A Long View · Atmospheric Carbon · June 2026

Where is it truly going?

Carbon dioxide has been far higher than today, and far lower. What is new is not the number but the speed — and the hand on the dial. A reading of where the carbon goes, whether it can return, and why the answer is physics in the air and justice on the ground.

PRE-INDUSTRIAL · 280 ppm TODAY · ~430 ppm LAST SEEN ABOVE THIS · ~14 million yrs ago
The band we lived in, and the wall we built
CO₂ through 800,000 years of ice-core record (ppm). For the entire span of the human story, carbon oscillated between roughly 180 and 300 ppm as ice ages came and went. The vertical stroke on the right is the last 250 years.
glacial–interglacial cycles the industrial spike
Source: ice-core composites (EPICA / Vostok) & NOAA Mauna Loa. The band is not a metaphor — it is the physical envelope within which every civilisation arose.

This is an attempt to build a bridge from the known to the unknown. It begins with the world as it is — the physics, the numbers, the unequal weight of them — then turns to a few real places where things were changed for the better, and from those builds, tentatively, some ideas for slowing the rate at which we add carbon to the air. It is a first map, not a last word.

There is a habit of consolation that runs through almost every conversation about carbon: the Earth has been hotter; CO₂ has been higher; this is natural variation. The first half is true; the second does not follow from it. What follows holds both halves at once — what deep time records, and what the physics dictates — before turning to the harder matter underneath: that this was never a question of personal virtue. It is a civilisational imperative, written in a ledger that the atmosphere keeps without regard for who signed it.

MOVEMENT I

The problem in deep time

CO₂ was once 1,600 parts per million and the poles were forests. The record is real. It is also the wrong clock.

Read across sixty-six million years and the consoling story looks true. Around fifty million years ago, in the hothouse of the early Eocene, atmospheric CO₂ reached as much as 1,600 ppm and global temperatures ran perhaps 12°C warmer than today. There were no ice sheets. Then carbon fell. By thirty-four million years ago it had dropped enough for the Antarctic ice sheet to begin forming; the last time CO₂ sat consistently above where it is now was about sixteen million years ago, near 480 ppm; by two and a half million years ago it had fallen to 270–280 ppm and the ice ages began their slow metronome. Modern humans appeared, around four hundred thousand years ago, into a world at or below that line — and stayed there until, about two hundred and fifty years ago, we began to move it ourselves.

So the deniers' premise is correct and their conclusion is a category error. The relevant variable is not the level but the rate, and the agency. Past CO₂ changes of this magnitude unfolded across hundreds of thousands to millions of years, slow enough for ocean chemistry, ice, vegetation and evolution to track them. We have done a comparable move in the time since the steam engine.

The giant caveat

Earth-system sensitivity describes climate over hundreds of thousands of years — melting ice, shifting forests, changing clouds. It consoles only those who do not have to live in the next two centuries. Deep time is not a reprieve. It is the measure of how long our two centuries will echo.

Sixty-six million years, descending — and the dot that is us
Reconstructed CO₂ (ppm) across the Cenozoic. A long cooling, the gift that let ice form and mammals diversify, ending in the narrow band of the human era — and a single near-vertical mark.
After the 66-million-year CO₂ synthesis (Columbia Climate School / CenCO₂PIP, 2023). Time axis compressed; the modern excursion is too fast to draw to scale without vanishing.

There is a second, quieter consequence that the temperature charts hide. About a quarter of the CO₂ we emit dissolves into the sea, where it becomes carbonic acid. Since the Industrial Revolution the surface ocean's pH has fallen from 8.21 to 8.10 — a 30% rise in acidity — eroding the ability of corals, molluscs and plankton to build the calcium structures the marine food web rests on. Even if warming stopped tomorrow, this chemistry would continue for as long as the carbon kept arriving. The sky is the famous crisis; the sea is the silent one.

MOVEMENT II

Where does it truly go — and can it return?

Most of us carry a tap-and-sink model: emit less, levels fall. It is the wrong model. Carbon is a stock, not a flow.

