Jul 15 2019
Clouds and Climate Change
A paper is making the rounds on climate denial sites that claims to debunk human-caused climate change in a single stroke. Predictably, the paper does nothing of the sort, but it does raise a complex issue regarding climate change that is worth reviewing. But first let’s get to the paper itself.
The paper, by J. Kauppinen and P. Malmi, is a pre-publication paper on the Arxiv. This means it is not peer-reviewed. Their central claim, from the abstract:
In this paper we will prove that GCM-models used in IPCC report AR5 fail to calculate the influences of the low cloud cover changes on the global temperature.
Right in the first sentence is a huge red flag – claiming to be able to “prove” that the IPCC report is false. That’s a bold claim, and suggests a less than rigorous intellectual approach. They also claim to rebuke a rather robust conclusion built on many lines of evidence with a single line of evidence – the single stroke approach. This is also a huge red flag.
The claim is built around one major line of reasoning, that if you compare low cloud cover with changes in global temperatures, you see a strong correlation. In fact, the authors argue, you can explain most of global warming as resulting from a decrease in low cloud cover, leaving almost nothing left for anthropogenic forcing. There is a great deal wrong with this claim. The site ClimateFeedback has helpfully curated much of the response from climate scientists, who eviscerate the Kauppinen paper, and I will give you a summary of their summary.
The main criticisms include:
- The authors do not give a reference for the dataset they are using, which appears to be wrong and/or out of date.
- They ignore datasets that show the opposite trend they claim.
- They are assuming cause and effect without giving any line of evidence or reasoning.
- They criticize the use of climate models, but then give their own highly flawed climate model.
- They falsely claim CO2 goes from the oceans to the atmosphere, when it’s the other way around.
- They dismiss without evidence or argument the entirety of climate science, while referencing their own (mostly unpublished) work.
- “It cites six references: one to the IPCC report and one scientific article, both of which they apparently did not read or understand; two of their own unpublished manuscripts and two of their own articles in questionable or predatory journals.”
- They present no hypothesis as to how cloud changes are being forced, and therefore no explanation for why the climate is changing.
In short, this paper is a great example of pseudoscience. It would not pass any serious peer-review, but it can seem persuasive to a lay public, especially one that is eager to believe its conclusions. In other words, this looks more like political fodder than a serious scientific paper.
There is, in reality, a complex relationship between climate change and cloud cover. There are several good expert summaries online, and again I will try to give you my own understanding of these summaries.
Clouds can both increase and decrease surface temperatures, depending mostly on their altitude. Clouds, because they mostly have a high albedo, reflect sunlight back up into space, which has a cooling effect by reducing the amount of heat from the sun that hits the surface. Clouds, however, also trap heat. As heat radiates from the surface that heat warms the clouds overhead until they also radiate that heat up into space. The amount that the clouds warm up before they are radiating the same amount of heat they are absorbing represents the degree to which they “trap” heat.
Which of these two opposite effects predominates is determined largely by their altitude for two reasons. Low lying clouds tend to be dense. The more dense the clouds, the more sunlight they reflect (the less they let through to the surface). So lower clouds tend to reflect more light than higher clouds.
Also – lower clouds tend to already be warmer than those higher up, which are cooler. Low clouds are almost the same temperature as the surface, so they trap very little heat. Higher clouds are much cooler, so they will trap much more heat radiated from the surface (again – because they have to heat up more before they radiate that same heat back out to space).
So low clouds trap little heat but reflect a lot of sun, and therefore have a cooling effect. Higher clouds tend to be thinner and cooler, so they reflect less sunlight but trap more heat and tend to have a net warming effect. There is also another variable – where on the Earth are the clouds. Clouds toward the poles have less of an effect on albedo, because snow already has a high albedo.
What happens, then, with global warming? This is still perhaps the most uncertain aspect of climate models, which is why some scientists think it is such a focus for climate change deniers. While acknowledging this uncertainty, the current best estimate is that the net effect of increasing surface temperature is to push clouds higher, making them cooler and wispier and shifting them from a net cooling to a net warming effect. Warming also seems to move clouds toward the poles, which decreases the cooling effect of their high albedo. This means that as the Earth warms, changes in cloud cover are likely to have a positive (meaning bad) feedback effect, leading to more warming.
So even if the correlation found in the paper is correct, validated climate models (as opposed to their unvalidated model) predict if anything the cause and effect is in the other direction – warming causes a decrease in low cloud cover, not the other way around (although it then does become a positive feedback mechanism).
All of this affects our estimates of climate sensitivity, which is defined as the amount of warming in degrees C the Earth will experience with a doubling of pre-industrial CO2 levels. Best current estimates are between 1.5 and 4.5 C, with results toward the middle of that range having higher probability. Values outside that range are extremely unlikely and inconsistent with some methods of measuring sensitivity. The actual net feedback effect of clouds is part of the uncertainty of that range.
The current paper adds nothing to our understanding of climate models and climate sensitivity. It is far from the rigor necessary to deal meaningfully with this complex data. It does provide an opportunity, however, to once again highlight the differences between science and pseudoscience.