Model Parameters

Carbon Cycle Model

This button allows the user to examine non-climate-related uncertainties in the carbon cycle model by changing the 1980s-mean value of net land-use change CO2 emissions. The default value for this variable is 1.1GtC (Gigatonnes of Carbon) per year, which is the IPCC Second Assessment Report (SAR) best estimate. There is considerable uncertainty, however, about the true value of these land-use change emissions. The IPCC SAR uncertainty range is from 0.4 to 1.8GtC per year. Choosing a low value leads to higher concentrations and vice versa. MAGICC runs the carbon cycle model three times,each with a different Dn80s value: 1.8, 1.1 and 0.4. These settings, and the associated CO2-fertilisation factor, determine the high and low ends of the range for the MAGICC display CO2 Concentration Graphs. It should be noted that the IPCC Third Assessment Report (TAR) suggests slightly different values for the above parameters. The MAGICC carbon cycle model, however, when climate feedbacks are included (the default case), gives CO2 concentrations that agree well with the IPCC TAR values (see Technical Report).

C-Cycle Climate Feedbacks

An important change from the SAR to the TAR was the inclusion and quantification of the effects of climate feedbacks on the carbon cycle. The net effect of these is a positive feedback, so their inclusion leads to higher concentrations than would otherwise be obtained. There is considerable uncertainty in the value of these feedbacks, and each of the three carbon cycle models used in the IPCC TAR [the Joos et al. (Bern), Jain et al.(ISAM) and MAGICC models] have different values. MAGICC uses feedback values that give a net climate feedback that is somewhat lower than the values from the other two models. These feedbacks can be turned on or off (default is on) so that users can judge the importance of the MAGICC climate feedbacks.

Aerosol Forcing

There are four components to the aerosol forcing used by MAGICC. All four components are subject to large uncertainties. The default values in MAGICC are those used in the IPCC TAR, but the user can select, as alternatives, high or low values for aerosol forcing.

Direct: this forcing is the direct, clear-sky effect of sulfate aerosols formed from fossil fuel combustion, i.e., aerosols derived from SO2 emissions. MAGICC uses a default current (1990) forcing estimate of -0.4Wm-2.

Indirect: the indirect forcing effect of aerosols is more uncertain than the direct effect. MAGICC uses a default value for the current indirect forcing of -0.8Wm-2.

Biospheric: this forcing is due to aerosols emitted from biomass burning, the sum of both organic and carbonaceous aerosol forcings. The default for the current value of this forcing component is -0.2Wm-2. Forcing is assumed to run parallel to gross land-use change emissions. (Note that, in the emissions scenarios, it is net land-use change emissions that are prescribed. Net emissions are gross emissions less regrowth. The MAGICC carbon cycle model calculates gross land-use change emissions internally.)

FOC: this acronym stands for Fossil plus Organic Carbonaceous aerosol forcing (often called 'soot'). The default for the current value of this forcing component is 0.1Wm-2, with forcing scaled to run parallel to SO2 emissions.

The separate components of aerosol forcing cannot be altered by the user. Instead, as a guide to the importance of aerosol forcing uncertainties, the user can choose high, default (mid) or low values of total aerosol forcing. The default total forcing in 1990 is -1.3Wm-2. The low and high values are -0.8Wm-2 and -1.8Wm-2.

Climate Sensitivity

The climate sensitivity defines the equilibrium response of global-mean surface air temperature to a doubling of the CO2 concentration. The IPCC SAR best-estimate of this parameter was 2.5°C, with a range from 1.5° to 4.5°C. For the TAR projections a central value of 2.6°C was used (the median for the seven AOGCMs whose results were used to calibrate MAGICC -- see below). In default mode MAGICC runs each emissions scenario three times through the climate model, each time with a different value for the climate sensitivity: 1.5°, 2.6° and 4.5°C. These three runs define the low and the high ends of the range of global temperature projections (approximately the 90% confidence interval) displayed in the Temperature Graphs.

It is possible to enter a user-defined value for the climate sensitivity. Click on the relevant box and enter your own value. This choice will now form part of your parameter selection for the User Model through which MAGICC also runs the two emissions scenarios. The recommended range from which to select the climate sensitivity is 0.5° to 5.5°C.

Thermohaline Circulation (THC)

An important determinant of temperature changes and oceanic thermal expansion is how the thermohaline circulation changes. The default case corresponds to a moderate slow-down of the THC as the globe warms, at a rate equal to the median of THC change results for the seven AOGCMs whose results were used to calibrate MAGICC. To illustrate the influence this slow-down has on temperature and sea level, the user can select either the default (varying THC) mode, or a constant THC.

Vertical Diffusivity (Kz)

Another important determinant of temperature changes and oceanic thermal expansion is how the speed with which oceanic mixing transports heat from the surface into the deeper ocean, parameterized in MAGICC by the vertical diffusivity. The default case (2.3cm2/s) corresponds to the median value for effective diffusivity for the seven AOGCMs whose results were used to calibrate MAGICC. To quantify the influence this choice has on temperature and sea level, the user can edit this value.

Ice Melt

Sea level rise is determined by oceanic thermal expansion, and ice melt and other water balance terms. In the TAR, uncertainty ranges were estimated for all non-expansion terms. The default in MAGICC uses all of the TAR central estimates. The user can choose to use, instead, either the high or low extremes (concatenating the two-sigma high and low limits for all non-expansion terms).

Model

In the TAR, global-mean temperature and sea level rise uncertainties were determined by simulating the results of seven AOGCMs. MAGICC was first calibrated to match the temperature and expansion outputs for these seven models forced with 1%/yr compound CO2 changes (from the CMIP2 data base). MAGICC was then run for each of the seven parameter sets for the full range of SRES emissions scenarios (an exercise that would have been computationally impossible to carry out using the original AOGCMs). The 'Model' button allows the user to simulate global-mean temperature and sea level changes for any of the seven models used in the TAR: GFDL, CSIRO, HadCM3, HadCM2, ECH4-OPYC, PCM, or CSM. The correct MAGICC parameter values and non-expansion sea-level model parameters to do this are hard-wired into the MAGICC code.