 These days, scientists used the world's fastest supercomputers to model the Earth's climate. But the first numerical model of the Earth was created over 100 years ago. It was done using a pen and paper by Swedish scientist Fonte Arranius. How good was this first attempt? Arranius performed a complex calculation and estimated that if the amount of carbon dioxide in the atmosphere were to double, the planet would warm about 5 to 6 degrees Celsius as a result. That result was remarkably close to estimates from today's complex global climate models. They find that a doubling of the amount of carbon dioxide in the atmosphere will warm the planet somewhere between 2 and 4.5 degrees Celsius. Arranius also predicted that if the amount of carbon dioxide in the atmosphere rose and warmed the planet, the atmosphere would be able to hold more water vapor. Water vapor is another greenhouse gas, so that would amplify global warming even further. By 1967, scientists Manabe and Weatherald had created a three-dimensional model of the Earth's atmosphere. They used that model to calculate that a doubling of the amount of carbon dioxide in the atmosphere would cause about 1.3 degrees Celsius warming by itself. That warming would increase the amount of water vapor in the atmosphere, which would cause another 1.1 degrees Celsius warming of global surface temperatures. Recent research has found that as the planet is warmed, the amount of water vapor in the atmosphere is increased as anticipated by climate models and by scientists in Sponte Arranius in the late 1800s. Global climate models have done well in projecting global temperature changes, too. In 1972, British meteorologist J. S. Sawyer projected that humans would increase the amount of carbon dioxide in the atmosphere by 25 percent above the 1850 levels by the year 2000. He estimated that the planet would warm about 0.6 degrees Celsius in response. Both projections were almost spot-on. In 1975, American climate scientist Wallace Broker coined the term global warming. He used a global climate model to estimate that due to rising carbon dioxide levels, between 1975 and 2015, the Earth's surface temperature would warm by about 1 degrees Celsius. His estimate was about 0.3 degrees Celsius too high, but that's pretty close for a simple climate model that he was using 40 years ago. In 1981, NASA climate scientist James Hansen had developed a somewhat more detailed model of the Earth's climates. He projected that between 1981 and 2015, the planet would warm about 0.5 degrees Celsius. In actuality, the planet warmed about 0.6 degrees Celsius during that time, so Hansen's project model was remarkably close with that projection. Hansen made another global warming projection in 1988 using a newer version of his climate model. However, this newer model was more sensitive to the increased greenhouse effect that is 1981 model and more sensitive than most of today's global climate models. Hansen projected that between 1988 and 2015, the Earth's average surface temperature would warm about 0.67 degrees Celsius, whereas it actually warmed about 0.5 degrees Celsius during that time. The Intergovernmental Panel on Climate Change has also used simulations from the most advanced climate models available to project global temperature changes in each of its reports. Those reports were published in 1990, 1995, 2001, 2007, and 2014. In each one, the global warming projections made by the climate models have turned out to be remarkably accurate. Climate models have made many more accurate projections too. Manava and weather held accurately predicted that while the lower atmosphere warmed, the upper atmosphere would cool. In 1975, they also predicted that the Arctic would warm faster than the rest of the planet, in part because of the decreased reflectivity due to melting ice. In 1989, Stofer, Manava, and Bryan predicted that the land surface would warm faster than the ocean surface. Models have predicted the geographic pattern of global warming, the loss of Arctic sea ice, the rising of sea levels, and so on. All of these predictions have come true, confirming that climate models are good representations of the Earth's climates. There's a common myth that climate models are useless. This myth stems from the fallacy of impossible expectations. Many people argue that climate models must be able to make perfect short-term predictions for us to rely on them. However, climate models actually do better in predicting long-term changes. That's because there are a lot of unpredictable factors like ocean and solar cycles that can have significant short-term influences on the climate. Sometimes they have short-term warming effects and sometimes cooling. Over the long-term, these effects average out, leaving long-term effects like the increased greenhouse effect to dominate climate change. That makes the long-term changes easier for climate models to predict. As climate scientists Tom Knutson and Robert Tollea said in 2005, if we had observations of the future, we obviously would trust them more than climate models. But unfortunately, observations of the future are not available at this time. There's no way to predict the future without a model. Like all models, global climate models are imperfect, but they've made some very accurate predictions for decades. On the other hand, predictions made by climate contrarians have been way off. They've mostly predicted global cooling while we've actually experienced global warming. That's a poor track record, not nearly as accurate as predictions made by climate models.