption/emission causes stratospheric warming rather than cooling.

    You can find a lot of online resources, plus ordering information, at

  8. I am delighted to see you signed up, Ray, and hope you stick around to rescue me when I get in trouble. (Ray has climatological superpowers, and I have long been his loyal but hapless Jimmy Olsen.)

    Intending to read that book has been on the docket for me for some time. And then, I actually lift it. That alone tends to slow me down, I'm afraid. But maybe this community will reach the point where it can support a study group.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.