Ask where a molecule of CO₂ goes and the answer is reassuring: it cycles out of the air within about five years, traded with the ocean surface and the leaves of plants. But that is turnover, not removal. When a slug of fossil carbon is added, the excess concentration does not leave on that schedule. It decays in stages. Roughly half is drawn down by ocean and land within decades — which is why, of all we have ever emitted, about 45% still hangs in the air. Then the drawdown stalls: the surface ocean's buffering chemistry saturates, and further uptake must wait on the slow overturning of the deep sea, across centuries. And a stubborn fifth to a quarter of the original pulse remains aloft for millennia, removable only by the weathering of silicate rock — the planet's geological thermostat — on a clock of a hundred thousand years.

The long tail: what becomes of a pulse of carbon
Fraction of an emitted pulse of CO₂ remaining in the atmosphere over time. The fast drawdown is real and partial; the tail is effectively permanent on any human timescale.
Schematic after Archer et al. — carbon-cycle pulse-response. "Net-zero" is the moment the curve stops climbing; it is not a moment the curve comes down.

This reframes the entire problem, and it answers the question directly. Can it truly go away? Not on any timescale that means anything to a person, a nation, or a civilisation — not without removal at a scale we cannot yet perform. The bulk reverses over centuries if emissions stop; a meaningful remainder is, for us, forever.

The corollary is easily missed. Because warming tracks the cumulative total of emissions — the running integral, not the annual rate — temperature stabilises only at net zero. Emitting less does not lower the temperature; it slows the rate at which the temperature keeps rising. The Keeling curve below is the visible proof: through every recession, treaty and clean-energy boom of the last seventy years, the line has bent its slope but never turned down.

The Keeling curve — seventy years, one direction
Monthly mean CO₂ at Mauna Loa (ppm), 1958–2026. The annual sawtooth is the northern forests breathing in and out. The trend has never paused.
Source: Scripps / NOAA Global Monitoring Laboratory. From 315 ppm in 1958 to ~430 ppm in 2026; the annual rise now runs near 2.4 ppm.
MOVEMENT III

The geography and justice of carbon

The physics is indifferent to distribution. The atmosphere mixes every tonne and warms everyone the same. The ethics, by contrast, is almost entirely a matter of distribution.

A tonne of carbon from a private jet and a tonne from a coal-fired power plant warm the planet identically; the sky keeps no separate accounts. And yet the entire ethical and political weight of the problem lives in an accounting the sky refuses to keep — who emitted, who benefited, who suffers. Holding those two facts together resolves much of the public confusion at once.

~10%
of people produce roughly half of all emissions — the global top decile by income
~50%
of people — the poorer half of humanity — produce around a tenth
−54%
UK territorial emissions since 1990 — the figure that gets quoted
−20%
UK emissions once imports are counted — the figure that is true
Who is the carbon for?
Share of global income (left) against share of global emissions (right), by income group. Emissions are a near-mirror of consumption — and consumption is the thing that is distributed, not need.
After World Inequality Lab / Chancel. The marginal tonne reflects rising consumption among the already-wealthy more than rising numbers among the poor.

This is why the most popular response — the consumer's response — is the least load-bearing. An electric car is a good thing and a small thing: it touches light passenger transport, a single-digit slice of global emissions, and only helps to the degree the grid charging it is clean. The very idea of a personal "carbon footprint" was popularised by an oil company's advertising, and it has the effect of relocating a structural problem into private guilt. It is not a personal choice. It is a civilisational imperative — and the unit of action is the nation, the industry, the energy system, not the shopper.

There is a sharper twist in the national accounts. Rich economies that "decoupled" growth from emissions did so partly by exporting their furnaces. Britain's headline 54% cut is a territorial figure; counting the carbon embedded in what it imports, the real reduction is closer to 20%. The steel was not cleaned — it was moved, often to a dirtier grid, so the same beam may now carry more global carbon while appearing on no British ledger. Part of the gap is also structural: a mature economy has already poured most of its cement and laid most of its steel — the carbon-heavy work of building roads, railways, grids and cities is largely behind it. Much of the developing world has yet to do that building, so a large share of its future emissions is not luxury but foundation. A services economy is not, in itself, a climate strategy; the shift is also an accounting move available to any one country precisely because it is unavailable to all of them at once. The planet cannot offshore to Mars. Somewhere, the thing still has to be made.

The atmosphere integrates the world's emissions into a single number. Justice is the work of remembering whose number it is.

