Vensim® PLE Equations for System Dynamics Climate Model

(01)CO2 emissions into Atmosphere=User Action CO2 emissions

Units: TonC/Year [0,1e+012]

(02)CO2 in Atmos= INTEG (CO2 emissions into Atmosphere-CO2 removals,

7.69e+011)

Units: TonC

(03)CO2 removals=(CO2 in Atmos-Preindustrial CO2)*Rate of CO2 Transfer

Units: TonC/Year

(04)FINAL TIME = 2100

Units: Year

The final time for the simulation.

(05)INITIAL TIME = 2000

Units: Year

The initial time for the simulation.

(06)Preindustrial CO2=6.77e+011

Units: TonC

(07)Rate of CO2 Transfer=0.012

Units: 1/Year

Rate of Storage of Atmospheric Greenhouse Gases [delta-m] (1/year) Inverse yields average residence time of gases (120 years).

(08)SAVEPER = 5

Units: Year [0,?]

The frequency with which output is stored.

(09)TIME STEP = 10

Units: Year [0,?]

The time step for the simulation.

(10)User Action CO2 emissions

Units: TonC/Year

We define different user inflow actions here (SRES Scenarios A2 and B1 were defined here for our model fits)

ISAM, SRES and Calibration details

As per (Jain et al. 1994), ISAM is a model that takes projections of human emissions of CO2 and other greenhouse gases and of atmospheric particulates and generates predictions of future greenhouse gas and aerosol concentrations, global climate change, and the impacts of climate change such as the expected rise in sea level. As per (Nakicenovic et al. 2000), the SRES take into account human activity for the future CO2 emissions. SRES are made so that climate models like ISAM can be simulated for such emission scenarios and formulate predictions of future concentrations of GHGs.

To calibrate our model with these scenarios, we converted the IPCC’s parts-per million by volume (ppmv) of CO2 concentration data into GtC units by using a factor of 2.083 GtC per ppmv (as per the method given in Oak Ridge National Laboratory 1990). We wanted to find the values of Rate of CO2 Transfer for the two extreme IPCC emission scenarios, given the best calibrationbetween our model values for CO2 in Atmos and the predictions of CO2 in Atmos from the ISAM climate model on the same emission scenarios. For the purpose of generating the calibrations,we simulated the climate model between years 2001 and 2100 and compared, using R2 and RMSE measures, the values of CO2 in Atmos predicted from our model to those predicted in the same time period from the ISAM model.

Method for Calculation of From and To Ranges in DCCS

For the calculation of the range of values, i.e.,From value and To value as required values of emissions for the Fossil Fuel Emissions (GtC/Year) and Deforestation Emissions(GtC/Year),we first categorized differentSpecial Report Emission Scenarios (SRES; Jain et al. 1994) as for better and for worseand this was done separately for Fossil Fuel Emissions and Deforestation Emissions. A for better scenario meantthat the yearly emissionsinto the atmosphere in that scenario had one of the lowest values compared to all other scenarios in year 2100. Similarly, for worse scenario meantthat the yearly emissionsinto the atmosphere in that scenario had one of the highest emission values compared to all other scenarios in year 2100.

Three lowest emission scenarios that had the lowest deforestation emissions in year 2100 (i.e., for better) were A1Fl, B1 and A1p; and, three emission scenarios that had the highest deforestation emissions in year 2100 (i.e., for worse) were A1B, A2p, B1p (Houghton et al. 2001). Similarly, three lowest emissionscenarios that had the lowest fossil-fuel emissions in year 2100 (i.e., for better) were A1T, B1 and B1p; and, three emission scenarios that had the highest fossil-fuel emissions in year 2100 (i.e., for worse) were A1Fl, A2 and A2p (Houghton et al. 2001).

We then found the average yearly percentage change in fossil-fuel and deforestation emissions between years 2000 and 2100 under for better and for worse scenarios (done separately for both kinds of scenarios). This yearly percentage change was defined as (new year emissions – old year emissions)/(old year emissions)*100). For example, for year 2000 to 2010 the average percentage change for fossil-fuel emissions was (((8.3-6.9)/6.9)+((8.5-6.9)/6.9)+((7.7-6.8)/6.8))/3*100. Similarly, average percentage changes were found for year 2010 to 2020 etc. till 2090 to 2100 for the 3 for better and the 3 for worse fossil-fuel and deforestation emissions. Among thepercentage values so computed we then picked the five lowest percentage change values each from the for better and for worst groups (done separately for both deforestation and fossil-fuel emissions). We then averaged the five lowest percentage values in each group.

Similarly, we picked the five highest percentage change values each from the for better and for worst groups and averaged the five highest percentage values in each group. Thus, in the end, this averaging exercise gave us two high averaged values of percentage change (one from the for better and the other from the for worse group) and two low averaged values of percentage change for emissions (again from the for better and for worse group) for both-fossil fuel and deforestation emission types (a total of 8 values for both kinds of emissions).We picked the maximum of the two high values as the To value and the minimum of the two low values as the From value done separately for the deforestation emissions and fossil-fuel emissions.

Participant instructions for slow-low Dynamic Condition

The change in text for instructions given in the rapid-high dynamic condition was the following: making the end year as 2100 (in place of 2200); the frequency of emission decisions to be every 2 years (in place of every 4 years); and, the absorptions to be 1.6% of the difference between the current CO2 concentration and the preindustrial CO2 concentration (in place of 1.2%).

1