by the Joint Program on the Science and Policy of Global Change of the Massachusetts Institute of Technology (MIT)
How effective and costly would a policy be in alleviating human-forced climate change? What are the advantages and risks of waiting for better scientific understanding? Which nations, regions and economic sectors face the greatest risks to unimpeded global change, and can we substantially reduce these risks through adaptation or mitigation?
Decisions made under these sorts of questions ultimately come down to an issue of risk. Policymakers, stakeholders, and local officials are increasingly relying on scientific climate information to help answer these questions. The MIT Integrated Global System Model (IGSM) is central to this effort of projecting the possible social, economic and environmental consequences of climate change.
The IGSM brings together the human, natural, and managed systems of our global environment. This “integrated” approach is critical because we often cannot directly measure the impacts of human development on the environment. Therefore, we must form computer models of the combined natural and human systems, compare the models with observations, and then apply the models in “numerical experiments” that assess the influence of human activities on the Earth system and how the response of the Earth system, in turn, will affect human systems.1
The IGSM framework has been developed and refined by the Joint Program on the Science and Policy of Global Change since the early 1990s. It’s currently being put to work in the developed world, and in developing nations through our work with the United Nations University-World Institute for Development Economics Research. From Zambezi, Africa, to the state of Colorado in the United States of America, its insights assist nations, sectors and communities in learning how to grow more efficiently and adapt to vital present and future challenges such as water management and energy resources.2
Integrated Assessment: Two Components
The IGSM is a “framework” of linked sub-models of varying complexity. Depending on the issues and specific research questions being addressed, users can choose which sub-models to use and add layers of complexity where needed.3
The two primary components are:
- The Emissions Predictions and Policy Analysis (EPPA) model, which analyzes human activity as it interacts with climate processes and assesses proposed policy measures; and
- An Earth system model that couples dynamic and chemical atmosphere, ocean, and natural biogeophysical and biogeochemical exchange models within a Global Land System framework. It analyzes the terrestrial biosphere interactions and feedbacks.
Economics, Emissions and Policy Analysis
The EPPA model is a multi-sector, multi-region computable general equilibrium (CGE) model of the world economy. It provides projections of world economic development and emissions, along with analysis of proposed emissions control measures. It is used to analyze the processes that produce emissions and to assess the consequences of policy proposals – providing estimates of the magnitude and distribution among nations of their costs and clarifying the ways that changes are mediated through international trade.
EPPA uses the Global Trade Analysis Project dataset (maintained at Purdue University), augmented by data on the emissions of greenhouse gases, aerosols and other relevant species, as well as taxes and details of selected economic sectors.4
The model projects economic variables – GDP, energy use, sectorial output, consumption, etc. – and emissions of greenhouse gases – CO2, CH4, N2O, HFCs, PFCs and SF6 – and other air pollutants – CO, VOC, NOx, SO2, NH3, black carbon, and organic carbon – from combustion of carbon-based fuels, industrial processes, waste handling, and agricultural activities. Different versions of the model have also been formulated for targeted studies to provide consistent treatment of feedbacks of climate change on the economy, such as effects on agriculture, forestry, biofuels and ecosystems, and interactions with urban air pollution and its health effects.
We utilize an efficient, flexible Earth system model with a hierarchy of complexities to facilitate investigations of feedbacks and uncertainties between model components and with human drivers and mitigation goals. It couples several submodels: atmospheric chemistry; atmospheric dynamics; oceanic dynamics; oceanic biogeochemistry; and terrestrial ecosystems. These model components are as close as possible to state-of-the-art – coupling together various configurations while maintaining computational efficiency, and enabling extensive testing of these phenomena. In one configuration, models of atmospheric dynamics and chemistry, thermodynamic sea-ice, land ecosystem and biogeochemistry, and a mixed-layer ocean representing the processes of heat and carbon uptake are combined. This configuration is MIT’s most computationally efficient Earth system model, and allows us to explore climate uncertainties by performing thousands of simulations. In another configuration, we employ a three-dimensional (3-D) model of ocean circulation, marine biology, and chemical processes that control the biogeochemical cycling of carbon, nutrients and alkalinity. In both of the configurations above, the Earth system component also includes an interactive atmospheric chemistry module, and an urban air chemistry component.
