Development and application of earth system models

Abstract

The global environment is a complex and dynamic system. Earth system modeling is needed to help understand changes in interacting subsystems, elucidate the influence of human activities, and explore possible future changes. Integrated assessment of environment and human development is arguably the most difficult and most important "systems" problem faced. To illustrate this approach, we present results from the integrated global system model (IGSM), which consists of coupled submodels addressing economic development, atmospheric chemistry, climate dynamics, and ecosystem processes. An uncertainty analysis implies that without mitigation policies, the global average surface temperature may rise between 3.5 °C and 7.4 °C from 1981-2000 to 2091-2100 (90% confidence limits). Polar temperatures, absent policy, are projected to rise from about 6.4 °C to 14 °C (90% confidence limits). Similar analysis of four increasingly stringent climate mitigation policy cases involving stabilization of greenhouse gases at various levels indicates that the greatest effect of these policies is to lower the probability of extreme changes. The IGSM is also used to elucidate potential unintended environmental consequences of renewable energy at large scales. There are significant reasons for attention to climate adaptation in addition to climate mitigation that earth system models can help inform. These models can also be applied to evaluate whether "climate engineering" is a viable option or a dangerous diversion. We must prepare young people to address this issue: The problem of preserving a habitable planet will engage present and future generations. Scientists must improve communication if research is to inform the public and policy makers better.

Fig. 1.

Framework and processes of the MIT IGSM. Feedbacks between the component models that are currently included, or proposed for inclusion in the next generation of the model, are shown as solid and dashed lines, respectively. GDP, gross domestic product. (Reproduced with permission from Massachusetts Institute of Technology.)

Fig. 2.

Probability distributions (frequencies) for global mean temperature change between the 1981–2000 average and the 2091–2100 average without a policy (leading to a 2091–2100 median 1,330 ppm-eq CO₂) and with median 560, 660, 780, or 890 ppm-eq CO₂ GHG stabilization policies (7, 8). The median and 95% confidence (2.5–97.5%) probability ranges are given in the legend for each case. Horizontal lines show 90% ranges (5–95%), and vertical lines indicate medians adapted from Webster et al. (8).

Fig. 3.

Wheels depicting the probabilities of increases in global average surface air temperature between the 1981–2000 and 2091–2100 decadal averages using the MIT IGSM without and with a stabilization policy, yielding a median 660 ppm-eq CO₂ in 2091–2100 (7, 8). Spinning versions of these wheels are available at http://globalchange.mit.edu. (Reproduced with permission from Massachusetts Institute of Technology.)