Atmospheric Sciences Editors' Vox

Managing Uncertainties in Climate Engineering

Control theory and climate engineering meet in a new special issue of JGR: Atmospheres.


Climate engineering, or geoengineering, looks at ways to temporarily offset climate change by deliberately modifying the climate system. One of the most studied ideas, which is the focus of a new special issue in the Journal of Geophysical Research: Atmospheres, involves stratospheric sulfate aerosols, mimicking the cooling experienced after large volcanic eruptions [Kravitz et al., 2018].

Previous research has shown that when volcanoes located at high latitudes erupt, sulfate aerosols tend to stay in a single hemisphere, whereas when volcanoes located in the tropics erupt, aerosols can be dispersed on a global scale. Therefore, to use climate engineering to cool the entire planet, injecting sulfur dioxide into the tropical atmosphere makes sense. However, the papers in this special issue begin to question those assumptions and, in doing so, broaden the conversation around climate engineering.

Simulations suggest that one of the side effects of such tropical injections may be that the tropics tend to be “overcooled” while the poles tend to be “undercooled.” Also, simulations show that the Northern hemisphere tends to cool more than the Southern hemisphere, which shifts tropical precipitation.

Rather than treating those as inevitable side effects, what if one chose a different location or multiple locations to inject? Or, instead of starting with the location of injection, what if one chose the climate objectives and then designed a climate engineering strategy to meet those objectives? These are some of the questions addressed by the papers in this special issue.

Climate engineering as a design problem

Simulated aerosol optical depth (colors) against observations (black) at three locations for the 1991 eruption of Mount Pinatubo. Credit: Mills et al. [2017], Figure 10
Simulating climate engineering with stratospheric sulfate aerosols requires having the right model. Mills et al. [2017] developed a version of the Whole Atmosphere Community Climate Model (WACCM) that captures complex processes like stratospheric chemistry or sulfate aerosol growth.

When tested, this model performed very well in reproducing atmospheric perturbations that were observed after the 1991 eruption of Mount Pinatubo.

The next stage was to understand the design space. How might the surface climate [Tilmes et al., 2017] and stratosphere [Richter et al., 2017] respond when sulfur dioxide is injected at different latitudes, altitudes, and magnitudes?

The group of scientists selected four locations for injecting sulfur dioxide into the stratosphere (30°S, 15°S, 15°N, and 30°N).

They found that if injection amounts are chosen correctly, climate engineering could offset changes due to greenhouse gases in global mean temperature, the inter-hemispheric temperature gradient, and the equator-to-pole temperature gradient [MacMartin et al., 2017].

Simulated global mean temperature (T0; top), interhemispheric temperature gradient (T1; middle) and equator-to-pole temperature gradient (T2; middle) for no geoengineering (RCP8.5) and with geoengineering (Feedback). Grey dashed lines represent the chosen temperature objectives. Bottom panel shows the annual injection amount at each of the four locations (colors) and the total injection amount (black). Credit: Kravitz et al. [2017], Figure 2
A key step is the introduction of control theory, which can be used to design a feedback strategy that would allow the objectives to be met in the presence of uncertainty, adjusting the amount of injection at each location independently every year of simulation.

The results of this effort were the first simulation in which stratospheric sulfate aerosol climate engineering was used to adjust multiple simultaneous injection locations to meet multiple simultaneous climate objectives [Kravitz et al., 2017].

This is an important step forward in understanding climate engineering, but this first effort was a demonstration that needs further development.

The stratospheric response, especially the effect on stratospheric winds, was investigated in this feedback simulation [Richter et al., 2018].

Another feedback simulation was conducted, with lower altitude injections, to understand the effects of injection altitude on surface climate and ozone [Tilmes et al., 2018a].

Of course, meeting the objectives requires detection; MacMartin et al. [2018] calculated detection time of changes for a variety of metrics.

What’s next?

These studies were limited in scope to modeling, but there is substantially more work to do. The design parameter space (latitude, altitude, and magnitude) has not been explored thoroughly, and this set of studies did not at all investigate the role of the time of year of injection. Moreover, these studies looked at broad-scale temperature features, which is a good first step, but there is a need to move toward more regional assessments and evaluation of impacts such as floods and droughts, as well as food, energy, and water security.

The Geoengineering Large Ensemble (GLENS) [Tilmes et al., 2018b] is a set of 20 stratospheric sulfate aerosol geoengineering simulations that has been made available to the scientific community as a resource to enable these sorts of analyses, further discussion, and build research capacity.

Understanding the potential role of climate engineering in addressing climate change is a daunting research task, and there is much more to do.

This special issue takes one step on what is hopefully a long road toward responsible evaluation of this topic.

—Ben Kravitz, Guest Editor of JGR: Atmospheres special issue, and Atmospheric Sciences and Global Change Division, Pacific Northwest National Laboratory; email: [email protected]

Citation: Kravitz, B. (2019), Managing uncertainties in climate engineering, Eos, 100, Published on 23 January 2019.
Text © 2019. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.