Today, we’re thrilled to publish our first-ever guest post on the ARC Initiative substack, from someone who has been a great collaborator and supporter of our work: Kelly Erhart.
Kelly is a close friend of ARC and the Climate Emergencies Forum (CEF), and a thoughtful voice working at the intersection of climate risk, planetary systems, and responsible innovation. Her career spans climate-tech entrepreneurship, nonprofit leadership, and now philanthropy at Outlier Projects, where she supports teams pushing forward the frontier of research on climate overshoot and catastrophic risk. Few people combine broad climate experience, systems thinking, and moral clarity the way she does.
Thanks for reading The ARC: Thoughts on a safe climate future! Subscribe for free to receive new posts and support my work.
In this piece, Kelly steps back to examine where we truly stand in the climate crisis, the “dashboard” of green, yellow, and red signals shaping our future, and explores how our current climate toolbox falls short of managing the destabilizing risks now emerging in the scientific literature. She lays out why mitigation, adaptation, and carbon removal remain essential and why they may not be enough on their own. Importantly, she makes the case for rigorously researching additional tools. Not as a replacement, but as climate insurance in a world where the stakes are rising sharply.
It’s a thoughtful, clear-eyed, and deeply human piece. Precisely the kind of analysis we hope to highlight more often.
We’re grateful to Kelly for opening this series with such a grounded and expansive exploration.
If you’ve spent time in conversations about climate change, the adage “there’s no silver bullet” will sound familiar. That’s because climate change is complex — it isn’t a single problem with a single solution. Climate change is predicated on a web of interlocking earth systems — oceans, ice sheets, forests, clouds, sunlight, atmospheric dynamics, energy systems, human systems, and more — each influencing the others in ways that can amplify risks or open up narrow windows of hope. The more science reveals about these feedback loops, the clearer it becomes that our understanding of climate change — and the impact that we as humans have on our planet — is a story that isn’t simple or linear. And it’s a story that is unfolding in real time.
I’ve been in awe of our planet for as long as I can remember. That awe made me a young environmentalist and compelled me to spend my career working on climate. After more than a decade working to advance climate solutions and research, I’m excited to share thoughts on where we are, the research pathways that intrigue me most, and musings on how I see the various earth, human, and system dynamics playing out in our current climate reality.
In this first post, I’ll explore:
Where we stand in climate today — introducing “the dashboard” concept
The most pressing climate risks we face in the future
An overview of current solutions — “the toolbox” and their efficacy at managing these risks
What an expanded “toolbox” for a safely managed climate might look like
Where We Stand: The Climate on a Knife’s Edge
The world’s warming trajectory has already passed the Paris Agreement’s 1.5 °C benchmark and is now dangerously tracking toward plausible scenarios of 2.5 - 3°C. Some call this “climate overshoot.” Long gone are the days when the consequences of climate overshoot were abstract — they’re increasingly visible, felt, and undeniable: in the record heat stifling communities, in the collapsing glaciers and ice sheets that are compromising arctic stability, in the wildfire smoke clogging lungs and skies around the world, and in the floods that are redrawing coastlines.
As I noted at the top, climate change is complex. Visualizing it through a metaphorical “dashboard” helps me hold that complexity. Just like a cockpit instrument panel, it lets us keep multiple signals in view at once: what’s stabilizing, what’s worsening, and what we still don’t understand. When we step back, the picture that emerges is mixed — some genuine wins, some wicked challenges, and some flashing alarms that demand urgent attention.
And then there are the red indicators — the signals that show the system itself is destabilizing, laid bare by the increasingly grim climate science that continues to surface. This category is driven by accelerants (climate feedback loops and potential tipping points) that act as multipliers, pushing warming further and faster — even if human emissions fall. These potentially catastrophic risks can transform a difficult century into an unsurvivable one. They’re not the incremental changes we can model easily, like slightly stronger storms or marginally more droughts. They’re the nonlinear shocks — runaway methane emissions from warming wetlands, permafrost and peatlands, abrupt ice-sheet collapse, cloud and albedo reductions, and possible ocean circulation slowdowns that could erase decades of progress in a single season. These shocks are increasingly referred to as “tipping points”* — a threshold beyond which key systems, like ice sheets, ocean currents, or rainforests, can rapidly shift into a new, often irreversible state. Climate science now warns that some tipping elements may already be close to their tipping threshold, or even crossed, raising the risk of harmful feedback loops (e.g., melting ice disrupting ocean currents, triggering more melting).
