The theory behind modeling weather and climate was developed in the early 20th century, but failed to make any large advances until the introduction of computers in the 1950s. Obtaining realistic predictions from GCMs required computation of large amounts of data, which only became feasible and reliable as computer power increased in the 1960s. As models became more sophisticated, scientists began to project climate data into the future. Early modeling efforts predicted an overall warming of global temperatures with the rise of CO2 levels in the atmosphere. This prediction remains in more advanced models. Today GCMs incorporate data on atmospheric processes, ecosystem attributes, and human activity to make future climate predictions under different projected scenarios.