Non-peer-reviewed manuscript falsely claims natural cloud changes can explain global warming

During the last hundred years the temperature is increased about 0.1°C because of carbon dioxide. The human contribution was about 0.01°C.
Flawed Reasoning: The authors' argument claims a correlation between cloud cover/relative humidity and global temperature proves that the former caused the latter without investigating whether they have the relationship backwards.
Inadequate support: The source of their claimed global cloud dataset is not given, and no research on their proposed mechanism for climate change is cited.
Fails to provide correct physical explanation: The manuscript incorrectly claims that the rise of atmospheric carbon dioxide is caused by release from ocean waters. It also provides no explanation for the claim that an increase in relative humidity causes global cooling.
Warming related to human activities is estimated to be around 1°C over the past century. This document claims to overturn decades of scientific findings but provides neither the source of the data it uses nor the physics responsible for the proposed relationship between clouds and global temperature.


CLAIM: Man-made Climate Change Doesn't Exist In Practice... During the last hundred years the temperature is increased [sic] about 0.1°C because of carbon dioxide. The human contribution was about 0.01°C.

Some news outlets are publishing articles stating that this claim is based on a new study. In reality, there is no new published study. The claim comes from a six-page document uploaded to arXiv, a website traditionally used by scientists to make manuscripts available before publication. This means that this article has not been peer-reviewed, so there is no guarantee to its credibility.

If the blogs that covered this as a new study had contacted independent scientists for insight, instead of accepting this short document as revolutionary science, they would have found that it does not have any scientific credibility.

As the scientists who examined this claim explained, the document relies on circular reasoning to claim that cloud cover and relative humidity have caused the change in global temperature, and ignores many additional factors affecting global temperature—including aerosol pollution, volcanic eruptions, and natural ocean oscillations. The published, peer-reviewed scientific research on this topic clearly shows that human activities are responsible for climate change.

Timothy Osborn, Professor, University of East Anglia, and Director of Research, Climatic Research Unit:
The unpublished paper by Kauppinen & Malmi is deeply flawed and the claims that (1) CO2 has caused only 0.1 degC of warming and that (2) only 10% (0.01 degC) of this warming is from human activity are both unsupported claims.

The paper should not be relied upon.

Their claims are based on a chain of reasoning with multiple flaws:
(1) They claim that climate models cannot be relied upon but do not demonstrate this.
(2) They instead make a new climate model (despite this being in contradiction of (1)).
(3) Their new climate model is unvalidated. It is based upon datasets of cloud and humidity without any sources given and which are not up-to-date. They provide no assessment of the accuracy of the data used—these variables are very difficult to measure on a global basis over the time period used. No physical basis is given for their new climate model (e.g. no process is given for how higher relative humidity can make the globe cool).
(4) They fail to consider cause and effect. For example, they assume without any support that a decrease in relative humidity is natural. They give no reasons why it would have decreased. They fail to consider whether climate change could have caused relative humidity to change.

(5) They state without any support that most of the atmospheric CO2 increase is due to emissions from the oceans. They ignore anthropogenic CO2 emissions which are more than large enough to explain the full increase. They ignore observational evidence that shows that the oceans are net sinks of CO2 at present, not net sources.
(6) They dismiss the entire body of climate science—especially that there is a significant greenhouse effect—and instead cite their own work (unpublished or published in journals outside the field).

In reality there is strong scientific evidence for conclusions in stark contrast to those of Kauppinen and Malmi, namely that (a) all of the CO2 rise is from human activity, (b) that 100% of the CO2-induced warming is therefore anthropogenic, and (c) that (together with anthropogenic emissions of other greenhouse gases like methane) the total anthropogenic warming is around 1 degC.

A published paper demonstrating (a) and (b) is Cawley (2011)1.

A body of evidence for (c) is Haustein et al (2017)2 and references therein.

Richard Betts, Professor, Met Office Hadley Centre & University of Exeter:
This document is not a proper scientific paper and would not pass peer review in an academic journal.  The crucial data sources (e.g. of the dataset claimed to be low cloud cover) are not provided, and the figure purporting to show changes in cloud cover is at odds with peer-reviewed papers like  Eastman et al1. That published scientific paper does not show the decline in low cloud cover claimed in this document.

graphs of trends in low, middle, and high cloud cover
Annual average daytime cloud cover. Source: Eastman et al (2011)

This document only cites 6 references, 4 of which are the authors’ own, and of these 2 are not actually published. Therefore I would not regard this document as having any scientific credibility.

