Computer Climate Models
Because there are so many possibilities for change, climatologists must know how clouds over the entire Earth will respond. Determining that response calls for computer models of the global climate that can explore changing conditions several decades into the future.
Climate models are sets of mathematical equations that describe the properties of Earth's atmosphere at descrete positions and times, along with the ways such properties can change. The challenge for climate models is to account for the most important physical processes and their complex interactions accurately enough to carry climatic predictions tens of years into the future. When contemporary models are given information about Earth's present condition - the size, shape and topography of the continents; the composition of the atmosphere; the amount of sunlight striking the globe - they create artificial climates that mathematically resemble the real one. They can forecast winds, rain, temperature, cloud cover and radiation balance to within 10 to 15 percent of the magnitude of their real-world counterparts.
Unfortunately, that margin of error is too large for making a reliable forecast about global warming. A doubling in atmospheric carbon dioxide (CO2), predicted to take place in the next 50 to 100 years, is expected to change the radiation balance at the surface by only about 2 percent. Yet according to current climate models, even such a small change could raise global temperatures between 2-5°C (4-9°F), with potentially dramatic consequences. If a 2 percent change makes that much difference, a climate model must be accurate to within at least half a percent to be useful. Thus today's models must be improved more than tenfold in accuracy, requiring much more and much better data for developing a better understanding of clouds.