Twenty-First-Century Climate: Future Projections and the Potential for Surprise (Invited Presentation)

Human emissions of greenhouse gases now overwhelm the influence of natural drivers on Earth’s climate. Existing concentrations already commit the world to at least an additional 0.6°C (1.1°F) of warming over this century, relative to the last few decades, and global mean atmospheric CO₂ concentration has now passed 400 ppm. This threshold was last experienced about 3 million years ago when, over millennial time scales, global mean temperature was approximately 2.0°–3.5°C (3.6°–6.3°F) above preindustrial, and sea level was somewhere between 20 ± 10 meters (66 ± 33 feet) higher than today. Continued growth in CO₂ emissions over this century and beyond would lead to concentrations not experienced in tens of millions of years, and the present-day emissions rate of nearly 10 gigatons of carbon per year suggests that there is no precise past climate analog for this century any time in at least the last 50 million years.

How will human emissions shape climate over the twenty-first century and beyond? Over the next two decades, global temperature is projected to increase an additional 0.3°C to 0.7°C (0.5°F to 1.3°F), depending primarily on natural variability and emissions of short-lived species, from CH₄ and black carbon that contribute to warming to SO₂ and other aerosols that have a net cooling effect. Beyond the next few decades, however, the magnitude of climate change depends primarily on cumulative emissions of greenhouse gases and the sensitivity of the climate system to those emissions.

For the period 2081-2100 relative to 1986-2005, projected changes in global temperature range from 2.6°–4.8°C (4.7°–8.6°F) under the higher RCP8.5 scenario to 0.3°–1.7°C (0.5°–1.3°F) under the lower RCP2.6 scenario. Although no probability is attached to any specific scenario, the observed acceleration in carbon emissions over the past 15–20 years has been consistent with higher scenarios. Since 2014, however, emission growth rates have slowed as economic growth has become less carbon-intensive: but not yet at a rate that would meet the goals of the Paris Agreement, which seeks to limit global temperature change to no more than 2°C (3.6°F), and preferably 1.5°C (2.7°F). Moreover, to stabilize climate—which is the long-term goal of the United Nations Framework Convention on Climate Change—it is not enough to halt the growth in carbon emissions. Global net carbon emissions will eventually need to reach zero, and to have a greater than 50% chance of limiting warming below 2°C, most recent economic scenarios require net emissions that are not just zero but negative (in other words, a net human uptake of CO2 from the atmosphere) before the end of the century.

Self-reinforcing cycles, often referred to as positive feedbacks, are the wildcard in the climate system when it comes to predicting the net response of the climate system to human forcing. Carbon and methane emissions from Arctic permafrost, for example, have the potential to significantly amplify both local and global warming, and reduce the amount of carbon from human activities that can still be emitted if the world is to meet global temperature targets such as the Paris Agreement. Other Earth system components, such as Arctic sea ice and the Greenland and Antarctic ice sheets, may exhibit thresholds beyond which self-reinforcing cycles can drive a component of the climate system, or even the entire system, into radically different states, such as ones with greatly diminished ice sheets or different large-scale patterns of atmosphere or ocean circulation. Although the probabilities of these state shifts may be difficult to assess, their consequences could be high.

While global climate models incorporate important climate processes that can be well quantified, they do not include all of the processes that can contribute to feedbacks and abrupt and/or irreversible changes. For this reason, future changes outside the range projected by climate models cannot be ruled out. In addition, the systematic tendency of climate models to underestimate temperature change during warm paleoclimates suggests that climate models are more likely to under- than to over-estimate the amount of long-term future change. Together, the range of plausible future projections and the potential for unforeseen consequences emphasize the importance of quantifying and limiting the influence of human forcing on climate.