climate (15_climate)

Description

The 15_climate module takes emissions and simulates the resulting climate variables (forcings, global mean temperature) using the MAGICC climate emulator. It may also be used within the optimization or between iterations in order to internalize climate damages or adjust the carbon price to meet a desired climate target

Interfaces

Interface plot missing!

Input

Output

Realizations

(A) magicc7_ar6

In this realization, concentration, forcing, and temperature values are calculated using a version of the MAGICC climate emulator. MAGICC has to be installed in a separate folder, which can be set in the magicc_template switch. MAGICC version 6.4 is used as default, but other versions with compatible path structure and variable names can be used as well as long as it has the proper interfacing R scripts.

The actual generation of MAGICC input data from REMIND emissions is not handled here, but in the core. Emissions that are not calculated by REMIND, such as F-gases, are taken from RCP trajectories.

With this module activated, MAGICC is run between iterations one time or several, depending on specific switches. By default, the global mean temperature change is simply read and passed on to other modules. Depending on the choice of carbon tax adjustment in cm_emiscen, the total carbon budget will also be adjusted based on the MAGICC radiation forcing or temperature outcomes between iterations. This can allow the carbon tax, which is adjusted based on the carbon budget, to be optimized for a given temperature target. If cm_magicc_temperatureImpulseResponse is on, a new Temperature Impulse Response Function (TIRF) is also generated between iterations. The TIRF is a smooth function that estimates the effect of an additional unit of emissions in global temperature, and it’s used by some damage modules that internalize climate damages (i.e. account for damages within the optimization). To derive a TIRF, MAGICC is run several times with pulse emissions added around the the emissions of that iteration.

Execute_unload 'fulldata_postsolve';
Execute "Rscript climate_assessment_run.R";
Execute_Loadpoint 'p15_forc_magicc'  p15_forc_magicc;
Execute_Loadpoint 'p15_magicc_temp' pm_globalMeanTemperature = pm_globalMeanTemperature;
pm_globalMeanTemperature(tall)$(tall.val gt 2100) = pm_globalMeanTemperature("2100");

Raw temperature (GMT anomaly) from MAGICC is further calibrated to match HADCRUT4 in 2000. This ensures that different MAGICC configurations start at the same observed temperature.

$ifthen.cm_magicc_calibrateTemperature2000 %cm_magicc_calibrateTemperature2000% == "HADCRUT4"

Calibrate temperature such that anomaly in 2006-2015 reference period is 0.97 (SR1.5 Table 2.2, footnote 1)

s15_tempOffset2010 = sum(tall$(tall.val gt 2005 and tall.val le 2015),pm_globalMeanTemperature(tall))/10; 
display s15_tempOffset2010;
pm_globalMeanTemperature(tall) = pm_globalMeanTemperature(tall) - s15_tempOffset2010 + 0.97;
display pm_globalMeanTemperature;
pm_gmt_conv = 100*smax(t,abs(pm_globalMeanTemperature(t)/max(p15_gmt0(t),1e-8) -1));
display pm_gmt_conv;
p15_gmt0(tall) = pm_globalMeanTemperature(tall);
$endif.cm_magicc_calibrateTemperature2000
pm_globalMeanTemperatureZeroed1900(tall)  = pm_globalMeanTemperature(tall) + 0.092; 

TIRF derivation

Derive temperature impulse response (TIRF) from MAGICC pulse scenarios

$ifthen.cm_magicc_tirf "%cm_magicc_temperatureImpulseResponse%" == "on"
if( ((iteration.val le 10) or ( mod(iteration.val,5 ) eq 0)) ,
    execute "Rscript climate_assessment_temperatureImpulseResponse.R";
    execute_loadpoint 'pm_magicc_temperatureImpulseResponse'  pm_temperatureImpulseResponseCO2 = pm_temperatureImpulseResponse;
);
$endif.cm_magicc_tirf

Iterative adjustment of budgets or carbon taxes to meet forcing target

if (cm_iterative_target_adj eq 2, !! otherwise adjustment happens in core/postsolve.gms 
    

Iterative adjustment for carbon tax runs: scale current tax pathway with the ratio of target forcing s15_gr_forc_os to current forcing p15_forc_magicc. The offset is only there to increase the speed of convergence, the values have no physical meaning. For low stabilization targets (rcp2.0, rcp2.6, rcp3.7) the target is the 2100 forcing target (s15_rcpCluster eq 1), for lower targets the forcing target is valid during the full century (s15_rcpCluster eq 0).

  if (cm_emiscen eq 9, 
  
    display pm_taxCO2eq, s15_gr_forc_os, p15_forc_magicc;
  
    if (s15_rcpCluster eq 1,
       pm_taxCO2eq(t,regi)
       = 
         pm_taxCO2eq(t,regi)
       * ((p15_forc_magicc("2100") - s15_forcing_budgetiterationoffset_tax - max(0,(cm_startyear-2020)/20))
       / (s15_gr_forc_os       - s15_forcing_budgetiterationoffset_tax - max(0,(cm_startyear-2020)/20)))**1.2
       ;
     
     elseif (s15_rcpCluster eq 0),
       pm_taxCO2eq(t,regi)
       = 
         pm_taxCO2eq(t,regi) 
       * ((smax(tall,p15_forc_magicc(tall)) - s15_forcing_budgetiterationoffset_tax - max(0,(cm_startyear-2020)/20))
       / (s15_gr_forc_nte       - s15_forcing_budgetiterationoffset_tax - max(0,(cm_startyear-2020)/20)))**1.2
       ;
      );
    pm_taxCO2eq(t,regi)$(t.val gt 2110) = pm_taxCO2eq("2110",regi); !! to prevent huge taxes after 2110 and the resulting convergence problems, set taxes after 2110 equal to 2110 value
    display pm_taxCO2eq;
    );
);

Limitations There are no known limitations.

(B) off

Limitations There are no known limitations.

Definitions

Objects

Sets

Authors

Jessica Strefler, Michaja Pehl, Christoph Bertram, Gabriel Abrahão, Tonn Rüter

See Also

80_optimization, core

References