On this page
Today, global warming is measured by thousands of stations and satellites. In 1938 one man did it with records from 147 stations and no computer. Guy Stewart Callendar was not a professional meteorologist, and most experts did not believe him. This post covers what he found, how his numbers hold up, a worked example from his paper, and what a student can learn from how he worked.
Who was Guy Callendar?
A 2013 note by Ed Hawkins and Phil Jones, On increasing global temperatures: 75 years after Callendar, gives his dates as 1898 to 1964. They call him an amateur climatologist who did much of his research in his spare time without access to a computer, and say he was meticulous; his notebooks are archived at the University of East Anglia. The American Institute of Physics history of the carbon dioxide greenhouse effect, by Spencer Weart, describes him as an English engineer and an expert on steam technology who apparently took up meteorology as a hobby.
What did his 1938 paper claim?
His paper, The artificial production of carbon dioxide and its influence on temperature, appeared in the Quarterly Journal of the Royal Meteorological Society in April 1938. Its summary makes three claims:
Burning fuel had added about 150,000 million tons of carbon dioxide to the air in the past half century, and about three quarters of it had stayed in the atmosphere.
Using the radiation absorption of carbon dioxide and water vapour, he estimated that this should raise mean temperature by about 0.003 °C per year.
Temperature records showed the world had actually warmed by about 0.005 °C per year over the past half century.
He knew he was going against opinion. Hawkins and Jones quote his opening: "Few of those familiar with the natural heat exchanges of the atmosphere, which go into the making of our climates and weather, would be prepared to admit that the activities of man could have any influence upon phenomena of so vast a scale."
How careful was his temperature work?
Hawkins and Jones compared his series with a modern land record, CRUTEM4. The two correlate at r = 0.81 over their common period, and his 1961 update, using many more stations, at r = 0.92. He also checked for the urban heat island effect by splitting city and rural stations, and found "no secular increase of temperature due to city influence", though his samples were small.
Worked example: Callendar's numbers, checked
Step 1: carbon dioxide that stayed. 150,000 million tons × 3/4 = 112,500 million tons, or about 112.5 billion tonnes (treating his tons as tonnes for this estimate).
Step 2: convert to parts per million. NASA's Earth fact sheet gives the atmosphere's mass as 5.1 × 10^18 kg and air's mean molecular weight as 28.97. Moles of air = 5.1 × 10^21 g ÷ 28.97 g/mol ≈ 1.76 × 10^20 mol. One ppm of CO2 = 1.76 × 10^14 mol × 44 g/mol ≈ 7.7 × 10^15 g, or about 7.7 billion tonnes. So 112.5 ÷ 7.7 ≈ 15 ppm added in 50 years, by his figures.
Step 3: compare predicted and observed warming over 50 years. Predicted: 0.003 × 50 = 0.15 °C. Observed: 0.005 × 50 = 0.25 °C. On his own numbers, carbon dioxide explained about 0.15 ÷ 0.25 = 60% of the warming.
For context, the Scripps CO2 Program records that Charles Keeling's analyser at Mauna Loa read 313 ppm on its first day in March 1958; see our post on the Keeling curve.
Why did experts reject him, and were they wrong?
The AIP history lists the objections: old carbon dioxide measurements were unreliable, water vapour was thought to block the same radiation already, and his calculation ignored heat carried upward by convection. Weart adds that later work showed the warming up to 1940 was mainly a natural fluctuation, as his critics suspected. Hawkins and Jones say its causes are still discussed and were probably several.
So Callendar was partly right for partly wrong reasons. His temperature series was excellent. His link to carbon dioxide was ahead of its evidence. He even thought warming would be "beneficial to mankind", delaying "the return of the deadly glaciers".
What mistakes do students make with this history?
Assuming the first person to be right had a complete proof. Callendar's physics was incomplete; the case was finished by later radiation calculations and measurements.
Mixing up carbon and carbon dioxide. 1 tonne of carbon becomes about 3.7 tonnes of CO2 (44 ÷ 12). Check which one a figure means.
Confusing rate with total. 0.005 °C per year sounds tiny until you multiply by 50 years.
What can a student learn from Callendar?
He gathered data himself, tested the obvious alternative, and published a clear, checkable number. Our post on Eunice Foote's experiments with carbon dioxide shows the same idea tested in a lab nearly a century earlier.
If unit conversions like moles to ppm slow you down, Fermi's AI tutor teaches rather than answers, noticing where you are stuck and checking your understanding as it goes.
Try it now: redo Step 2 with your own figure for one year of today's emissions and see what it adds in ppm.