Changes in land ecosystems due to climate changes are important considerations in policy discussions. Additionally, climate-driven changes in the terrestrial biosphere affect climate dynamics through feedbacks on both the carbon cycle and the natural emissions of trace gases. The terrestrial component of the IGSM includes hydrologic and ecologic models in a Global Land System framework. Hydrologic processes and surface-heat fluxes are represented by the Community Land Model (CLM), which is based on a multi-institutional collaboration of land models. Within the IGSM, CLM is dynamically linked to the global Terrestrial Ecosystems Model (TEM), developed by The Ecosystems Center at the Marine Biology Laboratory.
TEM is used to simulate the carbon dynamics of terrestrial ecosystems. Driven by dynamic inputs from both TEM and CLM, methane and nitrogen exchange are considered through the Natural Emissions Model (NEM). The coupled CLM/TEM/NEM model system represents the geographical distribution of global land cover and plant diversity through a mosaic approach in which all major land cover types and plant functional types are considered over a given domain, and are area-weighted to obtain aggregate fluxes and storages.
Central to the IGSM framework is the building in of uncertainty to account for key human influences, such as the growth of population and economic activity, the pace and direction of technical advance, and the response of the Earth system to these human drivers.
To investigate the feedbacks and uncertainties between model components and with human drivers and mitigation goals, the most efficient IGSM configuration of intermediate complexity is used and the model is run hundreds of times with each study. Every run is given slightly varied input parameters, selected so that each run has about an equal probability of being correct based on present observations and knowledge. Doing this gives a more realistic assessment of the range of potential future effects.
Putting this approach into practice, we, for example, analyse temperatures and find that the world could warm from 3.5˚ to as much as 6.7˚C by the end of the century.5 To illustrate temperature uncertainty, we’ve developed roulette-style wheels known as the Greenhouse Gamble Wheels. The face of each wheel is divided into colored slices, with the size of each slice representing the estimated probability of the temperature change in the year 2100 falling within that range. One wheel represents an unconstrained emissions (“no policy”) outcome, while the other depicts the outcome “with policy.”
In making these analyses, we are able to help decision-makers compare the value of various mitigation policies, energy technologies and adaptation strategies to lower the risk of global climate warming. We can also assess the costs for stabilization of greenhouse gases at various levels, and how these costs can be justified by the expected benefits from avoided damages.
Looking at emissions-control scenarios, for example, we’ve found that even relatively modest emissions-control measures can have a large impact on reducing the odds of the most extreme warming outcomes. If we immediately reduce global emissions there is about a 50-50 chance of stabilizing the climate at a level of no more than a few tenths above the 2˚C target – a level that is considered likely to be a tipping-point, above which potential severe effects from climate warming ensue.6
Even given this analysis, there is always a level of “deep uncertainty,” which describes physical relationships in the Earth system that are currently unknown. We cannot precisely predict some phenomena because the global environment involves complex and dynamic interacting processes that are not all fully understood, many of which have chaotic elements that fundamentally limit the predictability of the climate system. Even looking at the relationships we have expected and been measuring, we’ve experienced some surprises – such as the Arctic ice melting faster than any of the models predicted. Along with others in the field, we face the challenge that the changing climate may bring some significant costs that may not be evident until after we witness them.7
Incorporating a regional scale
From this discussion, one can see how the model is useful at the global scale. But as the threat of climate change grows, the importance of assessing the regional impacts grows along with it. As stated at the beginning of this article, local officials depend on such analysis to guide them through critical decisions.
Understanding the increased importance of determining the likelihood of regional climate effects, MIT has created a “hybridized” approach that widens the scope and flexibility of the analysis. By collecting emergent climate-change patterns from the community of climate-model projections analysed from the Coupled Model Intercomparison Project (CMIP) in conjunction with the International Panel on Climate Change (IPCC), MIT has combined these with the IGSM to develop hybrid frequency distributions (HFDs) that can quantify the likelihood of particular regional outcomes. To characterize the prevailing climate patterns that alter human emissions, we characterize each climate models’ spatial responses, relative to their zonal mean, from transient increases in trace-gas concentrations and then normalize these responses against their corresponding transient global temperature responses. This procedure allows for the construction of meta-ensembles of regional climate outcomes, by combining these patterns to the aforementioned ensembles of the MIT IGSM’s zonal climate outcomes – which then produce climate projections, with uncertainty, under different global climate policy scenarios – with regionally-resolved patterns. This hybridization of the climate-model longitudinal projections with the global and latitudinal patterns projected by the IGSM can, in principle, be applied to any given state or flux variable that has the sufficient observational and modeled information (from the CMIP archives). The approach consistently ties together the socio-economic data of different emission scenarios and various levels of uncertainty in the global and regional Earth-system response.