MOVEMENT IV

Survival, freedom, dignity

When the crisis is existential, very little matters but survival — and a large part of humanity still lives without the certainty of a daily meal. Any honest pathway begins there, or it is not honest.

There is a visceral truth underneath the data that the data can flatten. For perhaps two billion people, the climate question is not the first question. The first question is electricity that does not fail, a flame to cook on that does not fill the lungs with smoke, a harvest that holds, a child who lives. To tell them that the carbon budget is full — a budget spent overwhelmingly by others, over two centuries, to build the comfort from which the demand is now made — is, by any ordinary standard of fairness, indefensible. The right to a dignified life is not negotiable against a problem one did not create.

And yet the arithmetic, once examined, does not force the hard choice it seems to. Lifting the poorest several billion to a decent energy baseline adds a modest fraction to global emissions beside the consumption of the top tenth. The conflict between development and climate is, to a significant degree, a constructed one — and the fossil-fuel incumbent benefits from the belief in it. Three things weaken the binary:

Leapfrogging. The poor world need not re-walk the coal road any more than it laid copper wire before mobile phones. Distributed solar and storage now undercut grid-coal across much of Africa and South Asia; clean energy is frequently the faster and cheaper route to access, not its enemy.

The inversion. Climate is not a luxury concern set against Southern survival — it is a Southern survival concern, and disproportionately so. Wet-bulb heat crossing the limit of human endurance in the Indo-Gangetic plain, failing monsoons, drowning deltas: the people who burned the least inherit the most. A farmer whose land turns to dust has not lost an amenity. He has lost, in Amartya Sen's exact sense, a freedom.

The redefinition. Sen's insight is the one that dissolves the whole dilemma: development is the expansion of substantive freedoms — to be healthy, schooled, mobile, unafraid — and capability is not consumption. A high-dignity life is achievable at a fraction of the rich world's material throughput; Kerala's human development per tonne of carbon exceeds that of far wealthier places. The goal was never to have ten billion people consume at American levels, which the biosphere forecloses in any case. The goal is freedom — and freedom costs far less carbon than affluence does.

Sen — development as freedom

Poverty is the deprivation of capability, not merely of income. So is a flooded field, a poisoned cooking fire, a city above the wet-bulb line. Climate damage and underdevelopment are the same wound — the loss of what a person is substantively able to do and to be.

So the moral architecture is not "the South develops or the world decarbonises." It is: a carbon budget deliberately carved out and protected for the development of the poor; the burden of cutting, and the bill for clean leapfrogging, placed on those who already spent the budget; differentiated responsibility, climate finance and loss-and-damage treated as justice rather than charity. The poor world's emissions are not the problem to be managed. They are the claim to be honoured.

MOVEMENT V

Pathways as leverage, not gestures

Donella Meadows taught that not all interventions are equal: there is a hierarchy of places to push on a system, and we habitually push hardest on the weakest of them.

Before any pathway, the size of the task. Because warming tracks cumulative emissions, the room left is a finite quantity — and it is now small. To hold under 1.5°C on even odds, humanity can emit roughly 170 billion more tonnes of CO₂: about four years at the current rate of some 42 billion tonnes a year. The budget for 2°C buys perhaps twenty-five. These are not countdowns to begin acting; they are how little current emission the remaining room represents. Every pathway below is judged against that quantity.

The popular climate to-do list — switch the bulbs, sort the recycling, buy the electric car — sits at the very bottom of Meadows's ladder of leverage: tweaking parameters, adjusting numbers, the least powerful place to intervene in a system. The powerful places are higher and stranger: the rules, the goals, and at the summit, the paradigm — the shared idea out of which the whole system arises. Decarbonising the energy supply is a mid-ladder move of real force. Redefining the purpose of an economy away from throughput is the high-ladder move, the rarest and the most transformative.

LOW
ParametersLight-bulbs, recycling, the personal EV. Necessary, almost trivial. Where the public conversation lives.
MID
Flows & structureClean electricity, electrify everything reachable, then the hard core — steel, cement, ammonia, aviation, shipping. The real tonnage.
HIGH
Rules & goalsCarbon pricing, border adjustment (CBAM), the protected Southern budget, finance that flows downhill. Re-internalising the offshored tonne.
HIGHEST
ParadigmThe goal of the economy itself: throughput, or flourishing. Sen's freedom over consumption. The one lever that moves all the others.

