Climate Change Modeling

Climate change modeling is a scientific discipline that uses mathematically-based representations to enhance understanding and prediction of future climate change, and to assess strategies for climate change mitigation and adaptation. A wide variety of models are currently in use by policy-makers and researchers, for estimating future warming and its impacts, as well as costs of climate change mitigation and the role of technology and policy in reducing costs of mitigation. This review covers energy models, integrated assessment models, and Earth system models; historically, these three types of models have been developed independently by different groups, and used for different purposes. Recently, there has been an effort in the modeling communities to harmonize the assumptions of the different types of models¹, where possible, in order to produce more consistent information with a broader support base.

Energy models

Energy models are used in forecasting near- and medium-term energy supplies, demands, and prices, generally for informing governments and businesses about future energy prices and needs for capacity expansion in the energy sector. While these models do not directly address climate change, they are nevertheless relevant for future climate change due to the global energy system's heavy reliance on fossil fuels, and the established links between fossil fuel combustion and climate change. In the United States, the Department of Energy forecasts are based on the National Energy Modeling Systems (NEMS)², which uses detailed, technology-oriented linear optimization models. The International Energy Agency(IEA) also uses MARKAL³ models of national and regional energy systems in its forecasts, published annually in the World Energy Outlook.⁴

Several points bear mention with regard to the use of energy models. First, investments in the energy sector are typically capital-intensive and long-lived; a power plant built today will be producing power for decades. The decision to invest in new capacity is informed by expected future revenue, which is highly dependent on future energy prices. Therefore, energy models and forecasts of energy prices are important for present-day energy investment decisions that have implications decades into the future. Note also that historically, energy price forecasts have been notoriously inaccurate, particularly for oil prices.⁵ Second, carbon dioxide (CO2) emissions may be estimated from future primary energy consumption based on the carbon contents of the fuels, as the CO2 emissions intensity of fossil fuel combustion is relatively constant for all end uses. However, emissions of all other greenhouse gases (e.g. CH4 and N2O) are difficult to infer from the output of the energy models. Finally, energy models are limited in their application to climate change as the time frame for analysis is generally no more than three decades into the future.

Integrated assessment models

Integrated assessment models (IAMs) are used to explore the socioeconomic and technological drivers of greenhouse gas emissions and the policies for constraining these emissions from a long-term, global perspective. Integrated assessment models use scaled-down submodels from a variety of disciplines, placing representations of emissions-producing sectors (e.g. energy, agriculture) and representations of the Earth system into a single common framework. This allows analysis of scenarios of future technological improvement, policies, or socioeconomic developments, as they relate to climate change, climate change mitigation, and costs of mitigation. Prominent integrated assessment models based in the United States include EPRI's MERGE model⁶, PNNL's MiniCAM model⁷, and MIT's EPPA-IGSM.⁸ Internationally, other commonly cited integrated assessment models are the AIM model (Japan)⁹, the IMAGE model (the Netherlands)¹⁰, and the MESSAGE model (Austria).¹¹

Integrated assessment models represent greenhouse gas emissions from the energy and agricultural systems using a single, internally consistent framework that captures both bottom-up and top-down effects. For instance, models of consumer choice are used to economically represent decisions between different emissions-producing technologies, and these decisions are made within the context of changing fuel prices owing to diminishing fossil resource supplies, possible future carbon taxes, and technological change in the energy supply sectors. An important feature of this type of model is that future energy prices not only influence technology choice and levels of service demand by consumers; energy prices are themselves influenced by consumer demands. In contrast to marginal abatement cost curves (MACs), in which specific technological options for carbon mitigation are implicit to the curves assumed, integrated assessment models explicitly represent technological options for reducing carbon emissions. This allows whole-system analysis of the costs of greenhouse gas mitigation.

Integrated assessment models are generally linked with representations of the physical Earth system that are scaled down from actual Earth system models. These models, such as the MAGICC model¹², are designed to produce estimates of global average greenhouse gas concentrations and temperature change that are consistent with the full Earth system models, but with minimal computing requirements and little regional detail. However, for understanding and predicting future climate change in specified locations (e.g. planning for adaptation), more complex Earth system models are used.

Earth System Models (ESMs)

Earth system models (ESMs) are data-intensive, spatially explicit representations of the Earth's atmosphere, oceans, and land. They are used for estimating future atmospheric gas concentrations and climate change at specified locations over time. The future composition of the atmosphere will depend not only on greenhouse gas emissions, but on a variety of biological and chemical processes that are highly complex. The most prominent ESMs are run on supercomputers; examples are the models developed and maintained by the National Center for Atmospheric Research (USA)¹³, Meteorological Research Institute (Japan)¹⁴, or the Max Planck Institute for Meteorology (Germany).¹⁵ The IPCC maintains a full review of ESMs.¹⁶ More simplified Earth Models of Intermediate Complexity are generally capable of replicating the results of the full ESMs (though with coarser resolution), and are used for assessing vulnerability and strategies for adaptation.