...2.2 The Role of Emissions Scenarios
Because climate change is driven by emissions, a model run requires a detailed emissions scenario to be specified for the entire simulation period. The simulation will typically begin in the pre-industrial past, and use estimates of historical emissions and land-use change to "spin-up"
the model to the present-day climate. Future emissions are handled differently depending on whether the modeling goal is to perform an equilibrium experiment or a transient experiment. In equilibrium experiments, a specified change to some atmospheric constituent (e.g., a doubling of
CO2) is imposed and the model is run until the system comes to a new equilibrium. This type of experiment is used to quantify the climate forcing associated with different processes. It has also been used to estimate the widely quoted climate sensitivity, which is defined as the mean global
temperature increase associated with a doubling of CO2 from pre-industrial levels. The IPCC cites a best estimate for climate sensitivity of 3 degrees Celsius (°C) with a likely range of 2 to 4.5 °C. The multi-model distribution is asymmetric, so much higher values are possible, but much lower values are very unlikely [38]. Climate sensitivity is a measure of the global average
response of the climate to a set of forcings, and the computed ranges on sensitivity are not predictions of what the temperature change will be in the future.
Transient experiments are intended to mimic the more realistic situation of continuously changing emissions. In a typical set-up, CO2 concentrations are increased by 1% per year. The transient climate response (TCR), which is defined as the global average temperature change at the time of CO2 doubling, is used to quantify the change to the climate. This number can also be used as a model inter-comparison metric. There is less spread in predictions of TCR than for climate sensitivity, due in part to the fact that it is more constrained by observational data [10].
Edited by Gaspode:
Edited for rule 4.