It helps to know where the weight actually sits. In absolute terms the great blocks are power, industry, transport, agriculture and buildings — but the concentrated load, the part that is both large and genuinely hard to clean, lives in a short list of heavy industries. Measured by the carbon emitted per dollar of revenue — the dirtiness of the activity itself — the ranking is stark. Cement stands in a class of its own, more than double the intensity of power generation, because its carbon comes from the chemistry of turning limestone to lime, not merely from the furnace that heats it. Steel and aluminium follow, then oil and gas, then the utilities still burning coal and gas to make electricity. Far below sit the visible consumer economy — retail, software, finance, healthcare — among the least carbon-intensive activities there are.

The dirtiest dollar — carbon intensity by sector
Direct (Scope 1) CO₂ emitted per $1M of revenue, by sector. Not which sector emits the most in total, but which unit of economic activity is itself the most carbon-laden — the truer guide to where the hard work is.
Source: S&P Global, 2019–20 average Scope 1 intensity. Intensity is not total tonnage — but these heavy sectors are large in absolute terms too, which is what makes them load-bearing.

This is the quiet correction the chart delivers: the problem is not the consumer's phone or the office tower; it is the kiln, the smelter and the power station. These are the load-bearing industries — decarbonise this handful and most of the structural task is done; fail to, and nothing else adds up. They are also, not by accident, the same industries a services economy finds easiest to offshore and hardest to actually clean — the ones from Movement III that simply move when the accounting demands it.

The honest scorecard is neither despair nor triumphalism. What is working: the collapse in the cost of solar, wind and batteries — now the cheapest electricity humans have ever made — and a deployment surge led by China, at once the world's largest emitter and its largest installer of clean power. Thirty-five economies have now cut emissions for a decade while their economies grew — twice as many as the decade before. What is stuck: the hard-to-abate industrial core, where the chemistry itself (not just the fuel) emits carbon; land and agriculture, with their methane and nitrous oxide; and the political durability of any of it across electoral cycles. The future is therefore not foretold. It is a fan of paths.

What actually moves the hard middle of that ladder is not exhortation but price and policy. The lever is to make the cost of cleaning up part of the cost of the thing — a price on carbon, and a border adjustment so the cleaner producer is not undercut by the dirtier import (the device that re-internalises the offshored tonne from Movement III). Public money and public procurement then do for steel, cement and clean hydrogen what they did for solar: buy the cost curve down, through guaranteed-price contracts and first-of-a-kind plants, until the clean version is simply the cheaper one. And the Southern budget needs an enforceable plumbing of its own — concessional finance, technology transfer, the rich world paying for the leapfrog it can afford and the poor world cannot. None of this is mysterious. The obstacle is not knowledge; it is that the incumbents of the fossil economy benefit from delay, and the budget above is spent in the years it takes to overcome them. Which is why the decisive lever is finally political — and why the most consequential thing most people do about carbon is not what they buy, but how they vote.

The fan of futures — and the floor we cannot lift
Illustrative CO₂ pathways to 2100 (ppm). The gap between them is the whole of human choice. Beneath both runs the long tail: even on the steep path, the carbon already emitted does not return for millennia.
current trajectory rapid net-zero the committed floor
Illustrative, after IPCC AR6 scenario families. The point is not the exact line but the shape: the future is genuinely open above, and genuinely closed below.
MOVEMENT VI

Another way of living

Leverage in the abstract is easy to doubt. So here is leverage made real — two reversals from the world as it is, neither hypothetical, neither painless, both proof that the ledger can be turned.

Kitakyushu. The Japanese steel city on Kyushu grew through the early twentieth century into one of the country's great industrial zones, and by the 1960s its Dokai Bay was so fouled by factory effluent it was called the sea of death — water that could no longer hold even bacteria. The sky was no better. And then, within two decades, the blue came back and the fish returned to the bay of their own accord — while the factories kept running. The mechanism is this whole document in miniature. It began with citizens: women's associations and local residents forced the council to act. That pressure produced measurement; measurement produced pollution-prevention pacts with some thirty companies; the pacts produced redesigned industrial processes; and in 1970 the national parliament passed what came to be called the Pollution Diet. Not a single light-bulb was changed. The leverage was applied where Meadows said it must be — at the level of rules and goals — and it held.