In our initial study using this approach, we find that by the middle of this century – while some regions are affected by emission reduction measures more than others – when comparing business-as-usual with a greenhouse gas stabilization scenario, lowering emissions does reduce the odds of regional warming. In fact, the most extreme warming outcome from the business-as-usual case is eliminated entirely. At the same time, the odds of regional precipitation changes are seen to both increase and decrease by the middle of this century. However, when greenhouse gas concentrations are lowered through the stabilization scenario, the greatest likelihood of regional precipitation change moves toward more benign values by the end of the century. Stabilization also reduces the chances of more extreme precipitation changes.8
Specifically, these distributions of regional climate outcomes have been directly applied to assessments of climate risk for developing countries and have most recently focused on the Zambezi river basin. ii In this study, we consider the odds (i.e. the distribution) of changes that could be expected in important hydro-climate variables – precipitation (shown below) and surface-air temperature – under unconstrained emissions and global economic growth, as well as a modest stabilization scenario (Level 2 stabilization achieves 660 CO2 equivalent concentration by 2100). Changes in these quantities during the spring and summer have notable impacts on agricultural productivity as well as transportation infrastructure (i.e. roads, bridges, etc.). The unconstrained emissions outcome shows the “most likely” outcome (seen by the mode of the distribution) to convey a dry and warmer (not shown) situation – with a small chance (about 10 per cent) of conditions at least twice as dry as the most likely outcome. However, there is also a small chance that very wet conditions may occur – and these conditions present the greater risk in damage to the transportation infrastructure. In the stabilization case, the occurrence of these extreme outcomes are removed from the distributions – and the most likely outcome (nearly 50 per cent of the distribution, more than twice as likely in the unconstrained case) now lies at half of the drying (i.e. reduced precipitation) seen in the unconstrained case.
This hybridized approach is an immediate way to apply the full capabilities of the IGSM – that is, a probability analysis of the integrated natural and human systems – to a regional scale. Overall, this approach helps decision and policy makers make long-term decisions that will impact the future direction of planning in their nations. While the hybridized method brings much-needed progress in projecting the regional impacts of climate change, MIT’s ongoing improvements of the IGSM will include more explicit modelling of these regional features. We hope such added complexity will further build upon the capabilities toward fine-tuned regional assessments.
1 Prinn, R.G., 2012: Development and application of earth system models. Proceedings of the National Academy of Sciences, doi: 10.1073/pnas.1107470109. [back]
2 Arndt, C., P. Chinowsky, K. Strzepek, F. Tarp, and J. Thurlow, 2012: Economic Development under Climate Change. Review of Development Economics, Special Issue: Climate Change and Economic Development, 16(3): 369–377.[back]
3 Sokolov, A.P., P.H. Stone, C.E. Forest, R. Prinn, M.C. Sarofim, M. Webster, S. Paltsev, and C.A. Schlosser, 2009: Probabilistic Forecast for Twenty-First-Century Climate Based on Uncertainties in Emissions (Without Policy) and Climate Parameters. J. Climate, 22(19): 5175–5204.[back]
4 Paltsev, S., J.M. Reilly, H.D. Jacoby, R.S. Eckaus, J. McFarland, M. Sarofim, M. Asadoorian, and M. Babiker, 2005: The MIT Emissions Prediction and Policy Analysis (EPPA) Model: Version 4, MIT JPSPGC Report 125, August, 72 p.[back]
5 Joint Program on the Science and Policy of Global Change, 2012: 2012 Energy and Climate Outlook, MIT JPSPGC Special Report, March, 13 p.[back]
6 Webster, M., A.P. Sokolov, J.M. Reilly, C.E. Forest, S. Paltsev, A. Schlosser, C. Wang, D. Kicklighter, M. Sarofim, J. Melillo, R.G. Prinn, and H.D. Jacoby, 2012: Analysis of climate policy targets under uncertainty. Climatic Change, 112(3-4): 569–583.[back]
7 Reilly, J.M., S. Paltsev, K. Strzepek, N.E. Selin, Y. Cai, K.-M. Nam, E. Monier, S. Dutkiewicz, J. Scott, M. Webster, and A. Sokolov, 2012: Valuing Climate Impacts in Integrated Assessment Models: The MIT IGSM. Climatic Change, in press.[back]
8 Schlosser, C.A., X. Gao, K. Strzepek, A. Sokolov, C.E. Forest, S. Awadalla, W. Farmer, 2012: Quantifying the Likelihood of Regional Climate Change: A Hybridized Approach. J. Climate, doi: 10.1175/JCLI-D-11-00730.1.[back]