The tragic irony is that these cascading impacts don’t just directly harm ecosystems and humans. They also threaten our ability to keep decarbonizing and adapting. Mass displacement, food and water crises, an expansion of uninsurable properties in fire and flood zones, and infrastructure loss can drain political will and economic capacity right when we need more of both.
The reds on the dashboard are where the situation is most sobering. To avoid the outcomes that could be unsurvivable for our children or grandchildren, we need climate insurance policies that focus new attention where the stakes are highest. To build those, we need to explore potential additional “tools” for our climate management “toolbox”.
*Note: “tipping points” is a controversial, and not always accurate, way to talk about this suite of destabilizing risks for technical/definitional reasons. Outside semantics, what’s important to think about here are system changes that would cause outcomes to climate stability that would be irreversible on human timescales. In a future post, I’ll dive deeper into tipping points and overshoot risks: what they are, what’s at stake, and how they connect. But for now, you can explore an overview here.
Expanding the Toolbox
Tools 1 and 2: Mitigation and Adaptation
Note: This section covers current tools. If you’re interested in skipping directly to how I think we might address the red lights with research on an expanded toolset, scroll down to “Exploring New Tools.”
Drastically cutting emissions (mitigation) and helping the most vulnerable communities adapt to the locked-in impacts of climate change (adaptation) are non-negotiable global priorities that demand expanded support. Alongside this, let’s explore what’s possible with our current toolset.
CO2, the main driver of warming, is a gas that stays in the atmosphere for centuries. CO2 continues to build up in the atmosphere and, as a consequence, the needed emissions reductions for a safe climate have become steeper. At this point, the emissions reduction trajectories required to limit warming to 2 °C look highly implausible. While it is no longer possible to limit warming to 1.5 °C through emissions reductions alone, efforts to reduce emissions are more important than ever. They’re the foundation of climate action — the green and yellow lights we must keep pushing forward to flip the yellows to green as quickly as possible.
But if warming continues to outpace our plans, or if a catastrophic risk scenario materializes, disasters themselves could derail those very efforts. In that world, the long game of decarbonization becomes harder to sustain. Mass migration, political unrest, authoritarian backlashes, collapsing insurance markets, humanitarian crises — all of these can pull resources and political will away from decarbonization and conservation. In the face of destabilization, governments and societies could prioritize short-term crisis management over long-term emissions reduction.
Dashboard Detour: Risks that Threaten These Tools
Let’s explore one specific risk: catastrophic sea level rise, driven by glacier collapse.
If the West Antarctic Ice Sheet (or Thwaites Glacier) were to collapse, we could see one to two meters of sea level rise this century. That would displace hundreds of millions of people — more than the displacement of World War II — overwhelming governments, economies, and humanitarian systems. In that world, political will and resources would be consumed by migration management, refugee crises, and social instability. And, in the face of that kind of destabilization, it would likely become much harder to sustain long-term commitments to decarbonization, conservation, or carbon removal.
This is just one of the risk scenarios that underpins the urgent need to research additional stabilization tools: to help prevent runaway disasters from overwhelming our ability to continue cutting emissions in the first place. To be clear: This should not replace mitigation and adaptation, but instead help us see if there are safe and effective ways to build an insurance layer against the reds on the dashboard.
Stay tuned for individual pieces on other risks like this; and the research pathways that may help us to predict, slow, or prevent them.
Tool 3: Carbon Removal
I spent my career building climate solutions—from microbial sanitation systems to direct air capture and ocean-based carbon removal. After exploring carbon dioxide removal (CDR) solutions at depth, I cofounded Vesta to advance ocean alkalinity enhancement (OAE), which is one of (in my opinion) the most compelling pathways to do large-scale, durable CDR.
After working on CDR for a decade, I am heartened by the immense progress that’s been made in the field. But it’s also abundantly clear that the speed of CDR deployment is a drop in the bucket compared to what we need to achieve our climate goals. The scale is daunting: we’ll need ~10 gigatons of removal annually. Today, pilot projects have been deployed at only the 10,000s tonnes/year level. That’s like producing a few drops of CDR when we need an Olympic-sized swimming pool. Carbon dioxide removal is more necessary and urgent than ever, but if we’re thinking about the reds on the dashboard, it is not likely to scale fast enough to meet the pace or magnitude of the crisis alone.