The sobering half of the lesson is how rare the conditions are. Of 252 countries and territories, 158 still have no air-quality standard at all. The knowledge is not the bottleneck. The institutions, the civic pressure, and the political will are.

Windhoek. The second proof inverts a problem into a resource. Namibia is the driest country in sub-Saharan Africa, and its capital, short of water since the 1960s, did something almost no one else has dared: it drinks its own wastewater. The Goreangab plant has run direct potable reuse since 1968 — sewage treated nine or ten times into water cleaner than the ordinary tap. Singapore's NEWater now meets close to a third of national demand; Israel reuses nearly 90% of its effluent in agriculture. The sentence that matters is the reframing — wastewater is not a problem; it is an opportunity — set against the fact that some 80% of the world's wastewater is still simply discharged, while two billion people live without safe water.

A reciprocal obligation

The radical step the water case points toward is not virtue but price. Require those who draw on a shared system — beginning with industry — to return what they take, in the same form or better; and pay them enough that the cleaning is something they can afford to do. The whole economy is currently priced for extraction and blind to the cost of the extracted. Putting that cost back into the price is not a courtesy — it is a change in the goal of the system, the highest lever on Meadows's ladder.

The obvious objection must be met, because it is the right one: a price on carbon, left alone, falls hardest on the poor, for whom energy is a larger share of survival. But the answer is not to keep the fuel cheap — that mostly subsidises the rich, who burn the most; the broad fuel subsidy is a transfer to the comfortable presented as help for the poor. The answer is to price the carbon and return the money to people: a per-head dividend that, because the poor emit far less than the average, leaves the bottom of the distribution better off while the heavy consumer pays. Protect the survival tier — the cooking gas, the lifeline block of power — directly, and recall the arithmetic from Movement IV: the carbon cost of that floor is so small it is almost free to carve out. The price disciplines the carbon-heavy life at the top; it need not touch the dignity of the poor at all. Where this has been tried, the lesson is consistent — the compensation must arrive before the price rise, and visibly, or the reform tends to fail.

1980s
Kitakyushu's "sea of death" returns to fish — with the industry still running
158 / 252
countries & territories still without any air-quality standard
1968
Windhoek begins turning sewage into drinking water — and does so still
~80%
of the world's wastewater is discharged unreused — the opportunity untaken

Before drawing too much comfort, mark the difference. Kitakyushu and Windhoek were solvable for the very reason carbon is not: cause, effect and remedy sat in the same place. The citizens who forced the cleanup breathed the cleaned air; the city that recycled its water drank it. Local pollution is a commons where the sufferer is also the beneficiary, so self-interest and repair point the same way. Carbon is the exact inversion — the emitter is here and now, the harmed are elsewhere and unborn, and the one who gains from restraint is a stranger one will never meet. So these are not proof that carbon is easy. They are proof of something narrower and still vital: that the machine works — citizens, then measurement, then binding rules, then a price that tells the truth. The task is to build that same machine at planetary scale, and across exactly the displacement of cause from consequence that, until now, made it unnecessary to build.

What the cases share with carbon is one principle: return what you take. Put the bay back as you found it; put the water back, cleaner; stop adding to the air a gas that will outlast every institution now arguing about it. But hear it correctly. This is not a private courtesy owed by careful people — it is an instruction to the price of things: that what we pay should at last include the cost of cleaning up after what we buy. The world is built for extraction and silent on the true cost of the extracted, and changing that is a shift at the level of the system's goal — the rarest and highest leverage there is. It begins, as Kitakyushu began, with citizens insisting; and for most people that insistence is registered not at the till, but at the ballot.

MOVEMENT VII

The derangement, and the reimagining

Amitav Ghosh's charge is that this is, before it is anything else, a failure of imagination — that the forms through which modernity understands itself cannot quite hold the thing that is happening.

Ghosh calls it the Great Derangement: the strange inability of our dominant culture, our novels and economics and politics of the individual, to think a planetary, collective, slow-moving catastrophe at all. The same derangement produces the binary this document has spent its length dismantling — climate versus development — because it cannot imagine a flourishing that is not the Western consumption template, and so it offers the poor world a seat at a table where the food has already been eaten and the bill is about to arrive.