Even if the claimed observational cloud data turned out to be of good quality, the authors inaccurately describe figure 2 as “experimental observations”. “Experimental” would imply that it was derived by experiment: i.e., some sort of controlled scientific study, as opposed to observations of the uncontrolled natural world. (Unless they are claiming to have carried out an actual experiment on the Earth, which would be a bizarre claim!).  All they are doing is correlating two datasets (of unknown source). This does not “prove” anything, despite their claims that it does.

Their overall conclusion of small anthropogenic contribution to observed global warming is very different to the conclusions of numerous properly-documented scientific studies2, which have demonstrated beyond reasonable doubt that the observed warming in recent decades is due to human influence.

Chris Brierley, Senior Lecturer, University College London:
From a logic perspective, this effort makes two utterly unjustified assumptions:
(1) That any changes in low cloud cover are natural, rather than human-induced. The research discipline of aerosol-cloud interactions exists to explore this relationship; and the charlatans selling cloud-seeding would argue vociferously against all cloud changes being natural.
(2) Correlation = causality. Just because two time series show a strong correlation, does not mean that one causes the other. In fact, I’ve no idea why the authors think reducing cloud cover drives warming, rather than the over way around. The IPCC report, (Boucher et al1, cited by them) states that warming causes low cloud cover. An enlightening example to highlight this kind of error is the correlation between the number of storks and birth rate in Europe2.

But this article also misses some important hallmarks of real science:
(1) It gives only one reference to research by other scientists.
(2) Even this is a mis-application: they authors neglect to include any time-variation in their equation. This effectively assumes that the Earth responds instantaneously to any drivers.
(3) They explicitly state at the outset that they do not consider models as evidence.
(4) They do not explain where their data has come from (I guess though that the cloud cover has come from satellite irradiance, processed through a model).
(5) They infer meaning well-beyond the scope of their data, without any justification.

Mark Richardson, Research Associate, Colorado State University/NASA JPL:
Errors in this manuscript include:

(1) The climate model comparison shows the opposite of what Kaupinen & Malmo claim.

(2) Their assumed warming effects of CO2 are much smaller than we’ve measured.

(3) They say that clouds and humidity are causing all the temperature change but satellite measurements suggest, if anything, the opposite.

(4) Humans caused the CO2 rise and the oceans are absorbing CO2, this is changing ocean pH. Kaupinen and Malmi falsely say the opposite: that oceans are adding CO2 to the air.

There’s tons of observational evidence that human activity is driving global warming, and this data supports the projected range of ongoing and future global warming. The Kaupinen and Malmi conclusions are based on misrepresenting research, ignoring most of the evidence, correlating things then mixing up what causes what, and using false numbers.

Climate models simplify and apply the laws of physics to calculate Earth’s climate. Include human pollution since 1880 and they show global warming as observed, but if you only include natural changes (e.g. in the Sun and volcanoes) they calculate almost no warming. Kaupinen and Malmi’s article is totally confused and thinks this shows that the models can’t be trusted. It actually shows that if our physics is right then most observed warming is due to human activity.

The biggest single factor is increasing carbon dioxide (CO2) levels in the air. This is a gas that traps infrared heat trying to escape Earth and warms us up. Venus has many times more CO2, and it’s the main reason its surface is hot enough to melt lead. We’ve measured CO2 heating Earth1,2 and Planck’s Law tells us that this would directly warm Earth by about 300% more than assumed by Kaupinen and Malmi3.

Most of their article talks about changes in clouds and humidity. Physics tell us what to expect from clouds, and satellites have measured these changes4,5. We’ve also measured tropical clouds getting higher6, low clouds retreating when it warms7, and changes in ice and liquid mixtures in clouds8.

Newer work shows that cloud changes and how they insulate Earth and reflect sunlight can be calculated and predicted from changing temperature patterns9,10,11,12, i.e., the temperature patterns can mostly explain monthly cloud changes instead of the other way around.

Finally, after using a bunch of nonsense calculations to say that 0.1 °C warming is from CO2, they say that 90% of the change in CO2 is caused by the oceans. This violates conservation of mass from basic chemistry13: the oceans are actually absorbing CO214,15 which, again, is the complete opposite of what Kaupinen and Malmi claim. Without claiming the opposite of reality, their conclusions cannot be supported.

