In my previous article, “How to recognize when science becomes selective”[1], I shared strategies and tools to help evaluate if information is factual, false, or misleading.
Here, I explore the application of critical thinking to evaluate a real-world example that I found on social media, as well as its linked scientific publication.
Scientific articles are often complex as they are aimed at an audience of peers and the broader scientific community. These articles tend to be technical in their use of the language of their disciplines, choice of words, sentence construction and length, and use of acronyms.
For example, a cardiologist’s article intended for colleagues can also be readily understood by other physicians, nurses, cardiology researchers, and those with a strong interest in the subject. However, an average person would need a medical dictionary, some anatomy books, and patience to understand the content.
Because of introduced complexities, obfuscation of an article’s true content can occur. Possible consequences are that few people outside the peer group may be patient enough to wade through such articles, some stop reading due to reader fatigue, and others may back off thinking, “Well, they are the experts, so they must know what they are talking about, ergo it must be true.” End result: fewer consumers of the information ask questions. Consequently, the authors also ask less questions.
Let’s search for and evaluate what is actually communicated to us.
This example falls under the narrative of anthropogenic global warming (AGW), believed by some to be the major driver of the current phase of changing climates. The narrative’s argument is that carbon-dioxide (CO2) emissions, one of many variables affecting our atmosphere and climate, is the dominant – if not, the sole – variable causing warming.
A late-2025 post made on the LinkedIn social media platform about the prediction of heatwaves in a net-zero world was extensively based on modelling.[2]Excerpts from the post made by Australian climate scientist Sarah Perkins-Kirkpatrick follow:
“New research that came out overnight, published in Environmental Research: Climate … We found that delaying net zero will elevate heatwaves for at least the next millennium. Even delaying net zero by 5-10 years (e.g. 2050 instead of 2040) will result in a significantly different heatwave regime, where heatwaves are longer, hotter and occur more often. Also, despite when net zero is reached, heatwaves do not return to pre-industrial conditions … Our conclusions? reaching net zero sooner than 2050 is the best possible way to stop the worst heatwaves from occurring.
But failing that, we now know what is in store for us, and therefore can start the adaptation planning now. Not in 5-10 years, NOW.
It is also worth remembering that net zero is ABSOLUTELY IMPERATIVE—we have no option but to stop emitting greenhouse gases…”
Wow! Basically, if we don’t cut CO2 emissions right away, then we are toast. Literally. The post is laden with opinions with no indication of how the forward-looking postulations or forecasts, described as findings, were made. Along with some use of the caps-lock key to SHOUT OUT the person’s convictions.
If we stop reading at this point, then we risk accepting everything said at face value. Some of us might even repost this and add a few comments without reflecting further, even though a link to the related article, co-authored by this person and published on November 17, 2025, was included.[3]
Reading this post and taking a quick look at the article prompted me to react,
““ … heatwave projections from millennial scale simulations run with the … model…” ???””.
“Models built on models”, or a “house of cards” came to mind. And a lot of red flags jumped off the pages at me. Let’s quickly review the linked article to see what is and is not factual.
A lot of effort seems to have been invested in the “what if” scenarios described in this article, and the methodology used is discussed. Warning: adventurous readers of the entire article should be prepared to wade through very long paragraphs of long sentences and be tolerant of acronyms.
From the abstract:
“While historical and future increases in heatwave frequency, duration and intensity are well documented, no studies have yet examined heatwave changes after anthropogenic greenhouse gas emissions reach net zero. We address this by examining heatwave projections from millennial-scale simulations run with the Australian Earth System Model, ACCESS-ESM1-5. Each simulation branches off the SSP5-8.5 scenario at 5 year intervals between 2030–2060, from which point anthropogenic carbon dioxide emissions are set to net zero.”[4]
Two things jump out, with the first being that a number of 1,000-year-span simulations were run using one model (digging deeper in the article, one finds that 7 simulations were run using this single model). The second:
“historical and future increases in heatwave frequency, duration and intensity are well documented”. (in bold for emphasis)
Historical and future increases are well documented. Yes. You read that correctly.
How can future events be well documented if they have not yet happened?
For that matter, a similarly questionable statement is near the end of the Results section in the article (excerpt):
“While there are multiple physical mechanisms that initiate and sustain heatwave development …, the main driver of historical and future increases in heatwaves is anthropogenic climate change due to increasing greenhouse gas emissions (more bold text) …”
We can debate about what is “historical” (i.e., the 165-year period from 1850 through 2014 per the article as opposed to the entire ~4.6-billion-year age of our planet). Like when the authors say that anthropogenic climate change due to increasing greenhouse gas emissions was the main driver for historical heatwaves, they seem to ignore the well-known heatwaves that happened in North America during the 1930s, but I digress.
Regarding the future, it is fundamentally misleading to state that “future increases … are well documented” and that the main driver of “future increases in heatwaves is anthropogenic climate change…” Barring the emergence of an accurate form of soothsaying or tarot-card reading, of course.
The published article went through a vetting process, of which some elements were peer review and revision in 2025, between its receipt by the publisher on February 22 and when it was published on November 17. Amazingly, these statements made it through that process. How could this be?
Excerpt from the study’s conclusions:
“While we acknowledge that this assessment is based on just one climate model, it provides important insight that heatwaves could largely stabilise for at least a millennium based on their state once net zero is finally reached, should sustained and substantial net negative emissions be challenging to achieve.”[5]
The authors acknowledge the use of one model as the basis of their many simulations, and they defend the usefulness of their effort mentioning that it provides insight about something that could happen, subject to a condition. Remember that if a conditional verb like “could” is used, then it is valid to use the word “not” so it would read as “could not”. It is not cut and dried.
