Climate models

Scientists have studied global warming with computer models of the climate. These models are based on physical principles of fluid dynamics, radiative transfer, and other processes, with simplifications being necessary because of limitations in computer power and the complexity of the climate system. All modern climate models include an atmospheric model that is coupled to an ocean model and models for ice cover on land and sea. Some models also include treatments of chemical and biological processes. These models project a warmer climate due to increasing levels of greenhouse gases.However, even when the same assumptions of future greenhouse gas levels are used, there still remains a considerable range of climate sensitivity.

Including uncertainties in future greenhouse gas concentrations and climate modeling, the IPCC anticipates a warming of 1.1 °C to 6.4 °C (2.0 °F to 11.5 °F) by the end of the 21st century, relative to 1980–1999. Models have also been used to help investigate the causes of recent climate change by comparing the observed changes to those that the models project from various natural and human-derived causes.

Current climate models produce a good match to observations of global temperature changes over the last century, but do not simulate all aspects of climate. These models do not unambiguously attribute the warming that occurred from approximately 1910 to 1945 to either natural variation or human effects; however, they suggest that the warming since 1975 is dominated by man-made greenhouse gas emissions.

Global climate model projections of future climate are forced by imposed greenhouse gas emission scenarios, most often from the IPCC Special Report on Emissions Scenarios (SRES). Less commonly, models may also include a simulation of the carbon cycle; this generally shows a positive feedback, though this response is uncertain (under the A2 SRES scenario, responses vary between an extra 20 and 200 ppm of CO₂). Some observational studies also show a positive feedback.

In May 2008, it was predicted that "global surface temperature may not increase over the next decade, as natural climate variations in the North Atlantic and tropical Pacific temporarily offset the projected anthropogenic warming", based on the inclusion of ocean temperature observations.

The representation of clouds is one of the main sources of uncertainty in present-generation models, though progress is being made on this problem.

A minor issue in climate modeling is the perceived mismatch between actual conditions and those projected by the models. A 2007 study by David Douglass and colleagues compared the composite output of 22 leading global climate models with actual climate data and found that the models did not accurately project observed changes to the temperature profile in the tropical troposphere. The authors note that their conclusions contrast strongly with those of recent publications based on essentially the same data.^[92] A 2008 paper published by a 17-member team led by Ben Santer of Lawrence Livermore National Laboratory noted serious mathematical flaws in the Douglass study, and found instead that deviations between the models and observations were statistically insignificant.