Managed stratospheric aerosol injection phaseout under deep carbon removal: cross-model and aerosol-lifetime sensitivity
Using two reduced-complexity climate models and varying aerosol lifetimes, this study demonstrates that under a deep carbon dioxide removal pathway, a feedback-controlled phaseout of stratospheric aerosol injection successfully prevents dangerous temperature overshoots and ensures no positive temperature trends occur within 20 years after withdrawal.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine the Earth's atmosphere as a blanket that traps heat. When we burn fossil fuels, we thicken that blanket, causing the planet to warm. Scientists have long agreed that the only permanent solution is to thin the blanket by removing carbon dioxide from the air. However, because the planet is already warming, some researchers have proposed a temporary, emergency measure: spraying tiny particles into the upper atmosphere to reflect a small amount of sunlight back into space. This would act like a sunshade, cooling the planet quickly while we work on removing the carbon. The most dangerous part of this idea is not starting the cooling, but stopping it. If the sunshade is removed while the carbon is still thick in the atmosphere, the planet could warm up violently and rapidly, a scenario known as "termination shock." The critical question is whether we can design a system that slowly turns off the sunshade as we remove the carbon, avoiding that sudden spike in heat.
A new study by independent researcher Jonathan Karl Bily explores exactly this scenario using computer models. The research tests a specific strategy where a computer-controlled system manages the injection of these cooling particles. This system is designed to be a "feedback controller," meaning it constantly checks the Earth's temperature and adjusts the amount of cooling it provides. If the planet gets too hot, it adds more particles; if it gets too cool, it reduces them. Crucially, this experiment assumes a very aggressive and difficult path for removing carbon dioxide from the atmosphere, far beyond what current plans usually suggest. The goal was to see if this smart controller could naturally guide the cooling system to a complete stop without causing a dangerous rebound in temperature, even if the cooling particles linger in the sky for a while.
To test this, the researcher ran the same computer program in two different climate models, which are simplified versions of the Earth's complex climate system. These models were fed the same instructions: a steady, deep removal of carbon dioxide starting in the near future, rising to massive levels by the mid-21st century and continuing for over a hundred years. The controller was set to keep the global temperature below a specific limit, gradually lowering that limit over time. The study also tested a more realistic version of the cooling mechanism. In the simplest models, the cooling effect disappears the moment the injection stops. In reality, the particles stay in the sky for several months to a year and a half before falling out. The researcher added this delay to the simulation to see if the lingering effect would confuse the controller or cause problems when the system tried to shut down.
The results were remarkably consistent across all the tests. When the carbon removal process was deep and sustained, the controller successfully managed the transition. As the carbon dioxide in the atmosphere dropped, the need for the sunshade naturally decreased. The system reduced the injection of particles year by year until the cooling effect vanished completely. In every single simulation where the system had enough time to finish the process, the temperature did not spike upward after the cooling stopped. Instead, the temperature continued to fall or stayed steady. The study found that even when the particles lingered in the atmosphere for up to a year and a half, the controller compensated for this delay without needing to be reprogrammed or hitting its maximum limits. The temperature trajectory remained smooth, and the system effectively "phased out" the intervention rather than abruptly cutting it off.
However, the study also highlighted what this experiment did not prove. The success of the phaseout depended entirely on the assumption that humanity could remove carbon dioxide at a massive, sustained rate. The computer models showed that if the carbon removal stopped or slowed down, the cooling system would not be able to phase out safely. Furthermore, while the global average temperature behaved well in these simulations, the models used were too simple to predict what would happen to specific regions. They could not tell us if certain areas would get too much rain or too little, or how the ozone layer or ocean chemistry would react. The study also noted that the "over-cooling" observed in some runs—where the planet got slightly colder than the target before the system corrected itself—was a feature of the specific computer rules used, not a physical law.
The research concludes that a managed, gradual end to solar cooling is theoretically possible, but only under very specific and demanding conditions. It is not a standalone solution, nor is it a guarantee of safety. The study demonstrates that if we can successfully remove enough carbon from the atmosphere, a smart control system can turn off the sunshade without causing a sudden temperature shock. But this outcome is entirely conditional on the success of the carbon removal effort. Without that deep reduction in greenhouse gases, the experiment suggests that stopping the cooling would indeed lead to the rapid warming that scientists fear. The work serves as a proof of concept for the control logic, showing that the transition can be smooth in a computer model, but it leaves the heavy lifting of making that transition real to the difficult work of carbon removal and the complex, unresolved questions of regional climate impacts.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.