Is Climate Stabilization Still Possible?

Today's climate strategies (mitigation and adaptation) were built for a world of relative stability. But we're beyond the Paris 1.5°C target and heading toward a 2.5–3°C world. The rate of warming is accelerating faster than scientists predicted. Feedback loops are arriving sooner than modelled. The question is no longer just "how much will humans emit" but "how is the planet responding?" Climate is becoming a systems stability problem, and our institutions were built for the Holocene we no longer live in.

Earth used to reflect 30% of incoming sunlight. Over the past 25 years its reflectivity has fallen, driven by shrinking cloud cover, retreating ice, and aerosol pollution cleaned up for human health. The warming impact of that lost reflectivity is equivalent to the warming impact of every ton of CO₂ humanity has emitted since 1750, added to our planet again. NASA's CERES satellites confirm this. Climate models did not predict this. And critically: models that predict more future warming are the ones doing best at reproducing observed changes. We may need to revise warming projections substantially upwards.

The Thwaites Glacier is the keystone of the West Antarctic Ice Sheet. Its collapse alone could add up to 2 meters of sea-level rise: hundreds of millions of climate refugees, half living on less than $5 a day. 60–80% of all uncertainty in future SLR rides on this one ice sheet. The critical detail: warming was the trigger, but it is no longer the driver. Water is already under the bed, driving slippage. Even at net-zero emissions tomorrow, Thwaites keeps melting. The fuse is lit.

IPCC projections (used globally by governments, planners, and insurers) say sea levels will rise about half a meter by 2100 under 3°C warming. Data-driven extrapolation of observed trends over the past 60 years says SLR could be roughly 2× more. Current IPCC models even imply SLR slows under higher warming, which is physically inconsistent with what we observe. The gap in projections means tens of millions more displaced people. We're spending a lot to adapt to these outdated projections: New York's $52B storm surge barrier, Houston's $50B seawall, Jakarta's $40B. All of them are designed to outdated numbers.

There are multiple ways to reflect sunlight:

Surface albedo modification (painting roofs white, brightening deserts) works locally but barely touches global temperature.

Marine cloud brightening, spraying fine sea-salt into low clouds, might amplify their reflectivity, but models disagree by an order of magnitude and the precipitation side-effects could be large.

Space-based SRM puts reflective material between the earth and the sun to scatter light, but delivery would be extremely costly.

Stratospheric Aerosol Injection (SAI), releasing reflective particles into the stratosphere, appears most promising for even, global cooling.

Mt. Pinatubo as a model for SAI: June 1991, Pinatubo dumped 20 million tons of sulfur dioxide into the stratosphere. The planet cooled about 0.5°C, for a year. SAI is the most-studied sunlight-reflection candidate because we can see how it works at planetary scale, temporarily. But we don't know enough yet about SAI to understand the risks and benefits. Reflective is researching it deliberately, transparently, with governance. Research, not deployment: the wider the gap between accelerating climate risk and decision-grade SAI evidence, the higher the chance a panicked government deploys it badly.

The Arête Glacier Initiative's entire research program (better SLR forecasting plus testing whether glaciers can be refrozen to their beds with thermosyphons) costs $100M. Estimated deployment in Antarctica, if proven effective, is roughly $8B. Compare with one storm surge barrier in New York City: $52B. Houston: $50B. Jakarta: $40B. All sized to outdated projections. All reactive rather than working upstream. Climate stabilization research is asymmetrically cheap relative to the infrastructure being built to manage the failure mode.

Super pollutants like methane and nitrous oxide are responsible for roughly half of today's warming. Now rising temperatures are triggering more emissions from natural systems themselves: methane from tropical wetlands and permafrost, CO₂ and methane from wildfires, nitrous oxide from warming soils. These are climate feedbacks: the planet warming and emitting more greenhouse gas in response. They are not currently accounted for in climate policy or models. Like corporate expenses off the balance sheet. Without integrating them, mitigation plans are mis-budgeted. There may be ways to intervene in natural systems with different management practices to reduce the amount of emissions coming from them, and atmospheric methane removal is also being researched to enhance the atmosphere's natural processes of methane removal.

Every era of environmentalism has ruled the next era's tools taboo. Preservationists (1960s–1970s) rejected industrial decarbonization. Climate mitigators (1980s–2010s) rejected adaptation as surrender, methane abatement as distraction, carbon removal as fantasy. Adaptation got accepted only after climate impacts became undeniable. CDR and methane reduction got accepted only after the carbon budget was overshot. Each shift followed the same arc: scientific insight → institutional discomfort → mounting evidence → reframing → normalization. Stabilization is next on the line.

The deepest discomfort in this debate isn't whether to intervene. It's the assumption that intervention is something we might start doing. We are already engineering the planet. Burning fossil fuels alters atmospheric chemistry, ocean acidity, cloud formation. Cleaning up particulate pollution unmasks warming. Deforestation reshapes regional rainfall. Industrial aerosols cool the planet unevenly. Climate stabilization is the attempt to correct that history: to assess these risks on their merits and consider the most effective responses transparently, under governance.

Paradigm shifts (Thomas Kuhn) are changes in scientific understanding: they happen relatively fast, driven by scientists who notice anomalies in the current model. Episteme shifts (Michel Foucault) are deeper: changes in the unconscious structure of what counts as knowledge, what's culturally possible. They take generations. Copernicus moved the Earth in his lifetime. The episteme shift that followed, through Descartes, Bacon, and the Enlightenment, took centuries. The science of climate stabilization is the paradigm shift, already underway. The deeper move is the episteme shift: recognizing humanity as embedded inside a system of systems, not separate from it.

It's easy to feel climate is a lost cause. Erhart explicitly rejects optimism in favor of Rebecca Solnit's frame: blind hope (techno-optimism, degrowth utopianism, "AI's gonna fix it" or "we'll all move back to the land") and left despair ("it's so bad there's nothing we can do") both prevent honest engagement with the canvas we actually have. We don't have the world we wanted. We're not going to. But there is enormous space to decrease suffering and improve conditions for life on this planet.

Episteme shifts are not driven by scientists. They're driven by people willing to update their worldview in time for it to matter. There are now entire fields of Earth-system science, forecasting, stabilization research, governance, and risk reduction that barely existed a decade ago, and with them new chances to reduce suffering, cut catastrophic risk, and open the action window on climate. The opportunity is bigger than researching climate stabilization: it is to realign our incentives and institutions to meet our true interdependence with Earth systems. We have the chance to become wise ancestors.