Victor Venema, Scientist, University of Bonn, Germany:
This text may look like a scientific article to a lay-person, but I would not accept it as a bachelor thesis. It does not cite its data sources, it does not discuss the uncertainties in the data, nor does it discuss that other cloud data sets find the opposite trend. It does not explain sufficiently how computations were made to make the study reproducible and understandable. It does not discuss the conflict between its claimed low climate sensitivity and climatic changes in the (deep) past. 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.

Still even if we would grant this work to show that climate models do not give the right estimates of climate sensitivity, it would still not show that the Earth has a low climate sensitivity. There are several independent lines of evidence which give us estimates of the climate sensitivity, only one of which is climate models. A summary of this evidence can be found in the figure below from the last IPCC, which the authors cite.
[Additional comments available as annotations on the original pdf.]
figure showing studies of climate sensitivity based on models, observed temperature change, and paleoclimate, showing most fall between 1.5 and 4.5 C for a doubling of CO2
Estimates of climate sensitivity from studies of different types. Source: IPCC

Stephen Po-Chedley, Research Scientist, Lawrence Livermore National Laboratory:
It’s not clear how to respond to disingenuous summaries of an unpublished paper regarding global warming written by authors that appear to have limited training in climate science. The websites that have promoted this paper provide no counterpoint or basic fact checking on the bold claims made by the authors. The websites mislead readers regarding the well-documented scientific consensus that human activities have made a substantial contribution to the observed warming of the Earth’s surface. The paper itself is flawed: it doesn’t provide sufficient methodological details, including the datasets used in the study, misrepresents basic, well-accepted information about climate change, and ignores research studies undertaken by climate scientists.

The main claim is based on a correlation: that as the Earth warms, low clouds disappear. The authors’ narrative is that low clouds are decreasing due to some natural cause (no mechanism provided by the authors) and the disappearance of low clouds then results in surface warming. This is akin to claiming that increased ice cream sales leads to warmer temperatures. In reality, the feedback is a known and documented phenomenon and works the other way: as the surface of the Earth warms, low cloud coverage decreases, allowing more sunlight to reach and warm the Earth’s surface.

Global temperature datasets, developed by a number of independent research groups, show robust warming in the troposphere and at the Earth’s surface. The radiative effect of carbon dioxide has also been observed1. Considering multiple lines of evidence, the IPCC concluded that it is “extremely likely that human influence has been the dominant cause of the observed warming since the mid-20th century.” More recent analysis of satellite data shows that tropospheric warming from the satellite record is pronounced and cannot be explained by natural climate variability alone2.

Zeke Hausfather, Director of Climate and Energy, The Breakthrough Institute:
[This comment comes from a previous review of a similar claim.]
As we demonstrated in our recent Journal of Climate paper1, you don’t necessarily need internal variability to explain early 20th century warmth; a combination of increasing greenhouse gas concentrations, solar output, and a reduction in volcanic activity can explain most of the observed changes during that period:

graph of modeled and actual temperatures
Source: CarbonBrief

This is an area of active research, and other estimates (e.g. from Hegerl et al, 20182) suggest that natural variability could contribute around 50% of the warming during that period. But no one suggests that early 20th century warmth was solely due to natural variability. Many of the natural factors that played a role in early 20th century warmth, such as increased solar output, have been moving the other direction over the past 50 years. Natural factors alone would have resulted in cooling rather than warming over the past few decades:

graphs of two model simulations: one with human activities and one without
Observed temperature compared to (left) climate model simulations that include human activities and (right) climate model simulations with natural factors only. Source: US NCA

Patrick Brown, Assistant Professor, San Jose State University:
[This comment comes from a previous review of a similar claim.]
Careful analysis that attempts to take into account all major factors and their evolution in time indicates that anthropogenic increases in greenhouse gasses account for more than 100% of the observed warming on the century timescale (requiring cancellation from cooling influences). See the summary graphic from Carbon Brief, below.

graph of factors influencing climate over time
Source: Carbon Brief


Published on: 12 Jul 2019 | Editor:

Climate Feedback is a non-partisan, non-profit organization dedicated to science education. Our reviews are crowdsourced directly from a community of scientists with relevant expertise. We strive to explain whether and why information is or is not consistent with the science and to help readers know which news to trust.
Please get in touch if you have any comment or think there is an important claim or article that would need to be reviewed.