In 1896 Svante Arrhenius, a Swedish chemist, theorized that increasing or reducing the amount of carbon dioxide in the atmosphere might trigger a change in climate worldwide. Cutting the amount in half, he surmised, could lower average temperatures by 5 degrees C and might bring on an ice age. But, was such a big change possible?
A colleague, Arvid Hogbom, had studied the carbon cycle extensively, calculating the amounts of carbon from various sources including those from industrial emissions. Using Hogbom's numbers Arrhenius calculated that a doubling of carbon dioxide would increase global temperature by 5 degrees C, surprisingly close to modern estimates even though he was working under severe handicaps including relatively low calculating power--he used paper and pencil--and limited data.
He also estimated that it would take 2,000 years to get there. (The latest estimates place this event in the middle of the current century.) Of course, Arrhenius shouldn't be blamed for this underestimate given the impossibility of knowing what lay ahead for population growth and industrialization. But, his was the first in what has turned out to be a consistent string of underestimates concerning the pace and severity of global warming.
In 1938 Guy Stewart Callendar, a British engineer and amateur meteorologist, presented data to the Royal Meteorological Society demonstrating the arrival of global warming. He estimated that temperatures would rise by 1 degree C by 2200. Callendar wasn't very concerned about global warming for two reasons. The pace seemed so slow and he believed that increased levels of carbon dioxide would be good for agriculture.
Neither Arrhenius nor Callendar were well received. Climate theory and experimental evidence at the time seemed to demonstrate conclusively that what they were claiming could not happen.
Starting in 1940 the Earth embarked on a long cooling spell that lasted until 1970. This temporary hiatus in global warming led to a debate in the scientific community in the early 1970s about whether the Earth might be headed into a new ice age and not toward any continued warming. (It turned out later that the cooling trend may have been in part due to the huge amount of air pollution produced by the wartime and then postwar industrial boom, pollution which tended to reflect sunlight back into outer space.)
In 1979 the National Academy of Sciences offered an estimate that doubling of carbon dioxide levels in the atmosphere might lead to warming between 1.5 degrees C and 4.5 degrees C. In 2001 the Intergovernmental Panel on Climate Change (IPCC) estimated that the Earth could warm by between 1.4 degrees C and 5.8 degrees C by 2100. But, these estimates now seem too conservative. What we know today about the masking effects of particulate air pollution, our subsequent successful efforts to reduce them, and the rapid melting of polar ice indicates that the top end of the IPCC estimates will likely have to be raised, perhaps to somewhere between 7 degrees C to 11 degrees C.
Scientists are by nature inclined to claim only what their data and their models allow them to claim. When they get more data, they simply modify their models and plug in their new data. What the history of global warming modelling tells us so far is that we should be prepared for more surprises--on the upside.
(Some of the above information was taken from a delightful and lively history of global warming entitled, "The Discovery of Global Warming." Supplemental material regarding the book is available online in searchable form here.)