Notice, then, that the three thinkers are not three opinions. They are three doors into one room. Meadows says the highest leverage is the paradigm — the goal of the system. Sen says the goal was always freedom, and freedom is not consumption. Ghosh says the obstacle is that we cannot yet imagine such a goal as real. Put them together and the pathway stops being a list of technologies and becomes a single, demanding proposition: a conception of the good life that is generalisable to ten billion people. Carbon is the symptom that happens to carry a technical fix. The throughput definition of flourishing is the disease, and it does not.

And here is the turn the diagnosis demands. If the failure is one of imagination, the answer is not despair but invention — applied not to gadgets but to institutions. This is a problem humanity has genuinely never faced: a harm spread across the whole globe and deferred across centuries, with no precedent in law, economics or politics to copy. There is no template. It will be met, if it is met, by the same imaginative nerve the cases already show is possible — a city that chose to drink its own wastewater; a nation that paid every citizen a monthly dividend so that the true price of fuel could at last be felt without breaking the poor. Those were not technologies. They were acts of policy imagination, invented for problems that looked impossible until someone built the bridge across them. The work now is to imagine such instruments at planetary scale — and the burden of that imagining falls on the people we send to govern, which is to say, in the end, on the choices the rest of us make about whom to send.

The physics gives us a number the sky keeps for everyone. Justice is the work of remembering whose number it is. Imagination is the work of conceiving a life worth living that the number can afford.

The argument, in one breath

This is not a counsel of despair, and it is not the cheap optimism of the next gadget. It is the long view, which is the only honest one. The problem was made slowly, over two and a half centuries, by the steady accumulation of a single invisible gas. The response will also be long — measured not in news cycles but in the working lifetimes of institutions and the patient turning of a civilisation's idea of itself. The carbon already in the air will outlast everyone now alive. That is not a reason to look away. It is the reason the work is worth beginning, and beginning again.

The Record

Sources & data

Assembled from the notes collected at rough-notebook-6 · Atmospheric Carbon and the Another Way of Living series, extended with current readings. Charts are hand-drawn to the cited figures; where a series is illustrative it is marked as such.

DEEP TIME · 66-million-year CO₂ synthesis (CenCO₂PIP). Columbia Climate School — A New 66-Million-Year History of Carbon Dioxide.
THE MISCONCEPTION · Why "CO₂ was higher before" misleads. Time Scavengers — CO₂ Past, Present & Future.
MODERN SPIKE · ACIDIFICATION · 280→420+ ppm, +1.2°C, ocean pH 8.21→8.10. NOAA — Climate Change: Atmospheric Carbon Dioxide.
EMISSIONS GEOGRAPHY · Global emissions timelapse by nation. Global Carbon Atlas.
CURRENT CO₂ · ~430 ppm; 2026 annual ~429 ppm; rise ~2.4 ppm/yr. NOAA Global Monitoring Laboratory · Scripps Keeling Program.
THE BUDGET · ~170 GtCO₂ left for 1.5°C (~4 yrs); ~42 GtCO₂ emitted in 2025. Global Carbon Project — Global Carbon Budget 2025.
THE LONG TAIL · Multi-millennial persistence of a CO₂ pulse. After D. Archer, The Long Thaw; carbon-cycle pulse-response (IPCC AR6 WG1).
DISTRIBUTION · Top 10% ≈ ~50% of emissions; bottom 50% ≈ ~10%. World Inequality Lab (L. Chancel) — Climate Inequality Report.
OFFSHORED CARBON · UK territorial −54% vs consumption −~20% since 1990. UK ONS / DESNZ; Carbon Brief analysis (2024–26).
FUTURE PATHWAYS · Scenario families & net-zero commitment. IPCC Sixth Assessment, Working Group I & III.
A REVERSAL · AIR · Kitakyushu's recovery from the "sea of death," citizen-led, industry intact. Notebook — Another Way of Living: Reversing Air Pollution.
A REVERSAL · WATER · Windhoek & Singapore's potable reuse; the reciprocal-obligation principle. Notebook — Another Way of Living: Water & Wastewater.
THE FRAME · D. Meadows, Leverage Points · A. Sen, Development as Freedom · A. Ghosh, The Great Derangement.