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

Key takeaway

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.

Reviewed content

Verdict:

Incorrect

Claim:

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.

Verdict detail

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.

See our method of evaluation

Full 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 This means that this article has not been peer-reviewed, so there is no guarantee to its credibility.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.

• 1- Cawley (2011) On the Atmospheric Residence Time of Anthropogenically Sourced Carbon Dioxide,
  Energy & Fuels
• 2- Haustein et al (2017) A real-time Global Warming Index, Scientific Reports

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.

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.

• 1- Eastman et al (2011) Variations in Cloud Cover and Cloud Types over the Ocean from Surface Observations, 1954–2008, Journal of Climate
• 2- IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Chapter 10

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.

• 1- Boucher et al (2013) Clouds and aerosols. In Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
• 2- Matthews (2000) Storks deliver babies (p= 0.008), Teaching Statistics

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 Earth^1,2 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 changes^4,5. 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 patterns^9,10,11,12, 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 CO₂^14,15 which, again, is the complete opposite of what Kaupinen and Malmi claim. Without claiming the opposite of reality, their conclusions cannot be supported.

• 1- Feldman et al (2015) Observational determination of surface radiative forcing by CO₂ from 2000 to 2010, Nature
• 2- Harries et al (2001) Increases in greenhouse forcing inferred from the outgoing longwave radiation spectra of the Earth in 1970 and 1997, Nature
• 3- Soden et al (2008) Quantifying Climate Feedbacks Using Radiative Kernels, Journal of Climate
• 4- Yue et al (2019) Temporal and Spatial Characteristics of Short-term Cloud Feedback on Global and Local Interannual Climate Fluctuations from A-Train Observations, Journal of Climate
• 5- Zelinka et al (2016) Insights from a refined decomposition of cloud feedbacks, Geophysical Research Letters
• 6- Norris et al (2016) Evidence for climate change in the satellite cloud record, Nature
• 7- Brient and Schneider (2016) Constraints on Climate Sensitivity from Space-Based Measurements of Low-Cloud Reflection, Journal of Climate
• 8- Tan et al (2016) Observational constraints on mixed-phase clouds imply higher climate sensitivity, Science
• 9- Andrews et al (2018) Accounting for Changing Temperature Patterns Increases Historical Estimates of Climate Sensitivity, Geophysical Research Letters
• 10- Dong et al (2019) Attributing Historical and Future Evolution of Radiative Feedbacks to Regional Warming Patterns using a Green’s Function Approach: The Preeminence of the Western Pacific, Journal of Climate
• 11- Silvers et al (2018) The Diversity of Cloud Responses to Twentieth Century Sea Surface Temperatures, Geophysical Research Letters
• 12- Zhou et al (2016) Impact of decadal cloud variations on the Earth’s energy budget, Nature Geoscience
• 13- Richardson (2013) Comment on “The phase relation between atmospheric carbon dioxide and global temperature” by Humlum, Stordahl and Solheim, Global and Planetary Change
• 14- Hartfield et al (2018) State of the Climate in 2017, Bulletin of the American Meteorological Society
• 15- Lauvset et al (2015) Trends and drivers in global surface ocean pH over the past 3 decades, Biogeosciences

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 Hypothes.is annotations on the original pdf.]

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.

• 1-Feldman et al (2015) Observational determination of surface radiative forcing by CO₂ from 2000 to 2010, Nature
• 2-Santer et al (2017) Tropospheric Warming Over The Past Two Decades, Scientific Reports

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:

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:

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.