To reiterate: tools one, two, and three are critical and urgent. And, necessary as they are, alone or even deployed together, they will likely be insufficient to manage the catastrophic risks underpinning the reds on the dashboard.
Climate change is often thought of solely as a greenhouse gas emissions problem. While GHG emissions are one component of the challenge, climate change is driven by more than just an increase in GHGs. A necessary understanding is that climate change is also driven by the Earth’s overall energy balance and reflectivity (more on Earth’s energy imbalance in a future post), and by complex, interdependent processes that drive the feedback loops that can create dangerous, likely unadaptable, and in some cases perhaps irreversible, impacts to humans and ecosystems. The responsible next step is to ask whether and how an expanded set of tools can offer greater leverage to avert climate disasters.
Guess what: there may be. These potential tools are often pinned with the pejorative “geoengineering” (I’ll explore a “hot-take” — no pun intended — on this topic in a future post). Some of the approaches that could fit under this umbrella term are actually broadly accepted and non-controversial (think: carbon dioxide removal). Many are under-researched and carry significant risk. These tools can, rightly, provoke moral hazard concerns. But the times require us to rigorously research and evaluate them.
Unfortunately, even researching these tools is often criticized as hubristic. But humility and prudence require reckoning with uncomfortable possibilities. Political obstacles could continue to slow decarbonization, warming could continue to exceed expectations, and climate disasters could outpace our ability to adapt — or even derail our best efforts to decarbonize. While it is, indeed, hubris simply to wish away these possibilities, the truth is that humans have long been intervening in Earth’s systems — through fossil fuel combustion, deforestation, and pollution.
The burden of these inequitable interventions falls hardest on the poorest and most vulnerable, who contributed least and have the fewest resources to adapt. It’s not intellectually honest to ask whether humans should engineer the planet; we already are. The question is whether we can learn to do it responsibly: guided by feedback, transparency, and fairness. Expanding the toolbox with equity in mind means investing in knowledge that could prevent cascading harm to billions who cannot build seawalls or buy air conditioning. Irreversibility is coming either way; what matters is whether we do everything still within our power to minimize avoidable suffering while we can.
Adaptation measures were initially dismissed as “giving up on mitigation.” This may have been plausible when societal impacts were still minimal, but as they grew and became impossible to ignore, protecting communities became essential. In general, climate management tools that are potentially safe, effective, scalable, and affordable should be researched.
In the face of the compounding climate risks that we face today, if we are to be responsible planetary stewards, what additional tools might we explore? If climate science tells us anything, it’s that the baseline risks are deeper and more destabilizing than conventional wisdom suggests — but it also reveals something hopeful: the opportunity space for action may be bigger, more varied, and more actionable than we thought. There is a much wider, more varied climate toolbox that may be available to us if we choose to safely research and evaluate additional potential tools in time.
In a series of future posts, we’ll explore catastrophic climate risks in more depth, along with some of the tools that warrant future research, including: frontier methane and superpollutant abatement and removal, open-system carbon dioxide removal pathways, sunlight reflection, glacier stabilization, stabilizing the Arctic, critical ocean observations and forecasting, and pathways for preserving biosphere stability. Stay tuned for that post, thoughts on who should lead, and how to get this research done.
About the Author
Kelly Erhart funds and supports teams accelerating research on tools to address climate overshoot and catastrophic climate risk at Outlier Projects, a philanthropy. Previously, she was a climate non-profit founder and entrepreneur. Erhart’s background spans climate technology research, development, and commercialization; nonprofit leadership; entrepreneurship; and philanthropy.
Thanks for reading The ARC: Thoughts on a safe climate future! Subscribe for free to receive new posts and support my work.
Water vapor governs 70% of our atmosphere’s heat dynamics according to Walter Jenn. Restoration of the small water cycles, and the biotic pump, including restoring forests, grasslands, aquifers and rivers, is a powerful and safe tool - why is it that climate scientists like yourself barely mention water except for the words drought and flood?
Great post!
Water vapor governs 70% of our atmosphere’s heat dynamics according to Walter Jenn. Restoration of the small water cycles, and the biotic pump, including restoring forests, grasslands, aquifers and rivers, is a powerful and safe tool - why is it that climate scientists like yourself barely mention water except for the words drought and flood?