Let’s investigate some words
We already have enough information to see that something is seriously wrong. We can dig deeper by doing simple searches for the non-factual words listed in my first article under What Are and Are Not Facts.[6]
Using Acrobat Reader, I accessed the pdf version of the article and used the search function for the “not a fact” words[7] in the article’s main text, then counted the number of times these were used. Some examples:
Some interesting statistics showing many instances where these were used.
I then reviewed where these were used, of which many are in the Results section, to see how they may have contributed to any conclusions made. In many cases, definitive verbs were used instead of more appropriate qualitative wording such as “suggest” or “indicate” or “infer” or “could be”. For instance, in the article:
“For example, should net zero occur by 2040 (figure 3(c)), the average seasonal number of heatwave days around the middle of the millennium is still considerably less than delaying net zero to 2060.”[8]
Since the text refers to results from simulations, more accurate wording would be (in bold):
“For example, should net zero occur by 2040 (figure 3(c)), the simulations suggest that the average seasonal number of heatwave days around the middle of the millennium could be considerably less than delaying net zero to 2060.”
This is basic stuff that many of us learned when we were at university. As a young undergraduate geology student, I was taught:
- If something is proven, then use definitive words that convey this
- If something is unproven, then use words that properly convey the uncertainty of the statement
The scenarios in the article were often referred to as experiments. Really, the article summarizes a collection of “what if” scenarios. Models and scenarios are not experiments that test hypotheses.
There are fundamental differences between facts and models:
- A fact is a thing that is known to be true, especially when it is proved to be true[9]
- A model is a simplified construct, equation, or simulation designed by individuals, according to their criteria, in an attempt to mimic a system or process
Results from models are not facts, models are not facts, and models do not turn into facts.
The broader view and a cautionary alert
The model used by the authors was developed for Australia to participate in a larger international grouping of models called the Coupled Model Intercomparison Project Phase 6, for which there are presently 43 modelling centres worldwide that contribute to an archive of various climate models and historical simulations. The information from this archive provides the foundation for the Intergovernmental Panel on Climate Change’s (IPCC) Sixth Assessment Report,[10][11][12][13] which in turn can influence public policy because this panel is a United Nations body.
Of note are the words “models” and “simulations”, and that the model was developed to participate in the Coupled Model Intercomparison Project Phase 6. One must be aware and exercise caution because models and their ilk are moving targets, are evolving, and are not facts.
What this means
Serious concerns were raised from this quick evaluation of a social media post and its accompanying scientific article in an area of strong public interest:
- It became clear early in the review that something was seriously wrong.
- The study in the article was based on a model, upon which seven model runs (scenarios) were done.
- Results from models are not facts, models are not facts, and models do not turn into facts.
- Like a house of cards: if the initial model is proved wrong, then the base of the entire structure is weak, and the structure could easily collapse. Even if the initial model is correct, if any of the derivative models are wrong, the structure can still
- It is very disturbing that this article made it though a vetting process, in part peer review, “as-is”, most notably in unequivocally stating that future events, that have not yet happened, are well documented.
- Despite this, some definitive statements, including some expressing strong personal convictions, were made.
My biggest concern is that studies such as this may create bias as they risk being incorporated into larger ensembles, as was the Coupled Model Intercomparison Project in this case. The latter project is the foundation for reports completed by a United Nations body, the Intergovernmental Panel on Climate Change. Moreover, in this context, this could ultimately influence public policy worldwide, including investment and taxation, and could create financial hardship for individuals.
I am also worried that, over time, people may ignore that the Coupled Model Intercomparison Project is largely composed of models rather than facts. Because of this, future users of this information could refer to this as a database of factual information rather than the collection of models that it dominantly is, if this is not already happening.
References:
[1] How to recognize when science becomes selective, BIG Media
[2] Sarah Perkins-Kirkpatrick, Professor, Climate Science, Co-host Totally Cooked Podcast. The reader will need to be a LinkedIn user (free) for access.
[3] How to recognize when science becomes selective, BIG Media
[4] Heatwaves in a net zero World, S. Perkins-Kirkpatrick*, L. Palmer, A. King and T. Ziehn, Environmental Research, Climate, IOP
[5] Heatwaves in a net zero World, Sarah Perkins-Kirkpatrick*, Lucinda Palmer, Andrew King and Tilo Ziehn, Environmental Research, Climate, IOP
[6] How to recognize when science becomes selective, BIG Media
[7] How to recognize when science becomes selective, BIG Media
[8] Heatwaves in a net zero World, Sarah Perkins-Kirkpatrick*, Lucinda Palmer, Andrew King and Tilo Ziehn, Environmental Research, Climate, IOP
[9] Fact: Oxford Learner’s Dictionaries
[10] The Australian Earth System Model: ACCESS-ESM1.5, T. Ziehn, M.A. Chamberlain, R. M. Law, A. Lenton, R.W. Bodman, M. Dix, L. Stevens, Y-P Wang, J. Srbinovsky, Journal of Southern Hemisphere – Earth Systems Science
[11] Coupled Model Intercomparison Project
[12] CMIP6 Climate Projections Overview, Emergent Mind
[13] CMIP6 Frequently Asked Questions, Climate Data Canada
Editing provided by Patricia Scarr
(Peter Dorrins – BIG Media Ltd., 2026)
