MIT Composite Model

The 2009, 2010, and 2011 Climate CoLab contests used the MIT Composite Model.

You can try out the model to see how it works.

Component models #

The MIT Composite Model combines simulation models of four primary types:

  • climate simulator
  • mitigation costs
  • damage costs
  • physical impacts

Climate simulator #

The C-LEARN climate simulator is at the heart of the MIT Composite Model. C-LEARN projects the impact of regional emission reduction and land use targets on:

  • global emissions,
  • atmospheric concentration of CO2,
  • increase in global mean temperature (GMT), and
  • sea level rise.

Economic costs of climate change mitigation #

For the 2010 and 2011 contests, response surfaces were generated from stabilization scenarios created for Stanford Energy Modeling Forum EMF 22 exercise.

These response surfaces were generated from published data generated by a series of runs of integrated assessment models.

For more see EMF 22 response surfaces.

Mitigation models are only valid for some ranges. Please see additional information here.

Mitigation cost estimates for the 2009 contest are from response surfaces based on scenarios generated by 3 integrated assessment models as part of a study undertaken by the U.S. Climate Change Science Program (CCSP).

For more on these, see:

Economic costs of climate change damages #

Economic costs of climate change damages associated with temperature increases were based on response surfaces generated from the damage function of 2 integrated assessment models:

Physical impacts of climate change #

Impacts of termperature increases are based on 2 reports:

The physical impacts models project impact of temperature increases in 6 areas:

  • Water Availability,
  • Agricultural production
  • Health,
  • Land (for example, impact of flooding),
  • Environment/Ecosystems, and
  • Abrupt/Large Scale Impacts.

How the MIT Composite Model works #

The inputs to the MIT Composite model are the same as the inputs to the C-LEARN simulator:

  • greenhouse gas emission reductions targets for 3 major groups of countries (Developed, Rapidly developing, and Other developing)
  • targets for reduction of deforestation and increases in carbon dioxide sequestration due to aforestation

These inputs are sent to C-LEARN, which generates the following outputs through the year 2100:

  • emissions for each major group of countries (in gigatons or Gt)
  • atmospheric concentration of CO2 (in parts per million or ppm)
  • increase in global mean temperature or GMT (in degrees Celsius)
  • sea level increase (in milimeters or mm)

Temperature increase is used as an input to response surfaces generated from complex, computationally demanding models.

These response surfaces are the basis for additional algorithms that reside in the Climate CoLab, which generate the following outputs through the year 2100:

  • estimated mitigation costs
  • estimated costs of climate change damages

Finally, temperature increase is used as an input to tables created from from the IPCC's AR4 and studies undertaken by the Tyndall Centre to generate the final output:

  • estimated physical impacts of climate change.